Original Research

Anterior cervical discectomy and fusion (ACDF) has been performed with various techniques and devices for many years. Autologous iliac crest grafts were initially used for the Cloward^1,2 and Robinson and Smith³ techniques, but because of iliac crest graft site complications (eg, pain, infection, fracture, dystrophic scarring^4,5), the procedure was generally superseded by allograft implants. These implants were then supplemented with anterior locking plate devices. More recently, unitary devices combining a polyetheretherketone (PEEK) spacer with screw or blade fixation have been developed, such as the Zero P (Synthes, Inc.) and the ROI-C cervical cage (LDR). Bone graft is required to fill the cavity of these devices and to promote osseous union. Demineralized bone matrix,⁶ tricalcium phosphate,^7,8 and bone morphogenetic protein (BMP) have been used for these purposes, but they add expense to the procedure and have been associated with several complications (eg, neck swelling, dysphagia associated with BMP).⁹

Although multiple studies have demonstrated effective fusion rates and good outcomes for both iliac crest autograft and grafting/spacer constructs, the debate over cost and “added value” remains unresolved. One institution, which has published articles reviewing the spine literature and its own data, concluded that iliac crest autograft was the most cost-effective and consistently successful ACDF procedure.^5,10

The VA Portland Health Care System (VAPORHCS) has analyzed the use of local autograft sources at the surgical site to circumvent the need to make a second incision at the iliac crest and, theoretically, to decrease risks and expenses associated with iliac crest autograft, allograft bone, and artificial constructs. Given the paucity of data on this method, the case series presented here represents one of a few studies that analyze local autograft for promotion of arthrodesis in a PEEK spacer device.

This article will report on the prospectively collected results of consecutive cases performed by Dr. Ross using a ROI-C cervical cage for 1-level anterior cervical discectomy between August 2011 and November 2014. This study received institutional review board approval.

Methods

Neck disability index (NDI) forms were used to assess the impact of neck pain on patients’ ability to manage in everyday life. The NDI form was completed before surgery and 3 and 9 months after surgery.

Dr. Ross preferred to perform minimally invasive posterior cervical foraminotomy for unilateral radiculopathy. Therefore, all patients with radiculopathy had bilateral symptoms or a symptomatic midline disc protrusion not accessible from a posterior approach. Standard techniques were used to make a left-side approach to the anterior cervical spine except in cases in which a previous right-side approach could be reused. Under the microscope, the anterior longitudinal ligament and annulus were incised, and the anterior contents of the disc space were removed with curettes and pituitary rongeurs. Care was taken to remove all cartilage from beneath the anterior inferior lip of the rostral vertebral body and to remove a few millimeters of the anterior longitudinal ligament from the rostral vertebral body without use of monopolar cautery (Figure 1). A 2 mm Kerrison punch then was used to remove the anterior inferior lip of the rostral vertebral body, and this bone was saved for grafting. No bone wax was used within the disc space.

After all disc space cartilage was removed from the endplates, additional bone was obtained from the uncovertebral joints and posterior vertebral bodies as the decompression proceeded posteriorly. Occasionally, distraction posts were used if the disc space was too narrow for optimal visualization posteriorly. After decompression was achieved, a lordotic ROI-C cervical cage was packed in its lumen with the bone chips and impacted into the disc space under fluoroscopic guidance. The blades were impacted under fluoroscopic guidance as well. The wound was closed with absorbable suture.

Antibiotics were given for no more than 24 hours after surgery. Ketorolac was used for analgesia the night of the surgery, and patients were asked to not use nonsteroidal anti-inflammatory drugs for 3 months after surgery. Lateral radiographs were obtained 3 and 9 months after surgery and every 6 months thereafter until arthrodesis was detected.

Results

Seventy-seven consecutive patients underwent 1-level anterior cervical discectomy (Table 1). Twenty-four procedures were performed for radiculopathy, 52 for myelopathy, and 1 for central cord injury sustained in a fall by a patient with preexisting spinal stenosis. Surgery was performed at C3-C4 (25 cases), C4-C5 (11 cases), C5-C6 (15 cases), and C6-C7 (1 case) for patients with myelopathy. Surgery was performed at C3-C4 (2 cases), C4-C5 (3 cases), C5-C6 (9 cases), and C6-C7 (10 cases) for patients with radiculopathy.

Twenty-eight patients reported presurgery tobacco use. Although all tobacco-using patients agreed to cease use in the perioperative period, at least 9 admitted to resuming tobacco use immediately after surgery. Eighteen patients had diabetes mellitus. In 2 patients, a diagnosis of osteoporosis was made with dual-energy X-ray absorptiometry. One patient was a chronic user of steroids before and after surgery. Mean body mass index (BMI) was 30.6, and 13 patients were morbidly obese (BMI > 34).

In 2 cases, only a single blade was placed. The second blade could not be placed because of broken adjacent screws (1 case) or undetermined reason (1 case).

The mean time for follow-up was 17 months (range 3-34). Four patients were lost to follow-up: 3 after the 1-month postoperative visit and 1 with severe psychiatric problems after hospital discharge.

There were no new neurologic deficits, no wound infections, and no recurrent laryngeal nerve palsies in the 77 patients. Eight months after surgery, 1 patient with radiculopathy underwent foraminotomy at the index level for persisting foraminal stenosis. Two patients whose myelopathic symptoms persisted after surgery returned for minimally invasive posterior laminotomy to remove infolded ligamentum flavum. The presurgery and 3- and 9-month postsurgery NDI scores were available for 52 patients (Table 2). Before surgery the mean NDI score was 24 (range 8-40). Three months postsurgery the mean NDI score was 15 (range 2-27) for patients with myelopathy and 13 (range 2-28) for patients with radiculopathy. The patient with the highest NDI score (28) stated that though all his symptoms were relieved, he had gauged his responses to protect his disability claim. Nine months after surgery, the mean NDI scores were 9.5 (range 5-17) for patients with myelopathy and 6 (range 2-13) for patients with radiculopathy. No NDI score was higher postsurgery than presurgery.

Arthrodesis was defined as bony bridging between the adjacent vertebral bodies and the bone graft within the lumen of the device, anterior to the device, or posterior to the device. In Dr. Ross’ protocol, computed tomography (CT) scans or flexion-extension radiographs were obtained only if pseudarthrosis was suspected to avoid unnecessary radiation exposure. Sixty-six patients had at least the 3-month radiography follow-up available. All 52 patients with 9-month follow-up data achieved complete arthrodesis, as determined by plain film radiography. Bridging ossification was found anterior to the device in all but 9 patients. Trabeculated bone was growing through the lumen of the device in all cases (Figure 2). A broken blade without clinical correlation was noted on imaging for 1 patient.

The total cost of the ROI-C cervical cage (LDR) for VAPORHCS was $3,498, or $1,749 for the PEEK spacer plus $1,749 for 2 metal blades. In comparison, the total cost of a typical anterior locking plate would have been $6,700, or $3,200 for the plate plus $2,000 for 4 screws and $1,500 for an allograft fibular spacer. Demineralized bone matrix (1 mL) as used in cervical arthrodesis by other surgeons at VAPORHCS cost $279, or about $500 including shipping.

DISCUSSION

Anterior cervical discectomy with fusion is a very common and successful surgical procedure for cervical myelopathy, radiculopathy, and degenerative disease that has failed to be corrected with conservative therapy.¹⁰ Medicare data documented a 206% increase in 1-level fusion procedures for degenerative spine pathology performed between 1992 and 2005.¹¹ When a procedure is performed so often, it is appropriate to review methods and analyze efficacy, cost, and cost-effectiveness.

According to a 2007 meta-analysis, the fusion rates of 1-level ACDF arthrodesis at 1-year follow-up are 97.1% in patients treated with anterior plates and 92.1% in patients treated with noninstrumented fusion.¹² The rate disparity was larger for multiple-level fusion: 50% to 82.5% for instrumented cases^12,13 vs 3% to 42% for noninstrumented cases.^14-16 Given the higher fusion rates achieved with instrumentation, surgeons have favored its use in ACDF.

Computed Tomography Use

Computed tomography has long been considered the gold standard for assessing arthrodesis outcomes (eg, Siambanes and Mather).¹⁷ However, recent data on potential harm caused by CT-related ionizing radiation suggest a need for caution with routine CT use.^18,19 For cervical spine CT, Schonfeld and colleagues found that the risk for excess thyroid cancers ranged from 1 to 33 cases per 10,000 CT scans.²⁰ According to another report, “limiting neck CT scanning to a higher risk group would increase the gap between benefit and harm, whereas performing CT routinely on low-risk cases approaches a point where its harm equals or exceeds its benefit.”¹⁹ As some have questioned even routinepostoperative use of radiation in patients with unremarkable clinical courses—patients should be spared unnecessary exposure—CT scans or flexion-extensionradiographs were obtained at VAPORHCS only if clinical symptoms or radiographs were suggestive of pseudarthrosis.²¹ As none of the VAPORHCS patients had those symptoms, none underwent postoperative CT.

For anterior cervical arthrodesis, surgeon preference determines which of many different bone substrates can be used with instrumentation, which impacts the costs. Fusion substrates include structural autografts, structural allografts, morselized autografts, morselized allografts, demineralized allografts, porous ceramics and metals, and BMP. Given these many options, studies comparing the constructs are lacking, especially with regard to the cost of alternative fusion constructs that produce similar outcomes. The Centers for Disease Control and Prevention defines cost-benefit analysis as a “type of economic evaluation that measures both costs and benefits (ie, negative and positive consequences) associated with an intervention in dollar terms.”²² It has been reported that using iliac crest autografts with anterior plate instrumentation is the most cost-effective method, yet alternatives remain in use.^5,10

For ACDF, iliac crest bone is an ideal and widely used construct substrate. Structural grafts harvested from the crest provide significant stability due to their bicortical or tricortical configuration with interposed osteoinductive and osteogenic cancellous bone. Few graft complications (eg, graft resorption) and no immunogenic or infectious complications have been reported for iliac crest bone. However, autologous iliac crest increases operative time, and donor-site morbidity has been reported.^23,24 A retrospective questionnaire-based investigation by Silber and colleagues, who evaluated iliac crest bone graft site morbidity in 1-level ACDF, found that 26.1% of patients had pain at the iliac crest harvest site, and 15.7% had numbness.²⁴ Other complications, which occurred at lower rates, were bruising, hematoma, pelvic fracture, and poor cosmesis.^23,25 In addition, osteoporosis and comorbid conditions have made it a challenge to acquire iliac crest autograft, contributing to the popularity of alternative substrates.²⁵

Allografts

An alternative to autografts, allografts have the advantages of reduced operative time and reduced donor-site morbidity.²⁶ Major historical concerns with allografts have included risk for disease transmission, costs associated with sterilization and serologic screening of grafts, and lack of oversight, leading to human allografts being acquired from dubious sources and ending up in the operating room.^27,28 Two major types of allografts are available: mineralized and demineralized.

Arthrodesis rates are inferior for mineralized (structural) allografts with instrumentation than for autografts with instrumentation.²⁹ In addition, smoking and other comorbidities have influenced fusion rates more in allograft than autograft fusions.^30-33 However, allografts are being widely used because they avoid the donor-site morbidity associated with autografts and because they are load bearing, can provide structural stability and an osteoconductive matrix, and can be used off the shelf without adding much time to surgery.

Demineralized matrix substrates are commercial osteoconductive and osteoinductive biomaterials approved for filling bone gaps and extending graft when combined with autograft.^7,8 Despite their osteoinductive properties, these substrates have had a high degree of product inconsistency, in some cases leading to poor outcomes.³⁴ The lack of randomized studies with these constructs has made the determination of clear indications a challenge.

The initial enthusiasm over use of BMP, another bone-graft substitute for cervical fusion, was curtailed by reports of adverse events (AEs). Effective in anterior lumbar spine fusions, BMP was adapted to off-label use in the cervical spine a few years ago.³⁵ Initial studies by Baskin and colleagues and Bishop and colleagues showed its fusion rates superior to those of allograft.^31,32 Both studies reported no significant AEs. However, studies by Dickerman and colleagues and Smucker and colleagues demonstrated increased soft-tissue swelling leading to dysphagia and prolonged hospitalization, which were attributed to higher dosage (no study has identified a precise dose for individual patients).^36,37 In addition, the cost of BMP is higher than that of any other bone-graft option for ACDF.³ Osteolysis has also been reported with BMP use.^38-40 Carragee and colleagues highlighted the potential carcinogenicity of BMP, but this finding was not corroborated by Lad and colleagues.^41,42

Cost Considerations

In addition to surgical effectiveness, spine surgical device costs have come under increased scrutiny.^43-45 In 2012, plates were reported to cost (without overhead or profit margin to hospitals) between $1,015 and $3,601, and allograft spacers were estimated to cost between $1,220 and $3,640, cage costs ranged from $1,942 to $4,347, and PEEK spacers cost from $4,930 to $5,246.⁵ Individual surgeon instrumentation costs varied 10-fold based on the fusion constructs used.⁵

In a cost-effectiveness review of anterior cervical techniques, cage alone was the least expensive technique, disc arthroplasty or cage/plate/bone substitute groups were the next most expensive, and autograft alone was the most expensive option due to hip graft site morbidity.⁴³ In another study, operative time associated with harvesting an iliac crest graft was equivalent in cost to that of an interbody cage.⁴⁴ Other studies have compared the costs of various anterior cervical fusion constructs.^9,10,45,46 A limitation of these studies is that autologous bone often refers to iliac crest grafts rather than local autograft. Epstein reviewed data from these studies and concluded, “ACDF using dynamic plates and autografts are the most cost effective treatment for anterior cervical discectomy,” citing a cost of $1,015 for this construct.⁵ Although Epstein demonstrated the cost-effectiveness of autograft in an individual surgeon’s hands, the results also are significant in that the studies identified areas in which improvements can be made at other institutions. The ROI-C cervical cage and local autograft bone cost that the authors report is at the lower end of the range reported by Epstein.⁵

Device explant rates also can be a concern. Operative waste was well described in a retrospective analysis of 87 ACDF procedures.⁴⁷ The study found that the cost of explanting devices implanted during the same intraoperative period was equivalent to 9.2% of the cost of permanently implanted constructs. Epstein addressed operative waste by using educational modules to evaluate spine surgeons’ decision making before and after education. After the intervention, the institution noted a marked decline in costs related to explanted devices—from 20% in 2010 (before education) to 5.8% of the total cost of implanted devices in 2010 (after education).⁵

In the present study, the authors demonstrated that use of local morselized autograft with a PEEK spacer for 1-level ACDF had excellent arthrodesis rates and minimal complications. Of the 52 patients with 9 month postoperative data, all achieved arthrodesis regardless of tobacco use. This method compares favorably with other fusion options in terms of radiographic arthrodesis rates. In addition, it avoids the donor-site morbidity associated with autografts from an iliac site but maintains the benefits of the osteogenic, osteoconductive, and osteoinductive properties of autograft bone. Use of local autograft avoids the costs associated with iliac crest autograft, including increased operating and anesthesia time, additional operating room supplies (drapes, sutures, etc) needed for operating at a second site, and prolonged hospital stay due to pain at the donor site. Use of local autograft also obviates complications at a second surgical site; purchase, storage, and sterilization of allograft; and the neck swelling, possible carcinogenicity, and cost of purchase of BMP. Other than the occasional reuse of distraction posts, this method involves no other expensive explant supplies.

Autografts have osteogenic, osteoconductive, and osteoinductive properties, and autograft fusion rates are generally superior to allograft fusion rates. Bone morphogenetic protein fusion rates may be comparable to autograft fusion rates.^9,26,32 Shortcomings of iliac crest autografts include increased operative time, blood loss, and donor-site morbidity. Allografts are osteoconductive and osteoinductive, but their fusion rates are inferior to those of iliac crest autografts. Other shortcomings are infection transmission and immunogenicity risks, higher graft resorption and collapse rates, cost, and previous issues relating to provenance. Bone morphogenetic protein is the most osteoinductive material with fusion rates similar to those of autograft, but its use is associated with neck swelling, dysphagia, osteolysis, potential carcinogenicity, and high cost.⁹

Conclusion

Overall, use of local autograft with a PEEK spacer has all the advantages of iliac crest autograft along with the benefit of working within the same operative window as the ACDF, thus reducing the infection, bleeding, and pain risks that may be encountered with a second incision. This procedure is effective, inexpensive, and cost-effective compared with alternatives and may be preferable for 1-level ACDF. In a population of patients with high rates of tobacco use, diabetes mellitus, obesity, and other factors that negatively affect fusion rates, local autograft may be a good choice for efficacy and cost savings.

Acknowledgments
The authors thank Shirley McCartney, PhD, for editorial assistance and Andy Rekito, MS, for illustrative assistance.

Anterior cervical discectomy and fusion (ACDF) has been performed with various techniques and devices for many years. Autologous iliac crest grafts were initially used for the Cloward^1,2 and Robinson and Smith³ techniques, but because of iliac crest graft site complications (eg, pain, infection, fracture, dystrophic scarring^4,5), the procedure was generally superseded by allograft implants. These implants were then supplemented with anterior locking plate devices. More recently, unitary devices combining a polyetheretherketone (PEEK) spacer with screw or blade fixation have been developed, such as the Zero P (Synthes, Inc.) and the ROI-C cervical cage (LDR). Bone graft is required to fill the cavity of these devices and to promote osseous union. Demineralized bone matrix,⁶ tricalcium phosphate,^7,8 and bone morphogenetic protein (BMP) have been used for these purposes, but they add expense to the procedure and have been associated with several complications (eg, neck swelling, dysphagia associated with BMP).⁹

Although multiple studies have demonstrated effective fusion rates and good outcomes for both iliac crest autograft and grafting/spacer constructs, the debate over cost and “added value” remains unresolved. One institution, which has published articles reviewing the spine literature and its own data, concluded that iliac crest autograft was the most cost-effective and consistently successful ACDF procedure.^5,10

The VA Portland Health Care System (VAPORHCS) has analyzed the use of local autograft sources at the surgical site to circumvent the need to make a second incision at the iliac crest and, theoretically, to decrease risks and expenses associated with iliac crest autograft, allograft bone, and artificial constructs. Given the paucity of data on this method, the case series presented here represents one of a few studies that analyze local autograft for promotion of arthrodesis in a PEEK spacer device.

This article will report on the prospectively collected results of consecutive cases performed by Dr. Ross using a ROI-C cervical cage for 1-level anterior cervical discectomy between August 2011 and November 2014. This study received institutional review board approval.

Methods

Neck disability index (NDI) forms were used to assess the impact of neck pain on patients’ ability to manage in everyday life. The NDI form was completed before surgery and 3 and 9 months after surgery.

Dr. Ross preferred to perform minimally invasive posterior cervical foraminotomy for unilateral radiculopathy. Therefore, all patients with radiculopathy had bilateral symptoms or a symptomatic midline disc protrusion not accessible from a posterior approach. Standard techniques were used to make a left-side approach to the anterior cervical spine except in cases in which a previous right-side approach could be reused. Under the microscope, the anterior longitudinal ligament and annulus were incised, and the anterior contents of the disc space were removed with curettes and pituitary rongeurs. Care was taken to remove all cartilage from beneath the anterior inferior lip of the rostral vertebral body and to remove a few millimeters of the anterior longitudinal ligament from the rostral vertebral body without use of monopolar cautery (Figure 1). A 2 mm Kerrison punch then was used to remove the anterior inferior lip of the rostral vertebral body, and this bone was saved for grafting. No bone wax was used within the disc space.

After all disc space cartilage was removed from the endplates, additional bone was obtained from the uncovertebral joints and posterior vertebral bodies as the decompression proceeded posteriorly. Occasionally, distraction posts were used if the disc space was too narrow for optimal visualization posteriorly. After decompression was achieved, a lordotic ROI-C cervical cage was packed in its lumen with the bone chips and impacted into the disc space under fluoroscopic guidance. The blades were impacted under fluoroscopic guidance as well. The wound was closed with absorbable suture.

Antibiotics were given for no more than 24 hours after surgery. Ketorolac was used for analgesia the night of the surgery, and patients were asked to not use nonsteroidal anti-inflammatory drugs for 3 months after surgery. Lateral radiographs were obtained 3 and 9 months after surgery and every 6 months thereafter until arthrodesis was detected.

Results

Seventy-seven consecutive patients underwent 1-level anterior cervical discectomy (Table 1). Twenty-four procedures were performed for radiculopathy, 52 for myelopathy, and 1 for central cord injury sustained in a fall by a patient with preexisting spinal stenosis. Surgery was performed at C3-C4 (25 cases), C4-C5 (11 cases), C5-C6 (15 cases), and C6-C7 (1 case) for patients with myelopathy. Surgery was performed at C3-C4 (2 cases), C4-C5 (3 cases), C5-C6 (9 cases), and C6-C7 (10 cases) for patients with radiculopathy.

Twenty-eight patients reported presurgery tobacco use. Although all tobacco-using patients agreed to cease use in the perioperative period, at least 9 admitted to resuming tobacco use immediately after surgery. Eighteen patients had diabetes mellitus. In 2 patients, a diagnosis of osteoporosis was made with dual-energy X-ray absorptiometry. One patient was a chronic user of steroids before and after surgery. Mean body mass index (BMI) was 30.6, and 13 patients were morbidly obese (BMI > 34).

In 2 cases, only a single blade was placed. The second blade could not be placed because of broken adjacent screws (1 case) or undetermined reason (1 case).

The mean time for follow-up was 17 months (range 3-34). Four patients were lost to follow-up: 3 after the 1-month postoperative visit and 1 with severe psychiatric problems after hospital discharge.

There were no new neurologic deficits, no wound infections, and no recurrent laryngeal nerve palsies in the 77 patients. Eight months after surgery, 1 patient with radiculopathy underwent foraminotomy at the index level for persisting foraminal stenosis. Two patients whose myelopathic symptoms persisted after surgery returned for minimally invasive posterior laminotomy to remove infolded ligamentum flavum. The presurgery and 3- and 9-month postsurgery NDI scores were available for 52 patients (Table 2). Before surgery the mean NDI score was 24 (range 8-40). Three months postsurgery the mean NDI score was 15 (range 2-27) for patients with myelopathy and 13 (range 2-28) for patients with radiculopathy. The patient with the highest NDI score (28) stated that though all his symptoms were relieved, he had gauged his responses to protect his disability claim. Nine months after surgery, the mean NDI scores were 9.5 (range 5-17) for patients with myelopathy and 6 (range 2-13) for patients with radiculopathy. No NDI score was higher postsurgery than presurgery.

Arthrodesis was defined as bony bridging between the adjacent vertebral bodies and the bone graft within the lumen of the device, anterior to the device, or posterior to the device. In Dr. Ross’ protocol, computed tomography (CT) scans or flexion-extension radiographs were obtained only if pseudarthrosis was suspected to avoid unnecessary radiation exposure. Sixty-six patients had at least the 3-month radiography follow-up available. All 52 patients with 9-month follow-up data achieved complete arthrodesis, as determined by plain film radiography. Bridging ossification was found anterior to the device in all but 9 patients. Trabeculated bone was growing through the lumen of the device in all cases (Figure 2). A broken blade without clinical correlation was noted on imaging for 1 patient.

The total cost of the ROI-C cervical cage (LDR) for VAPORHCS was $3,498, or $1,749 for the PEEK spacer plus $1,749 for 2 metal blades. In comparison, the total cost of a typical anterior locking plate would have been $6,700, or $3,200 for the plate plus $2,000 for 4 screws and $1,500 for an allograft fibular spacer. Demineralized bone matrix (1 mL) as used in cervical arthrodesis by other surgeons at VAPORHCS cost $279, or about $500 including shipping.

DISCUSSION

Anterior cervical discectomy with fusion is a very common and successful surgical procedure for cervical myelopathy, radiculopathy, and degenerative disease that has failed to be corrected with conservative therapy.¹⁰ Medicare data documented a 206% increase in 1-level fusion procedures for degenerative spine pathology performed between 1992 and 2005.¹¹ When a procedure is performed so often, it is appropriate to review methods and analyze efficacy, cost, and cost-effectiveness.

According to a 2007 meta-analysis, the fusion rates of 1-level ACDF arthrodesis at 1-year follow-up are 97.1% in patients treated with anterior plates and 92.1% in patients treated with noninstrumented fusion.¹² The rate disparity was larger for multiple-level fusion: 50% to 82.5% for instrumented cases^12,13 vs 3% to 42% for noninstrumented cases.^14-16 Given the higher fusion rates achieved with instrumentation, surgeons have favored its use in ACDF.

Computed Tomography Use

Computed tomography has long been considered the gold standard for assessing arthrodesis outcomes (eg, Siambanes and Mather).¹⁷ However, recent data on potential harm caused by CT-related ionizing radiation suggest a need for caution with routine CT use.^18,19 For cervical spine CT, Schonfeld and colleagues found that the risk for excess thyroid cancers ranged from 1 to 33 cases per 10,000 CT scans.²⁰ According to another report, “limiting neck CT scanning to a higher risk group would increase the gap between benefit and harm, whereas performing CT routinely on low-risk cases approaches a point where its harm equals or exceeds its benefit.”¹⁹ As some have questioned even routinepostoperative use of radiation in patients with unremarkable clinical courses—patients should be spared unnecessary exposure—CT scans or flexion-extensionradiographs were obtained at VAPORHCS only if clinical symptoms or radiographs were suggestive of pseudarthrosis.²¹ As none of the VAPORHCS patients had those symptoms, none underwent postoperative CT.

For anterior cervical arthrodesis, surgeon preference determines which of many different bone substrates can be used with instrumentation, which impacts the costs. Fusion substrates include structural autografts, structural allografts, morselized autografts, morselized allografts, demineralized allografts, porous ceramics and metals, and BMP. Given these many options, studies comparing the constructs are lacking, especially with regard to the cost of alternative fusion constructs that produce similar outcomes. The Centers for Disease Control and Prevention defines cost-benefit analysis as a “type of economic evaluation that measures both costs and benefits (ie, negative and positive consequences) associated with an intervention in dollar terms.”²² It has been reported that using iliac crest autografts with anterior plate instrumentation is the most cost-effective method, yet alternatives remain in use.^5,10

For ACDF, iliac crest bone is an ideal and widely used construct substrate. Structural grafts harvested from the crest provide significant stability due to their bicortical or tricortical configuration with interposed osteoinductive and osteogenic cancellous bone. Few graft complications (eg, graft resorption) and no immunogenic or infectious complications have been reported for iliac crest bone. However, autologous iliac crest increases operative time, and donor-site morbidity has been reported.^23,24 A retrospective questionnaire-based investigation by Silber and colleagues, who evaluated iliac crest bone graft site morbidity in 1-level ACDF, found that 26.1% of patients had pain at the iliac crest harvest site, and 15.7% had numbness.²⁴ Other complications, which occurred at lower rates, were bruising, hematoma, pelvic fracture, and poor cosmesis.^23,25 In addition, osteoporosis and comorbid conditions have made it a challenge to acquire iliac crest autograft, contributing to the popularity of alternative substrates.²⁵

Allografts

An alternative to autografts, allografts have the advantages of reduced operative time and reduced donor-site morbidity.²⁶ Major historical concerns with allografts have included risk for disease transmission, costs associated with sterilization and serologic screening of grafts, and lack of oversight, leading to human allografts being acquired from dubious sources and ending up in the operating room.^27,28 Two major types of allografts are available: mineralized and demineralized.

Arthrodesis rates are inferior for mineralized (structural) allografts with instrumentation than for autografts with instrumentation.²⁹ In addition, smoking and other comorbidities have influenced fusion rates more in allograft than autograft fusions.^30-33 However, allografts are being widely used because they avoid the donor-site morbidity associated with autografts and because they are load bearing, can provide structural stability and an osteoconductive matrix, and can be used off the shelf without adding much time to surgery.

Demineralized matrix substrates are commercial osteoconductive and osteoinductive biomaterials approved for filling bone gaps and extending graft when combined with autograft.^7,8 Despite their osteoinductive properties, these substrates have had a high degree of product inconsistency, in some cases leading to poor outcomes.³⁴ The lack of randomized studies with these constructs has made the determination of clear indications a challenge.

The initial enthusiasm over use of BMP, another bone-graft substitute for cervical fusion, was curtailed by reports of adverse events (AEs). Effective in anterior lumbar spine fusions, BMP was adapted to off-label use in the cervical spine a few years ago.³⁵ Initial studies by Baskin and colleagues and Bishop and colleagues showed its fusion rates superior to those of allograft.^31,32 Both studies reported no significant AEs. However, studies by Dickerman and colleagues and Smucker and colleagues demonstrated increased soft-tissue swelling leading to dysphagia and prolonged hospitalization, which were attributed to higher dosage (no study has identified a precise dose for individual patients).^36,37 In addition, the cost of BMP is higher than that of any other bone-graft option for ACDF.³ Osteolysis has also been reported with BMP use.^38-40 Carragee and colleagues highlighted the potential carcinogenicity of BMP, but this finding was not corroborated by Lad and colleagues.^41,42

Cost Considerations

In addition to surgical effectiveness, spine surgical device costs have come under increased scrutiny.^43-45 In 2012, plates were reported to cost (without overhead or profit margin to hospitals) between $1,015 and $3,601, and allograft spacers were estimated to cost between $1,220 and $3,640, cage costs ranged from $1,942 to $4,347, and PEEK spacers cost from $4,930 to $5,246.⁵ Individual surgeon instrumentation costs varied 10-fold based on the fusion constructs used.⁵

In a cost-effectiveness review of anterior cervical techniques, cage alone was the least expensive technique, disc arthroplasty or cage/plate/bone substitute groups were the next most expensive, and autograft alone was the most expensive option due to hip graft site morbidity.⁴³ In another study, operative time associated with harvesting an iliac crest graft was equivalent in cost to that of an interbody cage.⁴⁴ Other studies have compared the costs of various anterior cervical fusion constructs.^9,10,45,46 A limitation of these studies is that autologous bone often refers to iliac crest grafts rather than local autograft. Epstein reviewed data from these studies and concluded, “ACDF using dynamic plates and autografts are the most cost effective treatment for anterior cervical discectomy,” citing a cost of $1,015 for this construct.⁵ Although Epstein demonstrated the cost-effectiveness of autograft in an individual surgeon’s hands, the results also are significant in that the studies identified areas in which improvements can be made at other institutions. The ROI-C cervical cage and local autograft bone cost that the authors report is at the lower end of the range reported by Epstein.⁵

Device explant rates also can be a concern. Operative waste was well described in a retrospective analysis of 87 ACDF procedures.⁴⁷ The study found that the cost of explanting devices implanted during the same intraoperative period was equivalent to 9.2% of the cost of permanently implanted constructs. Epstein addressed operative waste by using educational modules to evaluate spine surgeons’ decision making before and after education. After the intervention, the institution noted a marked decline in costs related to explanted devices—from 20% in 2010 (before education) to 5.8% of the total cost of implanted devices in 2010 (after education).⁵

In the present study, the authors demonstrated that use of local morselized autograft with a PEEK spacer for 1-level ACDF had excellent arthrodesis rates and minimal complications. Of the 52 patients with 9 month postoperative data, all achieved arthrodesis regardless of tobacco use. This method compares favorably with other fusion options in terms of radiographic arthrodesis rates. In addition, it avoids the donor-site morbidity associated with autografts from an iliac site but maintains the benefits of the osteogenic, osteoconductive, and osteoinductive properties of autograft bone. Use of local autograft avoids the costs associated with iliac crest autograft, including increased operating and anesthesia time, additional operating room supplies (drapes, sutures, etc) needed for operating at a second site, and prolonged hospital stay due to pain at the donor site. Use of local autograft also obviates complications at a second surgical site; purchase, storage, and sterilization of allograft; and the neck swelling, possible carcinogenicity, and cost of purchase of BMP. Other than the occasional reuse of distraction posts, this method involves no other expensive explant supplies.

Autografts have osteogenic, osteoconductive, and osteoinductive properties, and autograft fusion rates are generally superior to allograft fusion rates. Bone morphogenetic protein fusion rates may be comparable to autograft fusion rates.^9,26,32 Shortcomings of iliac crest autografts include increased operative time, blood loss, and donor-site morbidity. Allografts are osteoconductive and osteoinductive, but their fusion rates are inferior to those of iliac crest autografts. Other shortcomings are infection transmission and immunogenicity risks, higher graft resorption and collapse rates, cost, and previous issues relating to provenance. Bone morphogenetic protein is the most osteoinductive material with fusion rates similar to those of autograft, but its use is associated with neck swelling, dysphagia, osteolysis, potential carcinogenicity, and high cost.⁹

Conclusion

Overall, use of local autograft with a PEEK spacer has all the advantages of iliac crest autograft along with the benefit of working within the same operative window as the ACDF, thus reducing the infection, bleeding, and pain risks that may be encountered with a second incision. This procedure is effective, inexpensive, and cost-effective compared with alternatives and may be preferable for 1-level ACDF. In a population of patients with high rates of tobacco use, diabetes mellitus, obesity, and other factors that negatively affect fusion rates, local autograft may be a good choice for efficacy and cost savings.

Acknowledgments
The authors thank Shirley McCartney, PhD, for editorial assistance and Andy Rekito, MS, for illustrative assistance.

Anterior cervical discectomy and fusion (ACDF) has been performed with various techniques and devices for many years. Autologous iliac crest grafts were initially used for the Cloward^1,2 and Robinson and Smith³ techniques, but because of iliac crest graft site complications (eg, pain, infection, fracture, dystrophic scarring^4,5), the procedure was generally superseded by allograft implants. These implants were then supplemented with anterior locking plate devices. More recently, unitary devices combining a polyetheretherketone (PEEK) spacer with screw or blade fixation have been developed, such as the Zero P (Synthes, Inc.) and the ROI-C cervical cage (LDR). Bone graft is required to fill the cavity of these devices and to promote osseous union. Demineralized bone matrix,⁶ tricalcium phosphate,^7,8 and bone morphogenetic protein (BMP) have been used for these purposes, but they add expense to the procedure and have been associated with several complications (eg, neck swelling, dysphagia associated with BMP).⁹

Although multiple studies have demonstrated effective fusion rates and good outcomes for both iliac crest autograft and grafting/spacer constructs, the debate over cost and “added value” remains unresolved. One institution, which has published articles reviewing the spine literature and its own data, concluded that iliac crest autograft was the most cost-effective and consistently successful ACDF procedure.^5,10

The VA Portland Health Care System (VAPORHCS) has analyzed the use of local autograft sources at the surgical site to circumvent the need to make a second incision at the iliac crest and, theoretically, to decrease risks and expenses associated with iliac crest autograft, allograft bone, and artificial constructs. Given the paucity of data on this method, the case series presented here represents one of a few studies that analyze local autograft for promotion of arthrodesis in a PEEK spacer device.

This article will report on the prospectively collected results of consecutive cases performed by Dr. Ross using a ROI-C cervical cage for 1-level anterior cervical discectomy between August 2011 and November 2014. This study received institutional review board approval.

Methods

Neck disability index (NDI) forms were used to assess the impact of neck pain on patients’ ability to manage in everyday life. The NDI form was completed before surgery and 3 and 9 months after surgery.

Dr. Ross preferred to perform minimally invasive posterior cervical foraminotomy for unilateral radiculopathy. Therefore, all patients with radiculopathy had bilateral symptoms or a symptomatic midline disc protrusion not accessible from a posterior approach. Standard techniques were used to make a left-side approach to the anterior cervical spine except in cases in which a previous right-side approach could be reused. Under the microscope, the anterior longitudinal ligament and annulus were incised, and the anterior contents of the disc space were removed with curettes and pituitary rongeurs. Care was taken to remove all cartilage from beneath the anterior inferior lip of the rostral vertebral body and to remove a few millimeters of the anterior longitudinal ligament from the rostral vertebral body without use of monopolar cautery (Figure 1). A 2 mm Kerrison punch then was used to remove the anterior inferior lip of the rostral vertebral body, and this bone was saved for grafting. No bone wax was used within the disc space.

After all disc space cartilage was removed from the endplates, additional bone was obtained from the uncovertebral joints and posterior vertebral bodies as the decompression proceeded posteriorly. Occasionally, distraction posts were used if the disc space was too narrow for optimal visualization posteriorly. After decompression was achieved, a lordotic ROI-C cervical cage was packed in its lumen with the bone chips and impacted into the disc space under fluoroscopic guidance. The blades were impacted under fluoroscopic guidance as well. The wound was closed with absorbable suture.

Antibiotics were given for no more than 24 hours after surgery. Ketorolac was used for analgesia the night of the surgery, and patients were asked to not use nonsteroidal anti-inflammatory drugs for 3 months after surgery. Lateral radiographs were obtained 3 and 9 months after surgery and every 6 months thereafter until arthrodesis was detected.

Results

Seventy-seven consecutive patients underwent 1-level anterior cervical discectomy (Table 1). Twenty-four procedures were performed for radiculopathy, 52 for myelopathy, and 1 for central cord injury sustained in a fall by a patient with preexisting spinal stenosis. Surgery was performed at C3-C4 (25 cases), C4-C5 (11 cases), C5-C6 (15 cases), and C6-C7 (1 case) for patients with myelopathy. Surgery was performed at C3-C4 (2 cases), C4-C5 (3 cases), C5-C6 (9 cases), and C6-C7 (10 cases) for patients with radiculopathy.

Twenty-eight patients reported presurgery tobacco use. Although all tobacco-using patients agreed to cease use in the perioperative period, at least 9 admitted to resuming tobacco use immediately after surgery. Eighteen patients had diabetes mellitus. In 2 patients, a diagnosis of osteoporosis was made with dual-energy X-ray absorptiometry. One patient was a chronic user of steroids before and after surgery. Mean body mass index (BMI) was 30.6, and 13 patients were morbidly obese (BMI > 34).

In 2 cases, only a single blade was placed. The second blade could not be placed because of broken adjacent screws (1 case) or undetermined reason (1 case).

The mean time for follow-up was 17 months (range 3-34). Four patients were lost to follow-up: 3 after the 1-month postoperative visit and 1 with severe psychiatric problems after hospital discharge.

There were no new neurologic deficits, no wound infections, and no recurrent laryngeal nerve palsies in the 77 patients. Eight months after surgery, 1 patient with radiculopathy underwent foraminotomy at the index level for persisting foraminal stenosis. Two patients whose myelopathic symptoms persisted after surgery returned for minimally invasive posterior laminotomy to remove infolded ligamentum flavum. The presurgery and 3- and 9-month postsurgery NDI scores were available for 52 patients (Table 2). Before surgery the mean NDI score was 24 (range 8-40). Three months postsurgery the mean NDI score was 15 (range 2-27) for patients with myelopathy and 13 (range 2-28) for patients with radiculopathy. The patient with the highest NDI score (28) stated that though all his symptoms were relieved, he had gauged his responses to protect his disability claim. Nine months after surgery, the mean NDI scores were 9.5 (range 5-17) for patients with myelopathy and 6 (range 2-13) for patients with radiculopathy. No NDI score was higher postsurgery than presurgery.

Arthrodesis was defined as bony bridging between the adjacent vertebral bodies and the bone graft within the lumen of the device, anterior to the device, or posterior to the device. In Dr. Ross’ protocol, computed tomography (CT) scans or flexion-extension radiographs were obtained only if pseudarthrosis was suspected to avoid unnecessary radiation exposure. Sixty-six patients had at least the 3-month radiography follow-up available. All 52 patients with 9-month follow-up data achieved complete arthrodesis, as determined by plain film radiography. Bridging ossification was found anterior to the device in all but 9 patients. Trabeculated bone was growing through the lumen of the device in all cases (Figure 2). A broken blade without clinical correlation was noted on imaging for 1 patient.

The total cost of the ROI-C cervical cage (LDR) for VAPORHCS was $3,498, or $1,749 for the PEEK spacer plus $1,749 for 2 metal blades. In comparison, the total cost of a typical anterior locking plate would have been $6,700, or $3,200 for the plate plus $2,000 for 4 screws and $1,500 for an allograft fibular spacer. Demineralized bone matrix (1 mL) as used in cervical arthrodesis by other surgeons at VAPORHCS cost $279, or about $500 including shipping.

DISCUSSION

Anterior cervical discectomy with fusion is a very common and successful surgical procedure for cervical myelopathy, radiculopathy, and degenerative disease that has failed to be corrected with conservative therapy.¹⁰ Medicare data documented a 206% increase in 1-level fusion procedures for degenerative spine pathology performed between 1992 and 2005.¹¹ When a procedure is performed so often, it is appropriate to review methods and analyze efficacy, cost, and cost-effectiveness.

According to a 2007 meta-analysis, the fusion rates of 1-level ACDF arthrodesis at 1-year follow-up are 97.1% in patients treated with anterior plates and 92.1% in patients treated with noninstrumented fusion.¹² The rate disparity was larger for multiple-level fusion: 50% to 82.5% for instrumented cases^12,13 vs 3% to 42% for noninstrumented cases.^14-16 Given the higher fusion rates achieved with instrumentation, surgeons have favored its use in ACDF.

Computed Tomography Use

Computed tomography has long been considered the gold standard for assessing arthrodesis outcomes (eg, Siambanes and Mather).¹⁷ However, recent data on potential harm caused by CT-related ionizing radiation suggest a need for caution with routine CT use.^18,19 For cervical spine CT, Schonfeld and colleagues found that the risk for excess thyroid cancers ranged from 1 to 33 cases per 10,000 CT scans.²⁰ According to another report, “limiting neck CT scanning to a higher risk group would increase the gap between benefit and harm, whereas performing CT routinely on low-risk cases approaches a point where its harm equals or exceeds its benefit.”¹⁹ As some have questioned even routinepostoperative use of radiation in patients with unremarkable clinical courses—patients should be spared unnecessary exposure—CT scans or flexion-extensionradiographs were obtained at VAPORHCS only if clinical symptoms or radiographs were suggestive of pseudarthrosis.²¹ As none of the VAPORHCS patients had those symptoms, none underwent postoperative CT.

For anterior cervical arthrodesis, surgeon preference determines which of many different bone substrates can be used with instrumentation, which impacts the costs. Fusion substrates include structural autografts, structural allografts, morselized autografts, morselized allografts, demineralized allografts, porous ceramics and metals, and BMP. Given these many options, studies comparing the constructs are lacking, especially with regard to the cost of alternative fusion constructs that produce similar outcomes. The Centers for Disease Control and Prevention defines cost-benefit analysis as a “type of economic evaluation that measures both costs and benefits (ie, negative and positive consequences) associated with an intervention in dollar terms.”²² It has been reported that using iliac crest autografts with anterior plate instrumentation is the most cost-effective method, yet alternatives remain in use.^5,10

For ACDF, iliac crest bone is an ideal and widely used construct substrate. Structural grafts harvested from the crest provide significant stability due to their bicortical or tricortical configuration with interposed osteoinductive and osteogenic cancellous bone. Few graft complications (eg, graft resorption) and no immunogenic or infectious complications have been reported for iliac crest bone. However, autologous iliac crest increases operative time, and donor-site morbidity has been reported.^23,24 A retrospective questionnaire-based investigation by Silber and colleagues, who evaluated iliac crest bone graft site morbidity in 1-level ACDF, found that 26.1% of patients had pain at the iliac crest harvest site, and 15.7% had numbness.²⁴ Other complications, which occurred at lower rates, were bruising, hematoma, pelvic fracture, and poor cosmesis.^23,25 In addition, osteoporosis and comorbid conditions have made it a challenge to acquire iliac crest autograft, contributing to the popularity of alternative substrates.²⁵

Allografts

An alternative to autografts, allografts have the advantages of reduced operative time and reduced donor-site morbidity.²⁶ Major historical concerns with allografts have included risk for disease transmission, costs associated with sterilization and serologic screening of grafts, and lack of oversight, leading to human allografts being acquired from dubious sources and ending up in the operating room.^27,28 Two major types of allografts are available: mineralized and demineralized.

Arthrodesis rates are inferior for mineralized (structural) allografts with instrumentation than for autografts with instrumentation.²⁹ In addition, smoking and other comorbidities have influenced fusion rates more in allograft than autograft fusions.^30-33 However, allografts are being widely used because they avoid the donor-site morbidity associated with autografts and because they are load bearing, can provide structural stability and an osteoconductive matrix, and can be used off the shelf without adding much time to surgery.

Demineralized matrix substrates are commercial osteoconductive and osteoinductive biomaterials approved for filling bone gaps and extending graft when combined with autograft.^7,8 Despite their osteoinductive properties, these substrates have had a high degree of product inconsistency, in some cases leading to poor outcomes.³⁴ The lack of randomized studies with these constructs has made the determination of clear indications a challenge.

The initial enthusiasm over use of BMP, another bone-graft substitute for cervical fusion, was curtailed by reports of adverse events (AEs). Effective in anterior lumbar spine fusions, BMP was adapted to off-label use in the cervical spine a few years ago.³⁵ Initial studies by Baskin and colleagues and Bishop and colleagues showed its fusion rates superior to those of allograft.^31,32 Both studies reported no significant AEs. However, studies by Dickerman and colleagues and Smucker and colleagues demonstrated increased soft-tissue swelling leading to dysphagia and prolonged hospitalization, which were attributed to higher dosage (no study has identified a precise dose for individual patients).^36,37 In addition, the cost of BMP is higher than that of any other bone-graft option for ACDF.³ Osteolysis has also been reported with BMP use.^38-40 Carragee and colleagues highlighted the potential carcinogenicity of BMP, but this finding was not corroborated by Lad and colleagues.^41,42

Cost Considerations

In addition to surgical effectiveness, spine surgical device costs have come under increased scrutiny.^43-45 In 2012, plates were reported to cost (without overhead or profit margin to hospitals) between $1,015 and $3,601, and allograft spacers were estimated to cost between $1,220 and $3,640, cage costs ranged from $1,942 to $4,347, and PEEK spacers cost from $4,930 to $5,246.⁵ Individual surgeon instrumentation costs varied 10-fold based on the fusion constructs used.⁵

In a cost-effectiveness review of anterior cervical techniques, cage alone was the least expensive technique, disc arthroplasty or cage/plate/bone substitute groups were the next most expensive, and autograft alone was the most expensive option due to hip graft site morbidity.⁴³ In another study, operative time associated with harvesting an iliac crest graft was equivalent in cost to that of an interbody cage.⁴⁴ Other studies have compared the costs of various anterior cervical fusion constructs.^9,10,45,46 A limitation of these studies is that autologous bone often refers to iliac crest grafts rather than local autograft. Epstein reviewed data from these studies and concluded, “ACDF using dynamic plates and autografts are the most cost effective treatment for anterior cervical discectomy,” citing a cost of $1,015 for this construct.⁵ Although Epstein demonstrated the cost-effectiveness of autograft in an individual surgeon’s hands, the results also are significant in that the studies identified areas in which improvements can be made at other institutions. The ROI-C cervical cage and local autograft bone cost that the authors report is at the lower end of the range reported by Epstein.⁵

Device explant rates also can be a concern. Operative waste was well described in a retrospective analysis of 87 ACDF procedures.⁴⁷ The study found that the cost of explanting devices implanted during the same intraoperative period was equivalent to 9.2% of the cost of permanently implanted constructs. Epstein addressed operative waste by using educational modules to evaluate spine surgeons’ decision making before and after education. After the intervention, the institution noted a marked decline in costs related to explanted devices—from 20% in 2010 (before education) to 5.8% of the total cost of implanted devices in 2010 (after education).⁵

In the present study, the authors demonstrated that use of local morselized autograft with a PEEK spacer for 1-level ACDF had excellent arthrodesis rates and minimal complications. Of the 52 patients with 9 month postoperative data, all achieved arthrodesis regardless of tobacco use. This method compares favorably with other fusion options in terms of radiographic arthrodesis rates. In addition, it avoids the donor-site morbidity associated with autografts from an iliac site but maintains the benefits of the osteogenic, osteoconductive, and osteoinductive properties of autograft bone. Use of local autograft avoids the costs associated with iliac crest autograft, including increased operating and anesthesia time, additional operating room supplies (drapes, sutures, etc) needed for operating at a second site, and prolonged hospital stay due to pain at the donor site. Use of local autograft also obviates complications at a second surgical site; purchase, storage, and sterilization of allograft; and the neck swelling, possible carcinogenicity, and cost of purchase of BMP. Other than the occasional reuse of distraction posts, this method involves no other expensive explant supplies.

Autografts have osteogenic, osteoconductive, and osteoinductive properties, and autograft fusion rates are generally superior to allograft fusion rates. Bone morphogenetic protein fusion rates may be comparable to autograft fusion rates.^9,26,32 Shortcomings of iliac crest autografts include increased operative time, blood loss, and donor-site morbidity. Allografts are osteoconductive and osteoinductive, but their fusion rates are inferior to those of iliac crest autografts. Other shortcomings are infection transmission and immunogenicity risks, higher graft resorption and collapse rates, cost, and previous issues relating to provenance. Bone morphogenetic protein is the most osteoinductive material with fusion rates similar to those of autograft, but its use is associated with neck swelling, dysphagia, osteolysis, potential carcinogenicity, and high cost.⁹

Conclusion

Overall, use of local autograft with a PEEK spacer has all the advantages of iliac crest autograft along with the benefit of working within the same operative window as the ACDF, thus reducing the infection, bleeding, and pain risks that may be encountered with a second incision. This procedure is effective, inexpensive, and cost-effective compared with alternatives and may be preferable for 1-level ACDF. In a population of patients with high rates of tobacco use, diabetes mellitus, obesity, and other factors that negatively affect fusion rates, local autograft may be a good choice for efficacy and cost savings.

Acknowledgments
The authors thank Shirley McCartney, PhD, for editorial assistance and Andy Rekito, MS, for illustrative assistance.

Among the ever increasing number of post-9/11 veterans pursuing higher education are many who carry psychological injuries, which include depression, anxiety, and posttraumatic stress disorder (PTSD). The effects of these mental health issues can create acquired learning disabilities involving impairments in memory, attention, concentration, and abstract thinking.^1-4 Such learning disabilities can prevent a soldier from successfully transitioning to student-veteran.

Academic reasonable accommodations for veterans with psychiatric diagnoses can strategically enhance student-veteran role integration. Similar to reasonable accommodations for physical diagnoses, academic accommodations for psychiatric conditions enhance qualifying student-veterans’ abilities to successfully pursue higher education by enabling them to compensate for deficits in memory, recall, concentration, and abstract thinking. Such assistance for veterans with disabilities has been advocated in order to promote academic progression and student empowerment.^5,6 Although academic accommodations enable veterans to compensate for learning disabilities, such interventions are not routinely requested for a variety of reasons. There are several key factors influencing veterans’ decisions to request such accommodations.

To promote a healthy transition to the student-veteran role, health care providers (HCPs) should initiate conversations about potential acquired learning disabilities with post-9/11 veterans with psychiatric diagnoses who are or will become students. Unfortunately, the medical literature includes little information on this topic or on how to have these conversations. To date, there is no suggested theoretical framework for guiding such discussions.

As a foundation for such discussion, Part 1 of this article explained the implications of psychiatric diagnoses and other common factors that can significantly impede adult learning among post-9/11 veterans who are separated from service.⁷ Part 1 also addressed the fundamentals of academic reasonable accommodations, which are outlined in Table 1.

Through use of a theoretical model, part 2 of this study defines key factors influencing post-9/11 veterans’ decision to request academic reasonable accommodations for psychiatric diagnoses. It also provides practical advice for facilitating clinical conversations at each stage of the model to promote the acceptance of academic reasonable accommodations among eligible post-9/11 veterans.

Health Belief Model

The health belief model (HBM) can be adopted to understand the steps of veterans’ decision-making processes involving reasonable accommodations. The model outlines determinants of human behavior that influence the potential health care decision to deliberately mitigate harm from a perceived health threat.^8,9 The 6 primary components are perceived susceptibility, perceived severity, perceived benefits, perceived barriers, cues to action, and self-efficacy.^8,9 The HBM previously has been applied to a diverse range of health behaviors involving prevention, medical regimen adherence, and utilization of health care services.¹⁰ Its application to learning impairment and academic reasonable accommodations is outlined in Table 2.

When the HBM framework is applied to academic accommodations, the perceived health threat is acquired learning disability. The desired health care decision is the act of requesting academic reasonable accommodations. The targeted population at risk is the post-9/11 veteran cohort with symptomatic psychiatric diagnoses who are enrolled in, or who are considering, postsecondary education.

The initial perceived susceptibility step determines the degree to which these veterans judge themselves as being at risk for learning impairment because of psychiatric diagnoses. During this step, it is imperative that HCPs educate veterans on how mental health conditions can alter adult learning styles. Clinicians should describe the negative effects of psychiatric symptoms on memory, concentration, focus, attention, and abstract thinking. Insight is developed in this step as veterans recognize that their academic endeavors potentially could be affected by underlying mental health symptoms.

Perceived Severity

Recognition of the perceived severity of impaired learning is the next step in HBM. Veterans will need to self-evaluate their actual or potential academic performance based on their current state of memory, concentration, focus, and attention. Although many veterans might determine that the impact is transient or minimal, a significant number of veterans will observe that their learning abilities are greatly affected. If veterans identify with loss of those skills since the onset of serious mental health issues, there should be further discussion regarding the existence of academic accommodations that address any learning impairment expected to last longer than 6 months.

As discussed in part 1, mental health diagnoses involving mood, though possessing individually distinct diagnostic criteria, create potentially similar global learning impairments in terms of decreased memory, poor concentration, and slowed executive functioning.^1-4 Insight into the impact of any acquired learning disability from these mental health conditions and/or associated pharmacologic treatment can be encouraged if the clinician and client jointly review the client’s self-described premorbid learning style and compare it with the client’s current functioning in day-to-day activities requiring memory, concentration, and decision making. A clinician can use a gentle emphasis on the incongruities between premorbid learning ability and present-day impairments as a springboard for discussion about ways to compensate for learning impairments.

Additional insight can be elicited by providing practical examples of how other factors can accentuate the learning difficulties caused by serious or persistent psychiatric symptoms. By discussing these issues, clinicians can provide veterans with a more realistic understanding of potential obstacles in the postsecondary setting and the need for a strategic plan to address such challenges. For example, if a veteran takes prescription medications to manage underlying psychiatric conditions, a discussion regarding pertinent pharmacologic adverse effects (AEs) can highlight how academic performance might be affected. As outlined in part 1, fatigue, drowsiness, restlessness, mental grogginess, and insomnia are just a few medication AEs that may impair academic performance by negatively affecting memory, concentration, and executive functioning.

There are multiple circumstances that can increase the degree to which psychiatric symptoms impede recall, memory, insight, judgment, concentration, attention, organization, and abstract thinking. Impaired memory, poor concentration, irritability, and decreased attention can occur in the normal postmilitary transition period or as residual effects from mild-to-moderate traumatic brain injury. Multiple role responsibilities, such as being a spouse and parent, also can present significant mental distractions from academic endeavors. A physical impairment, such as tinnitus, hearing loss, or chronic pain, can impede classroom participation.

At this juncture, HCPs also should identify the academic consequences of impaired learning. Knowing these consequences will help veterans decide whether a course of action is needed to compensate for any learning disability that may be present. Inability to finish timed tests, difficulty taking notes, and inefficient studying are some of the more serious potential sequelae. Feared long-term consequences include a lack of progress through the required course load and, ultimately, failing courses.

A basic explanation of the potential financial effects of poor academic achievement provides another practical method for clinicians to outline negative consequences of an acquired learning disability. The student’s sole income for basic necessities is often the post-9/11 GI Bill, which pays for up to 36 months of education benefits and includes a living allowance and book stipend.

Unfortunately, given their financial dependence on the GI Bill, many veterans who withdraw from classes due to academic difficulties face economic uncertainty. If their withdrawal is not approved by the GI Bill program, these veterans must pay back all the money granted during the semester. Veterans who remain in school despite receiving failing marks cannot recover money spent on failed courses. This potentially results in veterans exceeding their entire GI Bill allotment before completing course requirements for their desired certificate or degree. Many veterans logically conclude that the potential financial devastation is a sufficiently severe consequence of impaired learning ability, and those who believe they have significantly impaired learning ability may become more motivated to reduce any risk of academic failure by pursuing academic accommodations.

In tandem with reviewing the potential severity of the problem, clinicians always should emphasize the availability of academic accommodations to circumvent the negative consequences of an acquired learning disability. Veterans who experience academic difficulties but are unaware of academic interventions may decide to forgo postsecondary education. By understanding basic details about accommodations, veterans can make the informed decision to pursue these interventions as part of a plan for academic success.

Perceived Benefits

Although identifying perceived susceptibility and perceived severity are necessary for veterans to consider academic reasonable accommodation use, eligible veterans still may not understand how these accommodations can apply to their situation. In the next step of HBM, veterans must view formal academic accommodations as a desirable solution to mitigate the effects of impaired learning ability. Veterans must appreciate the perceived benefits of such requests before they elect to pursue them.

At this point, HCPs should provide examples of academic accommodations to illustrate the simplicity and ease of such interventions. Tutoring, note-taking assistance, and providing additional time for testing are examples of a few types of accommodations featuring advantages that should be readily apparent to veterans returning to school. These measures not only lessen the likelihood of struggling academically, but also afford an opportunity to excel. By painting accommodations as a powerful method of self-advocacy, HCPs can inform veterans that accommodations enable a measure of control within the academic setting and assist with planning.

Perceived Barriers

Although identifying perceived benefits may be persuasive, discerning perceived barriers is an important HBM step that influences whether veterans will seek academic accommodations. Fortunately, many of the common barriers to accommodation requests are simply misconceptions that clinicians can address easily. For example, some veterans misconstrue reasonable accommodations as giving them an unfair advantage, which they find offensive to their personal integrity and pride. Clinicians should point out to these veterans that accommodations address deficits in learning abilities and merely level the academic playing field so the student-veteran is on par with those students without such impairments. The core work needed to pass the class remains unchanged by such accommodations.

Often a barrier is erected when veterans subscribe to the traditional military definition of disability, which is equated with having overwhelming physical injuries or paralyzing psychological states. These veterans are reluctant to request any formal accommodations, because they do not see themselves as having a disability under this restrictive definition. For these veterans, HCPs need to explain that the broad federal definition of disability does not imply veterans must be disabled in any other aspect of his or her life except for learning.

Some veterans do not want to draw attention to themselves either as a veteran or as a student with learning difficulties.^11,12 Aware of civilian stereotyping of veterans, they prefer to remain anonymous. In this instance, clinicians should emphasize that psychiatric diagnoses are confidential and that only the reasonable accommodations are shared with the professor—not the underlying medical problem. The clinician also should emphasize that the accommodations are open to all eligible adult students, not just student-veterans. Therefore, use of such accommodations is not a disclosure of veteran status.

In conjunction with addressing client fears about stereotyping of both veterans and students with learning disabilities, HCPs should be mindful that mental health stigma is a significant barrier to seeking mental health services among military personnel, post-9/11 veterans, and college students.^13,14 Therefore, clinicians should emphasize that academic accommodations for psychiatric diagnoses are not self-disclosing of psychiatric concerns and are usually the same accommodations used to address learning disabilities caused by other factors.

Veterans may believe that documentation obtained in support of reasonable accommodations is too intimidating or too personal to reveal. Not realizing that federal law prevents institutions from requesting in-depth documentation, veterans mistakenly believe that they must provide all medical documents in order to qualify for academic accommodations. To assuage these fears, clinicians should inform veterans that schools generally require only a documentation letter from a qualified provider and usually do not require other medical records.

To further alleviate veteran fears and promote a measure of client control, providers may find it beneficial to review the proposed medical documentation letter with the veteran and have the veteran approve the content. Figures 1 and 2 illustrate a basic medical documentation letter with optional institution-specific criteria. To ensure compliance with any applicable federal privacy regulations or local facility policy, clinicians should obtain an information release form from the veteran. The medical documentation letter can then be released to the veteran for hand delivery to the academic institution.

Veterans might be concerned about the potential lack of confidentiality regarding the diagnosis contributing to their learning disability. They also may worry that accommodations will prevent them from entering the field of their choice when they graduate, especially for law enforcement careers. These veterans can be reassured by informing them that use of academic accommodations is completely confidential during their school years and will not appear on their school graduation records. Recommending that veterans confirm the established confidentiality process with their schools may help allay fears about inadvertent release of private information by the institution.

Self-Efficacy and Cues to Action

Even after perceived benefits and barriers are identified, veterans still may not act unless they believe that they can intervene appropriately to address the problem. The HBM refers to this step as self-efficacy. Student-veterans must feel empowered to effectively make reasonable accommodation requests and negotiate any potential setbacks to the implementation of those accommodations. Health care providers should inform veterans about the availability of a disability resource center or other counseling service at each school that can help the student-veteran through the process of accommodation approval. Ideally, student-veterans also should receive guidance on how to approach professors regarding both the request for and the implementation of the approved reasonable accommodations.¹⁵ Counselors at the institution should offer this guidance and help veterans select the appropriate accommodations.

In the HBM, cues to action occur at every step. These cues consist of the influential factors promoting the desired behavior. Providing answers to common veteran questions about academic accommodations is one cue to action. Another is providing a written step-by-step guide explaining academic accommodations to veterans. (The author has created a veteran-centric guide to academic accommodations. The guide, which explains basic concepts and addresses common barriers to requesting such accommodations, is available upon request from [email protected]).

At all times, positive feedback from clinicians is important in motivating veterans to complete the entire process. Discussion may be stalled at any point if veterans overestimate current academic abilities or underestimate their level of impaired learning ability. Motivational interviewing techniques may help resolve this impasse. However, even if eligible veterans are not interested in pursuing academic accommodations, HCPs should leave the option open for consideration. Although interventions are most beneficial when instituted early in the student’s coursework, veterans can formally request academic accommodations at any stage of their academic career.

Conclusion

Formal academic accommodations are viable tools for cultivating academic success among student-veterans with significant psychiatric conditions. The adoption of such interventions requires understanding post-9/11 veterans’ motivation and concerns about formal academic accommodation requests. Application of the HBM can guide clinicians in their discussions with post-9/11 veterans. By understanding the veterans’ perspectives on the subject, HCPs can directly address the factors influencing the decision to seek academic accommodations.

Ensuring successful transition to the student-veteran role is of prime importance for veterans who bear emotional scars from military service. To this author’s knowledge, no structured educational programs currently exist that inform either post-9/11 veterans or their HCPs about pertinent aspects of academic accommodations for student-veterans with symptomatic psychiatric diagnoses that impede learning. Future endeavors need to include development of programs to inform veterans and providers about this important topic. Such programs should not only promote the dissemination of general information, but also explore specific ways to tailor accommodations to the cognitive needs of each veteran.

Among the ever increasing number of post-9/11 veterans pursuing higher education are many who carry psychological injuries, which include depression, anxiety, and posttraumatic stress disorder (PTSD). The effects of these mental health issues can create acquired learning disabilities involving impairments in memory, attention, concentration, and abstract thinking.^1-4 Such learning disabilities can prevent a soldier from successfully transitioning to student-veteran.

Academic reasonable accommodations for veterans with psychiatric diagnoses can strategically enhance student-veteran role integration. Similar to reasonable accommodations for physical diagnoses, academic accommodations for psychiatric conditions enhance qualifying student-veterans’ abilities to successfully pursue higher education by enabling them to compensate for deficits in memory, recall, concentration, and abstract thinking. Such assistance for veterans with disabilities has been advocated in order to promote academic progression and student empowerment.^5,6 Although academic accommodations enable veterans to compensate for learning disabilities, such interventions are not routinely requested for a variety of reasons. There are several key factors influencing veterans’ decisions to request such accommodations.

To promote a healthy transition to the student-veteran role, health care providers (HCPs) should initiate conversations about potential acquired learning disabilities with post-9/11 veterans with psychiatric diagnoses who are or will become students. Unfortunately, the medical literature includes little information on this topic or on how to have these conversations. To date, there is no suggested theoretical framework for guiding such discussions.

As a foundation for such discussion, Part 1 of this article explained the implications of psychiatric diagnoses and other common factors that can significantly impede adult learning among post-9/11 veterans who are separated from service.⁷ Part 1 also addressed the fundamentals of academic reasonable accommodations, which are outlined in Table 1.

Through use of a theoretical model, part 2 of this study defines key factors influencing post-9/11 veterans’ decision to request academic reasonable accommodations for psychiatric diagnoses. It also provides practical advice for facilitating clinical conversations at each stage of the model to promote the acceptance of academic reasonable accommodations among eligible post-9/11 veterans.

Health Belief Model

The health belief model (HBM) can be adopted to understand the steps of veterans’ decision-making processes involving reasonable accommodations. The model outlines determinants of human behavior that influence the potential health care decision to deliberately mitigate harm from a perceived health threat.^8,9 The 6 primary components are perceived susceptibility, perceived severity, perceived benefits, perceived barriers, cues to action, and self-efficacy.^8,9 The HBM previously has been applied to a diverse range of health behaviors involving prevention, medical regimen adherence, and utilization of health care services.¹⁰ Its application to learning impairment and academic reasonable accommodations is outlined in Table 2.

When the HBM framework is applied to academic accommodations, the perceived health threat is acquired learning disability. The desired health care decision is the act of requesting academic reasonable accommodations. The targeted population at risk is the post-9/11 veteran cohort with symptomatic psychiatric diagnoses who are enrolled in, or who are considering, postsecondary education.

The initial perceived susceptibility step determines the degree to which these veterans judge themselves as being at risk for learning impairment because of psychiatric diagnoses. During this step, it is imperative that HCPs educate veterans on how mental health conditions can alter adult learning styles. Clinicians should describe the negative effects of psychiatric symptoms on memory, concentration, focus, attention, and abstract thinking. Insight is developed in this step as veterans recognize that their academic endeavors potentially could be affected by underlying mental health symptoms.

Perceived Severity

Recognition of the perceived severity of impaired learning is the next step in HBM. Veterans will need to self-evaluate their actual or potential academic performance based on their current state of memory, concentration, focus, and attention. Although many veterans might determine that the impact is transient or minimal, a significant number of veterans will observe that their learning abilities are greatly affected. If veterans identify with loss of those skills since the onset of serious mental health issues, there should be further discussion regarding the existence of academic accommodations that address any learning impairment expected to last longer than 6 months.

As discussed in part 1, mental health diagnoses involving mood, though possessing individually distinct diagnostic criteria, create potentially similar global learning impairments in terms of decreased memory, poor concentration, and slowed executive functioning.^1-4 Insight into the impact of any acquired learning disability from these mental health conditions and/or associated pharmacologic treatment can be encouraged if the clinician and client jointly review the client’s self-described premorbid learning style and compare it with the client’s current functioning in day-to-day activities requiring memory, concentration, and decision making. A clinician can use a gentle emphasis on the incongruities between premorbid learning ability and present-day impairments as a springboard for discussion about ways to compensate for learning impairments.

Additional insight can be elicited by providing practical examples of how other factors can accentuate the learning difficulties caused by serious or persistent psychiatric symptoms. By discussing these issues, clinicians can provide veterans with a more realistic understanding of potential obstacles in the postsecondary setting and the need for a strategic plan to address such challenges. For example, if a veteran takes prescription medications to manage underlying psychiatric conditions, a discussion regarding pertinent pharmacologic adverse effects (AEs) can highlight how academic performance might be affected. As outlined in part 1, fatigue, drowsiness, restlessness, mental grogginess, and insomnia are just a few medication AEs that may impair academic performance by negatively affecting memory, concentration, and executive functioning.

There are multiple circumstances that can increase the degree to which psychiatric symptoms impede recall, memory, insight, judgment, concentration, attention, organization, and abstract thinking. Impaired memory, poor concentration, irritability, and decreased attention can occur in the normal postmilitary transition period or as residual effects from mild-to-moderate traumatic brain injury. Multiple role responsibilities, such as being a spouse and parent, also can present significant mental distractions from academic endeavors. A physical impairment, such as tinnitus, hearing loss, or chronic pain, can impede classroom participation.

At this juncture, HCPs also should identify the academic consequences of impaired learning. Knowing these consequences will help veterans decide whether a course of action is needed to compensate for any learning disability that may be present. Inability to finish timed tests, difficulty taking notes, and inefficient studying are some of the more serious potential sequelae. Feared long-term consequences include a lack of progress through the required course load and, ultimately, failing courses.

A basic explanation of the potential financial effects of poor academic achievement provides another practical method for clinicians to outline negative consequences of an acquired learning disability. The student’s sole income for basic necessities is often the post-9/11 GI Bill, which pays for up to 36 months of education benefits and includes a living allowance and book stipend.

Unfortunately, given their financial dependence on the GI Bill, many veterans who withdraw from classes due to academic difficulties face economic uncertainty. If their withdrawal is not approved by the GI Bill program, these veterans must pay back all the money granted during the semester. Veterans who remain in school despite receiving failing marks cannot recover money spent on failed courses. This potentially results in veterans exceeding their entire GI Bill allotment before completing course requirements for their desired certificate or degree. Many veterans logically conclude that the potential financial devastation is a sufficiently severe consequence of impaired learning ability, and those who believe they have significantly impaired learning ability may become more motivated to reduce any risk of academic failure by pursuing academic accommodations.

In tandem with reviewing the potential severity of the problem, clinicians always should emphasize the availability of academic accommodations to circumvent the negative consequences of an acquired learning disability. Veterans who experience academic difficulties but are unaware of academic interventions may decide to forgo postsecondary education. By understanding basic details about accommodations, veterans can make the informed decision to pursue these interventions as part of a plan for academic success.

Perceived Benefits

Although identifying perceived susceptibility and perceived severity are necessary for veterans to consider academic reasonable accommodation use, eligible veterans still may not understand how these accommodations can apply to their situation. In the next step of HBM, veterans must view formal academic accommodations as a desirable solution to mitigate the effects of impaired learning ability. Veterans must appreciate the perceived benefits of such requests before they elect to pursue them.

At this point, HCPs should provide examples of academic accommodations to illustrate the simplicity and ease of such interventions. Tutoring, note-taking assistance, and providing additional time for testing are examples of a few types of accommodations featuring advantages that should be readily apparent to veterans returning to school. These measures not only lessen the likelihood of struggling academically, but also afford an opportunity to excel. By painting accommodations as a powerful method of self-advocacy, HCPs can inform veterans that accommodations enable a measure of control within the academic setting and assist with planning.

Perceived Barriers

Although identifying perceived benefits may be persuasive, discerning perceived barriers is an important HBM step that influences whether veterans will seek academic accommodations. Fortunately, many of the common barriers to accommodation requests are simply misconceptions that clinicians can address easily. For example, some veterans misconstrue reasonable accommodations as giving them an unfair advantage, which they find offensive to their personal integrity and pride. Clinicians should point out to these veterans that accommodations address deficits in learning abilities and merely level the academic playing field so the student-veteran is on par with those students without such impairments. The core work needed to pass the class remains unchanged by such accommodations.

Often a barrier is erected when veterans subscribe to the traditional military definition of disability, which is equated with having overwhelming physical injuries or paralyzing psychological states. These veterans are reluctant to request any formal accommodations, because they do not see themselves as having a disability under this restrictive definition. For these veterans, HCPs need to explain that the broad federal definition of disability does not imply veterans must be disabled in any other aspect of his or her life except for learning.

Some veterans do not want to draw attention to themselves either as a veteran or as a student with learning difficulties.^11,12 Aware of civilian stereotyping of veterans, they prefer to remain anonymous. In this instance, clinicians should emphasize that psychiatric diagnoses are confidential and that only the reasonable accommodations are shared with the professor—not the underlying medical problem. The clinician also should emphasize that the accommodations are open to all eligible adult students, not just student-veterans. Therefore, use of such accommodations is not a disclosure of veteran status.

In conjunction with addressing client fears about stereotyping of both veterans and students with learning disabilities, HCPs should be mindful that mental health stigma is a significant barrier to seeking mental health services among military personnel, post-9/11 veterans, and college students.^13,14 Therefore, clinicians should emphasize that academic accommodations for psychiatric diagnoses are not self-disclosing of psychiatric concerns and are usually the same accommodations used to address learning disabilities caused by other factors.

Veterans may believe that documentation obtained in support of reasonable accommodations is too intimidating or too personal to reveal. Not realizing that federal law prevents institutions from requesting in-depth documentation, veterans mistakenly believe that they must provide all medical documents in order to qualify for academic accommodations. To assuage these fears, clinicians should inform veterans that schools generally require only a documentation letter from a qualified provider and usually do not require other medical records.

To further alleviate veteran fears and promote a measure of client control, providers may find it beneficial to review the proposed medical documentation letter with the veteran and have the veteran approve the content. Figures 1 and 2 illustrate a basic medical documentation letter with optional institution-specific criteria. To ensure compliance with any applicable federal privacy regulations or local facility policy, clinicians should obtain an information release form from the veteran. The medical documentation letter can then be released to the veteran for hand delivery to the academic institution.

Veterans might be concerned about the potential lack of confidentiality regarding the diagnosis contributing to their learning disability. They also may worry that accommodations will prevent them from entering the field of their choice when they graduate, especially for law enforcement careers. These veterans can be reassured by informing them that use of academic accommodations is completely confidential during their school years and will not appear on their school graduation records. Recommending that veterans confirm the established confidentiality process with their schools may help allay fears about inadvertent release of private information by the institution.

Self-Efficacy and Cues to Action

Even after perceived benefits and barriers are identified, veterans still may not act unless they believe that they can intervene appropriately to address the problem. The HBM refers to this step as self-efficacy. Student-veterans must feel empowered to effectively make reasonable accommodation requests and negotiate any potential setbacks to the implementation of those accommodations. Health care providers should inform veterans about the availability of a disability resource center or other counseling service at each school that can help the student-veteran through the process of accommodation approval. Ideally, student-veterans also should receive guidance on how to approach professors regarding both the request for and the implementation of the approved reasonable accommodations.¹⁵ Counselors at the institution should offer this guidance and help veterans select the appropriate accommodations.

In the HBM, cues to action occur at every step. These cues consist of the influential factors promoting the desired behavior. Providing answers to common veteran questions about academic accommodations is one cue to action. Another is providing a written step-by-step guide explaining academic accommodations to veterans. (The author has created a veteran-centric guide to academic accommodations. The guide, which explains basic concepts and addresses common barriers to requesting such accommodations, is available upon request from [email protected]).

At all times, positive feedback from clinicians is important in motivating veterans to complete the entire process. Discussion may be stalled at any point if veterans overestimate current academic abilities or underestimate their level of impaired learning ability. Motivational interviewing techniques may help resolve this impasse. However, even if eligible veterans are not interested in pursuing academic accommodations, HCPs should leave the option open for consideration. Although interventions are most beneficial when instituted early in the student’s coursework, veterans can formally request academic accommodations at any stage of their academic career.

Conclusion

Formal academic accommodations are viable tools for cultivating academic success among student-veterans with significant psychiatric conditions. The adoption of such interventions requires understanding post-9/11 veterans’ motivation and concerns about formal academic accommodation requests. Application of the HBM can guide clinicians in their discussions with post-9/11 veterans. By understanding the veterans’ perspectives on the subject, HCPs can directly address the factors influencing the decision to seek academic accommodations.

Ensuring successful transition to the student-veteran role is of prime importance for veterans who bear emotional scars from military service. To this author’s knowledge, no structured educational programs currently exist that inform either post-9/11 veterans or their HCPs about pertinent aspects of academic accommodations for student-veterans with symptomatic psychiatric diagnoses that impede learning. Future endeavors need to include development of programs to inform veterans and providers about this important topic. Such programs should not only promote the dissemination of general information, but also explore specific ways to tailor accommodations to the cognitive needs of each veteran.

Among the ever increasing number of post-9/11 veterans pursuing higher education are many who carry psychological injuries, which include depression, anxiety, and posttraumatic stress disorder (PTSD). The effects of these mental health issues can create acquired learning disabilities involving impairments in memory, attention, concentration, and abstract thinking.^1-4 Such learning disabilities can prevent a soldier from successfully transitioning to student-veteran.

Academic reasonable accommodations for veterans with psychiatric diagnoses can strategically enhance student-veteran role integration. Similar to reasonable accommodations for physical diagnoses, academic accommodations for psychiatric conditions enhance qualifying student-veterans’ abilities to successfully pursue higher education by enabling them to compensate for deficits in memory, recall, concentration, and abstract thinking. Such assistance for veterans with disabilities has been advocated in order to promote academic progression and student empowerment.^5,6 Although academic accommodations enable veterans to compensate for learning disabilities, such interventions are not routinely requested for a variety of reasons. There are several key factors influencing veterans’ decisions to request such accommodations.

To promote a healthy transition to the student-veteran role, health care providers (HCPs) should initiate conversations about potential acquired learning disabilities with post-9/11 veterans with psychiatric diagnoses who are or will become students. Unfortunately, the medical literature includes little information on this topic or on how to have these conversations. To date, there is no suggested theoretical framework for guiding such discussions.

As a foundation for such discussion, Part 1 of this article explained the implications of psychiatric diagnoses and other common factors that can significantly impede adult learning among post-9/11 veterans who are separated from service.⁷ Part 1 also addressed the fundamentals of academic reasonable accommodations, which are outlined in Table 1.

Through use of a theoretical model, part 2 of this study defines key factors influencing post-9/11 veterans’ decision to request academic reasonable accommodations for psychiatric diagnoses. It also provides practical advice for facilitating clinical conversations at each stage of the model to promote the acceptance of academic reasonable accommodations among eligible post-9/11 veterans.

Health Belief Model

The health belief model (HBM) can be adopted to understand the steps of veterans’ decision-making processes involving reasonable accommodations. The model outlines determinants of human behavior that influence the potential health care decision to deliberately mitigate harm from a perceived health threat.^8,9 The 6 primary components are perceived susceptibility, perceived severity, perceived benefits, perceived barriers, cues to action, and self-efficacy.^8,9 The HBM previously has been applied to a diverse range of health behaviors involving prevention, medical regimen adherence, and utilization of health care services.¹⁰ Its application to learning impairment and academic reasonable accommodations is outlined in Table 2.

When the HBM framework is applied to academic accommodations, the perceived health threat is acquired learning disability. The desired health care decision is the act of requesting academic reasonable accommodations. The targeted population at risk is the post-9/11 veteran cohort with symptomatic psychiatric diagnoses who are enrolled in, or who are considering, postsecondary education.

The initial perceived susceptibility step determines the degree to which these veterans judge themselves as being at risk for learning impairment because of psychiatric diagnoses. During this step, it is imperative that HCPs educate veterans on how mental health conditions can alter adult learning styles. Clinicians should describe the negative effects of psychiatric symptoms on memory, concentration, focus, attention, and abstract thinking. Insight is developed in this step as veterans recognize that their academic endeavors potentially could be affected by underlying mental health symptoms.

Perceived Severity

Recognition of the perceived severity of impaired learning is the next step in HBM. Veterans will need to self-evaluate their actual or potential academic performance based on their current state of memory, concentration, focus, and attention. Although many veterans might determine that the impact is transient or minimal, a significant number of veterans will observe that their learning abilities are greatly affected. If veterans identify with loss of those skills since the onset of serious mental health issues, there should be further discussion regarding the existence of academic accommodations that address any learning impairment expected to last longer than 6 months.

As discussed in part 1, mental health diagnoses involving mood, though possessing individually distinct diagnostic criteria, create potentially similar global learning impairments in terms of decreased memory, poor concentration, and slowed executive functioning.^1-4 Insight into the impact of any acquired learning disability from these mental health conditions and/or associated pharmacologic treatment can be encouraged if the clinician and client jointly review the client’s self-described premorbid learning style and compare it with the client’s current functioning in day-to-day activities requiring memory, concentration, and decision making. A clinician can use a gentle emphasis on the incongruities between premorbid learning ability and present-day impairments as a springboard for discussion about ways to compensate for learning impairments.

Additional insight can be elicited by providing practical examples of how other factors can accentuate the learning difficulties caused by serious or persistent psychiatric symptoms. By discussing these issues, clinicians can provide veterans with a more realistic understanding of potential obstacles in the postsecondary setting and the need for a strategic plan to address such challenges. For example, if a veteran takes prescription medications to manage underlying psychiatric conditions, a discussion regarding pertinent pharmacologic adverse effects (AEs) can highlight how academic performance might be affected. As outlined in part 1, fatigue, drowsiness, restlessness, mental grogginess, and insomnia are just a few medication AEs that may impair academic performance by negatively affecting memory, concentration, and executive functioning.

There are multiple circumstances that can increase the degree to which psychiatric symptoms impede recall, memory, insight, judgment, concentration, attention, organization, and abstract thinking. Impaired memory, poor concentration, irritability, and decreased attention can occur in the normal postmilitary transition period or as residual effects from mild-to-moderate traumatic brain injury. Multiple role responsibilities, such as being a spouse and parent, also can present significant mental distractions from academic endeavors. A physical impairment, such as tinnitus, hearing loss, or chronic pain, can impede classroom participation.

At this juncture, HCPs also should identify the academic consequences of impaired learning. Knowing these consequences will help veterans decide whether a course of action is needed to compensate for any learning disability that may be present. Inability to finish timed tests, difficulty taking notes, and inefficient studying are some of the more serious potential sequelae. Feared long-term consequences include a lack of progress through the required course load and, ultimately, failing courses.

A basic explanation of the potential financial effects of poor academic achievement provides another practical method for clinicians to outline negative consequences of an acquired learning disability. The student’s sole income for basic necessities is often the post-9/11 GI Bill, which pays for up to 36 months of education benefits and includes a living allowance and book stipend.

Unfortunately, given their financial dependence on the GI Bill, many veterans who withdraw from classes due to academic difficulties face economic uncertainty. If their withdrawal is not approved by the GI Bill program, these veterans must pay back all the money granted during the semester. Veterans who remain in school despite receiving failing marks cannot recover money spent on failed courses. This potentially results in veterans exceeding their entire GI Bill allotment before completing course requirements for their desired certificate or degree. Many veterans logically conclude that the potential financial devastation is a sufficiently severe consequence of impaired learning ability, and those who believe they have significantly impaired learning ability may become more motivated to reduce any risk of academic failure by pursuing academic accommodations.

In tandem with reviewing the potential severity of the problem, clinicians always should emphasize the availability of academic accommodations to circumvent the negative consequences of an acquired learning disability. Veterans who experience academic difficulties but are unaware of academic interventions may decide to forgo postsecondary education. By understanding basic details about accommodations, veterans can make the informed decision to pursue these interventions as part of a plan for academic success.

Perceived Benefits

Although identifying perceived susceptibility and perceived severity are necessary for veterans to consider academic reasonable accommodation use, eligible veterans still may not understand how these accommodations can apply to their situation. In the next step of HBM, veterans must view formal academic accommodations as a desirable solution to mitigate the effects of impaired learning ability. Veterans must appreciate the perceived benefits of such requests before they elect to pursue them.

At this point, HCPs should provide examples of academic accommodations to illustrate the simplicity and ease of such interventions. Tutoring, note-taking assistance, and providing additional time for testing are examples of a few types of accommodations featuring advantages that should be readily apparent to veterans returning to school. These measures not only lessen the likelihood of struggling academically, but also afford an opportunity to excel. By painting accommodations as a powerful method of self-advocacy, HCPs can inform veterans that accommodations enable a measure of control within the academic setting and assist with planning.

Perceived Barriers

Although identifying perceived benefits may be persuasive, discerning perceived barriers is an important HBM step that influences whether veterans will seek academic accommodations. Fortunately, many of the common barriers to accommodation requests are simply misconceptions that clinicians can address easily. For example, some veterans misconstrue reasonable accommodations as giving them an unfair advantage, which they find offensive to their personal integrity and pride. Clinicians should point out to these veterans that accommodations address deficits in learning abilities and merely level the academic playing field so the student-veteran is on par with those students without such impairments. The core work needed to pass the class remains unchanged by such accommodations.

Often a barrier is erected when veterans subscribe to the traditional military definition of disability, which is equated with having overwhelming physical injuries or paralyzing psychological states. These veterans are reluctant to request any formal accommodations, because they do not see themselves as having a disability under this restrictive definition. For these veterans, HCPs need to explain that the broad federal definition of disability does not imply veterans must be disabled in any other aspect of his or her life except for learning.

Some veterans do not want to draw attention to themselves either as a veteran or as a student with learning difficulties.^11,12 Aware of civilian stereotyping of veterans, they prefer to remain anonymous. In this instance, clinicians should emphasize that psychiatric diagnoses are confidential and that only the reasonable accommodations are shared with the professor—not the underlying medical problem. The clinician also should emphasize that the accommodations are open to all eligible adult students, not just student-veterans. Therefore, use of such accommodations is not a disclosure of veteran status.

In conjunction with addressing client fears about stereotyping of both veterans and students with learning disabilities, HCPs should be mindful that mental health stigma is a significant barrier to seeking mental health services among military personnel, post-9/11 veterans, and college students.^13,14 Therefore, clinicians should emphasize that academic accommodations for psychiatric diagnoses are not self-disclosing of psychiatric concerns and are usually the same accommodations used to address learning disabilities caused by other factors.

Veterans may believe that documentation obtained in support of reasonable accommodations is too intimidating or too personal to reveal. Not realizing that federal law prevents institutions from requesting in-depth documentation, veterans mistakenly believe that they must provide all medical documents in order to qualify for academic accommodations. To assuage these fears, clinicians should inform veterans that schools generally require only a documentation letter from a qualified provider and usually do not require other medical records.

To further alleviate veteran fears and promote a measure of client control, providers may find it beneficial to review the proposed medical documentation letter with the veteran and have the veteran approve the content. Figures 1 and 2 illustrate a basic medical documentation letter with optional institution-specific criteria. To ensure compliance with any applicable federal privacy regulations or local facility policy, clinicians should obtain an information release form from the veteran. The medical documentation letter can then be released to the veteran for hand delivery to the academic institution.

Veterans might be concerned about the potential lack of confidentiality regarding the diagnosis contributing to their learning disability. They also may worry that accommodations will prevent them from entering the field of their choice when they graduate, especially for law enforcement careers. These veterans can be reassured by informing them that use of academic accommodations is completely confidential during their school years and will not appear on their school graduation records. Recommending that veterans confirm the established confidentiality process with their schools may help allay fears about inadvertent release of private information by the institution.

Self-Efficacy and Cues to Action

Even after perceived benefits and barriers are identified, veterans still may not act unless they believe that they can intervene appropriately to address the problem. The HBM refers to this step as self-efficacy. Student-veterans must feel empowered to effectively make reasonable accommodation requests and negotiate any potential setbacks to the implementation of those accommodations. Health care providers should inform veterans about the availability of a disability resource center or other counseling service at each school that can help the student-veteran through the process of accommodation approval. Ideally, student-veterans also should receive guidance on how to approach professors regarding both the request for and the implementation of the approved reasonable accommodations.¹⁵ Counselors at the institution should offer this guidance and help veterans select the appropriate accommodations.

In the HBM, cues to action occur at every step. These cues consist of the influential factors promoting the desired behavior. Providing answers to common veteran questions about academic accommodations is one cue to action. Another is providing a written step-by-step guide explaining academic accommodations to veterans. (The author has created a veteran-centric guide to academic accommodations. The guide, which explains basic concepts and addresses common barriers to requesting such accommodations, is available upon request from [email protected]).

At all times, positive feedback from clinicians is important in motivating veterans to complete the entire process. Discussion may be stalled at any point if veterans overestimate current academic abilities or underestimate their level of impaired learning ability. Motivational interviewing techniques may help resolve this impasse. However, even if eligible veterans are not interested in pursuing academic accommodations, HCPs should leave the option open for consideration. Although interventions are most beneficial when instituted early in the student’s coursework, veterans can formally request academic accommodations at any stage of their academic career.

Conclusion

Formal academic accommodations are viable tools for cultivating academic success among student-veterans with significant psychiatric conditions. The adoption of such interventions requires understanding post-9/11 veterans’ motivation and concerns about formal academic accommodation requests. Application of the HBM can guide clinicians in their discussions with post-9/11 veterans. By understanding the veterans’ perspectives on the subject, HCPs can directly address the factors influencing the decision to seek academic accommodations.

Ensuring successful transition to the student-veteran role is of prime importance for veterans who bear emotional scars from military service. To this author’s knowledge, no structured educational programs currently exist that inform either post-9/11 veterans or their HCPs about pertinent aspects of academic accommodations for student-veterans with symptomatic psychiatric diagnoses that impede learning. Future endeavors need to include development of programs to inform veterans and providers about this important topic. Such programs should not only promote the dissemination of general information, but also explore specific ways to tailor accommodations to the cognitive needs of each veteran.

Mr. G is a 67-year-old veteran. During the Vietnam War, he had the “most dangerous job”: helicopter door gunner and mechanic. He served in multiple combat missions and was under constant threat of attack. On returning to the U.S., he experienced anger outbursts, nightmares, hypervigilance, and urges to engage in dangerous behavior, such as driving a motorcycle more than 100 mph. Then he began abusing alcohol and drugs. Mr. G’s behavior and substance abuse eventually led to strained family relationships, termination from a high-paying job, and homelessness.

In 2001, the Washington DC VAMC homeless and substance abuse staff provided outreach and services that helped him secure permanent subsidized housing, achieve and maintain sobriety, get treatment for posttraumatic stress disorder (PTSD), and get full-time employment. Mr. G has maintained permanent supported housing status for 9 years, regained a sense of purpose, remained drug- and alcohol-free since 2001, attended a Vietnam combat PTSD support group weekly, and been an exemplary employee for 8 years.

In 2009, on any given night in the U.S., more than 75,000 veterans were without a permanent home and living on the streets, as Mr. G had been.¹ Nearly 150,000 other veterans were spending the night in an emergency shelter or transitional housing. In early 2010, the U.S. Interagency Council on Homelessness (USICH) developed a strategic plan to align federal resources toward 4 key objectives, which included preventing and ending homelessness among veterans. Since then, the most dramatic reductions in homelessness have occurred among veterans, with an overall 33% decline in chronic homelessness nationwide.

According to the U.S. Department of Housing and Urban Development (HUD), permanent housing is defined as community-based housing without a designated length of stay in which formerly homeless individuals live as independently as possible.² Under permanent housing, a program participant must be the tenant on a lease for at least 1 year, and the lease is renewable and terminable only for cause. The federal definition of the chronically homeless is a person who is homeless and lives in a place not meant for human habitation, in a safe haven, or in an emergency shelter continuously for at least 1 year; or on at least 4 separate occasions within the past 3 years; and who can be diagnosed with ≥ 1 of the following conditions: substance use disorder, serious mental illness, developmental disability as defined in section 102 of the Developmental Disabilities Assistance Bill of Rights Act of 2000 (42 U.S.C. 15002), PTSD, cognitive impairments caused by brain injury, and chronic physical illness or disability.³

Ending homelessness makes sense from a variety of perspectives. From a moral perspective, no one should experience homelessness, but this is especially true for the men and women who have served in the U.S. military. From a health care and resources perspective, homelessness is associated with poorer medical outcomes; higher medical costs for emergency department visits and hospital admissions; longer stays (often for conditions that could be treated in ambulatory settings); and increased mortality.^4-7 From a societal perspective, homelessness is associated with costs for shelters and other forms of temporary housing and with higher justice system costs stemming from police, court, and jail involvement.⁸ The higher justice system costs are in part attributable to significantly longer incarcerations for homeless persons than for demographically similar inmates who have been similarly charged but have housing.⁹ According to recent studies, significant cost reductions have been achieved by addressing homelessness and providing permanent housing, particularly for the chronically homeless with mental illness.^10-14

This article describes the efforts that have been made, through collaborations and coalitions of government and community partners, to identify and address risk factors for homelessness in Washington, DC—and ultimately to end veteran homelessness in the nation’s capitol.

Historic Perspective on Veteran Homelessness

Although the problem was first described after the Revolutionary War and again after the Civil War, homelessness among U.S. veterans has only recently been academically studied. During the colonial period, men often were promised pensions or land grants in exchange for military service. Several states failed to deliver on these promises, throwing veterans into dire financial circumstances and leaving them homeless. By the end of the Civil War, the size of the veteran population (almost 2 million, counting only those who fought for the Union), combined with an unemployment rate of 40% and economic downturns, led to thousands of veterans becoming homeless.^15,16

Homelessness among veterans continued after World War I. In 1932, more than 15,000 homeless and disabled “Bonus Army,” World War I veterans, marched on Washington to demand payment of the financial benefits promised for their military service.

Although World War II and Korean War veterans did not experience homelessness as previous veterans had, the problem resurfaced after Vietnam—the combat veterans of that war were overrepresented among the homeless.^17,18 Those at highest risk ranged from age 35 to 44 years in the early years of the all-volunteer military. It has been suggested that their increased risk may reflect social selection—volunteers for military service came from poor economic backgrounds with fewer social supports.¹⁹ A more recent study found that 3.8% of more than 300,000 Iraq and Afghanistan veterans who were followed for 5 years after military discharge experienced a homeless episode.²⁰ Female veterans similarly are overrepresented among the homeless.²¹ Female veterans represent only 1% of the overall female population, yet 3% to 4% are homeless.

Homelessness has always been a social problem, but only during the 1970s and 1980s did homelessness increase in importance—the number of visibly homeless people rose during that period—and investigators began to study and address the issue. Experts have described several factors that contributed to the increase in homelessness during that time.^22,23 First, as part of the deinstitutionalization initiative, thousands of mentally ill persons were released from state mental hospitals without a plan in place for affordable or supervised housing. Second, single-room-occupancy dwellings in poor areas, where transient single men lived, were demolished, and affordable housing options decreased. Third, economic and social changes were factors, such as a decreased need for seasonal and unskilled labor; reduced likelihood that relatives will take in homeless family members; and decriminalized public intoxication, loitering, and vagrancy. Out of these conditions came an interest in studying the causes of and risk factors for veteran homelessness and in developing a multipronged approach to end veteran homelessness.

Demographics

Nationally, veterans account for 9% of the homeless population.²⁴ Predominantly, they are single men living in urban areas—only about 9% are women—and 43% served during the Vietnam era.²⁴ Among homeless veterans, minorities are overrepresented—45% are African American or Hispanic, as contrasted with 10% and 3%, respectively, of the general population. More than two-thirds served in the military for more than 3 years, and 76% have a history of mental illness or substance abuse. Compared with the general homeless population, homeless veterans are older, better educated, more likely to have been married, and more likely to have health insurance, primarily through the VA.^24,25

The Washington, DC, metropolitan area encompasses the District of Columbia, northern Virginia, the tricounty area of southern Maryland, and Montgomery and Prince George’s counties in Maryland. Demographics for veterans in this area vary somewhat from national figures. According to the 2015 Point-in-Time survey of the homeless, veterans accounted for 5% of the homeless population (less than the national percentage). Most homeless veterans were single men (11.6% were women) and African American (65% of single adults, 85% of adults in families). Forty-five percent reported being employed and 40% had a substance use disorder or a serious mental illness. A large proportion also had at least 1 physical disability, such as hypertension, hepatitis, arthritis, diabetes mellitus, or heart disease.²⁶

Risk Factors

Multiple studies and multivariate analyses have determined that veteran status is associated with an increased risk for homelessness for both male and female veterans.²⁷ Female veterans were 3 times and male veterans 2 times more likely than nonveterans to become homeless, even when poverty, age, race, and geographic variation were controlled. A recent systematic review of U.S. veterans found that the strongest and most consistent risk factors for homelessness were substance use disorders and mental illness, particularly psychotic disorders. Posttraumatic stress disorder was not more significant than other mental conditions but it was a risk factor. Low income, unemployment, and poor money management were also factors.

Social risk factors include lack of support from family and friends. Military service with multiple deployments, transfers, and on-base housing may contribute to interruption of social support and lead to social isolation, thus increasing veterans’ risk for homelessness. Some studies have found that veterans are more likely to report physical injury or medical problems as contributing to homelessness and more likely to have 2 or more chronic medical conditions. Last, history of incarceration and adverse childhood experiences (eg, behavioral problems, family instability, foster care, childhood abuse) also have been found to be risk factors for homelessness among veterans and nonveterans alike.

Understanding these risk factors is an important step in addressing homelessness. Homelessness prevention efforts can screen for these risk factors and then intervene as quickly as possible. Access to mental health and substance abuse services, employment assistance, disability benefits and other income supports, and social services may prevent initial and subsequent episodes of homelessness. The VA, as the largest integrated health system in the U.S., is a critical safety net for low-income and disabled veterans with complex psychosocial needs. One study found access to VA service-connected pensions was protective against homelessness.²⁸

Addressing Homelessness

There are 10 USICH-recommended strategies for ending veteran homelessness (Table 1). These strategies cannot be achieved by any single federal agency or exclusively by government agencies—they require multipronged approaches and private and public partnerships. In early 2011, the staff of the Washington DC VAMC homeless program identified a single point of contact who would regularly meet with each Continuum of Care local planning body and each Public Housing Authority (PHA). This contact could identify homeless veterans regardless of the agency from which they were requesting assistance. The contact also facilitated identification of specific bottleneck issues contributing to delays in housing veterans. One such issue was that veterans were filling out application forms by themselves, and in some cases, their information was incomplete, or supporting documents (eg, government-issued IDs) were missing. The VA team adjusted the procedures to better meet veterans’ needs. Now veterans identified as meeting requirements for housing assistance are enrolled in classes in which a caseworker reviews their completed applications to ensure they are complete and supporting documents included. If an ID is needed, the caseworker facilitates the process and then submits the veteran’s application to the local PHA for processing. The result has been no returned applications.

Another issue was that in some cases a veteran spoke with the PHA and indicated interest in an apartment and only later found out that the apartment failed inspection. There would then be back-and-forth communications between PHA and landlord to have repairs made so the unit could pass inspection, which often resulted in long delays. The solution was to have a stock of preinspected housing options. The Washington DC VAMC homeless program now has a housing specialist who identifies preinspected units and can expedite the lease to a veteran. In addition, the homeless program in partnership with the PHA now sets up regular meet-and-greet events for landlords and veterans so veterans can preview available rentals.

Housing-First Model

The homeless program readily adopted the housing-first model. This model focuses on helping individuals and families access housing as quickly as possible and remain permanently housed; this assistance is not time-limited but ongoing. After housing placement, the client can choose from an array of both time-limited and long-term services, which are individualized to promote housing stability and individual well-being. Most important, housing is not contingent on compliance with services. Instead, participants must comply with a standard lease agreement and are provided the services and support that can help them do so successfully. Services are determined by completing needs assessments. In addition, as a veteran is applying for housing, a caseworker actively uses motivational interviewing techniques to enhance the likelihood that the veteran will accept services.

Providing Core Service to Veterans

In 2012, the Washington DC VAMC opened the Community Resource and Referral Center (CRRC) as a one-stop shop for homeless and at-risk veterans in need of basic, core, wraparound services through the VA. Basic services include showers, laundry facilities, and a chapel. The CRRC has onsite psychiatric services that can engage veterans in mental health and substance abuse treatment. There is also an onsite primary care team who specialize in working with the homeless and can address any acute or chronic medical problems. For veterans who want to return to employment, there is an onsite Compensated Work Therapy program, which has uniquely partnered with the VAMC as well as community partners (eg, National Archives, Smithsonian Institution, Quantico National Cemetery, Arlington National Cemetery) to offer therapeutic job experiences that often lead to gainful full-time employment.

Veterans can use an onsite computer lab to complete resumes and apply for employment online, with the assistance of veteran peer specialists and vocational rehabilitation specialists. For veterans who are unable to work, the homeless program team assists with disability applications. Social Security, Veterans Benefits Administration representatives, and Legal Aid have office hours at the CRRC to provide one-stop shopping for veterans. Additional community partnerships have led to classes to help veterans manage income effectively, and veterans are assisted with completing tax returns.

Representatives from the Veterans Justice Outreach (VJO) and Health Care for Re-entry Veterans (HCRV) programs are also onsite. The VJO program provides outreach, assessment, and case management for justice-involved veterans to prevent unnecessary criminalization of mental illness and extended incarceration. A VJO program specialist works with courts, legal representatives, and jails and acts as a liaison to engage veterans in treatment and prevent incarcerations through jail diversion strategies. The HCRV program assists incarcerated veterans with reentry into the community through outreach and prerelease assessments, and referrals and links to medical, psychiatric, and social services, including employment services on release. This program also can provide short-term case management assistance on release.

In 2013, to further the implementation of the USICH strategies in the Washington, DC, metropolitan area, 12 local government and nonprofit agencies entered into the Veterans NOW coalition (Table 2). This collaboration enabled development of a coordinated effort to identify all area veterans experiencing homelessness, regardless of which agency the veterans contacted. Further, this team set up processes to assess veterans’ housing and service needs and to match each veteran to the most appropriate housing resource. There was consensus regarding use of the Service Prioritization Decision Assistance Tool (SPDAT) as the evidence-informed approach for prioritizing client needs and identifying areas in which support is most likely needed to prevent housing instability.²⁹ More than 50 staff members at 20 different agencies in the District of Columbia have now been trained and are using this tool.

Outcomes

In 2010, the Point-in-Time count identified 718 homeless veterans in the Washington, DC, metropolitan area. By 2016, that number had dropped to 326 (55% reduction). During this same period, the number of veterans served by the Washington DC VAMC homeless program more than tripled, from 2,100 individuals in 2010 to nearly 6,400 in 2015. The coalition has housed or prevented homelessness for nearly 1,300 veterans during the past 2 years alone. Veterans were housed through multiple programs and efforts, including VA Supportive Housing (VASH), Supportive Services for Veteran Families, and Washington, DC Department of Human Services Permanent Supportive Housing. During the past year, > 60% of veterans were successfully placed in VASH housing within 90 days of application submission. Table 3 lists the national targets for assessing performance measurement and success. Not only were the various performance measure benchmarks exceeded, but more important, > 90% of veterans in VASH and Health Care for Homeless Veterans were able to keep their housing stabilized. Using SPDAT, the most chronically homeless and vulnerable were housed first, which accounts for the lower numbers of homeless Washington, DC, area veterans with substance abuse and mental health problems identified in the Point-in-Time survey.

Discussion

At the Washington DC VAMC, HCHV program staff members used evidence-based and evidence-informed tools, collaborated with community partners, and implemented recommended best practices to end veteran homelessness by 2015. Permanent supportive housing through VASH is crucial for helping veterans overcome their lack of income and in providing mechanisms for engaging in mental health and substance abuse services as well as primary care and therapeutic employment opportunities.

When former VA Secretary Eric K. Shinseki first announced the goal of ending veteran homelessness by 2015, many people questioned this goal’s attainability and feasibility. However, through the adoption of the strategies recommended by USICH, the establishment of government and community partnerships (including faith-based groups) and the implementation of programs addressing substance abuse issues, mental and physical health, income limitations, and employment, this goal now seems possible. Overall veteran homelessness decreased by 36% since 2010, and unsheltered homelessness decreased by nearly 50%.³⁰ By the end of 2015, nearly 65,000 veterans are in permanent housing nationwide, and another 8,100 are in the process of obtaining permanent supportive housing. Also, 87% of unsheltered veterans were able to move to safe housing within 30 days of engagement. Last, Supportive Services for Veteran Families was able to assist more than 156,800 individuals (single veterans as well as their children and families).

Sustainability from a national perspective also depends on continued federal funding. Mr. G, described at the beginning of this article, served his country honorably but then experienced the factors that put him at risk for homelessness. Through a veteran-centric team approach, he was able to successfully address each of these factors. As there are another 500 homeless veterans in Washington, DC, much work still needs to be done. With important collaborations and partnerships now in place, the goal of ending veteran homelessness in the District of Columbia is within sight. When homelessness is a thing of the past, we will truly be able to Welcome Home each veteran.

Mr. G is a 67-year-old veteran. During the Vietnam War, he had the “most dangerous job”: helicopter door gunner and mechanic. He served in multiple combat missions and was under constant threat of attack. On returning to the U.S., he experienced anger outbursts, nightmares, hypervigilance, and urges to engage in dangerous behavior, such as driving a motorcycle more than 100 mph. Then he began abusing alcohol and drugs. Mr. G’s behavior and substance abuse eventually led to strained family relationships, termination from a high-paying job, and homelessness.

In 2001, the Washington DC VAMC homeless and substance abuse staff provided outreach and services that helped him secure permanent subsidized housing, achieve and maintain sobriety, get treatment for posttraumatic stress disorder (PTSD), and get full-time employment. Mr. G has maintained permanent supported housing status for 9 years, regained a sense of purpose, remained drug- and alcohol-free since 2001, attended a Vietnam combat PTSD support group weekly, and been an exemplary employee for 8 years.

In 2009, on any given night in the U.S., more than 75,000 veterans were without a permanent home and living on the streets, as Mr. G had been.¹ Nearly 150,000 other veterans were spending the night in an emergency shelter or transitional housing. In early 2010, the U.S. Interagency Council on Homelessness (USICH) developed a strategic plan to align federal resources toward 4 key objectives, which included preventing and ending homelessness among veterans. Since then, the most dramatic reductions in homelessness have occurred among veterans, with an overall 33% decline in chronic homelessness nationwide.

According to the U.S. Department of Housing and Urban Development (HUD), permanent housing is defined as community-based housing without a designated length of stay in which formerly homeless individuals live as independently as possible.² Under permanent housing, a program participant must be the tenant on a lease for at least 1 year, and the lease is renewable and terminable only for cause. The federal definition of the chronically homeless is a person who is homeless and lives in a place not meant for human habitation, in a safe haven, or in an emergency shelter continuously for at least 1 year; or on at least 4 separate occasions within the past 3 years; and who can be diagnosed with ≥ 1 of the following conditions: substance use disorder, serious mental illness, developmental disability as defined in section 102 of the Developmental Disabilities Assistance Bill of Rights Act of 2000 (42 U.S.C. 15002), PTSD, cognitive impairments caused by brain injury, and chronic physical illness or disability.³

Ending homelessness makes sense from a variety of perspectives. From a moral perspective, no one should experience homelessness, but this is especially true for the men and women who have served in the U.S. military. From a health care and resources perspective, homelessness is associated with poorer medical outcomes; higher medical costs for emergency department visits and hospital admissions; longer stays (often for conditions that could be treated in ambulatory settings); and increased mortality.^4-7 From a societal perspective, homelessness is associated with costs for shelters and other forms of temporary housing and with higher justice system costs stemming from police, court, and jail involvement.⁸ The higher justice system costs are in part attributable to significantly longer incarcerations for homeless persons than for demographically similar inmates who have been similarly charged but have housing.⁹ According to recent studies, significant cost reductions have been achieved by addressing homelessness and providing permanent housing, particularly for the chronically homeless with mental illness.^10-14

This article describes the efforts that have been made, through collaborations and coalitions of government and community partners, to identify and address risk factors for homelessness in Washington, DC—and ultimately to end veteran homelessness in the nation’s capitol.

Historic Perspective on Veteran Homelessness

Although the problem was first described after the Revolutionary War and again after the Civil War, homelessness among U.S. veterans has only recently been academically studied. During the colonial period, men often were promised pensions or land grants in exchange for military service. Several states failed to deliver on these promises, throwing veterans into dire financial circumstances and leaving them homeless. By the end of the Civil War, the size of the veteran population (almost 2 million, counting only those who fought for the Union), combined with an unemployment rate of 40% and economic downturns, led to thousands of veterans becoming homeless.^15,16

Homelessness among veterans continued after World War I. In 1932, more than 15,000 homeless and disabled “Bonus Army,” World War I veterans, marched on Washington to demand payment of the financial benefits promised for their military service.

Although World War II and Korean War veterans did not experience homelessness as previous veterans had, the problem resurfaced after Vietnam—the combat veterans of that war were overrepresented among the homeless.^17,18 Those at highest risk ranged from age 35 to 44 years in the early years of the all-volunteer military. It has been suggested that their increased risk may reflect social selection—volunteers for military service came from poor economic backgrounds with fewer social supports.¹⁹ A more recent study found that 3.8% of more than 300,000 Iraq and Afghanistan veterans who were followed for 5 years after military discharge experienced a homeless episode.²⁰ Female veterans similarly are overrepresented among the homeless.²¹ Female veterans represent only 1% of the overall female population, yet 3% to 4% are homeless.

Homelessness has always been a social problem, but only during the 1970s and 1980s did homelessness increase in importance—the number of visibly homeless people rose during that period—and investigators began to study and address the issue. Experts have described several factors that contributed to the increase in homelessness during that time.^22,23 First, as part of the deinstitutionalization initiative, thousands of mentally ill persons were released from state mental hospitals without a plan in place for affordable or supervised housing. Second, single-room-occupancy dwellings in poor areas, where transient single men lived, were demolished, and affordable housing options decreased. Third, economic and social changes were factors, such as a decreased need for seasonal and unskilled labor; reduced likelihood that relatives will take in homeless family members; and decriminalized public intoxication, loitering, and vagrancy. Out of these conditions came an interest in studying the causes of and risk factors for veteran homelessness and in developing a multipronged approach to end veteran homelessness.

Demographics

Nationally, veterans account for 9% of the homeless population.²⁴ Predominantly, they are single men living in urban areas—only about 9% are women—and 43% served during the Vietnam era.²⁴ Among homeless veterans, minorities are overrepresented—45% are African American or Hispanic, as contrasted with 10% and 3%, respectively, of the general population. More than two-thirds served in the military for more than 3 years, and 76% have a history of mental illness or substance abuse. Compared with the general homeless population, homeless veterans are older, better educated, more likely to have been married, and more likely to have health insurance, primarily through the VA.^24,25

The Washington, DC, metropolitan area encompasses the District of Columbia, northern Virginia, the tricounty area of southern Maryland, and Montgomery and Prince George’s counties in Maryland. Demographics for veterans in this area vary somewhat from national figures. According to the 2015 Point-in-Time survey of the homeless, veterans accounted for 5% of the homeless population (less than the national percentage). Most homeless veterans were single men (11.6% were women) and African American (65% of single adults, 85% of adults in families). Forty-five percent reported being employed and 40% had a substance use disorder or a serious mental illness. A large proportion also had at least 1 physical disability, such as hypertension, hepatitis, arthritis, diabetes mellitus, or heart disease.²⁶

Risk Factors

Multiple studies and multivariate analyses have determined that veteran status is associated with an increased risk for homelessness for both male and female veterans.²⁷ Female veterans were 3 times and male veterans 2 times more likely than nonveterans to become homeless, even when poverty, age, race, and geographic variation were controlled. A recent systematic review of U.S. veterans found that the strongest and most consistent risk factors for homelessness were substance use disorders and mental illness, particularly psychotic disorders. Posttraumatic stress disorder was not more significant than other mental conditions but it was a risk factor. Low income, unemployment, and poor money management were also factors.

Social risk factors include lack of support from family and friends. Military service with multiple deployments, transfers, and on-base housing may contribute to interruption of social support and lead to social isolation, thus increasing veterans’ risk for homelessness. Some studies have found that veterans are more likely to report physical injury or medical problems as contributing to homelessness and more likely to have 2 or more chronic medical conditions. Last, history of incarceration and adverse childhood experiences (eg, behavioral problems, family instability, foster care, childhood abuse) also have been found to be risk factors for homelessness among veterans and nonveterans alike.

Understanding these risk factors is an important step in addressing homelessness. Homelessness prevention efforts can screen for these risk factors and then intervene as quickly as possible. Access to mental health and substance abuse services, employment assistance, disability benefits and other income supports, and social services may prevent initial and subsequent episodes of homelessness. The VA, as the largest integrated health system in the U.S., is a critical safety net for low-income and disabled veterans with complex psychosocial needs. One study found access to VA service-connected pensions was protective against homelessness.²⁸

Addressing Homelessness

There are 10 USICH-recommended strategies for ending veteran homelessness (Table 1). These strategies cannot be achieved by any single federal agency or exclusively by government agencies—they require multipronged approaches and private and public partnerships. In early 2011, the staff of the Washington DC VAMC homeless program identified a single point of contact who would regularly meet with each Continuum of Care local planning body and each Public Housing Authority (PHA). This contact could identify homeless veterans regardless of the agency from which they were requesting assistance. The contact also facilitated identification of specific bottleneck issues contributing to delays in housing veterans. One such issue was that veterans were filling out application forms by themselves, and in some cases, their information was incomplete, or supporting documents (eg, government-issued IDs) were missing. The VA team adjusted the procedures to better meet veterans’ needs. Now veterans identified as meeting requirements for housing assistance are enrolled in classes in which a caseworker reviews their completed applications to ensure they are complete and supporting documents included. If an ID is needed, the caseworker facilitates the process and then submits the veteran’s application to the local PHA for processing. The result has been no returned applications.

Another issue was that in some cases a veteran spoke with the PHA and indicated interest in an apartment and only later found out that the apartment failed inspection. There would then be back-and-forth communications between PHA and landlord to have repairs made so the unit could pass inspection, which often resulted in long delays. The solution was to have a stock of preinspected housing options. The Washington DC VAMC homeless program now has a housing specialist who identifies preinspected units and can expedite the lease to a veteran. In addition, the homeless program in partnership with the PHA now sets up regular meet-and-greet events for landlords and veterans so veterans can preview available rentals.

Housing-First Model

The homeless program readily adopted the housing-first model. This model focuses on helping individuals and families access housing as quickly as possible and remain permanently housed; this assistance is not time-limited but ongoing. After housing placement, the client can choose from an array of both time-limited and long-term services, which are individualized to promote housing stability and individual well-being. Most important, housing is not contingent on compliance with services. Instead, participants must comply with a standard lease agreement and are provided the services and support that can help them do so successfully. Services are determined by completing needs assessments. In addition, as a veteran is applying for housing, a caseworker actively uses motivational interviewing techniques to enhance the likelihood that the veteran will accept services.

Providing Core Service to Veterans

In 2012, the Washington DC VAMC opened the Community Resource and Referral Center (CRRC) as a one-stop shop for homeless and at-risk veterans in need of basic, core, wraparound services through the VA. Basic services include showers, laundry facilities, and a chapel. The CRRC has onsite psychiatric services that can engage veterans in mental health and substance abuse treatment. There is also an onsite primary care team who specialize in working with the homeless and can address any acute or chronic medical problems. For veterans who want to return to employment, there is an onsite Compensated Work Therapy program, which has uniquely partnered with the VAMC as well as community partners (eg, National Archives, Smithsonian Institution, Quantico National Cemetery, Arlington National Cemetery) to offer therapeutic job experiences that often lead to gainful full-time employment.

Veterans can use an onsite computer lab to complete resumes and apply for employment online, with the assistance of veteran peer specialists and vocational rehabilitation specialists. For veterans who are unable to work, the homeless program team assists with disability applications. Social Security, Veterans Benefits Administration representatives, and Legal Aid have office hours at the CRRC to provide one-stop shopping for veterans. Additional community partnerships have led to classes to help veterans manage income effectively, and veterans are assisted with completing tax returns.

Representatives from the Veterans Justice Outreach (VJO) and Health Care for Re-entry Veterans (HCRV) programs are also onsite. The VJO program provides outreach, assessment, and case management for justice-involved veterans to prevent unnecessary criminalization of mental illness and extended incarceration. A VJO program specialist works with courts, legal representatives, and jails and acts as a liaison to engage veterans in treatment and prevent incarcerations through jail diversion strategies. The HCRV program assists incarcerated veterans with reentry into the community through outreach and prerelease assessments, and referrals and links to medical, psychiatric, and social services, including employment services on release. This program also can provide short-term case management assistance on release.

In 2013, to further the implementation of the USICH strategies in the Washington, DC, metropolitan area, 12 local government and nonprofit agencies entered into the Veterans NOW coalition (Table 2). This collaboration enabled development of a coordinated effort to identify all area veterans experiencing homelessness, regardless of which agency the veterans contacted. Further, this team set up processes to assess veterans’ housing and service needs and to match each veteran to the most appropriate housing resource. There was consensus regarding use of the Service Prioritization Decision Assistance Tool (SPDAT) as the evidence-informed approach for prioritizing client needs and identifying areas in which support is most likely needed to prevent housing instability.²⁹ More than 50 staff members at 20 different agencies in the District of Columbia have now been trained and are using this tool.

Outcomes

In 2010, the Point-in-Time count identified 718 homeless veterans in the Washington, DC, metropolitan area. By 2016, that number had dropped to 326 (55% reduction). During this same period, the number of veterans served by the Washington DC VAMC homeless program more than tripled, from 2,100 individuals in 2010 to nearly 6,400 in 2015. The coalition has housed or prevented homelessness for nearly 1,300 veterans during the past 2 years alone. Veterans were housed through multiple programs and efforts, including VA Supportive Housing (VASH), Supportive Services for Veteran Families, and Washington, DC Department of Human Services Permanent Supportive Housing. During the past year, > 60% of veterans were successfully placed in VASH housing within 90 days of application submission. Table 3 lists the national targets for assessing performance measurement and success. Not only were the various performance measure benchmarks exceeded, but more important, > 90% of veterans in VASH and Health Care for Homeless Veterans were able to keep their housing stabilized. Using SPDAT, the most chronically homeless and vulnerable were housed first, which accounts for the lower numbers of homeless Washington, DC, area veterans with substance abuse and mental health problems identified in the Point-in-Time survey.

Discussion

At the Washington DC VAMC, HCHV program staff members used evidence-based and evidence-informed tools, collaborated with community partners, and implemented recommended best practices to end veteran homelessness by 2015. Permanent supportive housing through VASH is crucial for helping veterans overcome their lack of income and in providing mechanisms for engaging in mental health and substance abuse services as well as primary care and therapeutic employment opportunities.

When former VA Secretary Eric K. Shinseki first announced the goal of ending veteran homelessness by 2015, many people questioned this goal’s attainability and feasibility. However, through the adoption of the strategies recommended by USICH, the establishment of government and community partnerships (including faith-based groups) and the implementation of programs addressing substance abuse issues, mental and physical health, income limitations, and employment, this goal now seems possible. Overall veteran homelessness decreased by 36% since 2010, and unsheltered homelessness decreased by nearly 50%.³⁰ By the end of 2015, nearly 65,000 veterans are in permanent housing nationwide, and another 8,100 are in the process of obtaining permanent supportive housing. Also, 87% of unsheltered veterans were able to move to safe housing within 30 days of engagement. Last, Supportive Services for Veteran Families was able to assist more than 156,800 individuals (single veterans as well as their children and families).

Sustainability from a national perspective also depends on continued federal funding. Mr. G, described at the beginning of this article, served his country honorably but then experienced the factors that put him at risk for homelessness. Through a veteran-centric team approach, he was able to successfully address each of these factors. As there are another 500 homeless veterans in Washington, DC, much work still needs to be done. With important collaborations and partnerships now in place, the goal of ending veteran homelessness in the District of Columbia is within sight. When homelessness is a thing of the past, we will truly be able to Welcome Home each veteran.

Mr. G is a 67-year-old veteran. During the Vietnam War, he had the “most dangerous job”: helicopter door gunner and mechanic. He served in multiple combat missions and was under constant threat of attack. On returning to the U.S., he experienced anger outbursts, nightmares, hypervigilance, and urges to engage in dangerous behavior, such as driving a motorcycle more than 100 mph. Then he began abusing alcohol and drugs. Mr. G’s behavior and substance abuse eventually led to strained family relationships, termination from a high-paying job, and homelessness.

In 2001, the Washington DC VAMC homeless and substance abuse staff provided outreach and services that helped him secure permanent subsidized housing, achieve and maintain sobriety, get treatment for posttraumatic stress disorder (PTSD), and get full-time employment. Mr. G has maintained permanent supported housing status for 9 years, regained a sense of purpose, remained drug- and alcohol-free since 2001, attended a Vietnam combat PTSD support group weekly, and been an exemplary employee for 8 years.

In 2009, on any given night in the U.S., more than 75,000 veterans were without a permanent home and living on the streets, as Mr. G had been.¹ Nearly 150,000 other veterans were spending the night in an emergency shelter or transitional housing. In early 2010, the U.S. Interagency Council on Homelessness (USICH) developed a strategic plan to align federal resources toward 4 key objectives, which included preventing and ending homelessness among veterans. Since then, the most dramatic reductions in homelessness have occurred among veterans, with an overall 33% decline in chronic homelessness nationwide.

According to the U.S. Department of Housing and Urban Development (HUD), permanent housing is defined as community-based housing without a designated length of stay in which formerly homeless individuals live as independently as possible.² Under permanent housing, a program participant must be the tenant on a lease for at least 1 year, and the lease is renewable and terminable only for cause. The federal definition of the chronically homeless is a person who is homeless and lives in a place not meant for human habitation, in a safe haven, or in an emergency shelter continuously for at least 1 year; or on at least 4 separate occasions within the past 3 years; and who can be diagnosed with ≥ 1 of the following conditions: substance use disorder, serious mental illness, developmental disability as defined in section 102 of the Developmental Disabilities Assistance Bill of Rights Act of 2000 (42 U.S.C. 15002), PTSD, cognitive impairments caused by brain injury, and chronic physical illness or disability.³

Ending homelessness makes sense from a variety of perspectives. From a moral perspective, no one should experience homelessness, but this is especially true for the men and women who have served in the U.S. military. From a health care and resources perspective, homelessness is associated with poorer medical outcomes; higher medical costs for emergency department visits and hospital admissions; longer stays (often for conditions that could be treated in ambulatory settings); and increased mortality.^4-7 From a societal perspective, homelessness is associated with costs for shelters and other forms of temporary housing and with higher justice system costs stemming from police, court, and jail involvement.⁸ The higher justice system costs are in part attributable to significantly longer incarcerations for homeless persons than for demographically similar inmates who have been similarly charged but have housing.⁹ According to recent studies, significant cost reductions have been achieved by addressing homelessness and providing permanent housing, particularly for the chronically homeless with mental illness.^10-14

This article describes the efforts that have been made, through collaborations and coalitions of government and community partners, to identify and address risk factors for homelessness in Washington, DC—and ultimately to end veteran homelessness in the nation’s capitol.

Historic Perspective on Veteran Homelessness

Although the problem was first described after the Revolutionary War and again after the Civil War, homelessness among U.S. veterans has only recently been academically studied. During the colonial period, men often were promised pensions or land grants in exchange for military service. Several states failed to deliver on these promises, throwing veterans into dire financial circumstances and leaving them homeless. By the end of the Civil War, the size of the veteran population (almost 2 million, counting only those who fought for the Union), combined with an unemployment rate of 40% and economic downturns, led to thousands of veterans becoming homeless.^15,16

Homelessness among veterans continued after World War I. In 1932, more than 15,000 homeless and disabled “Bonus Army,” World War I veterans, marched on Washington to demand payment of the financial benefits promised for their military service.

Although World War II and Korean War veterans did not experience homelessness as previous veterans had, the problem resurfaced after Vietnam—the combat veterans of that war were overrepresented among the homeless.^17,18 Those at highest risk ranged from age 35 to 44 years in the early years of the all-volunteer military. It has been suggested that their increased risk may reflect social selection—volunteers for military service came from poor economic backgrounds with fewer social supports.¹⁹ A more recent study found that 3.8% of more than 300,000 Iraq and Afghanistan veterans who were followed for 5 years after military discharge experienced a homeless episode.²⁰ Female veterans similarly are overrepresented among the homeless.²¹ Female veterans represent only 1% of the overall female population, yet 3% to 4% are homeless.

Homelessness has always been a social problem, but only during the 1970s and 1980s did homelessness increase in importance—the number of visibly homeless people rose during that period—and investigators began to study and address the issue. Experts have described several factors that contributed to the increase in homelessness during that time.^22,23 First, as part of the deinstitutionalization initiative, thousands of mentally ill persons were released from state mental hospitals without a plan in place for affordable or supervised housing. Second, single-room-occupancy dwellings in poor areas, where transient single men lived, were demolished, and affordable housing options decreased. Third, economic and social changes were factors, such as a decreased need for seasonal and unskilled labor; reduced likelihood that relatives will take in homeless family members; and decriminalized public intoxication, loitering, and vagrancy. Out of these conditions came an interest in studying the causes of and risk factors for veteran homelessness and in developing a multipronged approach to end veteran homelessness.

Demographics

Nationally, veterans account for 9% of the homeless population.²⁴ Predominantly, they are single men living in urban areas—only about 9% are women—and 43% served during the Vietnam era.²⁴ Among homeless veterans, minorities are overrepresented—45% are African American or Hispanic, as contrasted with 10% and 3%, respectively, of the general population. More than two-thirds served in the military for more than 3 years, and 76% have a history of mental illness or substance abuse. Compared with the general homeless population, homeless veterans are older, better educated, more likely to have been married, and more likely to have health insurance, primarily through the VA.^24,25

The Washington, DC, metropolitan area encompasses the District of Columbia, northern Virginia, the tricounty area of southern Maryland, and Montgomery and Prince George’s counties in Maryland. Demographics for veterans in this area vary somewhat from national figures. According to the 2015 Point-in-Time survey of the homeless, veterans accounted for 5% of the homeless population (less than the national percentage). Most homeless veterans were single men (11.6% were women) and African American (65% of single adults, 85% of adults in families). Forty-five percent reported being employed and 40% had a substance use disorder or a serious mental illness. A large proportion also had at least 1 physical disability, such as hypertension, hepatitis, arthritis, diabetes mellitus, or heart disease.²⁶

Risk Factors

Multiple studies and multivariate analyses have determined that veteran status is associated with an increased risk for homelessness for both male and female veterans.²⁷ Female veterans were 3 times and male veterans 2 times more likely than nonveterans to become homeless, even when poverty, age, race, and geographic variation were controlled. A recent systematic review of U.S. veterans found that the strongest and most consistent risk factors for homelessness were substance use disorders and mental illness, particularly psychotic disorders. Posttraumatic stress disorder was not more significant than other mental conditions but it was a risk factor. Low income, unemployment, and poor money management were also factors.

Social risk factors include lack of support from family and friends. Military service with multiple deployments, transfers, and on-base housing may contribute to interruption of social support and lead to social isolation, thus increasing veterans’ risk for homelessness. Some studies have found that veterans are more likely to report physical injury or medical problems as contributing to homelessness and more likely to have 2 or more chronic medical conditions. Last, history of incarceration and adverse childhood experiences (eg, behavioral problems, family instability, foster care, childhood abuse) also have been found to be risk factors for homelessness among veterans and nonveterans alike.

Understanding these risk factors is an important step in addressing homelessness. Homelessness prevention efforts can screen for these risk factors and then intervene as quickly as possible. Access to mental health and substance abuse services, employment assistance, disability benefits and other income supports, and social services may prevent initial and subsequent episodes of homelessness. The VA, as the largest integrated health system in the U.S., is a critical safety net for low-income and disabled veterans with complex psychosocial needs. One study found access to VA service-connected pensions was protective against homelessness.²⁸

Addressing Homelessness

There are 10 USICH-recommended strategies for ending veteran homelessness (Table 1). These strategies cannot be achieved by any single federal agency or exclusively by government agencies—they require multipronged approaches and private and public partnerships. In early 2011, the staff of the Washington DC VAMC homeless program identified a single point of contact who would regularly meet with each Continuum of Care local planning body and each Public Housing Authority (PHA). This contact could identify homeless veterans regardless of the agency from which they were requesting assistance. The contact also facilitated identification of specific bottleneck issues contributing to delays in housing veterans. One such issue was that veterans were filling out application forms by themselves, and in some cases, their information was incomplete, or supporting documents (eg, government-issued IDs) were missing. The VA team adjusted the procedures to better meet veterans’ needs. Now veterans identified as meeting requirements for housing assistance are enrolled in classes in which a caseworker reviews their completed applications to ensure they are complete and supporting documents included. If an ID is needed, the caseworker facilitates the process and then submits the veteran’s application to the local PHA for processing. The result has been no returned applications.

Another issue was that in some cases a veteran spoke with the PHA and indicated interest in an apartment and only later found out that the apartment failed inspection. There would then be back-and-forth communications between PHA and landlord to have repairs made so the unit could pass inspection, which often resulted in long delays. The solution was to have a stock of preinspected housing options. The Washington DC VAMC homeless program now has a housing specialist who identifies preinspected units and can expedite the lease to a veteran. In addition, the homeless program in partnership with the PHA now sets up regular meet-and-greet events for landlords and veterans so veterans can preview available rentals.

Housing-First Model

The homeless program readily adopted the housing-first model. This model focuses on helping individuals and families access housing as quickly as possible and remain permanently housed; this assistance is not time-limited but ongoing. After housing placement, the client can choose from an array of both time-limited and long-term services, which are individualized to promote housing stability and individual well-being. Most important, housing is not contingent on compliance with services. Instead, participants must comply with a standard lease agreement and are provided the services and support that can help them do so successfully. Services are determined by completing needs assessments. In addition, as a veteran is applying for housing, a caseworker actively uses motivational interviewing techniques to enhance the likelihood that the veteran will accept services.

Providing Core Service to Veterans

In 2012, the Washington DC VAMC opened the Community Resource and Referral Center (CRRC) as a one-stop shop for homeless and at-risk veterans in need of basic, core, wraparound services through the VA. Basic services include showers, laundry facilities, and a chapel. The CRRC has onsite psychiatric services that can engage veterans in mental health and substance abuse treatment. There is also an onsite primary care team who specialize in working with the homeless and can address any acute or chronic medical problems. For veterans who want to return to employment, there is an onsite Compensated Work Therapy program, which has uniquely partnered with the VAMC as well as community partners (eg, National Archives, Smithsonian Institution, Quantico National Cemetery, Arlington National Cemetery) to offer therapeutic job experiences that often lead to gainful full-time employment.

Veterans can use an onsite computer lab to complete resumes and apply for employment online, with the assistance of veteran peer specialists and vocational rehabilitation specialists. For veterans who are unable to work, the homeless program team assists with disability applications. Social Security, Veterans Benefits Administration representatives, and Legal Aid have office hours at the CRRC to provide one-stop shopping for veterans. Additional community partnerships have led to classes to help veterans manage income effectively, and veterans are assisted with completing tax returns.

Representatives from the Veterans Justice Outreach (VJO) and Health Care for Re-entry Veterans (HCRV) programs are also onsite. The VJO program provides outreach, assessment, and case management for justice-involved veterans to prevent unnecessary criminalization of mental illness and extended incarceration. A VJO program specialist works with courts, legal representatives, and jails and acts as a liaison to engage veterans in treatment and prevent incarcerations through jail diversion strategies. The HCRV program assists incarcerated veterans with reentry into the community through outreach and prerelease assessments, and referrals and links to medical, psychiatric, and social services, including employment services on release. This program also can provide short-term case management assistance on release.

In 2013, to further the implementation of the USICH strategies in the Washington, DC, metropolitan area, 12 local government and nonprofit agencies entered into the Veterans NOW coalition (Table 2). This collaboration enabled development of a coordinated effort to identify all area veterans experiencing homelessness, regardless of which agency the veterans contacted. Further, this team set up processes to assess veterans’ housing and service needs and to match each veteran to the most appropriate housing resource. There was consensus regarding use of the Service Prioritization Decision Assistance Tool (SPDAT) as the evidence-informed approach for prioritizing client needs and identifying areas in which support is most likely needed to prevent housing instability.²⁹ More than 50 staff members at 20 different agencies in the District of Columbia have now been trained and are using this tool.

Outcomes

In 2010, the Point-in-Time count identified 718 homeless veterans in the Washington, DC, metropolitan area. By 2016, that number had dropped to 326 (55% reduction). During this same period, the number of veterans served by the Washington DC VAMC homeless program more than tripled, from 2,100 individuals in 2010 to nearly 6,400 in 2015. The coalition has housed or prevented homelessness for nearly 1,300 veterans during the past 2 years alone. Veterans were housed through multiple programs and efforts, including VA Supportive Housing (VASH), Supportive Services for Veteran Families, and Washington, DC Department of Human Services Permanent Supportive Housing. During the past year, > 60% of veterans were successfully placed in VASH housing within 90 days of application submission. Table 3 lists the national targets for assessing performance measurement and success. Not only were the various performance measure benchmarks exceeded, but more important, > 90% of veterans in VASH and Health Care for Homeless Veterans were able to keep their housing stabilized. Using SPDAT, the most chronically homeless and vulnerable were housed first, which accounts for the lower numbers of homeless Washington, DC, area veterans with substance abuse and mental health problems identified in the Point-in-Time survey.

Discussion

At the Washington DC VAMC, HCHV program staff members used evidence-based and evidence-informed tools, collaborated with community partners, and implemented recommended best practices to end veteran homelessness by 2015. Permanent supportive housing through VASH is crucial for helping veterans overcome their lack of income and in providing mechanisms for engaging in mental health and substance abuse services as well as primary care and therapeutic employment opportunities.

When former VA Secretary Eric K. Shinseki first announced the goal of ending veteran homelessness by 2015, many people questioned this goal’s attainability and feasibility. However, through the adoption of the strategies recommended by USICH, the establishment of government and community partnerships (including faith-based groups) and the implementation of programs addressing substance abuse issues, mental and physical health, income limitations, and employment, this goal now seems possible. Overall veteran homelessness decreased by 36% since 2010, and unsheltered homelessness decreased by nearly 50%.³⁰ By the end of 2015, nearly 65,000 veterans are in permanent housing nationwide, and another 8,100 are in the process of obtaining permanent supportive housing. Also, 87% of unsheltered veterans were able to move to safe housing within 30 days of engagement. Last, Supportive Services for Veteran Families was able to assist more than 156,800 individuals (single veterans as well as their children and families).

Sustainability from a national perspective also depends on continued federal funding. Mr. G, described at the beginning of this article, served his country honorably but then experienced the factors that put him at risk for homelessness. Through a veteran-centric team approach, he was able to successfully address each of these factors. As there are another 500 homeless veterans in Washington, DC, much work still needs to be done. With important collaborations and partnerships now in place, the goal of ending veteran homelessness in the District of Columbia is within sight. When homelessness is a thing of the past, we will truly be able to Welcome Home each veteran.

A majority of men (70%) aged ≥ 70 years report erectile dysfunction (ED) in primary care settings.¹ Further, the cost of ED medication is increasing: nationally, the VA spent $71.7 million on ED medications in 2013, triple the amount from 2006,² despite a 2011 VA mandate limiting ED medication prescriptions to 4 doses per month per veteran.³ Unfortunately, although ED is common and costly, only about 12% of men in the community report being asked about their sexual health by their primary care provider (PCP) in the past 3 years.⁴ Further, little emphasis seems to be placed on preventive care. For example, men with ED in primary care clinics are unaware of ED risk factors such as hypertension, smoking, and obesity; indeed, only 17% of a large community sample could name 1 risk factor for ED.⁵ This is problematic because diet and exercise improve erectile functioning,yet men may not realize they can reduce ED through behavioral and lifestyle change.⁶

In addition, there is little research that investigates veterans’ satisfaction with ED treatment and its effectiveness. The taboo nature of talking about erections and sexual health may partially relate to the lack of research. When surveyed, PCPs noted that they do not talk about ED routinely with patients for reasons that include time constraints, lack of experience managing sexual problems, viewing ED medication as a lifestyle drug, perceiving ED as a nonserious concern, discomfort discussing the topic for both male and female PCPs, and viewing ED discussions as the responsibility of providers of the opposite gender.^7-9

Given the dearth of ED research within the veteran population, the purpose of the current study was to (1) explore the level of treatment satisfaction of veterans prescribed an ED medication, phosphodiesterase type 5 inhibitor (PDE5); (2) assess patients’ perception of discussions with their PCPs about sexual health concerns; and (3) provide preliminary data on veterans’ knowledge of ED risk factors and identify possible areas for preventive education. This study was intended to highlight areas for further investigation to improve ED treatment satisfaction among veterans.

Methods

The authors conducted an anonymous survey with veterans who were prescribed an ED medication within the previous 12 months. In 2012, researchers obtained 8,000 names of veterans prescribed a PDE5 medication at the Clement J. Zablocki VA Medical Center (CJZVAMC) in Milwaukee, Wisconsin, and randomly selected 1,000 persons to mail a research survey to be returned anonymously. Three hundred ten surveys were returned, a 31% response rate, which was similar (32%) to a comparable large ED survey study, in which the participants were randomly selected to participate and also were not recruited by their PCP.¹⁰ Because 13 participants were excluded due to incomplete surveys or obtaining primary medical care services outside the VA, the current sample consisted of 297 participants. The CJZVAMC institutional review board approved the study in March 2013, and de-identified data were collected from March 2013 to March 2014.

The authors assessed demographics and treatment information, including whether veterans had talked with their PCP about sexual concerns.

Of the 297 participants, 55% were aged > 65 years. Racial/ethnic groups reflected the veteran population at CJZVAMC, with 78% identifying as European American, 17% as African American, 2% as Hispanic American, 2% as biracial, and 1% as Asian American or American Indian. Eighty-one percent were identified as Christian, and 10% reported no religious preference. Sixty-seven percent reported having a current sexual partner.

Measures

The International Index of Erectile Function (IIEF-5), an abridged version of a longer, 15-item instrument, was administered to assess participants’ erectile function.^11,12 The IIEF-5 consists of 5 items that ask about participants’ erectile functioning over the past 6 months. Participants responded to items on a 1 to 5 scale ranging from “almost never/never” to “almost always/always”. Items were summed to create a total score that could range from 5 to 25. Total scores reflect erectile functioning and satisfaction, with low scores indicating greater dysfunction. This measure has shown high sensitivity (.98) and specificity (.88).¹¹ Cut scores for the current study were consistent with the literature: mild ED = 17-21; mild-to-moderate ED = 12-16; moderate ED = 8-11; and severe ED = 5-7.¹³ Reliability in this sample was α = .93.

The authors were unable to find a validated measure assessing men’s knowledge of ED risk factors in the literature. Therefore, participants’ knowledge of ED risk factors was assessed using an online nonvalidated questionnaire entitled “Impotence [Erectile Dysfunction] Quiz: Test Your Medical IQ of ED” from www.emedicinehelp.com.¹³ Questions assess knowledge of specific risk factors (eg, age, obesity, depression, prostate cancer), symptoms, incidence rate, treatments, normal erectile functioning, and implications of ED. The questionnaire contains 16 items (3 true/false and 13 multiple choice items), and the total score corresponds to the percentage correct. According to the online version, the average score is 11 items correct (69%).¹³

A single item asked participants to identify behavioral changes they had tried to improve their erectile functioning. Options included taking medications at a different time, and/or decreasing tobacco, caffeine, or alcohol consumption. The Erectile Dysfunction Inventory of Treatment Satisfaction – Patient Version (EDITS) is an 11-item questionnaire administered to assess participants’ satisfaction with their medical treatment of ED.¹⁴ Items assess treatment satisfaction, ease of use, confidence in ability to perform, partner satisfaction, and naturalness of erections achieved during treatment. These items are rated on a scale ranging from 0 (dissatisfaction) to 4 (high satisfaction) and then summed, with total scores ranging from 0 (extremely dissatisfied) to 100 (extremely satisfied). The measure displayed high internal consistency (α = .90) and high test-retest reliability (r = .98).¹⁴ Several studies have used cutoff scores of 0 = very dissatisfied; 25 = dissatisfied; 50 = neither satisfied nor dissatisfied; 75 = satisfied; and 100 = very satisfied.^15,16 These cut scores and classifiers were used in the current study; reliability was α = .92.

The authors further explored reasons for veteran dissatisfaction with ED treatment by asking participants to respond to a single item: “Why are you dissatisfied with your erectile dysfunction treatment?” They could indicate that they were satisfied or circle all options for dissatisfaction that applied (“I would like to receive more pills per month,” “The treatment does not work well,” or “I want more information about erectile dysfunction and treatment”), or write in a response. The authors inquired about the number of pills prescribed to ascertain whether dissatisfaction was due to VA-specific policies vs veterans’ understanding of ED and effectiveness of treatment, which providers have more ability to improve.

In addition to the quantitative data obtained from the completed surveys, unsolicited responses from participants to the principal investigator via phone calls, and letters regarding treatment satisfaction were gathered. The second author conducted a basic exploratory content analysis of these unsolicited responses to group them into themes related to this study, such as satisfaction or dissatisfaction with ED treatment.

Results

The authors first assessed levels of ED and satisfaction with treatment in the sample. On average, participants reported mild-to-moderate erectile dysfunction (M = 13.1; SD = 5.7), which is higher than that of the general population and consistent with samples of men referred for ED treatment.^17,18 Satisfaction levels were slightly above neutral on the EDITS questionnaire (M = 58.3%; SD = 24.5). In response to a separate single-item question regarding reasons for dissatisfaction, only 6.4% of veterans reported being satisfied with their ED treatment.

According to respondents, the primary reasons for dissatisfaction were wanting more medication (46%), finding the treatment ineffective (26.7%), and desiring more information (24%). Further, ED severity was negatively correlated with satisfaction with ED treatment (r = .72, P < .01; note that higher scores correspond to less severe ED on this measure). However, despite moderate-to-low levels of satisfaction, 79.2% of patients planned to continue with their ED treatment (59.3% very likely and 19.9% moderately likely).

The authors also assessed participants’ communication with PCPs about their sexual functioning. Twenty-five percent reported not talking with their PCP about sexual concerns (despite all having been prescribed an ED medication in the past year). In this sample, talking with one’s PCP was not related to increased knowledge of ED risk factors (t [294] = .32, ns). Those who talked to their PCP tended to be less satisfied with treatment (M = 56.2; SD = 24.5) than those who did not talk to their PCP (M = 64.7; SD = 23.3; t (213) = -2.2; P = .03), likely because those who felt their treatment was working for them felt less need to talk to their provider. Indeed, those who talked to their PCP trended to have more severe levels of ED (M = 12.7; SD = 5.8) than those who did not (M = 14.2; SD = 5.3; t [285] = -1.91; P = .057; note that higher scores correspond to less severe ED on this measure). Finally, adults aged > 65 years were less likely to talk to their PCP than were younger adults (69% vs 81%); χ² (1, N = 291) = 5.57; P = .018.

Generally, the level of knowledge of ED risk factors was lower than the average of respondents to the original online survey (62% vs 69%).¹³ Younger adults were slightly more knowledgeable (M = 64%; SD = 13) than were older adults (M = 60%; SD = 15), t (288.08) = 2.01; P = .046).

Finally, most veterans reported few attempted behavioral changes to address ED, such as taking medications at a different time or decreasing use of tobacco, caffeine, or alcohol (M = 1.3; SD = 1.1). Thirty percent had not tried any behavioral changes; 34.1% tried 1 change; and 35.9% had tried more than 1 behavioral change. In contrast, 89% of participants reported using a PDE5 medication. Eight-two percent of participants reported currently receiving ED treatment of some kind; within this group, 97.4% reported currently taking a PDE5 medication. Only 2.5% of veterans reported using other kinds of treatment, such as vacuum pump, suppository, over-the-counter medication, injections, and not using a PDE5 medication, whereas 6.7% were using other kinds of treatment as well as a PDE5 medication.

In addition to the quantitative responses, 48 participants wrote unsolicited comments about their experiences with ED treatment on their returned questionnaires. The principal investigator also received 9 telephone calls from intended study participants, who provided verbal feedback regarding their experience with ED treatment. Comments unrelated to the study were eliminated, and the remaining written and verbal responses were grouped into categories to identify themes. Mirroring the quantitative results, participants providing qualitative feedback were dissatisfied with their ED treatment. Specifically, 43% of the comments consisted of complaints regarding the ineffectiveness and/or undesirable adverse effects (AEs) of ED medications and other ED treatments, including physical AEs (eg, headaches), sentiments that treatment does not feel “natural,” and dissatisfaction with the quality and length of sexual encounters despite treatment. Yet 24% of the comments entailed requests for more and/or different ED medications. Less frequent, although significant, comments related to decreased sexual interest and performance because of other medical conditions, such as pain, prostate surgery, and hypertension (15%); desire for additional information about ED treatments from health care providers (9%); use of nonpharmacologic ED interventions (eg, vacuum pump, 7%); and concerns about their partners’ level of sexual dissatisfaction as a result of their ED (7%).

Discussion

The present study examined knowledge of ED risk factors and level of satisfaction with ED treatment in a veteran population. Pharmacologic interventions comprised the most prevalent form of ED treatment. Both quantitative and qualitative results indicated areas for improvement in veteran satisfaction with ED treatment. Overall, veterans reported being neither satisfied nor dissatisfied with their current ED treatment, although very few reported being satisfied in response to a single item. The discrepancy may be related to the negative wording of the latter question (“Why are you dissatisfied with your erectile dysfunction treatment?”), which potentially biased participants’ responses. Several veterans also provided many unsolicited comments regarding areas for improvement. Despite feeling neutral to dissatisfied with treatment, 80% planned to continue with treatment. Sources of dissatisfaction included restricted access to ED medication (eg, limiting pills to 4 per month), ineffectiveness of treatment (eg, poor quality of erection, lack of climax), physical AEs, a desire for more information about ED, and psychological and relational concerns (eg, partner sexual dissatisfaction). As one veteran in his 80s lamented in describing the apparent end to his sexual life despite current ED treatment, “Is that all there is? It is the end of the road.”

The authors identified several barriers to implementing potentially beneficial interventions other than ED medications. Specifically, despite receiving long-term treatment for ED, veteran participants showed average knowledge of information related to ED risk factors. Of concern, discussing sexual health concerns with a PCP was not associated with increased knowledge of ED risk factors. This may explain the finding that veterans plan to continue with medication treatment despite feeling only neutral to dissatisfied about their current ED treatment.

Veterans who talked to their PCP about ED were less satisfied with treatment than were those who did not talk to their PCP, likely because those who felt their treatment was working for them felt less need to talk to their provider. Indeed, those who talked to their PCP tended to have more severe ED than those who did not. It may be that veterans avoid discussing ED with their PCP until they reach advanced ED when it is too late for treatment to make a difference. The principal investigator’s receipt of unsolicited telephone calls from intended study participants desiring to discuss ED—something that has not occurred during the researchers’ involvement in dozens of prior health-related studies—illustrates the importance veterans place on sexual concerns and the need to encourage discussion about the topic in the context of health care appointments. Specifically, older adults would benefit from more conversations with PCPs as they reported less knowledge of ED risk factors and fewer conversations with PCPs about sexual concerns than did younger men.

Adverse Events

Given the AEs reported by veterans and the significant cost of ED medications within the VA system,² increased use of alternative nonpharmacologic and preventive behavioral approaches would be clinically and economically beneficial. For example, in one study, men with ED who engaged in a lifestyle program that focused on weight loss, diet, and exercise were found more likely to experience improvements in erectile functioning compared with men who did not participate.⁶ Yet in the current study, 30% of participants had not attempted behavioral changes to address ED.

The VA’s Health Promotion and Disease Prevention (HPDP) Program focuses on preventive services and behavioral interventions to reduce health risks within primary care settings.¹⁹ This program may provide a framework for efforts to prevent and ameliorate ED. Specifically, coaching and education by HPDP experts could reduce PCPs’ discomfort with sexual health discussions and normalize the value of such conversations for both providers and patients. Existing HPDP behavioral interventions targeting areas such as weight loss and smoking cessation also could emphasize the potential secondary benefit of improved sexual functioning. To that end, preventive health campaigns could include sexual health and ED prevention as topics on patient education materials. Including sexual functioning on telephone or in-person prescreening questionnaires prior to routine appointments with PCPs also may facilitate destigmatization of sex as an important health topic.

Limitations

Limitations of the current study include its correlational design, which precludes conclusions regarding casual relationships among the variables in question. The authors cannot speculate about how well their sample represents the general veteran population given its low response rate (although comparable to a similar study).¹⁰ In addition, the lack of a validated measure of ED risk-factor knowledge meant reliance on an online questionnaire with unknown psychometric properties. To identify alternatives to pharmacologic treatment for ED, it would be beneficial for future research to examine the reasons for dissatisfaction among veterans, assessing satisfaction changes after implementation of behavioral and/or preventive interventions.

Conclusion

This study deepens the understanding of ED treatment efficacy among veterans in light of the paucity of available information. Overall, veterans are neutral to dissatisfied with their ED treatment, yet plan to continue it in the context of limited alternatives and possible lack of knowledge of behavioral methods shown to improve erectile functioning. Future studies that examine the reasons for continuing medication despite neutral satisfaction would help explore this finding. Based on these results, the authors recommend increased attention and discussion of sexual health during PCP visits and enhanced efforts toward using behavioral strategies to prevent and reduce ED. Encouragement from PCPs to address sexual health concerns earlier in a veteran’s treatment course—and in the context of behavioral and lifestyle change—may assist in preventing veterans’ sexual lives from prematurely reaching “the end of the road.”

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the Clement J. Zablocki VA Medical Center in Milwaukee, Wisconsin.

A majority of men (70%) aged ≥ 70 years report erectile dysfunction (ED) in primary care settings.¹ Further, the cost of ED medication is increasing: nationally, the VA spent $71.7 million on ED medications in 2013, triple the amount from 2006,² despite a 2011 VA mandate limiting ED medication prescriptions to 4 doses per month per veteran.³ Unfortunately, although ED is common and costly, only about 12% of men in the community report being asked about their sexual health by their primary care provider (PCP) in the past 3 years.⁴ Further, little emphasis seems to be placed on preventive care. For example, men with ED in primary care clinics are unaware of ED risk factors such as hypertension, smoking, and obesity; indeed, only 17% of a large community sample could name 1 risk factor for ED.⁵ This is problematic because diet and exercise improve erectile functioning,yet men may not realize they can reduce ED through behavioral and lifestyle change.⁶

In addition, there is little research that investigates veterans’ satisfaction with ED treatment and its effectiveness. The taboo nature of talking about erections and sexual health may partially relate to the lack of research. When surveyed, PCPs noted that they do not talk about ED routinely with patients for reasons that include time constraints, lack of experience managing sexual problems, viewing ED medication as a lifestyle drug, perceiving ED as a nonserious concern, discomfort discussing the topic for both male and female PCPs, and viewing ED discussions as the responsibility of providers of the opposite gender.^7-9

Given the dearth of ED research within the veteran population, the purpose of the current study was to (1) explore the level of treatment satisfaction of veterans prescribed an ED medication, phosphodiesterase type 5 inhibitor (PDE5); (2) assess patients’ perception of discussions with their PCPs about sexual health concerns; and (3) provide preliminary data on veterans’ knowledge of ED risk factors and identify possible areas for preventive education. This study was intended to highlight areas for further investigation to improve ED treatment satisfaction among veterans.

Methods

The authors conducted an anonymous survey with veterans who were prescribed an ED medication within the previous 12 months. In 2012, researchers obtained 8,000 names of veterans prescribed a PDE5 medication at the Clement J. Zablocki VA Medical Center (CJZVAMC) in Milwaukee, Wisconsin, and randomly selected 1,000 persons to mail a research survey to be returned anonymously. Three hundred ten surveys were returned, a 31% response rate, which was similar (32%) to a comparable large ED survey study, in which the participants were randomly selected to participate and also were not recruited by their PCP.¹⁰ Because 13 participants were excluded due to incomplete surveys or obtaining primary medical care services outside the VA, the current sample consisted of 297 participants. The CJZVAMC institutional review board approved the study in March 2013, and de-identified data were collected from March 2013 to March 2014.

The authors assessed demographics and treatment information, including whether veterans had talked with their PCP about sexual concerns.

Of the 297 participants, 55% were aged > 65 years. Racial/ethnic groups reflected the veteran population at CJZVAMC, with 78% identifying as European American, 17% as African American, 2% as Hispanic American, 2% as biracial, and 1% as Asian American or American Indian. Eighty-one percent were identified as Christian, and 10% reported no religious preference. Sixty-seven percent reported having a current sexual partner.

Measures

The International Index of Erectile Function (IIEF-5), an abridged version of a longer, 15-item instrument, was administered to assess participants’ erectile function.^11,12 The IIEF-5 consists of 5 items that ask about participants’ erectile functioning over the past 6 months. Participants responded to items on a 1 to 5 scale ranging from “almost never/never” to “almost always/always”. Items were summed to create a total score that could range from 5 to 25. Total scores reflect erectile functioning and satisfaction, with low scores indicating greater dysfunction. This measure has shown high sensitivity (.98) and specificity (.88).¹¹ Cut scores for the current study were consistent with the literature: mild ED = 17-21; mild-to-moderate ED = 12-16; moderate ED = 8-11; and severe ED = 5-7.¹³ Reliability in this sample was α = .93.

The authors were unable to find a validated measure assessing men’s knowledge of ED risk factors in the literature. Therefore, participants’ knowledge of ED risk factors was assessed using an online nonvalidated questionnaire entitled “Impotence [Erectile Dysfunction] Quiz: Test Your Medical IQ of ED” from www.emedicinehelp.com.¹³ Questions assess knowledge of specific risk factors (eg, age, obesity, depression, prostate cancer), symptoms, incidence rate, treatments, normal erectile functioning, and implications of ED. The questionnaire contains 16 items (3 true/false and 13 multiple choice items), and the total score corresponds to the percentage correct. According to the online version, the average score is 11 items correct (69%).¹³

A single item asked participants to identify behavioral changes they had tried to improve their erectile functioning. Options included taking medications at a different time, and/or decreasing tobacco, caffeine, or alcohol consumption. The Erectile Dysfunction Inventory of Treatment Satisfaction – Patient Version (EDITS) is an 11-item questionnaire administered to assess participants’ satisfaction with their medical treatment of ED.¹⁴ Items assess treatment satisfaction, ease of use, confidence in ability to perform, partner satisfaction, and naturalness of erections achieved during treatment. These items are rated on a scale ranging from 0 (dissatisfaction) to 4 (high satisfaction) and then summed, with total scores ranging from 0 (extremely dissatisfied) to 100 (extremely satisfied). The measure displayed high internal consistency (α = .90) and high test-retest reliability (r = .98).¹⁴ Several studies have used cutoff scores of 0 = very dissatisfied; 25 = dissatisfied; 50 = neither satisfied nor dissatisfied; 75 = satisfied; and 100 = very satisfied.^15,16 These cut scores and classifiers were used in the current study; reliability was α = .92.

The authors further explored reasons for veteran dissatisfaction with ED treatment by asking participants to respond to a single item: “Why are you dissatisfied with your erectile dysfunction treatment?” They could indicate that they were satisfied or circle all options for dissatisfaction that applied (“I would like to receive more pills per month,” “The treatment does not work well,” or “I want more information about erectile dysfunction and treatment”), or write in a response. The authors inquired about the number of pills prescribed to ascertain whether dissatisfaction was due to VA-specific policies vs veterans’ understanding of ED and effectiveness of treatment, which providers have more ability to improve.

In addition to the quantitative data obtained from the completed surveys, unsolicited responses from participants to the principal investigator via phone calls, and letters regarding treatment satisfaction were gathered. The second author conducted a basic exploratory content analysis of these unsolicited responses to group them into themes related to this study, such as satisfaction or dissatisfaction with ED treatment.

Results

The authors first assessed levels of ED and satisfaction with treatment in the sample. On average, participants reported mild-to-moderate erectile dysfunction (M = 13.1; SD = 5.7), which is higher than that of the general population and consistent with samples of men referred for ED treatment.^17,18 Satisfaction levels were slightly above neutral on the EDITS questionnaire (M = 58.3%; SD = 24.5). In response to a separate single-item question regarding reasons for dissatisfaction, only 6.4% of veterans reported being satisfied with their ED treatment.

According to respondents, the primary reasons for dissatisfaction were wanting more medication (46%), finding the treatment ineffective (26.7%), and desiring more information (24%). Further, ED severity was negatively correlated with satisfaction with ED treatment (r = .72, P < .01; note that higher scores correspond to less severe ED on this measure). However, despite moderate-to-low levels of satisfaction, 79.2% of patients planned to continue with their ED treatment (59.3% very likely and 19.9% moderately likely).

The authors also assessed participants’ communication with PCPs about their sexual functioning. Twenty-five percent reported not talking with their PCP about sexual concerns (despite all having been prescribed an ED medication in the past year). In this sample, talking with one’s PCP was not related to increased knowledge of ED risk factors (t [294] = .32, ns). Those who talked to their PCP tended to be less satisfied with treatment (M = 56.2; SD = 24.5) than those who did not talk to their PCP (M = 64.7; SD = 23.3; t (213) = -2.2; P = .03), likely because those who felt their treatment was working for them felt less need to talk to their provider. Indeed, those who talked to their PCP trended to have more severe levels of ED (M = 12.7; SD = 5.8) than those who did not (M = 14.2; SD = 5.3; t [285] = -1.91; P = .057; note that higher scores correspond to less severe ED on this measure). Finally, adults aged > 65 years were less likely to talk to their PCP than were younger adults (69% vs 81%); χ² (1, N = 291) = 5.57; P = .018.

Generally, the level of knowledge of ED risk factors was lower than the average of respondents to the original online survey (62% vs 69%).¹³ Younger adults were slightly more knowledgeable (M = 64%; SD = 13) than were older adults (M = 60%; SD = 15), t (288.08) = 2.01; P = .046).

Finally, most veterans reported few attempted behavioral changes to address ED, such as taking medications at a different time or decreasing use of tobacco, caffeine, or alcohol (M = 1.3; SD = 1.1). Thirty percent had not tried any behavioral changes; 34.1% tried 1 change; and 35.9% had tried more than 1 behavioral change. In contrast, 89% of participants reported using a PDE5 medication. Eight-two percent of participants reported currently receiving ED treatment of some kind; within this group, 97.4% reported currently taking a PDE5 medication. Only 2.5% of veterans reported using other kinds of treatment, such as vacuum pump, suppository, over-the-counter medication, injections, and not using a PDE5 medication, whereas 6.7% were using other kinds of treatment as well as a PDE5 medication.

In addition to the quantitative responses, 48 participants wrote unsolicited comments about their experiences with ED treatment on their returned questionnaires. The principal investigator also received 9 telephone calls from intended study participants, who provided verbal feedback regarding their experience with ED treatment. Comments unrelated to the study were eliminated, and the remaining written and verbal responses were grouped into categories to identify themes. Mirroring the quantitative results, participants providing qualitative feedback were dissatisfied with their ED treatment. Specifically, 43% of the comments consisted of complaints regarding the ineffectiveness and/or undesirable adverse effects (AEs) of ED medications and other ED treatments, including physical AEs (eg, headaches), sentiments that treatment does not feel “natural,” and dissatisfaction with the quality and length of sexual encounters despite treatment. Yet 24% of the comments entailed requests for more and/or different ED medications. Less frequent, although significant, comments related to decreased sexual interest and performance because of other medical conditions, such as pain, prostate surgery, and hypertension (15%); desire for additional information about ED treatments from health care providers (9%); use of nonpharmacologic ED interventions (eg, vacuum pump, 7%); and concerns about their partners’ level of sexual dissatisfaction as a result of their ED (7%).

Discussion

The present study examined knowledge of ED risk factors and level of satisfaction with ED treatment in a veteran population. Pharmacologic interventions comprised the most prevalent form of ED treatment. Both quantitative and qualitative results indicated areas for improvement in veteran satisfaction with ED treatment. Overall, veterans reported being neither satisfied nor dissatisfied with their current ED treatment, although very few reported being satisfied in response to a single item. The discrepancy may be related to the negative wording of the latter question (“Why are you dissatisfied with your erectile dysfunction treatment?”), which potentially biased participants’ responses. Several veterans also provided many unsolicited comments regarding areas for improvement. Despite feeling neutral to dissatisfied with treatment, 80% planned to continue with treatment. Sources of dissatisfaction included restricted access to ED medication (eg, limiting pills to 4 per month), ineffectiveness of treatment (eg, poor quality of erection, lack of climax), physical AEs, a desire for more information about ED, and psychological and relational concerns (eg, partner sexual dissatisfaction). As one veteran in his 80s lamented in describing the apparent end to his sexual life despite current ED treatment, “Is that all there is? It is the end of the road.”

The authors identified several barriers to implementing potentially beneficial interventions other than ED medications. Specifically, despite receiving long-term treatment for ED, veteran participants showed average knowledge of information related to ED risk factors. Of concern, discussing sexual health concerns with a PCP was not associated with increased knowledge of ED risk factors. This may explain the finding that veterans plan to continue with medication treatment despite feeling only neutral to dissatisfied about their current ED treatment.

Veterans who talked to their PCP about ED were less satisfied with treatment than were those who did not talk to their PCP, likely because those who felt their treatment was working for them felt less need to talk to their provider. Indeed, those who talked to their PCP tended to have more severe ED than those who did not. It may be that veterans avoid discussing ED with their PCP until they reach advanced ED when it is too late for treatment to make a difference. The principal investigator’s receipt of unsolicited telephone calls from intended study participants desiring to discuss ED—something that has not occurred during the researchers’ involvement in dozens of prior health-related studies—illustrates the importance veterans place on sexual concerns and the need to encourage discussion about the topic in the context of health care appointments. Specifically, older adults would benefit from more conversations with PCPs as they reported less knowledge of ED risk factors and fewer conversations with PCPs about sexual concerns than did younger men.

Adverse Events

Given the AEs reported by veterans and the significant cost of ED medications within the VA system,² increased use of alternative nonpharmacologic and preventive behavioral approaches would be clinically and economically beneficial. For example, in one study, men with ED who engaged in a lifestyle program that focused on weight loss, diet, and exercise were found more likely to experience improvements in erectile functioning compared with men who did not participate.⁶ Yet in the current study, 30% of participants had not attempted behavioral changes to address ED.

The VA’s Health Promotion and Disease Prevention (HPDP) Program focuses on preventive services and behavioral interventions to reduce health risks within primary care settings.¹⁹ This program may provide a framework for efforts to prevent and ameliorate ED. Specifically, coaching and education by HPDP experts could reduce PCPs’ discomfort with sexual health discussions and normalize the value of such conversations for both providers and patients. Existing HPDP behavioral interventions targeting areas such as weight loss and smoking cessation also could emphasize the potential secondary benefit of improved sexual functioning. To that end, preventive health campaigns could include sexual health and ED prevention as topics on patient education materials. Including sexual functioning on telephone or in-person prescreening questionnaires prior to routine appointments with PCPs also may facilitate destigmatization of sex as an important health topic.

Limitations

Limitations of the current study include its correlational design, which precludes conclusions regarding casual relationships among the variables in question. The authors cannot speculate about how well their sample represents the general veteran population given its low response rate (although comparable to a similar study).¹⁰ In addition, the lack of a validated measure of ED risk-factor knowledge meant reliance on an online questionnaire with unknown psychometric properties. To identify alternatives to pharmacologic treatment for ED, it would be beneficial for future research to examine the reasons for dissatisfaction among veterans, assessing satisfaction changes after implementation of behavioral and/or preventive interventions.

Conclusion

This study deepens the understanding of ED treatment efficacy among veterans in light of the paucity of available information. Overall, veterans are neutral to dissatisfied with their ED treatment, yet plan to continue it in the context of limited alternatives and possible lack of knowledge of behavioral methods shown to improve erectile functioning. Future studies that examine the reasons for continuing medication despite neutral satisfaction would help explore this finding. Based on these results, the authors recommend increased attention and discussion of sexual health during PCP visits and enhanced efforts toward using behavioral strategies to prevent and reduce ED. Encouragement from PCPs to address sexual health concerns earlier in a veteran’s treatment course—and in the context of behavioral and lifestyle change—may assist in preventing veterans’ sexual lives from prematurely reaching “the end of the road.”

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the Clement J. Zablocki VA Medical Center in Milwaukee, Wisconsin.

A majority of men (70%) aged ≥ 70 years report erectile dysfunction (ED) in primary care settings.¹ Further, the cost of ED medication is increasing: nationally, the VA spent $71.7 million on ED medications in 2013, triple the amount from 2006,² despite a 2011 VA mandate limiting ED medication prescriptions to 4 doses per month per veteran.³ Unfortunately, although ED is common and costly, only about 12% of men in the community report being asked about their sexual health by their primary care provider (PCP) in the past 3 years.⁴ Further, little emphasis seems to be placed on preventive care. For example, men with ED in primary care clinics are unaware of ED risk factors such as hypertension, smoking, and obesity; indeed, only 17% of a large community sample could name 1 risk factor for ED.⁵ This is problematic because diet and exercise improve erectile functioning,yet men may not realize they can reduce ED through behavioral and lifestyle change.⁶

In addition, there is little research that investigates veterans’ satisfaction with ED treatment and its effectiveness. The taboo nature of talking about erections and sexual health may partially relate to the lack of research. When surveyed, PCPs noted that they do not talk about ED routinely with patients for reasons that include time constraints, lack of experience managing sexual problems, viewing ED medication as a lifestyle drug, perceiving ED as a nonserious concern, discomfort discussing the topic for both male and female PCPs, and viewing ED discussions as the responsibility of providers of the opposite gender.^7-9

Given the dearth of ED research within the veteran population, the purpose of the current study was to (1) explore the level of treatment satisfaction of veterans prescribed an ED medication, phosphodiesterase type 5 inhibitor (PDE5); (2) assess patients’ perception of discussions with their PCPs about sexual health concerns; and (3) provide preliminary data on veterans’ knowledge of ED risk factors and identify possible areas for preventive education. This study was intended to highlight areas for further investigation to improve ED treatment satisfaction among veterans.

Methods

The authors conducted an anonymous survey with veterans who were prescribed an ED medication within the previous 12 months. In 2012, researchers obtained 8,000 names of veterans prescribed a PDE5 medication at the Clement J. Zablocki VA Medical Center (CJZVAMC) in Milwaukee, Wisconsin, and randomly selected 1,000 persons to mail a research survey to be returned anonymously. Three hundred ten surveys were returned, a 31% response rate, which was similar (32%) to a comparable large ED survey study, in which the participants were randomly selected to participate and also were not recruited by their PCP.¹⁰ Because 13 participants were excluded due to incomplete surveys or obtaining primary medical care services outside the VA, the current sample consisted of 297 participants. The CJZVAMC institutional review board approved the study in March 2013, and de-identified data were collected from March 2013 to March 2014.

The authors assessed demographics and treatment information, including whether veterans had talked with their PCP about sexual concerns.

Of the 297 participants, 55% were aged > 65 years. Racial/ethnic groups reflected the veteran population at CJZVAMC, with 78% identifying as European American, 17% as African American, 2% as Hispanic American, 2% as biracial, and 1% as Asian American or American Indian. Eighty-one percent were identified as Christian, and 10% reported no religious preference. Sixty-seven percent reported having a current sexual partner.

Measures

The International Index of Erectile Function (IIEF-5), an abridged version of a longer, 15-item instrument, was administered to assess participants’ erectile function.^11,12 The IIEF-5 consists of 5 items that ask about participants’ erectile functioning over the past 6 months. Participants responded to items on a 1 to 5 scale ranging from “almost never/never” to “almost always/always”. Items were summed to create a total score that could range from 5 to 25. Total scores reflect erectile functioning and satisfaction, with low scores indicating greater dysfunction. This measure has shown high sensitivity (.98) and specificity (.88).¹¹ Cut scores for the current study were consistent with the literature: mild ED = 17-21; mild-to-moderate ED = 12-16; moderate ED = 8-11; and severe ED = 5-7.¹³ Reliability in this sample was α = .93.

The authors were unable to find a validated measure assessing men’s knowledge of ED risk factors in the literature. Therefore, participants’ knowledge of ED risk factors was assessed using an online nonvalidated questionnaire entitled “Impotence [Erectile Dysfunction] Quiz: Test Your Medical IQ of ED” from www.emedicinehelp.com.¹³ Questions assess knowledge of specific risk factors (eg, age, obesity, depression, prostate cancer), symptoms, incidence rate, treatments, normal erectile functioning, and implications of ED. The questionnaire contains 16 items (3 true/false and 13 multiple choice items), and the total score corresponds to the percentage correct. According to the online version, the average score is 11 items correct (69%).¹³

A single item asked participants to identify behavioral changes they had tried to improve their erectile functioning. Options included taking medications at a different time, and/or decreasing tobacco, caffeine, or alcohol consumption. The Erectile Dysfunction Inventory of Treatment Satisfaction – Patient Version (EDITS) is an 11-item questionnaire administered to assess participants’ satisfaction with their medical treatment of ED.¹⁴ Items assess treatment satisfaction, ease of use, confidence in ability to perform, partner satisfaction, and naturalness of erections achieved during treatment. These items are rated on a scale ranging from 0 (dissatisfaction) to 4 (high satisfaction) and then summed, with total scores ranging from 0 (extremely dissatisfied) to 100 (extremely satisfied). The measure displayed high internal consistency (α = .90) and high test-retest reliability (r = .98).¹⁴ Several studies have used cutoff scores of 0 = very dissatisfied; 25 = dissatisfied; 50 = neither satisfied nor dissatisfied; 75 = satisfied; and 100 = very satisfied.^15,16 These cut scores and classifiers were used in the current study; reliability was α = .92.

The authors further explored reasons for veteran dissatisfaction with ED treatment by asking participants to respond to a single item: “Why are you dissatisfied with your erectile dysfunction treatment?” They could indicate that they were satisfied or circle all options for dissatisfaction that applied (“I would like to receive more pills per month,” “The treatment does not work well,” or “I want more information about erectile dysfunction and treatment”), or write in a response. The authors inquired about the number of pills prescribed to ascertain whether dissatisfaction was due to VA-specific policies vs veterans’ understanding of ED and effectiveness of treatment, which providers have more ability to improve.

In addition to the quantitative data obtained from the completed surveys, unsolicited responses from participants to the principal investigator via phone calls, and letters regarding treatment satisfaction were gathered. The second author conducted a basic exploratory content analysis of these unsolicited responses to group them into themes related to this study, such as satisfaction or dissatisfaction with ED treatment.

Results

The authors first assessed levels of ED and satisfaction with treatment in the sample. On average, participants reported mild-to-moderate erectile dysfunction (M = 13.1; SD = 5.7), which is higher than that of the general population and consistent with samples of men referred for ED treatment.^17,18 Satisfaction levels were slightly above neutral on the EDITS questionnaire (M = 58.3%; SD = 24.5). In response to a separate single-item question regarding reasons for dissatisfaction, only 6.4% of veterans reported being satisfied with their ED treatment.

According to respondents, the primary reasons for dissatisfaction were wanting more medication (46%), finding the treatment ineffective (26.7%), and desiring more information (24%). Further, ED severity was negatively correlated with satisfaction with ED treatment (r = .72, P < .01; note that higher scores correspond to less severe ED on this measure). However, despite moderate-to-low levels of satisfaction, 79.2% of patients planned to continue with their ED treatment (59.3% very likely and 19.9% moderately likely).

The authors also assessed participants’ communication with PCPs about their sexual functioning. Twenty-five percent reported not talking with their PCP about sexual concerns (despite all having been prescribed an ED medication in the past year). In this sample, talking with one’s PCP was not related to increased knowledge of ED risk factors (t [294] = .32, ns). Those who talked to their PCP tended to be less satisfied with treatment (M = 56.2; SD = 24.5) than those who did not talk to their PCP (M = 64.7; SD = 23.3; t (213) = -2.2; P = .03), likely because those who felt their treatment was working for them felt less need to talk to their provider. Indeed, those who talked to their PCP trended to have more severe levels of ED (M = 12.7; SD = 5.8) than those who did not (M = 14.2; SD = 5.3; t [285] = -1.91; P = .057; note that higher scores correspond to less severe ED on this measure). Finally, adults aged > 65 years were less likely to talk to their PCP than were younger adults (69% vs 81%); χ² (1, N = 291) = 5.57; P = .018.

Generally, the level of knowledge of ED risk factors was lower than the average of respondents to the original online survey (62% vs 69%).¹³ Younger adults were slightly more knowledgeable (M = 64%; SD = 13) than were older adults (M = 60%; SD = 15), t (288.08) = 2.01; P = .046).

Finally, most veterans reported few attempted behavioral changes to address ED, such as taking medications at a different time or decreasing use of tobacco, caffeine, or alcohol (M = 1.3; SD = 1.1). Thirty percent had not tried any behavioral changes; 34.1% tried 1 change; and 35.9% had tried more than 1 behavioral change. In contrast, 89% of participants reported using a PDE5 medication. Eight-two percent of participants reported currently receiving ED treatment of some kind; within this group, 97.4% reported currently taking a PDE5 medication. Only 2.5% of veterans reported using other kinds of treatment, such as vacuum pump, suppository, over-the-counter medication, injections, and not using a PDE5 medication, whereas 6.7% were using other kinds of treatment as well as a PDE5 medication.

In addition to the quantitative responses, 48 participants wrote unsolicited comments about their experiences with ED treatment on their returned questionnaires. The principal investigator also received 9 telephone calls from intended study participants, who provided verbal feedback regarding their experience with ED treatment. Comments unrelated to the study were eliminated, and the remaining written and verbal responses were grouped into categories to identify themes. Mirroring the quantitative results, participants providing qualitative feedback were dissatisfied with their ED treatment. Specifically, 43% of the comments consisted of complaints regarding the ineffectiveness and/or undesirable adverse effects (AEs) of ED medications and other ED treatments, including physical AEs (eg, headaches), sentiments that treatment does not feel “natural,” and dissatisfaction with the quality and length of sexual encounters despite treatment. Yet 24% of the comments entailed requests for more and/or different ED medications. Less frequent, although significant, comments related to decreased sexual interest and performance because of other medical conditions, such as pain, prostate surgery, and hypertension (15%); desire for additional information about ED treatments from health care providers (9%); use of nonpharmacologic ED interventions (eg, vacuum pump, 7%); and concerns about their partners’ level of sexual dissatisfaction as a result of their ED (7%).

Discussion

The present study examined knowledge of ED risk factors and level of satisfaction with ED treatment in a veteran population. Pharmacologic interventions comprised the most prevalent form of ED treatment. Both quantitative and qualitative results indicated areas for improvement in veteran satisfaction with ED treatment. Overall, veterans reported being neither satisfied nor dissatisfied with their current ED treatment, although very few reported being satisfied in response to a single item. The discrepancy may be related to the negative wording of the latter question (“Why are you dissatisfied with your erectile dysfunction treatment?”), which potentially biased participants’ responses. Several veterans also provided many unsolicited comments regarding areas for improvement. Despite feeling neutral to dissatisfied with treatment, 80% planned to continue with treatment. Sources of dissatisfaction included restricted access to ED medication (eg, limiting pills to 4 per month), ineffectiveness of treatment (eg, poor quality of erection, lack of climax), physical AEs, a desire for more information about ED, and psychological and relational concerns (eg, partner sexual dissatisfaction). As one veteran in his 80s lamented in describing the apparent end to his sexual life despite current ED treatment, “Is that all there is? It is the end of the road.”

The authors identified several barriers to implementing potentially beneficial interventions other than ED medications. Specifically, despite receiving long-term treatment for ED, veteran participants showed average knowledge of information related to ED risk factors. Of concern, discussing sexual health concerns with a PCP was not associated with increased knowledge of ED risk factors. This may explain the finding that veterans plan to continue with medication treatment despite feeling only neutral to dissatisfied about their current ED treatment.

Veterans who talked to their PCP about ED were less satisfied with treatment than were those who did not talk to their PCP, likely because those who felt their treatment was working for them felt less need to talk to their provider. Indeed, those who talked to their PCP tended to have more severe ED than those who did not. It may be that veterans avoid discussing ED with their PCP until they reach advanced ED when it is too late for treatment to make a difference. The principal investigator’s receipt of unsolicited telephone calls from intended study participants desiring to discuss ED—something that has not occurred during the researchers’ involvement in dozens of prior health-related studies—illustrates the importance veterans place on sexual concerns and the need to encourage discussion about the topic in the context of health care appointments. Specifically, older adults would benefit from more conversations with PCPs as they reported less knowledge of ED risk factors and fewer conversations with PCPs about sexual concerns than did younger men.

Adverse Events

Given the AEs reported by veterans and the significant cost of ED medications within the VA system,² increased use of alternative nonpharmacologic and preventive behavioral approaches would be clinically and economically beneficial. For example, in one study, men with ED who engaged in a lifestyle program that focused on weight loss, diet, and exercise were found more likely to experience improvements in erectile functioning compared with men who did not participate.⁶ Yet in the current study, 30% of participants had not attempted behavioral changes to address ED.

The VA’s Health Promotion and Disease Prevention (HPDP) Program focuses on preventive services and behavioral interventions to reduce health risks within primary care settings.¹⁹ This program may provide a framework for efforts to prevent and ameliorate ED. Specifically, coaching and education by HPDP experts could reduce PCPs’ discomfort with sexual health discussions and normalize the value of such conversations for both providers and patients. Existing HPDP behavioral interventions targeting areas such as weight loss and smoking cessation also could emphasize the potential secondary benefit of improved sexual functioning. To that end, preventive health campaigns could include sexual health and ED prevention as topics on patient education materials. Including sexual functioning on telephone or in-person prescreening questionnaires prior to routine appointments with PCPs also may facilitate destigmatization of sex as an important health topic.

Limitations

Limitations of the current study include its correlational design, which precludes conclusions regarding casual relationships among the variables in question. The authors cannot speculate about how well their sample represents the general veteran population given its low response rate (although comparable to a similar study).¹⁰ In addition, the lack of a validated measure of ED risk-factor knowledge meant reliance on an online questionnaire with unknown psychometric properties. To identify alternatives to pharmacologic treatment for ED, it would be beneficial for future research to examine the reasons for dissatisfaction among veterans, assessing satisfaction changes after implementation of behavioral and/or preventive interventions.

Conclusion

This study deepens the understanding of ED treatment efficacy among veterans in light of the paucity of available information. Overall, veterans are neutral to dissatisfied with their ED treatment, yet plan to continue it in the context of limited alternatives and possible lack of knowledge of behavioral methods shown to improve erectile functioning. Future studies that examine the reasons for continuing medication despite neutral satisfaction would help explore this finding. Based on these results, the authors recommend increased attention and discussion of sexual health during PCP visits and enhanced efforts toward using behavioral strategies to prevent and reduce ED. Encouragement from PCPs to address sexual health concerns earlier in a veteran’s treatment course—and in the context of behavioral and lifestyle change—may assist in preventing veterans’ sexual lives from prematurely reaching “the end of the road.”

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the Clement J. Zablocki VA Medical Center in Milwaukee, Wisconsin.

Multiple sclerosis (MS) is an autoimmune disorder in which the myelin of the brain and spinal cord is selectively targeted by immune-system cells. As a result, nerve transmission is disrupted, leading to a variety of unpredictable symptoms from weakness and a lack of balance to blindness and paralysis of the body. Clinically, MS can take 4 courses, including relapsing-remitting (RRMS), primary-progressive (PPMS), secondary-progressive, and progressive-relapsing.¹ At onset, 85% of diagnosed patients have RRMS, and 10% to 15% have PPMS.² If untreated, patients with RRMS become secondary-progressive, with progressive disability and indiscrete relapses.³ Hence, disease-modifying therapies are targeted toward decreasing the relapse rate as well as slowing the progression of the disease.⁴

Annually, about 16,000 veterans with MS receive health care services from the VHA.⁵ The C.W. Bill Young Bay Pines VA Healthcare System (BPVAHCS) is a level 1 facility that annually serves more than 105,000 veterans. The BPVAMC sees veterans with a wide variety of neurologic illnesses and has 5 full-time neurologists with subspecialty training. The BPVAHCS facility has outpatient clinics and a 200 inpatient bed facility. The Neurology Department sees 125 outpatients per week and consults on about 30 inpatients per week.

Methods

A retrospective review of BPVAHCS patients diagnosed with MS from January 2009 to July 2014 was performed with institutional review board approval. Patient data were collected from ICD-9-CM codes and kept confidential. A list of patients was collected from Neurology Clinic patient visits with “Multiple Sclerosis” on the problem list.

Patient medical records were reviewed to collect the following information: presence of rigorous diagnosis of MS, clinical course of MS in patient, presence or absence of disease-modifying therapy, and disease-modifying agents (DMAs) used.

Determining factors for DMA treatment included increasing tiredness, weakness, visual symptoms, and radiologic evidence (magnetic resonance imaging) of recurrent, active lesions. Each patient was examined on a case-by-case basis to assess whether or not the patient actually had MS and if so, whether they were being treated with DMAs. Only patients with RRMS were included. Patients were excluded from the study if they were deceased, not currently under BPVACHS care, or had symptoms of optic neuritis but were not fully indicative of MS. Patients with clinically isolated syndrome, probable diagnosis of MS, or PPMS also were excluded from the study.

Exclusion from this study was based on 2 additional premises. Patients were excluded if they discontinued an initial ABC (interferon beta-1a, interferon beta-1b, glatiramer acetate) due to DMA treatment relapse or adverse effects (AEs), such as injection site reactions, flulike symptoms, or depression. Additionally, patients who were not willing to take more DMA medications were excluded if they felt they were relatively stable (had infrequent relapses) and believed that additional medication was not worth the risk of potential AEs.

The study patients were seen and followed up by the neurologists. All the data for this study were based on interactions with the neurologists and not primary care providers (PCPs). Because MS treatment is complex, PCPs have little involvement in its management. The percentage of patients not on any DMAs was calculated from the list of BPVAHCS patients with RRMS.

The results were compared with a similar retrospective cohort study conducted using the Commercial Claims database and Medicare Supplemental and Coordination of Benefits database to identify individuals newly diagnosed with MS.⁶ This study was chosen because it was similar in methodology but investigated a comparable non-VA group. A 2-tailed difference between proportions test was then performed to determine whether the BPVAHCS patients with MS who were not treated with DMAs were significantly different from those from this non-VA population. Additionally, data from VA patients who were receiving DMAs were further examined and presented.

Results

At the BPVAHCS, 262 patients were diagnosed with MS and 43% were not treated with DMAs. Margolis and colleagues found that about 60% of its 11,061 newly diagnosed non-VA patients with MS remained untreated.⁶ Although the latter proportion is higher, a 2-tailed difference between proportions test indicates that the proportion of patients with MS being treated at the VA was significantly lower (P < .01).

Among the 148 patients who were diagnosed with MS and treated with DMA at BPVAHCS, 5 different DMAs were identified (Table). The most commonly prescribed regimen was glatiramer acetate, which was used by 56 of 148 patients (37.8%). Fifty-two patients (35.1%) used interferon beta-1a. Of the 2 interferon DMAs, beta-1a was twice as popular as beta-1b, which was prescribed to 22 (14.9%) of patients. Dimethyl fumarate (6.8%) and fingolimod (5.4%) were used sparingly, because they were new to the market (cost and availability also were factors). With time, increased efficacy and objective assessment of benefit in the reduction of the T2 lesion load may result in a greater use of these oral DMAs.^7–9

Based on this evaluation, 43% of patients who were diagnosed with MS were untreated at BPVAHCS. Concern over treatment AEs, the inconvenience of injectable dosing, and patients who were not 100% service-connected and lost to follow-up because of the cost may have contributed to the poor rate of treatment.

Discussion

Injected-based DMAs, such as interferon beta-1a, interferon beta-1b, and glatiramer acetate, were first introduced in the 1990s, but these proved to be inconvenient and triggered AEs, including injection site reactions. Overall, their efficacy was about 30%, with interferon beta-1a showing a 27% reduction in relapses.¹⁰ In 2010, oral DMAs, such as fingolimod, were FDA approved. These oral DMAs were a significant improvement over injectable DMAs but still had AEs. Hence, their use was restricted to neurologists by the BPVAHCS, and rightfully so.

Still, newer and more effective oral DMAs are showing promise, such as dimethyl fumarate, teriflunomide, and alemtuzumab. These new DMAs have significantly impacted the treatment of MS as they are not only easier for patients to adhere to and for neurologists to prescribe, but most significantly, have had a 50% decrease in the rate of relapse.¹⁰ Yet, the newer oral DMAs were less commonly prescribed than the older treatments at BPVAHCS.

Since this study did not demonstrate increased use of oral DMAs at the BPVAHCS, more PCP and neurologist-focused educational programs on the use of DMAs may be beneficial. Educational programs should lead to a reevaluation of patients with MS to consider oral DMAs, which offer better efficacy and fewer AEs. The newer oral DMAs have shown a higher reduction of T2 lesions, and the significantly decreased incidence of relapses in many other medical facilities is quite promising for the BPVAHCS.^7-9

The data collected at BPVAHCS were part of a quality improvement (QI) study that will be used by the Neurology Department to follow up on the patients with MS in order to implement DMA therapies. A questionnaire was developed for following up with BPVAHCS patients with MS. The primary purpose of the questionnaire is to help neurologists identify the reasons patients avoid DMA therapies and to reduce the number of BPVAHCS patients not on the most efficacious MS DMA treatment.

Conclusion

Multiple sclerosis is a disease without a cure. Current treatment strategies focus on modifying the course of the disease and managing its symptoms. However, even as promising new treatments emerge, the current literature suggests that a significant number of patients diagnosed with MS are not receiving DMAs and may not be receiving optimal treatment.¹¹

Findings from this study indicate that although DMAs are optimal for patients with MS, they may not be prescribed as frequently at BPVAHCS as they are at a non-VA care facility. It is unclear whether this finding is explained by an educational gap, clinical differences between non-VA and VA patients, organizational factors, or a combination of these variables. Further study is warranted to examine the use of DMAs among veterans with MS and factors that facilitate or impede optimal practice. The BPVAHCS will use data from this retrospective cohort study in a QI initiative for patients with MS. Findings from the QI initiative will be reported using the Standards for Quality Improvement Reporting Excellence.^12,13

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the Bay Pines VA Healthcare System.

Multiple sclerosis (MS) is an autoimmune disorder in which the myelin of the brain and spinal cord is selectively targeted by immune-system cells. As a result, nerve transmission is disrupted, leading to a variety of unpredictable symptoms from weakness and a lack of balance to blindness and paralysis of the body. Clinically, MS can take 4 courses, including relapsing-remitting (RRMS), primary-progressive (PPMS), secondary-progressive, and progressive-relapsing.¹ At onset, 85% of diagnosed patients have RRMS, and 10% to 15% have PPMS.² If untreated, patients with RRMS become secondary-progressive, with progressive disability and indiscrete relapses.³ Hence, disease-modifying therapies are targeted toward decreasing the relapse rate as well as slowing the progression of the disease.⁴

Annually, about 16,000 veterans with MS receive health care services from the VHA.⁵ The C.W. Bill Young Bay Pines VA Healthcare System (BPVAHCS) is a level 1 facility that annually serves more than 105,000 veterans. The BPVAMC sees veterans with a wide variety of neurologic illnesses and has 5 full-time neurologists with subspecialty training. The BPVAHCS facility has outpatient clinics and a 200 inpatient bed facility. The Neurology Department sees 125 outpatients per week and consults on about 30 inpatients per week.

Methods

A retrospective review of BPVAHCS patients diagnosed with MS from January 2009 to July 2014 was performed with institutional review board approval. Patient data were collected from ICD-9-CM codes and kept confidential. A list of patients was collected from Neurology Clinic patient visits with “Multiple Sclerosis” on the problem list.

Patient medical records were reviewed to collect the following information: presence of rigorous diagnosis of MS, clinical course of MS in patient, presence or absence of disease-modifying therapy, and disease-modifying agents (DMAs) used.

Determining factors for DMA treatment included increasing tiredness, weakness, visual symptoms, and radiologic evidence (magnetic resonance imaging) of recurrent, active lesions. Each patient was examined on a case-by-case basis to assess whether or not the patient actually had MS and if so, whether they were being treated with DMAs. Only patients with RRMS were included. Patients were excluded from the study if they were deceased, not currently under BPVACHS care, or had symptoms of optic neuritis but were not fully indicative of MS. Patients with clinically isolated syndrome, probable diagnosis of MS, or PPMS also were excluded from the study.

Exclusion from this study was based on 2 additional premises. Patients were excluded if they discontinued an initial ABC (interferon beta-1a, interferon beta-1b, glatiramer acetate) due to DMA treatment relapse or adverse effects (AEs), such as injection site reactions, flulike symptoms, or depression. Additionally, patients who were not willing to take more DMA medications were excluded if they felt they were relatively stable (had infrequent relapses) and believed that additional medication was not worth the risk of potential AEs.

The study patients were seen and followed up by the neurologists. All the data for this study were based on interactions with the neurologists and not primary care providers (PCPs). Because MS treatment is complex, PCPs have little involvement in its management. The percentage of patients not on any DMAs was calculated from the list of BPVAHCS patients with RRMS.

The results were compared with a similar retrospective cohort study conducted using the Commercial Claims database and Medicare Supplemental and Coordination of Benefits database to identify individuals newly diagnosed with MS.⁶ This study was chosen because it was similar in methodology but investigated a comparable non-VA group. A 2-tailed difference between proportions test was then performed to determine whether the BPVAHCS patients with MS who were not treated with DMAs were significantly different from those from this non-VA population. Additionally, data from VA patients who were receiving DMAs were further examined and presented.

Results

At the BPVAHCS, 262 patients were diagnosed with MS and 43% were not treated with DMAs. Margolis and colleagues found that about 60% of its 11,061 newly diagnosed non-VA patients with MS remained untreated.⁶ Although the latter proportion is higher, a 2-tailed difference between proportions test indicates that the proportion of patients with MS being treated at the VA was significantly lower (P < .01).

Among the 148 patients who were diagnosed with MS and treated with DMA at BPVAHCS, 5 different DMAs were identified (Table). The most commonly prescribed regimen was glatiramer acetate, which was used by 56 of 148 patients (37.8%). Fifty-two patients (35.1%) used interferon beta-1a. Of the 2 interferon DMAs, beta-1a was twice as popular as beta-1b, which was prescribed to 22 (14.9%) of patients. Dimethyl fumarate (6.8%) and fingolimod (5.4%) were used sparingly, because they were new to the market (cost and availability also were factors). With time, increased efficacy and objective assessment of benefit in the reduction of the T2 lesion load may result in a greater use of these oral DMAs.^7–9

Based on this evaluation, 43% of patients who were diagnosed with MS were untreated at BPVAHCS. Concern over treatment AEs, the inconvenience of injectable dosing, and patients who were not 100% service-connected and lost to follow-up because of the cost may have contributed to the poor rate of treatment.

Discussion

Injected-based DMAs, such as interferon beta-1a, interferon beta-1b, and glatiramer acetate, were first introduced in the 1990s, but these proved to be inconvenient and triggered AEs, including injection site reactions. Overall, their efficacy was about 30%, with interferon beta-1a showing a 27% reduction in relapses.¹⁰ In 2010, oral DMAs, such as fingolimod, were FDA approved. These oral DMAs were a significant improvement over injectable DMAs but still had AEs. Hence, their use was restricted to neurologists by the BPVAHCS, and rightfully so.

Still, newer and more effective oral DMAs are showing promise, such as dimethyl fumarate, teriflunomide, and alemtuzumab. These new DMAs have significantly impacted the treatment of MS as they are not only easier for patients to adhere to and for neurologists to prescribe, but most significantly, have had a 50% decrease in the rate of relapse.¹⁰ Yet, the newer oral DMAs were less commonly prescribed than the older treatments at BPVAHCS.

Since this study did not demonstrate increased use of oral DMAs at the BPVAHCS, more PCP and neurologist-focused educational programs on the use of DMAs may be beneficial. Educational programs should lead to a reevaluation of patients with MS to consider oral DMAs, which offer better efficacy and fewer AEs. The newer oral DMAs have shown a higher reduction of T2 lesions, and the significantly decreased incidence of relapses in many other medical facilities is quite promising for the BPVAHCS.^7-9

The data collected at BPVAHCS were part of a quality improvement (QI) study that will be used by the Neurology Department to follow up on the patients with MS in order to implement DMA therapies. A questionnaire was developed for following up with BPVAHCS patients with MS. The primary purpose of the questionnaire is to help neurologists identify the reasons patients avoid DMA therapies and to reduce the number of BPVAHCS patients not on the most efficacious MS DMA treatment.

Conclusion

Multiple sclerosis is a disease without a cure. Current treatment strategies focus on modifying the course of the disease and managing its symptoms. However, even as promising new treatments emerge, the current literature suggests that a significant number of patients diagnosed with MS are not receiving DMAs and may not be receiving optimal treatment.¹¹

Findings from this study indicate that although DMAs are optimal for patients with MS, they may not be prescribed as frequently at BPVAHCS as they are at a non-VA care facility. It is unclear whether this finding is explained by an educational gap, clinical differences between non-VA and VA patients, organizational factors, or a combination of these variables. Further study is warranted to examine the use of DMAs among veterans with MS and factors that facilitate or impede optimal practice. The BPVAHCS will use data from this retrospective cohort study in a QI initiative for patients with MS. Findings from the QI initiative will be reported using the Standards for Quality Improvement Reporting Excellence.^12,13

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the Bay Pines VA Healthcare System.

Multiple sclerosis (MS) is an autoimmune disorder in which the myelin of the brain and spinal cord is selectively targeted by immune-system cells. As a result, nerve transmission is disrupted, leading to a variety of unpredictable symptoms from weakness and a lack of balance to blindness and paralysis of the body. Clinically, MS can take 4 courses, including relapsing-remitting (RRMS), primary-progressive (PPMS), secondary-progressive, and progressive-relapsing.¹ At onset, 85% of diagnosed patients have RRMS, and 10% to 15% have PPMS.² If untreated, patients with RRMS become secondary-progressive, with progressive disability and indiscrete relapses.³ Hence, disease-modifying therapies are targeted toward decreasing the relapse rate as well as slowing the progression of the disease.⁴

Annually, about 16,000 veterans with MS receive health care services from the VHA.⁵ The C.W. Bill Young Bay Pines VA Healthcare System (BPVAHCS) is a level 1 facility that annually serves more than 105,000 veterans. The BPVAMC sees veterans with a wide variety of neurologic illnesses and has 5 full-time neurologists with subspecialty training. The BPVAHCS facility has outpatient clinics and a 200 inpatient bed facility. The Neurology Department sees 125 outpatients per week and consults on about 30 inpatients per week.

Methods

A retrospective review of BPVAHCS patients diagnosed with MS from January 2009 to July 2014 was performed with institutional review board approval. Patient data were collected from ICD-9-CM codes and kept confidential. A list of patients was collected from Neurology Clinic patient visits with “Multiple Sclerosis” on the problem list.

Patient medical records were reviewed to collect the following information: presence of rigorous diagnosis of MS, clinical course of MS in patient, presence or absence of disease-modifying therapy, and disease-modifying agents (DMAs) used.

Determining factors for DMA treatment included increasing tiredness, weakness, visual symptoms, and radiologic evidence (magnetic resonance imaging) of recurrent, active lesions. Each patient was examined on a case-by-case basis to assess whether or not the patient actually had MS and if so, whether they were being treated with DMAs. Only patients with RRMS were included. Patients were excluded from the study if they were deceased, not currently under BPVACHS care, or had symptoms of optic neuritis but were not fully indicative of MS. Patients with clinically isolated syndrome, probable diagnosis of MS, or PPMS also were excluded from the study.

Exclusion from this study was based on 2 additional premises. Patients were excluded if they discontinued an initial ABC (interferon beta-1a, interferon beta-1b, glatiramer acetate) due to DMA treatment relapse or adverse effects (AEs), such as injection site reactions, flulike symptoms, or depression. Additionally, patients who were not willing to take more DMA medications were excluded if they felt they were relatively stable (had infrequent relapses) and believed that additional medication was not worth the risk of potential AEs.

The study patients were seen and followed up by the neurologists. All the data for this study were based on interactions with the neurologists and not primary care providers (PCPs). Because MS treatment is complex, PCPs have little involvement in its management. The percentage of patients not on any DMAs was calculated from the list of BPVAHCS patients with RRMS.

The results were compared with a similar retrospective cohort study conducted using the Commercial Claims database and Medicare Supplemental and Coordination of Benefits database to identify individuals newly diagnosed with MS.⁶ This study was chosen because it was similar in methodology but investigated a comparable non-VA group. A 2-tailed difference between proportions test was then performed to determine whether the BPVAHCS patients with MS who were not treated with DMAs were significantly different from those from this non-VA population. Additionally, data from VA patients who were receiving DMAs were further examined and presented.

Results

At the BPVAHCS, 262 patients were diagnosed with MS and 43% were not treated with DMAs. Margolis and colleagues found that about 60% of its 11,061 newly diagnosed non-VA patients with MS remained untreated.⁶ Although the latter proportion is higher, a 2-tailed difference between proportions test indicates that the proportion of patients with MS being treated at the VA was significantly lower (P < .01).

Among the 148 patients who were diagnosed with MS and treated with DMA at BPVAHCS, 5 different DMAs were identified (Table). The most commonly prescribed regimen was glatiramer acetate, which was used by 56 of 148 patients (37.8%). Fifty-two patients (35.1%) used interferon beta-1a. Of the 2 interferon DMAs, beta-1a was twice as popular as beta-1b, which was prescribed to 22 (14.9%) of patients. Dimethyl fumarate (6.8%) and fingolimod (5.4%) were used sparingly, because they were new to the market (cost and availability also were factors). With time, increased efficacy and objective assessment of benefit in the reduction of the T2 lesion load may result in a greater use of these oral DMAs.^7–9

Based on this evaluation, 43% of patients who were diagnosed with MS were untreated at BPVAHCS. Concern over treatment AEs, the inconvenience of injectable dosing, and patients who were not 100% service-connected and lost to follow-up because of the cost may have contributed to the poor rate of treatment.

Discussion

Injected-based DMAs, such as interferon beta-1a, interferon beta-1b, and glatiramer acetate, were first introduced in the 1990s, but these proved to be inconvenient and triggered AEs, including injection site reactions. Overall, their efficacy was about 30%, with interferon beta-1a showing a 27% reduction in relapses.¹⁰ In 2010, oral DMAs, such as fingolimod, were FDA approved. These oral DMAs were a significant improvement over injectable DMAs but still had AEs. Hence, their use was restricted to neurologists by the BPVAHCS, and rightfully so.

Still, newer and more effective oral DMAs are showing promise, such as dimethyl fumarate, teriflunomide, and alemtuzumab. These new DMAs have significantly impacted the treatment of MS as they are not only easier for patients to adhere to and for neurologists to prescribe, but most significantly, have had a 50% decrease in the rate of relapse.¹⁰ Yet, the newer oral DMAs were less commonly prescribed than the older treatments at BPVAHCS.

Since this study did not demonstrate increased use of oral DMAs at the BPVAHCS, more PCP and neurologist-focused educational programs on the use of DMAs may be beneficial. Educational programs should lead to a reevaluation of patients with MS to consider oral DMAs, which offer better efficacy and fewer AEs. The newer oral DMAs have shown a higher reduction of T2 lesions, and the significantly decreased incidence of relapses in many other medical facilities is quite promising for the BPVAHCS.^7-9

The data collected at BPVAHCS were part of a quality improvement (QI) study that will be used by the Neurology Department to follow up on the patients with MS in order to implement DMA therapies. A questionnaire was developed for following up with BPVAHCS patients with MS. The primary purpose of the questionnaire is to help neurologists identify the reasons patients avoid DMA therapies and to reduce the number of BPVAHCS patients not on the most efficacious MS DMA treatment.

Conclusion

Multiple sclerosis is a disease without a cure. Current treatment strategies focus on modifying the course of the disease and managing its symptoms. However, even as promising new treatments emerge, the current literature suggests that a significant number of patients diagnosed with MS are not receiving DMAs and may not be receiving optimal treatment.¹¹

Findings from this study indicate that although DMAs are optimal for patients with MS, they may not be prescribed as frequently at BPVAHCS as they are at a non-VA care facility. It is unclear whether this finding is explained by an educational gap, clinical differences between non-VA and VA patients, organizational factors, or a combination of these variables. Further study is warranted to examine the use of DMAs among veterans with MS and factors that facilitate or impede optimal practice. The BPVAHCS will use data from this retrospective cohort study in a QI initiative for patients with MS. Findings from the QI initiative will be reported using the Standards for Quality Improvement Reporting Excellence.^12,13

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the Bay Pines VA Healthcare System.