Orthopedics

The 2 cardinal properties of stem cells are the ability to self-renew and the ability to differentiate into distinctive end-stage cell types. The work of Caplan¹ captured our early attention, with cells cultured from bone marrow differentiating into a number of different cell types of orthopedic interest. Our latest attention has been captured by the additional abilities of these cells to mobilize, monitor, and interact with their surrounding environment.^2-4 In response to their environment, stem cells are able to release a broad spectrum of macromolecules with trophic, chemotactic, and immunomodulatory potential, which allows them to participate in injury response, tissue healing, and tissue regeneration.⁴ These cells are innate to the body’s monitoring, maintenance, repair, and stress response systems.^2,4-11 Basic science and animal studies have illustrated the potential of cells with stem potential regardless of their environment/source of harvest.

Where Can We Get Stem Cells?

Cells with stem properties are present in many environmental niches, including the bone marrow, peripheral circulatory system, adipose tissue, synovial tissue, muscle tissue, and tendon tissue.^12-15 A number of cell types with stem properties populate the bone marrow niche, including hematopoietic stem/progenitor cells (HSPC), perivascular stromal cells (PSC), endothelial stem cells (ESC), and immature cells with qualities like embryonal stem cells termed very small embryonal-like stem cells (VESL).^12,15-19 All of these cells have stem properties and have been shown to differentiate to tissues of orthopedic interest.The interplay, interaction, and potential of these cell types is complex and incompletely understood.^12,15-19 When bone marrow is aspirated for culturing purposes, it is unclear which cell line produces the plastic-adherent multipotent cells grown in culture, which are often referred to as mesenchymal stem cells (MSCs). Researches propose that HSPC and/or VESL circulate peripherally in small numbers but leave the bone marrow in certain mobilization instances and are important for the monitoring and maintenance of the majority of tissues in our bodies.^5,16 Current clinical utilization of these cell types by the orthopedic community primarily utilizes point-of-care bone marrow aspiration and concentration, while the hematology oncology community mobilizes cells from the bone marrow to the blood stream with pharmaceutical agents and harvests cells via apheresis. Bone marrow aspiration produces variable numbers of stem cells, with studies ranging from 1 stem cell per mL of tissue collected to 300,000 stem cells per mL of tissue collected.20Mobilization and apheresis can produce large volumes of peripheral blood-derived cells with 600,000 HSPC per mL and 2.32 million PSC per mL of tissue collected.²¹

In adipose tissue, cells adherent to the abluminal side of blood vessels known as pericytes also carry stem qualities. Aspiration and processing of adipose tissue can access these stem cells, producing a product often referred to as stromal vascular fraction (SVF). Processing of lipoaspirate to create stromal vascular fraction requires mechanical or enzymatic processing. This also produces variable numbers of stem cells, with quantitative studies ranging from 5000 to 1.5 million stem cells per mL of tissue collected.²⁰ Similar to adipose-derived stem cells, synovial-derived and muscle-derived stem cells also require mechanical or enzymatic processing. For applications where it is believed that a large number of cells is necessary, investigators often utilize culturing techniques for all sources with the exception of mobilization and apheresis harvest. As clinicians, 3 challenges have proven more important than which cell type to utilize: 1) patient-care logistics regarding collection and application; 2) the undefined dose-response curve regarding stem cell treatments; and 3) evolving government/community regulation.

Regulation of Stem Cell Therapies

The regulation of stem cell technologies is a double-edged sword for development. While loose regulation encourages clinical application and experimentation, patient safety and efficacy concerns are raised, and a technology’s worth is not proven before clinical application. Tight regulation temporarily hampers progress, yet ensures the proof of safety and efficacy prior to widespread implementation. Within the United States, the Food and Drug Administration (FDA) has tightened regulation, established precedent, and intervened in the ability of clinicians to utilize stem cell therapies in humans, through “warning letters,” “untitled letters,” and industry guidance documents.^22-30

The FDA categorizes stem cell therapies as human cells, tissues, and cellular- and tissue-based products (HCT/Ps). Section 361 of the Public Health Safety (PHS) Act established and outlined the authority of the FDA to regulate low-risk HCT/Ps in order to prevent the introduction, transmission, and spread of communicable disease. Section 361 provided standards for safety without requiring preclinical development. The FDA established 4 principles to determine the risk of HCT/Ps: the extent of manipulation involved in manufacture, the metabolic activity/autologous nature of the product, whether the product represents a tissue combined with another product, and whether the product is utilized in a fashion homologous with its original function (Figure 1). If a product/therapy meets requirements around all 4 of these principles, then it is deemed a low-risk product and regulated under Section 361 alone. If a product/therapy does not meet requirements around all 4 of these principles, then the FDA regulates the product/therapy under additional codes including Section 351 of the PHS Act. Section 351 outlines a developmental process including preclinical animal trials, phased clinical study, and premarket review by the FDA prior to offering the product/treatment in clinical practice. The developmental process requires investigators and/or industry developers to initiate an Investigational New Drug (IND) program whose end goal is to present data from all developmental study and obtain a Biologic License Application (BLA) approval to market the product.^22-23 To establish safety and efficacy, the traditional IND program involves a preclinical animal study, a small pilot human study (Phase I), a small initial randomized controlled trial (Phase II), followed by a large multicenter randomized controlled trial (Phase III) (Figure 2). The FDA has recognized little to no stem cell treatments as products regulated by Section 361 alone. Additionally, the FDA has established precedent regarding allograft stem cells, cells obtained from fat harvest, amniotic/placental products, and cultured cells, suggesting that these products are not low risk and require an IND pathway outlined in Section 351.^24-30

Bone Marrow Aspiration

Surprisingly, the FDA has not moved to regulate the point-of-care use of bone marrow aspirate or platelet-rich plasma and has labeled these as “not HCTPs.” The stem cell concentration of bone marrow aspirate is technique-dependent, declines with age, and has been found to be an important factor for clinical benefit.³¹ While it is possible to aspirate from multiple sites, posterior iliac crest harvest produces the highest stem cell yield.^32-34 Hernigou and colleagues^35-36 have outlined safe zones for trocar placement and illustrated that strong aspiration with small-volume syringes, 10-mL syringes, optimizes stem cell harvest. Additionally, studies by Hernigou and colleagues^31,37-38 involving tibial nonunion, avascular necrosis of the femur, and augmentation of rotator cuff repair are guideposts to clinicians utilizing bone marrow aspirate.

Amniotic Stem Cell Technologies and Adipose-Derived Stem Cells

While some argue that there is regulatory confusion around amniotic/placental-derived tissues and adipose-derived products, the FDA has clearly established precedent establishing these as products requiring Section 351 development.^26-29 Companies are marketing products derived from perinatal byproducts, yet there are multiple FDA letters suggesting that these are not products regulated solely under PHS Act 361 because they do not meet the criteria of homologous use and are not autologous.^28-29 Use of these products places risk upon the clinician and the patient. Some argue that adipose-derived stem cell products are 361 products. While the FDA has approved devices for the mechanical processing of lipoaspirate, they have established precedent suggesting that they consider orthopedic applications nonhomologous and any processing that “alters the original relevant characteristics of adipose tissue relating to the tissue’s utility for reconstruction, repair, or replacement” as more than minimal manipulation.^26,27 The FDA originally planned an open forum for discussion with clinicians and industry for April 2016. This open forum was delayed due to the volume of interest, and a workshop has been planned for Fall 2016.

Future Regulation of Stem Cell Technologies

While many countries have mirrored the FDA with tight regulatory mechanisms, a few countries have established modern regulatory mechanisms aimed at the promotion of conscientious development, including South Korea, Japan, and England. For example, in 2014 Japan labeled stem cell technologies as “regenerative medicine products,” setting them apart from pharmaceuticals, and implemented a new approval system allowing early observed commercialization with reimbursement after less stringent safety and efficacy milestones.²²The observed commercialization lowers time and financial hurdles for development while still requiring the proof of the technology’s worth. Countries that have effected change have positioned themselves to be pioneers in this emerging field.

In March 2016, the Reliable and Effective Growth for Regenerative Health Options that Improve Wellness (REGROW) Act of 2016 (S. 2689 / H.R. 4762) was introduced into the United States Congress. This bipartisan, bicameral legislation was introduced, read twice, and referred to subcommittee. Its goal is to reduce barriers and accelerate development of biologic therapies while keeping the frame work set forth under Sections 351 and 361 of the PHS Act.³⁹ Similar to the pathway in Japan, the REGROW Act would establish a conditional approval pathway that would ensure products are safe and effective while also evolving the regulatory pathway towards progress (Figure 3). Development would still require an IND application after preclinical animal study. However, after safety was established with human Phase I data and preliminary evidence of efficacy with Phase II data, patients could be treated with the investigational therapies and reimbursement collected for a limited period of time (5 years) prior to a large Phase III human clinical trial. Patients treated with the new therapy would be monitored closely. All results would be reported to the FDA in a BLA. This change in legislation would lower but not remove regulatory hurdles necessary for development.

Conclusion

The future of stem cell treatments hinges upon the creation of new favorable regulatory mechanisms that will promote clinical application while ensuring that safety and efficacy milestones are reached. Clinical researchers require freedom to develop these technologies while protecting patients and ensuring the validity of treatments. The coordination of research and regulatory affairs on a global level is necessary focusing on the harmonization of guidelines, regulations, and mechanisms for simultaneous adoption in different countries. The global orthopedic community has made strides regarding the science of stem cell technologies; it is time for us to initiate progressive change regarding regulation so that we can determine what is effective clinically.

The 2 cardinal properties of stem cells are the ability to self-renew and the ability to differentiate into distinctive end-stage cell types. The work of Caplan¹ captured our early attention, with cells cultured from bone marrow differentiating into a number of different cell types of orthopedic interest. Our latest attention has been captured by the additional abilities of these cells to mobilize, monitor, and interact with their surrounding environment.^2-4 In response to their environment, stem cells are able to release a broad spectrum of macromolecules with trophic, chemotactic, and immunomodulatory potential, which allows them to participate in injury response, tissue healing, and tissue regeneration.⁴ These cells are innate to the body’s monitoring, maintenance, repair, and stress response systems.^2,4-11 Basic science and animal studies have illustrated the potential of cells with stem potential regardless of their environment/source of harvest.

Where Can We Get Stem Cells?

Cells with stem properties are present in many environmental niches, including the bone marrow, peripheral circulatory system, adipose tissue, synovial tissue, muscle tissue, and tendon tissue.^12-15 A number of cell types with stem properties populate the bone marrow niche, including hematopoietic stem/progenitor cells (HSPC), perivascular stromal cells (PSC), endothelial stem cells (ESC), and immature cells with qualities like embryonal stem cells termed very small embryonal-like stem cells (VESL).^12,15-19 All of these cells have stem properties and have been shown to differentiate to tissues of orthopedic interest.The interplay, interaction, and potential of these cell types is complex and incompletely understood.^12,15-19 When bone marrow is aspirated for culturing purposes, it is unclear which cell line produces the plastic-adherent multipotent cells grown in culture, which are often referred to as mesenchymal stem cells (MSCs). Researches propose that HSPC and/or VESL circulate peripherally in small numbers but leave the bone marrow in certain mobilization instances and are important for the monitoring and maintenance of the majority of tissues in our bodies.^5,16 Current clinical utilization of these cell types by the orthopedic community primarily utilizes point-of-care bone marrow aspiration and concentration, while the hematology oncology community mobilizes cells from the bone marrow to the blood stream with pharmaceutical agents and harvests cells via apheresis. Bone marrow aspiration produces variable numbers of stem cells, with studies ranging from 1 stem cell per mL of tissue collected to 300,000 stem cells per mL of tissue collected.20Mobilization and apheresis can produce large volumes of peripheral blood-derived cells with 600,000 HSPC per mL and 2.32 million PSC per mL of tissue collected.²¹

In adipose tissue, cells adherent to the abluminal side of blood vessels known as pericytes also carry stem qualities. Aspiration and processing of adipose tissue can access these stem cells, producing a product often referred to as stromal vascular fraction (SVF). Processing of lipoaspirate to create stromal vascular fraction requires mechanical or enzymatic processing. This also produces variable numbers of stem cells, with quantitative studies ranging from 5000 to 1.5 million stem cells per mL of tissue collected.²⁰ Similar to adipose-derived stem cells, synovial-derived and muscle-derived stem cells also require mechanical or enzymatic processing. For applications where it is believed that a large number of cells is necessary, investigators often utilize culturing techniques for all sources with the exception of mobilization and apheresis harvest. As clinicians, 3 challenges have proven more important than which cell type to utilize: 1) patient-care logistics regarding collection and application; 2) the undefined dose-response curve regarding stem cell treatments; and 3) evolving government/community regulation.

Regulation of Stem Cell Therapies

The regulation of stem cell technologies is a double-edged sword for development. While loose regulation encourages clinical application and experimentation, patient safety and efficacy concerns are raised, and a technology’s worth is not proven before clinical application. Tight regulation temporarily hampers progress, yet ensures the proof of safety and efficacy prior to widespread implementation. Within the United States, the Food and Drug Administration (FDA) has tightened regulation, established precedent, and intervened in the ability of clinicians to utilize stem cell therapies in humans, through “warning letters,” “untitled letters,” and industry guidance documents.^22-30

The FDA categorizes stem cell therapies as human cells, tissues, and cellular- and tissue-based products (HCT/Ps). Section 361 of the Public Health Safety (PHS) Act established and outlined the authority of the FDA to regulate low-risk HCT/Ps in order to prevent the introduction, transmission, and spread of communicable disease. Section 361 provided standards for safety without requiring preclinical development. The FDA established 4 principles to determine the risk of HCT/Ps: the extent of manipulation involved in manufacture, the metabolic activity/autologous nature of the product, whether the product represents a tissue combined with another product, and whether the product is utilized in a fashion homologous with its original function (Figure 1). If a product/therapy meets requirements around all 4 of these principles, then it is deemed a low-risk product and regulated under Section 361 alone. If a product/therapy does not meet requirements around all 4 of these principles, then the FDA regulates the product/therapy under additional codes including Section 351 of the PHS Act. Section 351 outlines a developmental process including preclinical animal trials, phased clinical study, and premarket review by the FDA prior to offering the product/treatment in clinical practice. The developmental process requires investigators and/or industry developers to initiate an Investigational New Drug (IND) program whose end goal is to present data from all developmental study and obtain a Biologic License Application (BLA) approval to market the product.^22-23 To establish safety and efficacy, the traditional IND program involves a preclinical animal study, a small pilot human study (Phase I), a small initial randomized controlled trial (Phase II), followed by a large multicenter randomized controlled trial (Phase III) (Figure 2). The FDA has recognized little to no stem cell treatments as products regulated by Section 361 alone. Additionally, the FDA has established precedent regarding allograft stem cells, cells obtained from fat harvest, amniotic/placental products, and cultured cells, suggesting that these products are not low risk and require an IND pathway outlined in Section 351.^24-30

Bone Marrow Aspiration

Surprisingly, the FDA has not moved to regulate the point-of-care use of bone marrow aspirate or platelet-rich plasma and has labeled these as “not HCTPs.” The stem cell concentration of bone marrow aspirate is technique-dependent, declines with age, and has been found to be an important factor for clinical benefit.³¹ While it is possible to aspirate from multiple sites, posterior iliac crest harvest produces the highest stem cell yield.^32-34 Hernigou and colleagues^35-36 have outlined safe zones for trocar placement and illustrated that strong aspiration with small-volume syringes, 10-mL syringes, optimizes stem cell harvest. Additionally, studies by Hernigou and colleagues^31,37-38 involving tibial nonunion, avascular necrosis of the femur, and augmentation of rotator cuff repair are guideposts to clinicians utilizing bone marrow aspirate.

Amniotic Stem Cell Technologies and Adipose-Derived Stem Cells

While some argue that there is regulatory confusion around amniotic/placental-derived tissues and adipose-derived products, the FDA has clearly established precedent establishing these as products requiring Section 351 development.^26-29 Companies are marketing products derived from perinatal byproducts, yet there are multiple FDA letters suggesting that these are not products regulated solely under PHS Act 361 because they do not meet the criteria of homologous use and are not autologous.^28-29 Use of these products places risk upon the clinician and the patient. Some argue that adipose-derived stem cell products are 361 products. While the FDA has approved devices for the mechanical processing of lipoaspirate, they have established precedent suggesting that they consider orthopedic applications nonhomologous and any processing that “alters the original relevant characteristics of adipose tissue relating to the tissue’s utility for reconstruction, repair, or replacement” as more than minimal manipulation.^26,27 The FDA originally planned an open forum for discussion with clinicians and industry for April 2016. This open forum was delayed due to the volume of interest, and a workshop has been planned for Fall 2016.

Future Regulation of Stem Cell Technologies

While many countries have mirrored the FDA with tight regulatory mechanisms, a few countries have established modern regulatory mechanisms aimed at the promotion of conscientious development, including South Korea, Japan, and England. For example, in 2014 Japan labeled stem cell technologies as “regenerative medicine products,” setting them apart from pharmaceuticals, and implemented a new approval system allowing early observed commercialization with reimbursement after less stringent safety and efficacy milestones.²²The observed commercialization lowers time and financial hurdles for development while still requiring the proof of the technology’s worth. Countries that have effected change have positioned themselves to be pioneers in this emerging field.

In March 2016, the Reliable and Effective Growth for Regenerative Health Options that Improve Wellness (REGROW) Act of 2016 (S. 2689 / H.R. 4762) was introduced into the United States Congress. This bipartisan, bicameral legislation was introduced, read twice, and referred to subcommittee. Its goal is to reduce barriers and accelerate development of biologic therapies while keeping the frame work set forth under Sections 351 and 361 of the PHS Act.³⁹ Similar to the pathway in Japan, the REGROW Act would establish a conditional approval pathway that would ensure products are safe and effective while also evolving the regulatory pathway towards progress (Figure 3). Development would still require an IND application after preclinical animal study. However, after safety was established with human Phase I data and preliminary evidence of efficacy with Phase II data, patients could be treated with the investigational therapies and reimbursement collected for a limited period of time (5 years) prior to a large Phase III human clinical trial. Patients treated with the new therapy would be monitored closely. All results would be reported to the FDA in a BLA. This change in legislation would lower but not remove regulatory hurdles necessary for development.

Conclusion

The future of stem cell treatments hinges upon the creation of new favorable regulatory mechanisms that will promote clinical application while ensuring that safety and efficacy milestones are reached. Clinical researchers require freedom to develop these technologies while protecting patients and ensuring the validity of treatments. The coordination of research and regulatory affairs on a global level is necessary focusing on the harmonization of guidelines, regulations, and mechanisms for simultaneous adoption in different countries. The global orthopedic community has made strides regarding the science of stem cell technologies; it is time for us to initiate progressive change regarding regulation so that we can determine what is effective clinically.

The 2 cardinal properties of stem cells are the ability to self-renew and the ability to differentiate into distinctive end-stage cell types. The work of Caplan¹ captured our early attention, with cells cultured from bone marrow differentiating into a number of different cell types of orthopedic interest. Our latest attention has been captured by the additional abilities of these cells to mobilize, monitor, and interact with their surrounding environment.^2-4 In response to their environment, stem cells are able to release a broad spectrum of macromolecules with trophic, chemotactic, and immunomodulatory potential, which allows them to participate in injury response, tissue healing, and tissue regeneration.⁴ These cells are innate to the body’s monitoring, maintenance, repair, and stress response systems.^2,4-11 Basic science and animal studies have illustrated the potential of cells with stem potential regardless of their environment/source of harvest.

Where Can We Get Stem Cells?

Cells with stem properties are present in many environmental niches, including the bone marrow, peripheral circulatory system, adipose tissue, synovial tissue, muscle tissue, and tendon tissue.^12-15 A number of cell types with stem properties populate the bone marrow niche, including hematopoietic stem/progenitor cells (HSPC), perivascular stromal cells (PSC), endothelial stem cells (ESC), and immature cells with qualities like embryonal stem cells termed very small embryonal-like stem cells (VESL).^12,15-19 All of these cells have stem properties and have been shown to differentiate to tissues of orthopedic interest.The interplay, interaction, and potential of these cell types is complex and incompletely understood.^12,15-19 When bone marrow is aspirated for culturing purposes, it is unclear which cell line produces the plastic-adherent multipotent cells grown in culture, which are often referred to as mesenchymal stem cells (MSCs). Researches propose that HSPC and/or VESL circulate peripherally in small numbers but leave the bone marrow in certain mobilization instances and are important for the monitoring and maintenance of the majority of tissues in our bodies.^5,16 Current clinical utilization of these cell types by the orthopedic community primarily utilizes point-of-care bone marrow aspiration and concentration, while the hematology oncology community mobilizes cells from the bone marrow to the blood stream with pharmaceutical agents and harvests cells via apheresis. Bone marrow aspiration produces variable numbers of stem cells, with studies ranging from 1 stem cell per mL of tissue collected to 300,000 stem cells per mL of tissue collected.20Mobilization and apheresis can produce large volumes of peripheral blood-derived cells with 600,000 HSPC per mL and 2.32 million PSC per mL of tissue collected.²¹

In adipose tissue, cells adherent to the abluminal side of blood vessels known as pericytes also carry stem qualities. Aspiration and processing of adipose tissue can access these stem cells, producing a product often referred to as stromal vascular fraction (SVF). Processing of lipoaspirate to create stromal vascular fraction requires mechanical or enzymatic processing. This also produces variable numbers of stem cells, with quantitative studies ranging from 5000 to 1.5 million stem cells per mL of tissue collected.²⁰ Similar to adipose-derived stem cells, synovial-derived and muscle-derived stem cells also require mechanical or enzymatic processing. For applications where it is believed that a large number of cells is necessary, investigators often utilize culturing techniques for all sources with the exception of mobilization and apheresis harvest. As clinicians, 3 challenges have proven more important than which cell type to utilize: 1) patient-care logistics regarding collection and application; 2) the undefined dose-response curve regarding stem cell treatments; and 3) evolving government/community regulation.

Regulation of Stem Cell Therapies

The regulation of stem cell technologies is a double-edged sword for development. While loose regulation encourages clinical application and experimentation, patient safety and efficacy concerns are raised, and a technology’s worth is not proven before clinical application. Tight regulation temporarily hampers progress, yet ensures the proof of safety and efficacy prior to widespread implementation. Within the United States, the Food and Drug Administration (FDA) has tightened regulation, established precedent, and intervened in the ability of clinicians to utilize stem cell therapies in humans, through “warning letters,” “untitled letters,” and industry guidance documents.^22-30

The FDA categorizes stem cell therapies as human cells, tissues, and cellular- and tissue-based products (HCT/Ps). Section 361 of the Public Health Safety (PHS) Act established and outlined the authority of the FDA to regulate low-risk HCT/Ps in order to prevent the introduction, transmission, and spread of communicable disease. Section 361 provided standards for safety without requiring preclinical development. The FDA established 4 principles to determine the risk of HCT/Ps: the extent of manipulation involved in manufacture, the metabolic activity/autologous nature of the product, whether the product represents a tissue combined with another product, and whether the product is utilized in a fashion homologous with its original function (Figure 1). If a product/therapy meets requirements around all 4 of these principles, then it is deemed a low-risk product and regulated under Section 361 alone. If a product/therapy does not meet requirements around all 4 of these principles, then the FDA regulates the product/therapy under additional codes including Section 351 of the PHS Act. Section 351 outlines a developmental process including preclinical animal trials, phased clinical study, and premarket review by the FDA prior to offering the product/treatment in clinical practice. The developmental process requires investigators and/or industry developers to initiate an Investigational New Drug (IND) program whose end goal is to present data from all developmental study and obtain a Biologic License Application (BLA) approval to market the product.^22-23 To establish safety and efficacy, the traditional IND program involves a preclinical animal study, a small pilot human study (Phase I), a small initial randomized controlled trial (Phase II), followed by a large multicenter randomized controlled trial (Phase III) (Figure 2). The FDA has recognized little to no stem cell treatments as products regulated by Section 361 alone. Additionally, the FDA has established precedent regarding allograft stem cells, cells obtained from fat harvest, amniotic/placental products, and cultured cells, suggesting that these products are not low risk and require an IND pathway outlined in Section 351.^24-30

Bone Marrow Aspiration

Surprisingly, the FDA has not moved to regulate the point-of-care use of bone marrow aspirate or platelet-rich plasma and has labeled these as “not HCTPs.” The stem cell concentration of bone marrow aspirate is technique-dependent, declines with age, and has been found to be an important factor for clinical benefit.³¹ While it is possible to aspirate from multiple sites, posterior iliac crest harvest produces the highest stem cell yield.^32-34 Hernigou and colleagues^35-36 have outlined safe zones for trocar placement and illustrated that strong aspiration with small-volume syringes, 10-mL syringes, optimizes stem cell harvest. Additionally, studies by Hernigou and colleagues^31,37-38 involving tibial nonunion, avascular necrosis of the femur, and augmentation of rotator cuff repair are guideposts to clinicians utilizing bone marrow aspirate.

Amniotic Stem Cell Technologies and Adipose-Derived Stem Cells

While some argue that there is regulatory confusion around amniotic/placental-derived tissues and adipose-derived products, the FDA has clearly established precedent establishing these as products requiring Section 351 development.^26-29 Companies are marketing products derived from perinatal byproducts, yet there are multiple FDA letters suggesting that these are not products regulated solely under PHS Act 361 because they do not meet the criteria of homologous use and are not autologous.^28-29 Use of these products places risk upon the clinician and the patient. Some argue that adipose-derived stem cell products are 361 products. While the FDA has approved devices for the mechanical processing of lipoaspirate, they have established precedent suggesting that they consider orthopedic applications nonhomologous and any processing that “alters the original relevant characteristics of adipose tissue relating to the tissue’s utility for reconstruction, repair, or replacement” as more than minimal manipulation.^26,27 The FDA originally planned an open forum for discussion with clinicians and industry for April 2016. This open forum was delayed due to the volume of interest, and a workshop has been planned for Fall 2016.

Future Regulation of Stem Cell Technologies

While many countries have mirrored the FDA with tight regulatory mechanisms, a few countries have established modern regulatory mechanisms aimed at the promotion of conscientious development, including South Korea, Japan, and England. For example, in 2014 Japan labeled stem cell technologies as “regenerative medicine products,” setting them apart from pharmaceuticals, and implemented a new approval system allowing early observed commercialization with reimbursement after less stringent safety and efficacy milestones.²²The observed commercialization lowers time and financial hurdles for development while still requiring the proof of the technology’s worth. Countries that have effected change have positioned themselves to be pioneers in this emerging field.

In March 2016, the Reliable and Effective Growth for Regenerative Health Options that Improve Wellness (REGROW) Act of 2016 (S. 2689 / H.R. 4762) was introduced into the United States Congress. This bipartisan, bicameral legislation was introduced, read twice, and referred to subcommittee. Its goal is to reduce barriers and accelerate development of biologic therapies while keeping the frame work set forth under Sections 351 and 361 of the PHS Act.³⁹ Similar to the pathway in Japan, the REGROW Act would establish a conditional approval pathway that would ensure products are safe and effective while also evolving the regulatory pathway towards progress (Figure 3). Development would still require an IND application after preclinical animal study. However, after safety was established with human Phase I data and preliminary evidence of efficacy with Phase II data, patients could be treated with the investigational therapies and reimbursement collected for a limited period of time (5 years) prior to a large Phase III human clinical trial. Patients treated with the new therapy would be monitored closely. All results would be reported to the FDA in a BLA. This change in legislation would lower but not remove regulatory hurdles necessary for development.

Conclusion

The future of stem cell treatments hinges upon the creation of new favorable regulatory mechanisms that will promote clinical application while ensuring that safety and efficacy milestones are reached. Clinical researchers require freedom to develop these technologies while protecting patients and ensuring the validity of treatments. The coordination of research and regulatory affairs on a global level is necessary focusing on the harmonization of guidelines, regulations, and mechanisms for simultaneous adoption in different countries. The global orthopedic community has made strides regarding the science of stem cell technologies; it is time for us to initiate progressive change regarding regulation so that we can determine what is effective clinically.

The sun has finally set on Kobe Bryant’s magnificent career. After all the tributes and tearful goodbyes, he has finally played his last game and become a part of basketball history. Ever since his field trip to Germany for interleukin-1 receptor antagonist protein (IRAP) treatments to his knee, and his subsequent return to high-level play, I’ve been under siege in the office by patients who “want what Kobe had.” I’ve had to explain, time and time again, that IRAP treatment is not available in the United States and that platelet-rich plasma (PRP) is the closest alternative treatment, convince them that PRP may be even better, and then let them know that it’s considered experimental and not covered by insurance.

In the last issue, we discussed the future of orthopedics, which in my opinion will rely heavily on the biologic therapies now considered experimental. In this issue, we will look into our crystal balls and imagine what that future might look like. To do so, we should first consider what we hope to accomplish through the incorporation of biologic therapies.

The regeneration of articular cartilage, acceleration of fracture and tissue healing, and faster incorporation of tendon grafts to bone have long been considered the Holy Grail of Orthopedics. In his best seller, The Da Vinci Code, Dan Brown makes a compelling argument that the Holy Grail, the chalice thought to have held the blood of Christ, was in fact a mistranslated reference to his living descendants. Whenever I have a visitor or student in the operating room, I focus the scope on the synovial capillaries so they can see the individual red blood cells passing single-file through the vessels on their way to supply cells with the nutrients they need.

Perhaps, like in The Da Vinci Code, the solution to our greatest biologic challenges lies in the blood, already there, just waiting to be unlocked.

PRP has been utilized for everything from tendinopathy to arthropathy, with varied results in the literature. The lack of standardization of PRP preparations, which vary in inclusion of white cells and absolute platelet count, confounds these results even further. In this issue, we review its use in sports medicine and knee arthritis, taking a closer look at partial ulnar collateral ligament tears in baseball players.

In “Tips of the Trade,” we present a technique for “superior capsular reconstruction” that provides a novel solution for patients with pseudoparalysis from massive rotator cuff tears with little other options beside reverse total shoulder arthroplasty.

The one absolute statement I can make regarding biologics is that we currently have more questions than answers, and every hypothesis we prove simply begets more questions. More randomized controlled studies are needed in virtually every aspect of biologics, and we should all consider taking part. While the solutions our patients crave may not arrive during our careers, or even our lifetimes, the groundwork we do now will set the stage for future generations to enjoy biologically enhanced outcomes.

The sun has finally set on Kobe Bryant’s magnificent career. After all the tributes and tearful goodbyes, he has finally played his last game and become a part of basketball history. Ever since his field trip to Germany for interleukin-1 receptor antagonist protein (IRAP) treatments to his knee, and his subsequent return to high-level play, I’ve been under siege in the office by patients who “want what Kobe had.” I’ve had to explain, time and time again, that IRAP treatment is not available in the United States and that platelet-rich plasma (PRP) is the closest alternative treatment, convince them that PRP may be even better, and then let them know that it’s considered experimental and not covered by insurance.

In the last issue, we discussed the future of orthopedics, which in my opinion will rely heavily on the biologic therapies now considered experimental. In this issue, we will look into our crystal balls and imagine what that future might look like. To do so, we should first consider what we hope to accomplish through the incorporation of biologic therapies.

The regeneration of articular cartilage, acceleration of fracture and tissue healing, and faster incorporation of tendon grafts to bone have long been considered the Holy Grail of Orthopedics. In his best seller, The Da Vinci Code, Dan Brown makes a compelling argument that the Holy Grail, the chalice thought to have held the blood of Christ, was in fact a mistranslated reference to his living descendants. Whenever I have a visitor or student in the operating room, I focus the scope on the synovial capillaries so they can see the individual red blood cells passing single-file through the vessels on their way to supply cells with the nutrients they need.

Perhaps, like in The Da Vinci Code, the solution to our greatest biologic challenges lies in the blood, already there, just waiting to be unlocked.

PRP has been utilized for everything from tendinopathy to arthropathy, with varied results in the literature. The lack of standardization of PRP preparations, which vary in inclusion of white cells and absolute platelet count, confounds these results even further. In this issue, we review its use in sports medicine and knee arthritis, taking a closer look at partial ulnar collateral ligament tears in baseball players.

In “Tips of the Trade,” we present a technique for “superior capsular reconstruction” that provides a novel solution for patients with pseudoparalysis from massive rotator cuff tears with little other options beside reverse total shoulder arthroplasty.

The one absolute statement I can make regarding biologics is that we currently have more questions than answers, and every hypothesis we prove simply begets more questions. More randomized controlled studies are needed in virtually every aspect of biologics, and we should all consider taking part. While the solutions our patients crave may not arrive during our careers, or even our lifetimes, the groundwork we do now will set the stage for future generations to enjoy biologically enhanced outcomes.

The sun has finally set on Kobe Bryant’s magnificent career. After all the tributes and tearful goodbyes, he has finally played his last game and become a part of basketball history. Ever since his field trip to Germany for interleukin-1 receptor antagonist protein (IRAP) treatments to his knee, and his subsequent return to high-level play, I’ve been under siege in the office by patients who “want what Kobe had.” I’ve had to explain, time and time again, that IRAP treatment is not available in the United States and that platelet-rich plasma (PRP) is the closest alternative treatment, convince them that PRP may be even better, and then let them know that it’s considered experimental and not covered by insurance.

In the last issue, we discussed the future of orthopedics, which in my opinion will rely heavily on the biologic therapies now considered experimental. In this issue, we will look into our crystal balls and imagine what that future might look like. To do so, we should first consider what we hope to accomplish through the incorporation of biologic therapies.

The regeneration of articular cartilage, acceleration of fracture and tissue healing, and faster incorporation of tendon grafts to bone have long been considered the Holy Grail of Orthopedics. In his best seller, The Da Vinci Code, Dan Brown makes a compelling argument that the Holy Grail, the chalice thought to have held the blood of Christ, was in fact a mistranslated reference to his living descendants. Whenever I have a visitor or student in the operating room, I focus the scope on the synovial capillaries so they can see the individual red blood cells passing single-file through the vessels on their way to supply cells with the nutrients they need.

Perhaps, like in The Da Vinci Code, the solution to our greatest biologic challenges lies in the blood, already there, just waiting to be unlocked.

PRP has been utilized for everything from tendinopathy to arthropathy, with varied results in the literature. The lack of standardization of PRP preparations, which vary in inclusion of white cells and absolute platelet count, confounds these results even further. In this issue, we review its use in sports medicine and knee arthritis, taking a closer look at partial ulnar collateral ligament tears in baseball players.

In “Tips of the Trade,” we present a technique for “superior capsular reconstruction” that provides a novel solution for patients with pseudoparalysis from massive rotator cuff tears with little other options beside reverse total shoulder arthroplasty.

The one absolute statement I can make regarding biologics is that we currently have more questions than answers, and every hypothesis we prove simply begets more questions. More randomized controlled studies are needed in virtually every aspect of biologics, and we should all consider taking part. While the solutions our patients crave may not arrive during our careers, or even our lifetimes, the groundwork we do now will set the stage for future generations to enjoy biologically enhanced outcomes.

A new index can be used to predict which women will experience faster bone loss while transitioning to menopause, according to a study in the Journal of Clinical Endocrinology & Metabolism.

To create the new index, called the Bone Balance Index, researchers used data from a cohort of 685 women ages 42 to 52 as they went through menopause. The women were either premenopausal or in early perimenopause when they enrolled in the study, and all participants had their final menstrual period during follow-up.

Urine and blood samples were taken from the women to measure bone turnover markers. The women also had their bone mineral density measured every year.

The investigators combined measurements of bone breakdown and bone formation to determine each woman’s net bone balance before their final menstrual period. The study authors found that compared to a measurement of bone breakdown alone, the Bone Balance Index was a stronger predictor of bone loss from 2 years before the final menstrual period to 3 to 4 years later.

A new index can be used to predict which women will experience faster bone loss while transitioning to menopause, according to a study in the Journal of Clinical Endocrinology & Metabolism.

To create the new index, called the Bone Balance Index, researchers used data from a cohort of 685 women ages 42 to 52 as they went through menopause. The women were either premenopausal or in early perimenopause when they enrolled in the study, and all participants had their final menstrual period during follow-up.

Urine and blood samples were taken from the women to measure bone turnover markers. The women also had their bone mineral density measured every year.

The investigators combined measurements of bone breakdown and bone formation to determine each woman’s net bone balance before their final menstrual period. The study authors found that compared to a measurement of bone breakdown alone, the Bone Balance Index was a stronger predictor of bone loss from 2 years before the final menstrual period to 3 to 4 years later.

A new index can be used to predict which women will experience faster bone loss while transitioning to menopause, according to a study in the Journal of Clinical Endocrinology & Metabolism.

To create the new index, called the Bone Balance Index, researchers used data from a cohort of 685 women ages 42 to 52 as they went through menopause. The women were either premenopausal or in early perimenopause when they enrolled in the study, and all participants had their final menstrual period during follow-up.

Urine and blood samples were taken from the women to measure bone turnover markers. The women also had their bone mineral density measured every year.

The investigators combined measurements of bone breakdown and bone formation to determine each woman’s net bone balance before their final menstrual period. The study authors found that compared to a measurement of bone breakdown alone, the Bone Balance Index was a stronger predictor of bone loss from 2 years before the final menstrual period to 3 to 4 years later.

A substantial percentage of patients who receive opioid medications before undergoing arthroplasty continue to take them up to 6 months after surgery, according to a study published in Pain.

Researchers analyzed opioid use in 574 patients who underwent arthroplasty. Patients were followed up at 1, 3, and 6 months after surgery to assess rates of long-term opioid use and risk factors for long-term opioid use. About 30% of patients were taking opioids prior to their joint replacement surgery. Of this group, 53% of knee-replacement patients and 35% of hip replacement patients continued taking opioids 6 months after surgery.

Patients who were not taking opioids prior to surgery were less likely to report persistent opioid use. About 8% in the knee replacement group and 4% in the hip replacement group were still taking opioids at the 6-month follow-up. Patients who were taking the highest doses of opioids before surgery were most likely to continue to take them for 6 months.

Among patients not previously taking opioids, those with higher pain scores the day of surgery were more likely to report persistent opioid use at 6 months. However, improvement in knee or hip pain after arthroplasty did not reduce the likelihood of long-term opioid use.

A substantial percentage of patients who receive opioid medications before undergoing arthroplasty continue to take them up to 6 months after surgery, according to a study published in Pain.

Researchers analyzed opioid use in 574 patients who underwent arthroplasty. Patients were followed up at 1, 3, and 6 months after surgery to assess rates of long-term opioid use and risk factors for long-term opioid use. About 30% of patients were taking opioids prior to their joint replacement surgery. Of this group, 53% of knee-replacement patients and 35% of hip replacement patients continued taking opioids 6 months after surgery.

Patients who were not taking opioids prior to surgery were less likely to report persistent opioid use. About 8% in the knee replacement group and 4% in the hip replacement group were still taking opioids at the 6-month follow-up. Patients who were taking the highest doses of opioids before surgery were most likely to continue to take them for 6 months.

Among patients not previously taking opioids, those with higher pain scores the day of surgery were more likely to report persistent opioid use at 6 months. However, improvement in knee or hip pain after arthroplasty did not reduce the likelihood of long-term opioid use.

A substantial percentage of patients who receive opioid medications before undergoing arthroplasty continue to take them up to 6 months after surgery, according to a study published in Pain.

Researchers analyzed opioid use in 574 patients who underwent arthroplasty. Patients were followed up at 1, 3, and 6 months after surgery to assess rates of long-term opioid use and risk factors for long-term opioid use. About 30% of patients were taking opioids prior to their joint replacement surgery. Of this group, 53% of knee-replacement patients and 35% of hip replacement patients continued taking opioids 6 months after surgery.

Patients who were not taking opioids prior to surgery were less likely to report persistent opioid use. About 8% in the knee replacement group and 4% in the hip replacement group were still taking opioids at the 6-month follow-up. Patients who were taking the highest doses of opioids before surgery were most likely to continue to take them for 6 months.

Among patients not previously taking opioids, those with higher pain scores the day of surgery were more likely to report persistent opioid use at 6 months. However, improvement in knee or hip pain after arthroplasty did not reduce the likelihood of long-term opioid use.

The American Academy of Orthopedic Surgeons has introduced apps that orthopedic surgeons can use to explain musculoskeletal problems and procedures to their patients. The Guides to Orthopedic Surgery cover total knee replacement, total hip replacement, and ACL reconstruction. These apps can be loaded onto exam room desktops or used on an iPad.

The apps also provide ways to create custom educational information for patients, and may be set up with certain electronic medical records to support Meaningful Use requirements. A free trial of the apps is available until June 30. More information: www.aaosnotice.org/Ortho_App/.

The American Academy of Orthopedic Surgeons has introduced apps that orthopedic surgeons can use to explain musculoskeletal problems and procedures to their patients. The Guides to Orthopedic Surgery cover total knee replacement, total hip replacement, and ACL reconstruction. These apps can be loaded onto exam room desktops or used on an iPad.

The apps also provide ways to create custom educational information for patients, and may be set up with certain electronic medical records to support Meaningful Use requirements. A free trial of the apps is available until June 30. More information: www.aaosnotice.org/Ortho_App/.

The American Academy of Orthopedic Surgeons has introduced apps that orthopedic surgeons can use to explain musculoskeletal problems and procedures to their patients. The Guides to Orthopedic Surgery cover total knee replacement, total hip replacement, and ACL reconstruction. These apps can be loaded onto exam room desktops or used on an iPad.

The apps also provide ways to create custom educational information for patients, and may be set up with certain electronic medical records to support Meaningful Use requirements. A free trial of the apps is available until June 30. More information: www.aaosnotice.org/Ortho_App/.

People with osteoarthritis who go on to have a total or partial knee replacement do not appear to have an increased risk of all-cause mortality, but the jury is still out on whether they gain any improvement, a study showed.

In their research published in the Annals of the Rheumatic Diseases [2016 May 17. doi: 10.1136/annrheumdis-2016-209167], Dr. Devyani Misra of Boston University and colleagues noted that knee replacement (KR) was thought to decrease long-term mortality risk because of the relief from pain and improvement in function that typically comes with surgery. However, studies on the topic had been conflicting, largely because of the challenges associated with studying mortality with KR surgery in observational settings.

©Nandyphotos/Thinkstock

In the current study the research team sought to evaluate the relation of KR to the risk of all-cause mortality among subjects with knee OA, while at the same time giving particular attention to “potential sources of confounding bias that may account for [the] effect of KR on mortality.”

Using patient data from the U.K. primary care electronic database THIN, the investigators compared the risk of mortality among 14,042 subjects who had OA, were aged 50-89 years old, and had had or had not had KR.

They discovered a strong protective effect of KR on all-cause long-term mortality risk, particularly among the adults over 63 years of age.

For example, people who had undergone KR had a 28% lower risk of mortality than did non-KR subjects (hazard ratio, 0.72; 95% confidence interval, 0.66-0.78).

In the overall propensity score–matched study sample, crude mortality per 1,000 person-years (total person-years) for the KR and non-KR cohorts were 19 (61,015) and 25 (58,294), respectively.

However, despite their best efforts, the researchers said the results showed evidence of residual confounding.

“For example, the observation of improved survival immediately after KR, despite the expectation of potential short-term increased postoperative mortality risk supports the presence of residual confounding,” they wrote.

Another finding suggestive of confounding was that the protective effect was seen only in older patients (over 63) when the authors stratified study participants by age.

“While it is possible that survival benefit seen in older patients with KR is a true effect because it is in this group that greater physical activity is particularly important to survival, more likely it is a result of residual confounding because subject selection is rigorous in this age group due to vulnerability,” the authors wrote.

They concluded that knee replacement “did not appear to be associated with an increased risk of all-cause mortality.”

“While we cannot rule out that KR may potentially reduce the risk of mortality over the long term, the true extent of that potential benefit is difficult to discern due to confounding by indication in observational studies using administrative data or electronic health records,” they added.

This study was funded by the Arthritis Foundation Postdoctoral Fellowship Award, the ACR Rheumatology Research Foundation Investigator Award, and a Boston University scholarship grant.

People with osteoarthritis who go on to have a total or partial knee replacement do not appear to have an increased risk of all-cause mortality, but the jury is still out on whether they gain any improvement, a study showed.

In their research published in the Annals of the Rheumatic Diseases [2016 May 17. doi: 10.1136/annrheumdis-2016-209167], Dr. Devyani Misra of Boston University and colleagues noted that knee replacement (KR) was thought to decrease long-term mortality risk because of the relief from pain and improvement in function that typically comes with surgery. However, studies on the topic had been conflicting, largely because of the challenges associated with studying mortality with KR surgery in observational settings.

©Nandyphotos/Thinkstock

In the current study the research team sought to evaluate the relation of KR to the risk of all-cause mortality among subjects with knee OA, while at the same time giving particular attention to “potential sources of confounding bias that may account for [the] effect of KR on mortality.”

Using patient data from the U.K. primary care electronic database THIN, the investigators compared the risk of mortality among 14,042 subjects who had OA, were aged 50-89 years old, and had had or had not had KR.

They discovered a strong protective effect of KR on all-cause long-term mortality risk, particularly among the adults over 63 years of age.

For example, people who had undergone KR had a 28% lower risk of mortality than did non-KR subjects (hazard ratio, 0.72; 95% confidence interval, 0.66-0.78).

In the overall propensity score–matched study sample, crude mortality per 1,000 person-years (total person-years) for the KR and non-KR cohorts were 19 (61,015) and 25 (58,294), respectively.

However, despite their best efforts, the researchers said the results showed evidence of residual confounding.

“For example, the observation of improved survival immediately after KR, despite the expectation of potential short-term increased postoperative mortality risk supports the presence of residual confounding,” they wrote.

Another finding suggestive of confounding was that the protective effect was seen only in older patients (over 63) when the authors stratified study participants by age.

“While it is possible that survival benefit seen in older patients with KR is a true effect because it is in this group that greater physical activity is particularly important to survival, more likely it is a result of residual confounding because subject selection is rigorous in this age group due to vulnerability,” the authors wrote.

They concluded that knee replacement “did not appear to be associated with an increased risk of all-cause mortality.”

“While we cannot rule out that KR may potentially reduce the risk of mortality over the long term, the true extent of that potential benefit is difficult to discern due to confounding by indication in observational studies using administrative data or electronic health records,” they added.

This study was funded by the Arthritis Foundation Postdoctoral Fellowship Award, the ACR Rheumatology Research Foundation Investigator Award, and a Boston University scholarship grant.

People with osteoarthritis who go on to have a total or partial knee replacement do not appear to have an increased risk of all-cause mortality, but the jury is still out on whether they gain any improvement, a study showed.

In their research published in the Annals of the Rheumatic Diseases [2016 May 17. doi: 10.1136/annrheumdis-2016-209167], Dr. Devyani Misra of Boston University and colleagues noted that knee replacement (KR) was thought to decrease long-term mortality risk because of the relief from pain and improvement in function that typically comes with surgery. However, studies on the topic had been conflicting, largely because of the challenges associated with studying mortality with KR surgery in observational settings.

©Nandyphotos/Thinkstock

In the current study the research team sought to evaluate the relation of KR to the risk of all-cause mortality among subjects with knee OA, while at the same time giving particular attention to “potential sources of confounding bias that may account for [the] effect of KR on mortality.”

Using patient data from the U.K. primary care electronic database THIN, the investigators compared the risk of mortality among 14,042 subjects who had OA, were aged 50-89 years old, and had had or had not had KR.

They discovered a strong protective effect of KR on all-cause long-term mortality risk, particularly among the adults over 63 years of age.

For example, people who had undergone KR had a 28% lower risk of mortality than did non-KR subjects (hazard ratio, 0.72; 95% confidence interval, 0.66-0.78).

In the overall propensity score–matched study sample, crude mortality per 1,000 person-years (total person-years) for the KR and non-KR cohorts were 19 (61,015) and 25 (58,294), respectively.

However, despite their best efforts, the researchers said the results showed evidence of residual confounding.

“For example, the observation of improved survival immediately after KR, despite the expectation of potential short-term increased postoperative mortality risk supports the presence of residual confounding,” they wrote.

Another finding suggestive of confounding was that the protective effect was seen only in older patients (over 63) when the authors stratified study participants by age.

“While it is possible that survival benefit seen in older patients with KR is a true effect because it is in this group that greater physical activity is particularly important to survival, more likely it is a result of residual confounding because subject selection is rigorous in this age group due to vulnerability,” the authors wrote.

They concluded that knee replacement “did not appear to be associated with an increased risk of all-cause mortality.”

“While we cannot rule out that KR may potentially reduce the risk of mortality over the long term, the true extent of that potential benefit is difficult to discern due to confounding by indication in observational studies using administrative data or electronic health records,” they added.

This study was funded by the Arthritis Foundation Postdoctoral Fellowship Award, the ACR Rheumatology Research Foundation Investigator Award, and a Boston University scholarship grant.

Reverse total shoulder arthroplasty (RTSA) has become a reliable treatment option for many pathologic conditions of the shoulder, including rotator cuff arthropathy, proximal humerus fractures, and others.^1-4 While the treatment outcomes have generally been reported as good, some concern exists over the postoperative range of motion (ROM) in patients following RTSA, including external rotation.^5-7 The original RTSA design was introduced by Neer in the 1970s and has undergone many modifications since that time.^1,2 The original Grammont-style prosthesis involved medialization of the glenoid, inferiorizing the center of rotation (with increased deltoid tensioning), and a neck-shaft angle of 155°.^1,8 While clinical results of the 155° design were encouraging, concerns arose over the significance of the common finding of scapular notching, or contact between the scapular neck and inferior portion of the humeral polyethylene when the arm is adducted.^9,10

To address this concern, a prosthesis design with a 135° neck-shaft angle was introduced.¹¹ This new design did significantly decrease the rate of scapular notching, and although some reported a concern over implant stability with the 135° prosthesis, recent data has shown no difference in dislocation rates between the 135° and 155° prostheses.³ A different variable that has not been evaluated between these prostheses is the active ROM that is achieved postoperatively, and the change in ROM from pre- to post-RTSA.^12,13 As active ROM plays a significant role in shoulder function and patient satisfaction, the question of whether a significant difference exists in postoperative ROM between the 135° and 155° prostheses must be addressed.

The purpose of this study was to perform a systematic review investigating active ROM following RTSA to determine if active postoperative ROM following RTSA differs between the 135° and 155° humeral inclination prostheses, and to determine if there is a significant difference between the change in preoperative and postoperative ROM between the 135° and 155° prostheses. The authors hypothesize that there will be no significant difference in active postoperative ROM between the 135° and 155° prostheses, and that the difference between preoperative and postoperative ROM (that is, the amount of motion gained by the surgery) will not significantly differ between the 135° and 155° prostheses.

Methods

A systematic review was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines using a PRISMA checklist.¹⁵ Systematic review registration was performed using the PROSPERO international prospective register of systematic reviews (registration date 3/9/15, registration number CRD42015017367).¹⁶ Two reviewers independently conducted the search on March 7, 2015 using the following databases: Medline, Cochrane Central Register of Controlled Trials, SportDiscus, and CINAHL. The electronic search citation algorithm utilized was: (((((reverse[Title/Abstract]) AND shoulder[Title/Abstract]) AND arthroplasty[Title/Abstract]) NOT arthroscopic[Title/Abstract]) NOT cadaver[Title/Abstract]) NOT biomechanical[Title/Abstract]. English language Level I-IV evidence (2011 update by the Oxford Centre for Evidence-Based Medicine¹⁷) clinical studies that reported the type of RTSA prosthesis that was used as well as postoperative ROM with at least 12 months follow-up were eligible. All references within included studies were cross-referenced for inclusion if missed by the initial search. If duplicate subject publications were discovered, the study with the longer duration of follow-up or larger number of patients was included. Level V evidence reviews, letters to the editor, basic science, biomechanical studies, arthroscopic shoulder surgery, imaging, surgical technique, and classification studies were excluded. Studies were excluded if both a 135° and 155° prosthesis were utilized and the outcomes were not stratified by the humeral inclination. Studies that did not report ROM were excluded.

A total of 456 studies were located, and, after implementation of the exclusion criteria, 65 studies from 2005-2015 were included in the final analysis (Figure). Subjects of interest in this systematic review underwent a RTSA. Studies were not excluded based on the surgical indications (rotator cuff tear arthropathy, proximal humerus fractures, osteoarthritis) and there was no minimum follow-up or rehabilitation requirement. Study and subject demographic parameters analyzed included year of publication, journal of publication, country and continent of publication, years of subject enrollment, presence of study financial conflict of interest, number of subjects and shoulders, gender, age, the manufacturer and type of prosthesis used, and the degree of the humeral inclination (135° vs 155° humeral cup). Preoperative ROM, including forward elevation, abduction, external rotation with the arm adducted, and external rotation with the arm at 90° of abduction, were recorded. The same ROM measurements were recorded for the final follow-up visit that was reported. Internal rotation was recorded, but because of the variability with how this measurement was reported, it was not analyzed. Clinical outcome scores and complications were not assessed. Study methodological quality was evaluated using the Modified Coleman Methodology Score (MCMS).¹⁸

Statistical Analysis

Descriptive statistics were calculated, including mean ± standard deviation for quantitative continuous data and frequencies with percentages for qualitative categorical data. ROM comparisons between 135° and 155° components (pre- vs postoperative for each and postoperative between the 2) were made using 2 proportion z-test calculator (http://in-silico.net/tools/statistics/ztest) using alpha .05 because of the difference in sample sizes between compared groups.

Results

Sixty-five studies with 3302 patients (3434 shoulders) were included in this study. There was a total of 1211 shoulders in the 135° lateralized glenosphere group and 2223 shoulders in the 155° group. The studies had an average MCMS of 40.4 ± 8.2 (poor), 48% of studies reported a conflict of interest, 32% had no conflict of interest, and 20% did not report whether a conflict of interest existed or not. The majority of studies included were level IV evidence (85%). Mean patient age was 71.1 ± 7.6 years; 29% of patients were male and 71% were female. No significant difference existed between patient age at the time of surgery; the average age of patients in the 135° lateralized glenosphere group was 71.67 ± 3.8 years, while the average patient age of patients in the 155° group was 70.97 ± 8.8 years. Mean follow-up for all patients included in this study was 37.2 ± 16.5 months. Of the 65 studies included, 3 were published from Asia, 4 were published from Australia, 24 were from North America, and 34 were from Europe. Of the individual countries whose studies were included, the United States had 23 included studies, France had 13 included studies, and Italy had 4 included studies. All other countries had <4 studies included.

Patients who received either a 135° or a 155° prosthesis showed significant improvements in external rotation with the arm at the side (P < .05), forward elevation (P < .05), and abduction (P < .05) following surgery (Table). When comparing the 135° and 155° groups, patients who received a 135° prosthesis showed significantly greater improvements in external rotation with the arm at the side (P < .001) and had significantly more overall external rotation postoperatively (P < .001) than patients who received a 155° prosthesis. The only preoperative ROM difference between groups was the 155° group started with significantly more forward elevation than the 135°group prior to surgery (P = .002).

Discussion

RTSA is indicated in patients with rotator cuff tear arthropathy, pseudoparalysis, and a functional deltoid.^1,2,4 The purpose of this systematic review was to determine if active ROM following RTSA differs between the 135° and 155° humeral inclination prostheses, and to determine if there is a significant difference between the change in preoperative and postoperative ROM between the 135° and 155° prostheses. Forward elevation, abduction, and external rotation all significantly improved following surgery in both groups, with no significant difference between groups in motion or amount of motion improvement, mostly confirming the study hypotheses. However, patients in the 135° group had significantly greater postoperative external rotation and greater amount of external rotation improvement compared to the 155° group.

Two of the frequently debated issues regarding implant geometry is stability and scapular notching between the 135° and 155° humeral inclination designs. Erickson and colleagues3 recently evaluated the rate of scapular notching and dislocations between the 135° and 155° RTSA prostheses. The authors found that the 135° prosthesis had a significantly lower incidence of scapular notching vs the 155° group and that the rate of dislocations was not significantly different between groups.³ In the latter systematic review, the authors attempted to evaluate ROM between the 135° and 155° prostheses, but as the inclusion criteria of the study was reporting on scapular notching and dislocation rates, many studies reporting solely on ROM were excluded, and the influence of humeral inclination on ROM was inconclusive.³ Furthermore, there have been no studies that have directly compared ROM following RTSA between the 135° and 155° prostheses. While studies evaluating each prosthesis on an individual level have shown an improvement in ROM from pre- to postsurgery, there have been no large studies that have compared the postoperative ROM and change in pre- to postoperative ROM between the 135° and 155° prostheses.^11,13,19,20

One study by Valenti and colleagues²¹ evaluated a group of 30 patients with an average age of 69.5 years who underwent RTSA using either a 135° or a 155° prosthesis. Although the study did not directly compare the 2 types of prostheses, it did report the separate outcomes for each prosthesis. At an average follow-up of 36.4 months, the authors found that patients who had the 135° prosthesis implanted had a mean increase in forward elevation and external rotation of 53° and 9°, while patients who had the 155° showed an increase of 56° in forward elevation and a loss of 1° of external rotation. Both prostheses showed a significant increase in forward elevation, but neither had a significant increase in external rotation. Furthermore, scapular notching was seen in 4 patients in the 155° group, while no patients in the 135° group had evidence of notching.

The results of the current study were similar in that both the 135° and 155° prosthesis showed improvements in forward elevation following surgery, and the 135° group showed a significantly greater gain in external rotation than the 155° group. A significant component of shoulder function and patient satisfaction following RTSA is active ROM. However, this variable has not explicitly been evaluated in the literature until now. The clinical significance of this finding is unclear. Patients with adequate external rotation prior to surgery likely would not see a functional difference between prostheses, while those patients who were borderline on a functional amount of external rotation would see a clinically significant benefit with the 135° prosthesis. Studies have shown that the 135° prosthesis is more anatomic than the 155°, and this could explain the difference seen in ROM outcomes between the 2 prostheses.¹⁹ Ladermann and colleagues²² recently created and evaluated a 3-dimensional computer model to evaluate possible differences between the 135° and 155° prosthesis. The authors found a significant increase in external rotation of the 135° compared to the 155°, likely related to a difference in acromiohumeral distance as well as inlay vs onlay humeral trays between the 2 prostheses. The results of this study parallel the computer model, thereby validating these experimental results.

It is important to understand what the minimum functional ROM of the shoulder is (in other words, the ROM necessary to complete activities of daily living (ADLs).²³ Namdari and colleagues²⁴ used motion analysis software to evaluate the shoulder ROM necessary to complete 10 different ADLs, including combing hair, washing the back of the opposite shoulder, and reaching a shelf above their head without bending their elbow in 20 patients with a mean age of 29.2 years. They found that patients required 121° ± 6.7° of flexion, 46° ± 5.3° of extension, 128° ± 7.9° of abduction, 116° ± 9.1° of cross-body adduction, 59° ± 10° of external rotation with the arm 90° abducted, and 102° ± 7.7° of internal rotation with the arm at the side (external rotation with the arm at the side was not well defined).²⁴ Hence, while abduction and forward elevation seem comparable, the results from the current study do raise concerns about the amount of external rotation obtained following RTSA as it relates to a patients’ ability to perform ADLs, specifically in the 155° prosthesis, as the average postoperative external rotation in this group was 20.5°. Therefore, based on the results of this study, it appears that, while both the 135° and 155° RTSA prostheses provide similar gain in forward elevation and abduction ROM as well as overall forward elevation and abduction, the 135° prosthesis provides significantly more external rotation with the arm at the side than the 155° prosthesis.

Limitations

Although this study attempted to look at all studies that reported active ROM in patients following a RTSA, and 2 authors performed the search, there is a possibility that some studies were missed, introducing study selection bias. Furthermore, the mean follow-up was over 3 years following surgery, but the minimum follow-up requirement for studies to be included was only 12 months. Hence, this transfer bias introduces the possibility that the patient’s ROM would have changed had they been followed for a standard period of time. There are many variables that come into play in evaluating ROM, and although the study attempted to control for these, there are some that could not be controlled for due to lack of reporting by some studies. Glenosphere size and humeral retroversion were not recorded, as they were not reliably reported in all studies, so motion outcomes based on these variables was not evaluated. Complications and clinical outcomes were not assessed in this review and as such, conclusions regarding these variables cannot be drawn from this study. Finally, indications for surgery were not reliably reported in the studies included in this paper, so differences may have existed between surgical indications of the 135° and 155° groups that could have affected outcomes.

Conclusion

Patients who receive a 135° RTSA gain significantly more external rotation from pre- to postsurgery and have an overall greater amount of external rotation than patients who receive a 155° prosthesis. Both groups show improvements in forward elevation, external rotation, and abduction following surgery.

Reverse total shoulder arthroplasty (RTSA) has become a reliable treatment option for many pathologic conditions of the shoulder, including rotator cuff arthropathy, proximal humerus fractures, and others.^1-4 While the treatment outcomes have generally been reported as good, some concern exists over the postoperative range of motion (ROM) in patients following RTSA, including external rotation.^5-7 The original RTSA design was introduced by Neer in the 1970s and has undergone many modifications since that time.^1,2 The original Grammont-style prosthesis involved medialization of the glenoid, inferiorizing the center of rotation (with increased deltoid tensioning), and a neck-shaft angle of 155°.^1,8 While clinical results of the 155° design were encouraging, concerns arose over the significance of the common finding of scapular notching, or contact between the scapular neck and inferior portion of the humeral polyethylene when the arm is adducted.^9,10

To address this concern, a prosthesis design with a 135° neck-shaft angle was introduced.¹¹ This new design did significantly decrease the rate of scapular notching, and although some reported a concern over implant stability with the 135° prosthesis, recent data has shown no difference in dislocation rates between the 135° and 155° prostheses.³ A different variable that has not been evaluated between these prostheses is the active ROM that is achieved postoperatively, and the change in ROM from pre- to post-RTSA.^12,13 As active ROM plays a significant role in shoulder function and patient satisfaction, the question of whether a significant difference exists in postoperative ROM between the 135° and 155° prostheses must be addressed.

The purpose of this study was to perform a systematic review investigating active ROM following RTSA to determine if active postoperative ROM following RTSA differs between the 135° and 155° humeral inclination prostheses, and to determine if there is a significant difference between the change in preoperative and postoperative ROM between the 135° and 155° prostheses. The authors hypothesize that there will be no significant difference in active postoperative ROM between the 135° and 155° prostheses, and that the difference between preoperative and postoperative ROM (that is, the amount of motion gained by the surgery) will not significantly differ between the 135° and 155° prostheses.

Methods

A systematic review was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines using a PRISMA checklist.¹⁵ Systematic review registration was performed using the PROSPERO international prospective register of systematic reviews (registration date 3/9/15, registration number CRD42015017367).¹⁶ Two reviewers independently conducted the search on March 7, 2015 using the following databases: Medline, Cochrane Central Register of Controlled Trials, SportDiscus, and CINAHL. The electronic search citation algorithm utilized was: (((((reverse[Title/Abstract]) AND shoulder[Title/Abstract]) AND arthroplasty[Title/Abstract]) NOT arthroscopic[Title/Abstract]) NOT cadaver[Title/Abstract]) NOT biomechanical[Title/Abstract]. English language Level I-IV evidence (2011 update by the Oxford Centre for Evidence-Based Medicine¹⁷) clinical studies that reported the type of RTSA prosthesis that was used as well as postoperative ROM with at least 12 months follow-up were eligible. All references within included studies were cross-referenced for inclusion if missed by the initial search. If duplicate subject publications were discovered, the study with the longer duration of follow-up or larger number of patients was included. Level V evidence reviews, letters to the editor, basic science, biomechanical studies, arthroscopic shoulder surgery, imaging, surgical technique, and classification studies were excluded. Studies were excluded if both a 135° and 155° prosthesis were utilized and the outcomes were not stratified by the humeral inclination. Studies that did not report ROM were excluded.

A total of 456 studies were located, and, after implementation of the exclusion criteria, 65 studies from 2005-2015 were included in the final analysis (Figure). Subjects of interest in this systematic review underwent a RTSA. Studies were not excluded based on the surgical indications (rotator cuff tear arthropathy, proximal humerus fractures, osteoarthritis) and there was no minimum follow-up or rehabilitation requirement. Study and subject demographic parameters analyzed included year of publication, journal of publication, country and continent of publication, years of subject enrollment, presence of study financial conflict of interest, number of subjects and shoulders, gender, age, the manufacturer and type of prosthesis used, and the degree of the humeral inclination (135° vs 155° humeral cup). Preoperative ROM, including forward elevation, abduction, external rotation with the arm adducted, and external rotation with the arm at 90° of abduction, were recorded. The same ROM measurements were recorded for the final follow-up visit that was reported. Internal rotation was recorded, but because of the variability with how this measurement was reported, it was not analyzed. Clinical outcome scores and complications were not assessed. Study methodological quality was evaluated using the Modified Coleman Methodology Score (MCMS).¹⁸

Statistical Analysis

Descriptive statistics were calculated, including mean ± standard deviation for quantitative continuous data and frequencies with percentages for qualitative categorical data. ROM comparisons between 135° and 155° components (pre- vs postoperative for each and postoperative between the 2) were made using 2 proportion z-test calculator (http://in-silico.net/tools/statistics/ztest) using alpha .05 because of the difference in sample sizes between compared groups.

Results

Sixty-five studies with 3302 patients (3434 shoulders) were included in this study. There was a total of 1211 shoulders in the 135° lateralized glenosphere group and 2223 shoulders in the 155° group. The studies had an average MCMS of 40.4 ± 8.2 (poor), 48% of studies reported a conflict of interest, 32% had no conflict of interest, and 20% did not report whether a conflict of interest existed or not. The majority of studies included were level IV evidence (85%). Mean patient age was 71.1 ± 7.6 years; 29% of patients were male and 71% were female. No significant difference existed between patient age at the time of surgery; the average age of patients in the 135° lateralized glenosphere group was 71.67 ± 3.8 years, while the average patient age of patients in the 155° group was 70.97 ± 8.8 years. Mean follow-up for all patients included in this study was 37.2 ± 16.5 months. Of the 65 studies included, 3 were published from Asia, 4 were published from Australia, 24 were from North America, and 34 were from Europe. Of the individual countries whose studies were included, the United States had 23 included studies, France had 13 included studies, and Italy had 4 included studies. All other countries had <4 studies included.

Patients who received either a 135° or a 155° prosthesis showed significant improvements in external rotation with the arm at the side (P < .05), forward elevation (P < .05), and abduction (P < .05) following surgery (Table). When comparing the 135° and 155° groups, patients who received a 135° prosthesis showed significantly greater improvements in external rotation with the arm at the side (P < .001) and had significantly more overall external rotation postoperatively (P < .001) than patients who received a 155° prosthesis. The only preoperative ROM difference between groups was the 155° group started with significantly more forward elevation than the 135°group prior to surgery (P = .002).

Discussion

RTSA is indicated in patients with rotator cuff tear arthropathy, pseudoparalysis, and a functional deltoid.^1,2,4 The purpose of this systematic review was to determine if active ROM following RTSA differs between the 135° and 155° humeral inclination prostheses, and to determine if there is a significant difference between the change in preoperative and postoperative ROM between the 135° and 155° prostheses. Forward elevation, abduction, and external rotation all significantly improved following surgery in both groups, with no significant difference between groups in motion or amount of motion improvement, mostly confirming the study hypotheses. However, patients in the 135° group had significantly greater postoperative external rotation and greater amount of external rotation improvement compared to the 155° group.

Two of the frequently debated issues regarding implant geometry is stability and scapular notching between the 135° and 155° humeral inclination designs. Erickson and colleagues3 recently evaluated the rate of scapular notching and dislocations between the 135° and 155° RTSA prostheses. The authors found that the 135° prosthesis had a significantly lower incidence of scapular notching vs the 155° group and that the rate of dislocations was not significantly different between groups.³ In the latter systematic review, the authors attempted to evaluate ROM between the 135° and 155° prostheses, but as the inclusion criteria of the study was reporting on scapular notching and dislocation rates, many studies reporting solely on ROM were excluded, and the influence of humeral inclination on ROM was inconclusive.³ Furthermore, there have been no studies that have directly compared ROM following RTSA between the 135° and 155° prostheses. While studies evaluating each prosthesis on an individual level have shown an improvement in ROM from pre- to postsurgery, there have been no large studies that have compared the postoperative ROM and change in pre- to postoperative ROM between the 135° and 155° prostheses.^11,13,19,20

One study by Valenti and colleagues²¹ evaluated a group of 30 patients with an average age of 69.5 years who underwent RTSA using either a 135° or a 155° prosthesis. Although the study did not directly compare the 2 types of prostheses, it did report the separate outcomes for each prosthesis. At an average follow-up of 36.4 months, the authors found that patients who had the 135° prosthesis implanted had a mean increase in forward elevation and external rotation of 53° and 9°, while patients who had the 155° showed an increase of 56° in forward elevation and a loss of 1° of external rotation. Both prostheses showed a significant increase in forward elevation, but neither had a significant increase in external rotation. Furthermore, scapular notching was seen in 4 patients in the 155° group, while no patients in the 135° group had evidence of notching.

The results of the current study were similar in that both the 135° and 155° prosthesis showed improvements in forward elevation following surgery, and the 135° group showed a significantly greater gain in external rotation than the 155° group. A significant component of shoulder function and patient satisfaction following RTSA is active ROM. However, this variable has not explicitly been evaluated in the literature until now. The clinical significance of this finding is unclear. Patients with adequate external rotation prior to surgery likely would not see a functional difference between prostheses, while those patients who were borderline on a functional amount of external rotation would see a clinically significant benefit with the 135° prosthesis. Studies have shown that the 135° prosthesis is more anatomic than the 155°, and this could explain the difference seen in ROM outcomes between the 2 prostheses.¹⁹ Ladermann and colleagues²² recently created and evaluated a 3-dimensional computer model to evaluate possible differences between the 135° and 155° prosthesis. The authors found a significant increase in external rotation of the 135° compared to the 155°, likely related to a difference in acromiohumeral distance as well as inlay vs onlay humeral trays between the 2 prostheses. The results of this study parallel the computer model, thereby validating these experimental results.

It is important to understand what the minimum functional ROM of the shoulder is (in other words, the ROM necessary to complete activities of daily living (ADLs).²³ Namdari and colleagues²⁴ used motion analysis software to evaluate the shoulder ROM necessary to complete 10 different ADLs, including combing hair, washing the back of the opposite shoulder, and reaching a shelf above their head without bending their elbow in 20 patients with a mean age of 29.2 years. They found that patients required 121° ± 6.7° of flexion, 46° ± 5.3° of extension, 128° ± 7.9° of abduction, 116° ± 9.1° of cross-body adduction, 59° ± 10° of external rotation with the arm 90° abducted, and 102° ± 7.7° of internal rotation with the arm at the side (external rotation with the arm at the side was not well defined).²⁴ Hence, while abduction and forward elevation seem comparable, the results from the current study do raise concerns about the amount of external rotation obtained following RTSA as it relates to a patients’ ability to perform ADLs, specifically in the 155° prosthesis, as the average postoperative external rotation in this group was 20.5°. Therefore, based on the results of this study, it appears that, while both the 135° and 155° RTSA prostheses provide similar gain in forward elevation and abduction ROM as well as overall forward elevation and abduction, the 135° prosthesis provides significantly more external rotation with the arm at the side than the 155° prosthesis.

Limitations

Although this study attempted to look at all studies that reported active ROM in patients following a RTSA, and 2 authors performed the search, there is a possibility that some studies were missed, introducing study selection bias. Furthermore, the mean follow-up was over 3 years following surgery, but the minimum follow-up requirement for studies to be included was only 12 months. Hence, this transfer bias introduces the possibility that the patient’s ROM would have changed had they been followed for a standard period of time. There are many variables that come into play in evaluating ROM, and although the study attempted to control for these, there are some that could not be controlled for due to lack of reporting by some studies. Glenosphere size and humeral retroversion were not recorded, as they were not reliably reported in all studies, so motion outcomes based on these variables was not evaluated. Complications and clinical outcomes were not assessed in this review and as such, conclusions regarding these variables cannot be drawn from this study. Finally, indications for surgery were not reliably reported in the studies included in this paper, so differences may have existed between surgical indications of the 135° and 155° groups that could have affected outcomes.

Conclusion

Patients who receive a 135° RTSA gain significantly more external rotation from pre- to postsurgery and have an overall greater amount of external rotation than patients who receive a 155° prosthesis. Both groups show improvements in forward elevation, external rotation, and abduction following surgery.

Reverse total shoulder arthroplasty (RTSA) has become a reliable treatment option for many pathologic conditions of the shoulder, including rotator cuff arthropathy, proximal humerus fractures, and others.^1-4 While the treatment outcomes have generally been reported as good, some concern exists over the postoperative range of motion (ROM) in patients following RTSA, including external rotation.^5-7 The original RTSA design was introduced by Neer in the 1970s and has undergone many modifications since that time.^1,2 The original Grammont-style prosthesis involved medialization of the glenoid, inferiorizing the center of rotation (with increased deltoid tensioning), and a neck-shaft angle of 155°.^1,8 While clinical results of the 155° design were encouraging, concerns arose over the significance of the common finding of scapular notching, or contact between the scapular neck and inferior portion of the humeral polyethylene when the arm is adducted.^9,10

To address this concern, a prosthesis design with a 135° neck-shaft angle was introduced.¹¹ This new design did significantly decrease the rate of scapular notching, and although some reported a concern over implant stability with the 135° prosthesis, recent data has shown no difference in dislocation rates between the 135° and 155° prostheses.³ A different variable that has not been evaluated between these prostheses is the active ROM that is achieved postoperatively, and the change in ROM from pre- to post-RTSA.^12,13 As active ROM plays a significant role in shoulder function and patient satisfaction, the question of whether a significant difference exists in postoperative ROM between the 135° and 155° prostheses must be addressed.

The purpose of this study was to perform a systematic review investigating active ROM following RTSA to determine if active postoperative ROM following RTSA differs between the 135° and 155° humeral inclination prostheses, and to determine if there is a significant difference between the change in preoperative and postoperative ROM between the 135° and 155° prostheses. The authors hypothesize that there will be no significant difference in active postoperative ROM between the 135° and 155° prostheses, and that the difference between preoperative and postoperative ROM (that is, the amount of motion gained by the surgery) will not significantly differ between the 135° and 155° prostheses.

Methods

A systematic review was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines using a PRISMA checklist.¹⁵ Systematic review registration was performed using the PROSPERO international prospective register of systematic reviews (registration date 3/9/15, registration number CRD42015017367).¹⁶ Two reviewers independently conducted the search on March 7, 2015 using the following databases: Medline, Cochrane Central Register of Controlled Trials, SportDiscus, and CINAHL. The electronic search citation algorithm utilized was: (((((reverse[Title/Abstract]) AND shoulder[Title/Abstract]) AND arthroplasty[Title/Abstract]) NOT arthroscopic[Title/Abstract]) NOT cadaver[Title/Abstract]) NOT biomechanical[Title/Abstract]. English language Level I-IV evidence (2011 update by the Oxford Centre for Evidence-Based Medicine¹⁷) clinical studies that reported the type of RTSA prosthesis that was used as well as postoperative ROM with at least 12 months follow-up were eligible. All references within included studies were cross-referenced for inclusion if missed by the initial search. If duplicate subject publications were discovered, the study with the longer duration of follow-up or larger number of patients was included. Level V evidence reviews, letters to the editor, basic science, biomechanical studies, arthroscopic shoulder surgery, imaging, surgical technique, and classification studies were excluded. Studies were excluded if both a 135° and 155° prosthesis were utilized and the outcomes were not stratified by the humeral inclination. Studies that did not report ROM were excluded.

A total of 456 studies were located, and, after implementation of the exclusion criteria, 65 studies from 2005-2015 were included in the final analysis (Figure). Subjects of interest in this systematic review underwent a RTSA. Studies were not excluded based on the surgical indications (rotator cuff tear arthropathy, proximal humerus fractures, osteoarthritis) and there was no minimum follow-up or rehabilitation requirement. Study and subject demographic parameters analyzed included year of publication, journal of publication, country and continent of publication, years of subject enrollment, presence of study financial conflict of interest, number of subjects and shoulders, gender, age, the manufacturer and type of prosthesis used, and the degree of the humeral inclination (135° vs 155° humeral cup). Preoperative ROM, including forward elevation, abduction, external rotation with the arm adducted, and external rotation with the arm at 90° of abduction, were recorded. The same ROM measurements were recorded for the final follow-up visit that was reported. Internal rotation was recorded, but because of the variability with how this measurement was reported, it was not analyzed. Clinical outcome scores and complications were not assessed. Study methodological quality was evaluated using the Modified Coleman Methodology Score (MCMS).¹⁸

Statistical Analysis

Descriptive statistics were calculated, including mean ± standard deviation for quantitative continuous data and frequencies with percentages for qualitative categorical data. ROM comparisons between 135° and 155° components (pre- vs postoperative for each and postoperative between the 2) were made using 2 proportion z-test calculator (http://in-silico.net/tools/statistics/ztest) using alpha .05 because of the difference in sample sizes between compared groups.

Results

Sixty-five studies with 3302 patients (3434 shoulders) were included in this study. There was a total of 1211 shoulders in the 135° lateralized glenosphere group and 2223 shoulders in the 155° group. The studies had an average MCMS of 40.4 ± 8.2 (poor), 48% of studies reported a conflict of interest, 32% had no conflict of interest, and 20% did not report whether a conflict of interest existed or not. The majority of studies included were level IV evidence (85%). Mean patient age was 71.1 ± 7.6 years; 29% of patients were male and 71% were female. No significant difference existed between patient age at the time of surgery; the average age of patients in the 135° lateralized glenosphere group was 71.67 ± 3.8 years, while the average patient age of patients in the 155° group was 70.97 ± 8.8 years. Mean follow-up for all patients included in this study was 37.2 ± 16.5 months. Of the 65 studies included, 3 were published from Asia, 4 were published from Australia, 24 were from North America, and 34 were from Europe. Of the individual countries whose studies were included, the United States had 23 included studies, France had 13 included studies, and Italy had 4 included studies. All other countries had <4 studies included.

Patients who received either a 135° or a 155° prosthesis showed significant improvements in external rotation with the arm at the side (P < .05), forward elevation (P < .05), and abduction (P < .05) following surgery (Table). When comparing the 135° and 155° groups, patients who received a 135° prosthesis showed significantly greater improvements in external rotation with the arm at the side (P < .001) and had significantly more overall external rotation postoperatively (P < .001) than patients who received a 155° prosthesis. The only preoperative ROM difference between groups was the 155° group started with significantly more forward elevation than the 135°group prior to surgery (P = .002).

Discussion

RTSA is indicated in patients with rotator cuff tear arthropathy, pseudoparalysis, and a functional deltoid.^1,2,4 The purpose of this systematic review was to determine if active ROM following RTSA differs between the 135° and 155° humeral inclination prostheses, and to determine if there is a significant difference between the change in preoperative and postoperative ROM between the 135° and 155° prostheses. Forward elevation, abduction, and external rotation all significantly improved following surgery in both groups, with no significant difference between groups in motion or amount of motion improvement, mostly confirming the study hypotheses. However, patients in the 135° group had significantly greater postoperative external rotation and greater amount of external rotation improvement compared to the 155° group.

Two of the frequently debated issues regarding implant geometry is stability and scapular notching between the 135° and 155° humeral inclination designs. Erickson and colleagues3 recently evaluated the rate of scapular notching and dislocations between the 135° and 155° RTSA prostheses. The authors found that the 135° prosthesis had a significantly lower incidence of scapular notching vs the 155° group and that the rate of dislocations was not significantly different between groups.³ In the latter systematic review, the authors attempted to evaluate ROM between the 135° and 155° prostheses, but as the inclusion criteria of the study was reporting on scapular notching and dislocation rates, many studies reporting solely on ROM were excluded, and the influence of humeral inclination on ROM was inconclusive.³ Furthermore, there have been no studies that have directly compared ROM following RTSA between the 135° and 155° prostheses. While studies evaluating each prosthesis on an individual level have shown an improvement in ROM from pre- to postsurgery, there have been no large studies that have compared the postoperative ROM and change in pre- to postoperative ROM between the 135° and 155° prostheses.^11,13,19,20

One study by Valenti and colleagues²¹ evaluated a group of 30 patients with an average age of 69.5 years who underwent RTSA using either a 135° or a 155° prosthesis. Although the study did not directly compare the 2 types of prostheses, it did report the separate outcomes for each prosthesis. At an average follow-up of 36.4 months, the authors found that patients who had the 135° prosthesis implanted had a mean increase in forward elevation and external rotation of 53° and 9°, while patients who had the 155° showed an increase of 56° in forward elevation and a loss of 1° of external rotation. Both prostheses showed a significant increase in forward elevation, but neither had a significant increase in external rotation. Furthermore, scapular notching was seen in 4 patients in the 155° group, while no patients in the 135° group had evidence of notching.

The results of the current study were similar in that both the 135° and 155° prosthesis showed improvements in forward elevation following surgery, and the 135° group showed a significantly greater gain in external rotation than the 155° group. A significant component of shoulder function and patient satisfaction following RTSA is active ROM. However, this variable has not explicitly been evaluated in the literature until now. The clinical significance of this finding is unclear. Patients with adequate external rotation prior to surgery likely would not see a functional difference between prostheses, while those patients who were borderline on a functional amount of external rotation would see a clinically significant benefit with the 135° prosthesis. Studies have shown that the 135° prosthesis is more anatomic than the 155°, and this could explain the difference seen in ROM outcomes between the 2 prostheses.¹⁹ Ladermann and colleagues²² recently created and evaluated a 3-dimensional computer model to evaluate possible differences between the 135° and 155° prosthesis. The authors found a significant increase in external rotation of the 135° compared to the 155°, likely related to a difference in acromiohumeral distance as well as inlay vs onlay humeral trays between the 2 prostheses. The results of this study parallel the computer model, thereby validating these experimental results.

It is important to understand what the minimum functional ROM of the shoulder is (in other words, the ROM necessary to complete activities of daily living (ADLs).²³ Namdari and colleagues²⁴ used motion analysis software to evaluate the shoulder ROM necessary to complete 10 different ADLs, including combing hair, washing the back of the opposite shoulder, and reaching a shelf above their head without bending their elbow in 20 patients with a mean age of 29.2 years. They found that patients required 121° ± 6.7° of flexion, 46° ± 5.3° of extension, 128° ± 7.9° of abduction, 116° ± 9.1° of cross-body adduction, 59° ± 10° of external rotation with the arm 90° abducted, and 102° ± 7.7° of internal rotation with the arm at the side (external rotation with the arm at the side was not well defined).²⁴ Hence, while abduction and forward elevation seem comparable, the results from the current study do raise concerns about the amount of external rotation obtained following RTSA as it relates to a patients’ ability to perform ADLs, specifically in the 155° prosthesis, as the average postoperative external rotation in this group was 20.5°. Therefore, based on the results of this study, it appears that, while both the 135° and 155° RTSA prostheses provide similar gain in forward elevation and abduction ROM as well as overall forward elevation and abduction, the 135° prosthesis provides significantly more external rotation with the arm at the side than the 155° prosthesis.

Limitations

Although this study attempted to look at all studies that reported active ROM in patients following a RTSA, and 2 authors performed the search, there is a possibility that some studies were missed, introducing study selection bias. Furthermore, the mean follow-up was over 3 years following surgery, but the minimum follow-up requirement for studies to be included was only 12 months. Hence, this transfer bias introduces the possibility that the patient’s ROM would have changed had they been followed for a standard period of time. There are many variables that come into play in evaluating ROM, and although the study attempted to control for these, there are some that could not be controlled for due to lack of reporting by some studies. Glenosphere size and humeral retroversion were not recorded, as they were not reliably reported in all studies, so motion outcomes based on these variables was not evaluated. Complications and clinical outcomes were not assessed in this review and as such, conclusions regarding these variables cannot be drawn from this study. Finally, indications for surgery were not reliably reported in the studies included in this paper, so differences may have existed between surgical indications of the 135° and 155° groups that could have affected outcomes.

Conclusion

Patients who receive a 135° RTSA gain significantly more external rotation from pre- to postsurgery and have an overall greater amount of external rotation than patients who receive a 155° prosthesis. Both groups show improvements in forward elevation, external rotation, and abduction following surgery.

Developing a high-efficiency operating room (OR) is both a challenging and rewarding goal for any healthcare system. The OR is traditionally a high-cost/high-revenue environment¹ and operative efficacy has been correlated with low complication rates and surgical success.² An efficient OR is one that maximizes utilization while providing safe, reproducible, cost-effective, high-quality care. Total joint arthroplasty (TJA) has occupied the center stage for OR efficiency research, in part due to increasing demands from our aging population³ and economic pressures related to high implant costs, decreased reimbursement, and competition for market shares when OR time and space are limited.

A PubMed search on OR efficiency in TJA shows a disproportionately high focus on surgical technique, such as use of patient-specific instrumentation (PSI), computer-assisted surgery (CAS), minimally invasive surgery, and closure with barbed suture. In a retrospective review of 352 TKA patients who had PSI vs conventional instrumentation, DeHaan and colleagues⁴ found that PSI was associated with significantly decreased operative and room turnover times (20.4 minutes and 6.4 minutes, respectively). In another prospective multicenter study, Mont and colleagues⁵ showed a reduction in surgical time by 8.90 min for navigated total knee arthroplasty (TKA) performed with single-use instruments, cutting blocks, and trials. Other investigators compared PSI to CAS in TKA and found PSI to be 1.45 times more profitable than CAS, with 3 PSI cases performed in an 8-hour OR day compared to 2 CAS cases.⁶

There is no question that improved surgical technique can enhance OR efficiency. However, this model, while promising, is difficult to implement on a wide scale due to surgeon preferences, vendor limitations, and added costs related to the advanced preoperative imaging studies, manufacturing of the custom guides, and maintenance of navigation equipment. In addition, while interventions such as the use of barbed suture have the potential for speeding closure time, the time saved (4.7 minutes in one randomized trial)⁷ may not be enough to affect major utilization differences per OR per day. These technologies are also frequently employed by high-volume surgeons with high-volume teams and institutions.

Ideally, we need investment in the human capital and a collective change in work cultures to produce high-quality, well-choreographed, easily reproducible routines. An efficient OR requires the synchronous involvement of a large team of individuals, including hospital administrators, surgery schedulers, surgeons, anesthesiologists, preoperative holding area staff, OR nurses, surgical attendants, sterile processing personnel, and recovery room nurses. Case schedulers should match allocated block time with time required for surgery based on the historical performance of the individual surgeon, preferably scheduling similar cases on the same day. Preoperative work-up and medical clearance should be completed prior to scheduling to avoid last-minute cancellations. Patient reminders and accommodations for those traveling from long distances can further minimize late arrivals. Prompt initiation of the perioperative clinical pathway upon a patient’s check-in is important. The surgical site should be marked and the anesthesia plan confirmed upon arrival in the preoperative holding area. Necessary products need to be ready and/or administrated in time for transfer to the OR. These include prophylactic antibiotics, coagulation factors (eg, tranexamic acid), and blood products as indicated. Spinal anesthesia, regional nerve blocks, and intravenous (IV) lines should be completed before transfer to the OR. A “block room” close to the OR can allow concurrent induction of anesthesia and has been shown to increase the number of surgical cases performed during a regular workday.⁸ Hair clipping within the surgical site and pre-scrubbing of the operative extremity should also be performed prior to transfer to the OR in order to minimize micro-organisms and dispersal of loose hair onto the sterile field.

Upon arrival of the patient to the OR, instrument tables based on the surgeon preference cards should be opened, instrument count and implant templating completed, necessary imaging displayed, and OR staff ready with specific responsibilities assigned to each member. Small and colleagues⁹ showed that using dedicated orthopedic staff familiar with the surgical routine decreased operative time by 19 minutes per procedure, or 1.25 hours for a surgeon performing 4 primary TJAs per day. Practices such as routine placement of a urinary catheter should be seriously scrutinized. In a randomized prospective study of patients undergoing total hip arthroplasty under spinal anesthesia, Miller and colleagues¹⁰ found no benefit for indwelling catheters in preventing urinary retention. In another randomized prospective study, Huang and colleagues¹¹ found the prevalence of urinary tract infections was significantly higher in TJA patients who received indwelling urinary catheters.

A scrub nurse familiar with the instruments, their assembly, and the sequence of events can ensure efficient surgical flow. The scrub nurse needs to anticipate missing or defective tools and call for them, ideally before the incision is made. Direct comparison studies are needed to assess the efficacy of routine intraoperative imaging vs commercially available universal cup alignment guides or clinical examinations in determining acceptable component positioning and limb length. Following component implantation and before wound closure, the circulating nurse should initiate the process of acquisition of a recovery room bed, make sure dressing supplies and necessary equipment are available, and call for surgical attendants. Lack of surgical attendants, delayed transfer from the OR table to hospital bed, and prolonged acquisition of a recovery room bed have been identified as major OR inefficiencies in a retrospective study by Attarian and colleagues.¹²

In summary, time is the OR’s most valuable resource.¹³ We believe that a consistent, almost automated attitude to the above procedures decreases variability and improves efficiency. By providing clear communication of the surgical needs with the team, having consistent anesthesia and nursing staff, implementing consistent perioperative protocols, and insuring that all necessary instruments and modalities are available prior to starting the procedure, we were able to sustainably increase OR throughput in a large teaching hospital.^9,14 This process, however, requires constant review to identify and eliminate new gaps, with each member of the team sharing a frank desire to improve. In this regard, hospital administrators share the duty to facilitate the implementation of any necessary changes, allocation of needed resources, and rewarding good effort, which could ultimately increase staff satisfaction and retention. Because efficiency is the ratio of benefits (eg, revenue, safety, etc.) to investment (eg, implant costs, wages, etc.), raises the question: what would be the effect of transitioning from hourly-wage to a salary-based system for key support staff? Unlike hourly-wage personnel, who have no incentive for productivity, a salaried employee assigned to a high-efficiency OR will inherently strive for improvement, employing higher organizational skills to accomplish a common goal. To our knowledge, there is no published data on this topic.

Developing a high-efficiency operating room (OR) is both a challenging and rewarding goal for any healthcare system. The OR is traditionally a high-cost/high-revenue environment¹ and operative efficacy has been correlated with low complication rates and surgical success.² An efficient OR is one that maximizes utilization while providing safe, reproducible, cost-effective, high-quality care. Total joint arthroplasty (TJA) has occupied the center stage for OR efficiency research, in part due to increasing demands from our aging population³ and economic pressures related to high implant costs, decreased reimbursement, and competition for market shares when OR time and space are limited.

A PubMed search on OR efficiency in TJA shows a disproportionately high focus on surgical technique, such as use of patient-specific instrumentation (PSI), computer-assisted surgery (CAS), minimally invasive surgery, and closure with barbed suture. In a retrospective review of 352 TKA patients who had PSI vs conventional instrumentation, DeHaan and colleagues⁴ found that PSI was associated with significantly decreased operative and room turnover times (20.4 minutes and 6.4 minutes, respectively). In another prospective multicenter study, Mont and colleagues⁵ showed a reduction in surgical time by 8.90 min for navigated total knee arthroplasty (TKA) performed with single-use instruments, cutting blocks, and trials. Other investigators compared PSI to CAS in TKA and found PSI to be 1.45 times more profitable than CAS, with 3 PSI cases performed in an 8-hour OR day compared to 2 CAS cases.⁶

There is no question that improved surgical technique can enhance OR efficiency. However, this model, while promising, is difficult to implement on a wide scale due to surgeon preferences, vendor limitations, and added costs related to the advanced preoperative imaging studies, manufacturing of the custom guides, and maintenance of navigation equipment. In addition, while interventions such as the use of barbed suture have the potential for speeding closure time, the time saved (4.7 minutes in one randomized trial)⁷ may not be enough to affect major utilization differences per OR per day. These technologies are also frequently employed by high-volume surgeons with high-volume teams and institutions.

Ideally, we need investment in the human capital and a collective change in work cultures to produce high-quality, well-choreographed, easily reproducible routines. An efficient OR requires the synchronous involvement of a large team of individuals, including hospital administrators, surgery schedulers, surgeons, anesthesiologists, preoperative holding area staff, OR nurses, surgical attendants, sterile processing personnel, and recovery room nurses. Case schedulers should match allocated block time with time required for surgery based on the historical performance of the individual surgeon, preferably scheduling similar cases on the same day. Preoperative work-up and medical clearance should be completed prior to scheduling to avoid last-minute cancellations. Patient reminders and accommodations for those traveling from long distances can further minimize late arrivals. Prompt initiation of the perioperative clinical pathway upon a patient’s check-in is important. The surgical site should be marked and the anesthesia plan confirmed upon arrival in the preoperative holding area. Necessary products need to be ready and/or administrated in time for transfer to the OR. These include prophylactic antibiotics, coagulation factors (eg, tranexamic acid), and blood products as indicated. Spinal anesthesia, regional nerve blocks, and intravenous (IV) lines should be completed before transfer to the OR. A “block room” close to the OR can allow concurrent induction of anesthesia and has been shown to increase the number of surgical cases performed during a regular workday.⁸ Hair clipping within the surgical site and pre-scrubbing of the operative extremity should also be performed prior to transfer to the OR in order to minimize micro-organisms and dispersal of loose hair onto the sterile field.

Upon arrival of the patient to the OR, instrument tables based on the surgeon preference cards should be opened, instrument count and implant templating completed, necessary imaging displayed, and OR staff ready with specific responsibilities assigned to each member. Small and colleagues⁹ showed that using dedicated orthopedic staff familiar with the surgical routine decreased operative time by 19 minutes per procedure, or 1.25 hours for a surgeon performing 4 primary TJAs per day. Practices such as routine placement of a urinary catheter should be seriously scrutinized. In a randomized prospective study of patients undergoing total hip arthroplasty under spinal anesthesia, Miller and colleagues¹⁰ found no benefit for indwelling catheters in preventing urinary retention. In another randomized prospective study, Huang and colleagues¹¹ found the prevalence of urinary tract infections was significantly higher in TJA patients who received indwelling urinary catheters.

A scrub nurse familiar with the instruments, their assembly, and the sequence of events can ensure efficient surgical flow. The scrub nurse needs to anticipate missing or defective tools and call for them, ideally before the incision is made. Direct comparison studies are needed to assess the efficacy of routine intraoperative imaging vs commercially available universal cup alignment guides or clinical examinations in determining acceptable component positioning and limb length. Following component implantation and before wound closure, the circulating nurse should initiate the process of acquisition of a recovery room bed, make sure dressing supplies and necessary equipment are available, and call for surgical attendants. Lack of surgical attendants, delayed transfer from the OR table to hospital bed, and prolonged acquisition of a recovery room bed have been identified as major OR inefficiencies in a retrospective study by Attarian and colleagues.¹²

In summary, time is the OR’s most valuable resource.¹³ We believe that a consistent, almost automated attitude to the above procedures decreases variability and improves efficiency. By providing clear communication of the surgical needs with the team, having consistent anesthesia and nursing staff, implementing consistent perioperative protocols, and insuring that all necessary instruments and modalities are available prior to starting the procedure, we were able to sustainably increase OR throughput in a large teaching hospital.^9,14 This process, however, requires constant review to identify and eliminate new gaps, with each member of the team sharing a frank desire to improve. In this regard, hospital administrators share the duty to facilitate the implementation of any necessary changes, allocation of needed resources, and rewarding good effort, which could ultimately increase staff satisfaction and retention. Because efficiency is the ratio of benefits (eg, revenue, safety, etc.) to investment (eg, implant costs, wages, etc.), raises the question: what would be the effect of transitioning from hourly-wage to a salary-based system for key support staff? Unlike hourly-wage personnel, who have no incentive for productivity, a salaried employee assigned to a high-efficiency OR will inherently strive for improvement, employing higher organizational skills to accomplish a common goal. To our knowledge, there is no published data on this topic.

Developing a high-efficiency operating room (OR) is both a challenging and rewarding goal for any healthcare system. The OR is traditionally a high-cost/high-revenue environment¹ and operative efficacy has been correlated with low complication rates and surgical success.² An efficient OR is one that maximizes utilization while providing safe, reproducible, cost-effective, high-quality care. Total joint arthroplasty (TJA) has occupied the center stage for OR efficiency research, in part due to increasing demands from our aging population³ and economic pressures related to high implant costs, decreased reimbursement, and competition for market shares when OR time and space are limited.

A PubMed search on OR efficiency in TJA shows a disproportionately high focus on surgical technique, such as use of patient-specific instrumentation (PSI), computer-assisted surgery (CAS), minimally invasive surgery, and closure with barbed suture. In a retrospective review of 352 TKA patients who had PSI vs conventional instrumentation, DeHaan and colleagues⁴ found that PSI was associated with significantly decreased operative and room turnover times (20.4 minutes and 6.4 minutes, respectively). In another prospective multicenter study, Mont and colleagues⁵ showed a reduction in surgical time by 8.90 min for navigated total knee arthroplasty (TKA) performed with single-use instruments, cutting blocks, and trials. Other investigators compared PSI to CAS in TKA and found PSI to be 1.45 times more profitable than CAS, with 3 PSI cases performed in an 8-hour OR day compared to 2 CAS cases.⁶

There is no question that improved surgical technique can enhance OR efficiency. However, this model, while promising, is difficult to implement on a wide scale due to surgeon preferences, vendor limitations, and added costs related to the advanced preoperative imaging studies, manufacturing of the custom guides, and maintenance of navigation equipment. In addition, while interventions such as the use of barbed suture have the potential for speeding closure time, the time saved (4.7 minutes in one randomized trial)⁷ may not be enough to affect major utilization differences per OR per day. These technologies are also frequently employed by high-volume surgeons with high-volume teams and institutions.

Ideally, we need investment in the human capital and a collective change in work cultures to produce high-quality, well-choreographed, easily reproducible routines. An efficient OR requires the synchronous involvement of a large team of individuals, including hospital administrators, surgery schedulers, surgeons, anesthesiologists, preoperative holding area staff, OR nurses, surgical attendants, sterile processing personnel, and recovery room nurses. Case schedulers should match allocated block time with time required for surgery based on the historical performance of the individual surgeon, preferably scheduling similar cases on the same day. Preoperative work-up and medical clearance should be completed prior to scheduling to avoid last-minute cancellations. Patient reminders and accommodations for those traveling from long distances can further minimize late arrivals. Prompt initiation of the perioperative clinical pathway upon a patient’s check-in is important. The surgical site should be marked and the anesthesia plan confirmed upon arrival in the preoperative holding area. Necessary products need to be ready and/or administrated in time for transfer to the OR. These include prophylactic antibiotics, coagulation factors (eg, tranexamic acid), and blood products as indicated. Spinal anesthesia, regional nerve blocks, and intravenous (IV) lines should be completed before transfer to the OR. A “block room” close to the OR can allow concurrent induction of anesthesia and has been shown to increase the number of surgical cases performed during a regular workday.⁸ Hair clipping within the surgical site and pre-scrubbing of the operative extremity should also be performed prior to transfer to the OR in order to minimize micro-organisms and dispersal of loose hair onto the sterile field.

Upon arrival of the patient to the OR, instrument tables based on the surgeon preference cards should be opened, instrument count and implant templating completed, necessary imaging displayed, and OR staff ready with specific responsibilities assigned to each member. Small and colleagues⁹ showed that using dedicated orthopedic staff familiar with the surgical routine decreased operative time by 19 minutes per procedure, or 1.25 hours for a surgeon performing 4 primary TJAs per day. Practices such as routine placement of a urinary catheter should be seriously scrutinized. In a randomized prospective study of patients undergoing total hip arthroplasty under spinal anesthesia, Miller and colleagues¹⁰ found no benefit for indwelling catheters in preventing urinary retention. In another randomized prospective study, Huang and colleagues¹¹ found the prevalence of urinary tract infections was significantly higher in TJA patients who received indwelling urinary catheters.

A scrub nurse familiar with the instruments, their assembly, and the sequence of events can ensure efficient surgical flow. The scrub nurse needs to anticipate missing or defective tools and call for them, ideally before the incision is made. Direct comparison studies are needed to assess the efficacy of routine intraoperative imaging vs commercially available universal cup alignment guides or clinical examinations in determining acceptable component positioning and limb length. Following component implantation and before wound closure, the circulating nurse should initiate the process of acquisition of a recovery room bed, make sure dressing supplies and necessary equipment are available, and call for surgical attendants. Lack of surgical attendants, delayed transfer from the OR table to hospital bed, and prolonged acquisition of a recovery room bed have been identified as major OR inefficiencies in a retrospective study by Attarian and colleagues.¹²

In summary, time is the OR’s most valuable resource.¹³ We believe that a consistent, almost automated attitude to the above procedures decreases variability and improves efficiency. By providing clear communication of the surgical needs with the team, having consistent anesthesia and nursing staff, implementing consistent perioperative protocols, and insuring that all necessary instruments and modalities are available prior to starting the procedure, we were able to sustainably increase OR throughput in a large teaching hospital.^9,14 This process, however, requires constant review to identify and eliminate new gaps, with each member of the team sharing a frank desire to improve. In this regard, hospital administrators share the duty to facilitate the implementation of any necessary changes, allocation of needed resources, and rewarding good effort, which could ultimately increase staff satisfaction and retention. Because efficiency is the ratio of benefits (eg, revenue, safety, etc.) to investment (eg, implant costs, wages, etc.), raises the question: what would be the effect of transitioning from hourly-wage to a salary-based system for key support staff? Unlike hourly-wage personnel, who have no incentive for productivity, a salaried employee assigned to a high-efficiency OR will inherently strive for improvement, employing higher organizational skills to accomplish a common goal. To our knowledge, there is no published data on this topic.

Since Neer and colleagues¹ first reported in 1982, glenoid loosening persists as a common cause of anatomic total shoulder arthroplasty (TSA) failure.^1-4 Currently, cemented, all-polyethylene glenoid components are the gold standard, and minimum clinical survival of 10 to 15 years is expected.^3,5 Several clinical studies^5-9 and in vitro biomechanical studies¹⁰ have suggested an advantage of pegged over keeled glenoid components, but glenoid component loosening remains a frequent complication,¹¹ with the cement–implant interface suggested as the weak link of fixation.^10,12 In addition to mechanical loosening, poor cement penetration and heat-induced necrosis have been postulated as contributing to glenoid component loosening.^13,14

Because of these potential complications, there is a growing consideration to minimize or abandon cement fixation and rely on biological fixation to polyethylene for long-term component stability.¹⁵ A newer pegged glenoid component design consists of traditional, peripherally located pegs designed for cement fixation as well as a central, uncemented, fluted, interference-fit peg that allows for bony ingrowth. Short-term clinical studies have shown that bony ingrowth into the space between the flutes can be achieved with a hybrid cementation technique and that, when that occurs, excellent outcomes are likely.^13,16-19 The immediate in vivo stability of this implant design upon initial implantation, before the cement has cured, has prompted some surgeons to consider implanting the device without cement. In a recent series in which this implant design was used without cement, clinical and radiographic results were promising.¹⁵

Despite the widespread clinical use, little biomechanical work has been done to characterize initial fixation of all-polyethylene pegged glenoid implants. We conducted a study to compare glenoid micromotion in an all-polyethylene, centrally fluted pegged glenoid component as a function of 3 fixation techniques: cementless interference-fit fixation, hybrid partial cementation based on manufacturer recommendations, and full cementation to simulate a gold-standard, traditional, cemented, pegged design.

Materials and MethodsBiomechanical Testing

The biomechanical testing methodology used in this study was based on previous studies20-23 and on ASTM standard F2028-1224 using polyurethane bone substitute 0.24 g/cm³ (Pacific Research Laboratories) with ultimate strength of 4.9 MPa and compressive modulus of 123 MPa for component implantation. This material was selected because its mechanical properties are similar to those of cancellous glenoid bone in primary shoulder arthroplasty,25 and it minimizes variability with use of cadaveric specimens. Components were mounted on an MTS 858 Mini-Bionix II materials testing frame (Figure 1). A static compressive load of 756 N (170 lb) was applied via a mass-pulley system simulating the joint compressive force the shoulder is likely to experience during higher load activities.^24,26 The glenoid component was positioned on a linear bearing to allow for joint compression.

Test Groups and Cement Fixation Techniques

All-polyethylene pegged glenoid components (Anchor Peg Glenoid, size 44; DePuy Orthopaedics) were used for biomechanical testing (Figure 1). Polyurethane blocks were reamed with a size 44 reamer until the superior-inferior distance reached 33 mm, ensuring complete seating of implant. Three fixation-technique groups were formed: interference-fit, hybrid cement, and fully cemented. Interference-fit fixation was done without polymethylmethacrylate (PMMA) cement. In hybrid fixation, 2 cm³ of PMMA (SpeedSet Cement, Stryker Orthopaedics) was injected (using a catheter tip syringe) into the peripheral peg holes and manually pressurized; the central peg was press-fit into polyurethane bone substitute. In the fully cemented group, both peripheral and central peg holes received PMMA; the peripheral peg holes were cemented as in hybrid fixation, and the central peg hole was injected with 3 cm³ of PMMA, which was then manually pressurized. The humeral head component (Global Advantage, 44×18 mm; DePuy Synthes) was mounted on the test frame actuator and centrally located within the glenoid at the start of the test.

Determination of Humeral Head Translation via Subluxation Testing

Humeral head subluxation distance, simulating a humeral head rim loading event, was calculated on the basis of preliminary tests outlined in the ASTM standard.²⁴ Three glenoids (1 per fixation technique) were mounted on the test frame with a humeral head positioned centrally within the glenoid. After the joint compressive force was applied, the humeral head was translated along the true superior axis of the glenoid at a rate of 50 mm/min. Testing software was used to record humeral head displacement and load data at a frequency of 100 Hz. Humeral head subluxation displacement was determined at the end of the linear region of the force versus displacement response. This distance, averaged from the 3 subluxation tests, was used as the subluxation distance during cyclic testing.

Determination of Glenoid Component Motion via Cyclic Testing

After subluxation displacement was determined, glenoid components were mounted on the test frame (5 per fixation technique) and subjected to 50,000 cycles of humeral head translation at a frequency of 2 Hz. Amplitude of the humeral head displacement against the glenoid component followed a sinusoidal pattern with maxima and minima represented by the subluxation displacement (positive and negative, respectively). Glenoid edge compression/distraction of the superior edge and glenoid inferior/superior translation were monitored with 2 variable resistance reluctance transducers (Microminiature DVRT; 4.5-µm resolution; MicroStrain) secured to the glenoid component and testing fixture.

Microminiature DVRT measurements of glenoid motion were taken for 5 consecutive cycles at cycles 1, 20, 100, 500, 1000, 5000, 10,000, 15,000, 20,000, 30,000, 40,000, and 50,000. Distraction-compression displacement and superior-inferior translation measurements were recorded relative to the glenoid position with the humeral head at the neutral position at a given cycle. Final glenoid micromotion data were calculated from the mean of consecutive cycles at each cycle time point.

Statistical Analysis

Glenoid motion results are reported as means and standard deviations. Comparisons with 2 factors of fixation technique and number of cycles for glenoid distraction, glenoid compression, and absolute glenoid translation were characterized with 2-way analysis of variance (SigmaPlot Version 11.0; Systat Software), with the Holm-Šídák test used for post hoc determination of significant relationships.

Results

Under subluxation testing, the humeral head translation distance at the end of the linear region was determined to be 0.50 mm. Subsequently for cyclic testing, the humeral head was then translated 0.50 mm from the neutral position of the humeral head along both the superior and inferior axes of the glenoid. All glenoids successfully completed the entire 50,000 cycles of testing. For the glenoid component, Figure 2 depicts distraction and compression, and Figure 3 depicts superior-inferior translation.

Glenoid Component Distraction

Overall, mean (SD) glenoid distraction was significantly higher for interference-fit fixation, 0.21 (0.10) mm, than for hybrid cement fixation, 0.16 (0.05) mm (P < .001), and fully cemented fixation, 0.09 (0.07) mm (P < .001). It was also significantly higher for hybrid fixation than fully cemented fixation (P < .001). From cycle 1000 to cycle 50,000, distraction was significantly higher for interference-fit fixation than for fully cemented fixation at each time point (P < .05).

Glenoid Component Compression

Mean (SD) compression was significantly higher for hybrid cement fixation, 0.31 (0.13) mm, than for interference-fit fixation, 0.17 (0.07) mm (P < .001), and fully cemented fixation, 0.17 (0.08) mm (P < .001). No significant difference was found between interference-fit and fully cemented fixation (P = .793) (Figure 2). At cycles 1, 20, 100, and 500, compression was significantly higher for hybrid fixation than for fully cemented fixation (P < .05). In addition, at cycle 500, it was significantly higher for hybrid fixation than for interference-fit fixation (P < .05).

Glenoid Component Translation

Mean (SD) glenoid translation was significantly lower for fully cemented fixation, 0.10 (0.04) mm, than for interference-fit fixation, 0.13 (0.04) mm (P < .001), and hybrid cement fixation, 0.13 (0.03) mm (P < .001), with all time points considered. There was no significant difference between interference-fit and hybrid fixation (P = .343). Initial translation at cycle 1 was significantly higher for interference-fit and hybid fixation than for fully cemented fixation.

Discussion

Despite advances in glenoid component design, glenoid loosening remains the most common cause of anatomical TSA failure. Recent implants have been designed to take advantage of an all-polyethylene component while providing long-lasting fixation through bony ingrowth into a central peg. In a study of the hybrid cementation technique drescribed here, Groh¹⁷ found no glenoid loosening or radiolucent lines but discovered fingerlike projections of bone between the flanges of the implant in 24 (29%) of 83 cases. Churchill and colleagues¹⁶ also reported bony ingrowth into the central peg in 15 (75%) of 20 patients. Furthermore, Arnold and colleagues13 reported complete bony ingrowth (6/6 inter-fin compartments) in 23 (71%) of 35 shoulders at a mean of 43 months. Wirth and colleagues¹⁹ reported increased radiodensity between the flanges of the central peg in 30 of 44 cases (68%) and osteolysis around the central peg in 3 of 44 cases (7%) at 3 years.

There are also reports of successful bony ingrowth associated with all-polyethylene components implanted without cement. In a canine study using an early ingrowth implant design, Wirth and colleagues²⁷ showed that, though initial fixation was superior with a cemented, keeled implant, pullout strength of the uncemented, pegged implant improved over time and eventually far surpassed that of the cemented, keeled implant owing to both the loosening of the cemented component and the bony ingrowth into the central peg component. Furthermore, Anglin and colleagues¹⁰ confirmed that component micromotion was lower with pegged glenoid components than with keeled components in a biomechanical model. De Wilde and colleagues¹⁵ recently reported on a series of uncemented, central fluted peg glenoids implanted in 34 patients followed clinically and with computed tomography for a minimum of 24 months. The investigators found bony ingrowth into the central peg in 27 (79%) of 34 patients and no signs of loosening in 30 (88%) of 34 patients. Incomplete lucencies around 1 or 2 peripheral pegs were found in 2 (6%) of 34 patients, and complete lucencies around 2 or more peripheral pegs were found in 2 (6%) of 34 patients. However, there was no statistical difference in clinical outcome between patients with and without loosening.

With this type of implant, initial fixation that provides stability while minimizing micromotion under biomechanical loading likely is crucial for attaining bony growth within the central peg flanges. To our knowledge, this is the first biomechanical study to compare micromotion using 3 different fixation methods with a central fluted peg glenoid component design. Of all these fixation methods, fully cemented fixation yielded the most stable glenoid throughout testing with respect to the evaluated parameters. However, this method is not necessarily clinically applicable, as a fully cemented glenoid would inhibit any bony growth within the central flange, which is necessary for long-term biological fixation. Our data showed that, though glenoid distraction was significantly lower with hybrid cement fixation, this fixation method exhibited significantly higher glenoid compression. In addition, there were no significant differences between glenoid components with hybrid fixation and glenoid components with interference-fit fixation with respect to component translation in the superior-inferior direction. These findings may indicate that initial fixation is not significantly improved by the addition of cement to the peripheral pegs in a glenoid component with a central fluted peg design.

The interference fit of the central peg is primarily responsible for conferring long-term implant stability,^13,27 which is ultimately achieved by bony formation between the flutes of the peg. Other authors have reported that, for bony ingrowth to occur, micromotion between the bone–implant interface must not exceed 20 to 150 µm.^28-30 Other than for interference-fit distraction at more than 1000 cycles and hybrid cement fixation compression at each time point throughout testing, our data fall within the reported upper limits of micromotion to support bony ingrowth. Increased micromotion in the interference-fit fixation group is seen at later time points and may be caused by the inability to simulate the potential fixation gained from bony ingrowth allowed with this surgical technique. Research is needed to further explain this increase in distraction.

Results from this study must be interpreted with caution because of limitations of the in vitro testing methodology. This biomechanical model using bone substitute characterizes glenoid fixation at time zero, directly after implantation, followed by repetitive cyclic loading simulating 5 years of implant service. This differs from the clinical scenario in which the shoulder undergoes postoperative immobilization or protected motion during which the early phases of bony remodeling are likely occurring. Furthermore, simulation of 5 years of implant service may not be necessary for an implant that is expected to achieve ultimate fixation by bony ingrowth within the first several months after implantation. Use of this implant without cement is classified off-label, and surgeons should take this into consideration during implantation. Last, this study could not simulate the effect of bony ingrowth on fixation, though our experimental technique of cementing the central peg may be a gross approximation of a fully ingrown central peg and its expected rigid fixation.

Fully cemented fixation of a polyethylene glenoid is superior to hybrid cement fixation and interference-fit fixation with respect to early glenoid micromotion. However, the long-term stability of a fully cemented polyethylene glenoid component remains a clinical concern, as fixation is achieved by bony ingrowth around the central fluted peg of the implant. In this study, interference-fit and hybrid fixation had equivocal component micromotion in biomechanical testing. Our findings suggest that cementation of the peripheral pegs confers no additional initial stability over an uncemented interference-fit technique in a biomechanical model. More research is needed to further evaluate interference-fit fixation as a viable option for implantation of a central fluted, all-polyethylene glenoid component.

Since Neer and colleagues¹ first reported in 1982, glenoid loosening persists as a common cause of anatomic total shoulder arthroplasty (TSA) failure.^1-4 Currently, cemented, all-polyethylene glenoid components are the gold standard, and minimum clinical survival of 10 to 15 years is expected.^3,5 Several clinical studies^5-9 and in vitro biomechanical studies¹⁰ have suggested an advantage of pegged over keeled glenoid components, but glenoid component loosening remains a frequent complication,¹¹ with the cement–implant interface suggested as the weak link of fixation.^10,12 In addition to mechanical loosening, poor cement penetration and heat-induced necrosis have been postulated as contributing to glenoid component loosening.^13,14

Because of these potential complications, there is a growing consideration to minimize or abandon cement fixation and rely on biological fixation to polyethylene for long-term component stability.¹⁵ A newer pegged glenoid component design consists of traditional, peripherally located pegs designed for cement fixation as well as a central, uncemented, fluted, interference-fit peg that allows for bony ingrowth. Short-term clinical studies have shown that bony ingrowth into the space between the flutes can be achieved with a hybrid cementation technique and that, when that occurs, excellent outcomes are likely.^13,16-19 The immediate in vivo stability of this implant design upon initial implantation, before the cement has cured, has prompted some surgeons to consider implanting the device without cement. In a recent series in which this implant design was used without cement, clinical and radiographic results were promising.¹⁵

Despite the widespread clinical use, little biomechanical work has been done to characterize initial fixation of all-polyethylene pegged glenoid implants. We conducted a study to compare glenoid micromotion in an all-polyethylene, centrally fluted pegged glenoid component as a function of 3 fixation techniques: cementless interference-fit fixation, hybrid partial cementation based on manufacturer recommendations, and full cementation to simulate a gold-standard, traditional, cemented, pegged design.

Materials and MethodsBiomechanical Testing

The biomechanical testing methodology used in this study was based on previous studies20-23 and on ASTM standard F2028-1224 using polyurethane bone substitute 0.24 g/cm³ (Pacific Research Laboratories) with ultimate strength of 4.9 MPa and compressive modulus of 123 MPa for component implantation. This material was selected because its mechanical properties are similar to those of cancellous glenoid bone in primary shoulder arthroplasty,25 and it minimizes variability with use of cadaveric specimens. Components were mounted on an MTS 858 Mini-Bionix II materials testing frame (Figure 1). A static compressive load of 756 N (170 lb) was applied via a mass-pulley system simulating the joint compressive force the shoulder is likely to experience during higher load activities.^24,26 The glenoid component was positioned on a linear bearing to allow for joint compression.

Test Groups and Cement Fixation Techniques

All-polyethylene pegged glenoid components (Anchor Peg Glenoid, size 44; DePuy Orthopaedics) were used for biomechanical testing (Figure 1). Polyurethane blocks were reamed with a size 44 reamer until the superior-inferior distance reached 33 mm, ensuring complete seating of implant. Three fixation-technique groups were formed: interference-fit, hybrid cement, and fully cemented. Interference-fit fixation was done without polymethylmethacrylate (PMMA) cement. In hybrid fixation, 2 cm³ of PMMA (SpeedSet Cement, Stryker Orthopaedics) was injected (using a catheter tip syringe) into the peripheral peg holes and manually pressurized; the central peg was press-fit into polyurethane bone substitute. In the fully cemented group, both peripheral and central peg holes received PMMA; the peripheral peg holes were cemented as in hybrid fixation, and the central peg hole was injected with 3 cm³ of PMMA, which was then manually pressurized. The humeral head component (Global Advantage, 44×18 mm; DePuy Synthes) was mounted on the test frame actuator and centrally located within the glenoid at the start of the test.

Determination of Humeral Head Translation via Subluxation Testing

Humeral head subluxation distance, simulating a humeral head rim loading event, was calculated on the basis of preliminary tests outlined in the ASTM standard.²⁴ Three glenoids (1 per fixation technique) were mounted on the test frame with a humeral head positioned centrally within the glenoid. After the joint compressive force was applied, the humeral head was translated along the true superior axis of the glenoid at a rate of 50 mm/min. Testing software was used to record humeral head displacement and load data at a frequency of 100 Hz. Humeral head subluxation displacement was determined at the end of the linear region of the force versus displacement response. This distance, averaged from the 3 subluxation tests, was used as the subluxation distance during cyclic testing.

Determination of Glenoid Component Motion via Cyclic Testing

After subluxation displacement was determined, glenoid components were mounted on the test frame (5 per fixation technique) and subjected to 50,000 cycles of humeral head translation at a frequency of 2 Hz. Amplitude of the humeral head displacement against the glenoid component followed a sinusoidal pattern with maxima and minima represented by the subluxation displacement (positive and negative, respectively). Glenoid edge compression/distraction of the superior edge and glenoid inferior/superior translation were monitored with 2 variable resistance reluctance transducers (Microminiature DVRT; 4.5-µm resolution; MicroStrain) secured to the glenoid component and testing fixture.

Microminiature DVRT measurements of glenoid motion were taken for 5 consecutive cycles at cycles 1, 20, 100, 500, 1000, 5000, 10,000, 15,000, 20,000, 30,000, 40,000, and 50,000. Distraction-compression displacement and superior-inferior translation measurements were recorded relative to the glenoid position with the humeral head at the neutral position at a given cycle. Final glenoid micromotion data were calculated from the mean of consecutive cycles at each cycle time point.

Statistical Analysis

Glenoid motion results are reported as means and standard deviations. Comparisons with 2 factors of fixation technique and number of cycles for glenoid distraction, glenoid compression, and absolute glenoid translation were characterized with 2-way analysis of variance (SigmaPlot Version 11.0; Systat Software), with the Holm-Šídák test used for post hoc determination of significant relationships.

Results

Under subluxation testing, the humeral head translation distance at the end of the linear region was determined to be 0.50 mm. Subsequently for cyclic testing, the humeral head was then translated 0.50 mm from the neutral position of the humeral head along both the superior and inferior axes of the glenoid. All glenoids successfully completed the entire 50,000 cycles of testing. For the glenoid component, Figure 2 depicts distraction and compression, and Figure 3 depicts superior-inferior translation.

Glenoid Component Distraction

Overall, mean (SD) glenoid distraction was significantly higher for interference-fit fixation, 0.21 (0.10) mm, than for hybrid cement fixation, 0.16 (0.05) mm (P < .001), and fully cemented fixation, 0.09 (0.07) mm (P < .001). It was also significantly higher for hybrid fixation than fully cemented fixation (P < .001). From cycle 1000 to cycle 50,000, distraction was significantly higher for interference-fit fixation than for fully cemented fixation at each time point (P < .05).

Glenoid Component Compression

Mean (SD) compression was significantly higher for hybrid cement fixation, 0.31 (0.13) mm, than for interference-fit fixation, 0.17 (0.07) mm (P < .001), and fully cemented fixation, 0.17 (0.08) mm (P < .001). No significant difference was found between interference-fit and fully cemented fixation (P = .793) (Figure 2). At cycles 1, 20, 100, and 500, compression was significantly higher for hybrid fixation than for fully cemented fixation (P < .05). In addition, at cycle 500, it was significantly higher for hybrid fixation than for interference-fit fixation (P < .05).

Glenoid Component Translation

Mean (SD) glenoid translation was significantly lower for fully cemented fixation, 0.10 (0.04) mm, than for interference-fit fixation, 0.13 (0.04) mm (P < .001), and hybrid cement fixation, 0.13 (0.03) mm (P < .001), with all time points considered. There was no significant difference between interference-fit and hybrid fixation (P = .343). Initial translation at cycle 1 was significantly higher for interference-fit and hybid fixation than for fully cemented fixation.

Discussion

Despite advances in glenoid component design, glenoid loosening remains the most common cause of anatomical TSA failure. Recent implants have been designed to take advantage of an all-polyethylene component while providing long-lasting fixation through bony ingrowth into a central peg. In a study of the hybrid cementation technique drescribed here, Groh¹⁷ found no glenoid loosening or radiolucent lines but discovered fingerlike projections of bone between the flanges of the implant in 24 (29%) of 83 cases. Churchill and colleagues¹⁶ also reported bony ingrowth into the central peg in 15 (75%) of 20 patients. Furthermore, Arnold and colleagues13 reported complete bony ingrowth (6/6 inter-fin compartments) in 23 (71%) of 35 shoulders at a mean of 43 months. Wirth and colleagues¹⁹ reported increased radiodensity between the flanges of the central peg in 30 of 44 cases (68%) and osteolysis around the central peg in 3 of 44 cases (7%) at 3 years.

There are also reports of successful bony ingrowth associated with all-polyethylene components implanted without cement. In a canine study using an early ingrowth implant design, Wirth and colleagues²⁷ showed that, though initial fixation was superior with a cemented, keeled implant, pullout strength of the uncemented, pegged implant improved over time and eventually far surpassed that of the cemented, keeled implant owing to both the loosening of the cemented component and the bony ingrowth into the central peg component. Furthermore, Anglin and colleagues¹⁰ confirmed that component micromotion was lower with pegged glenoid components than with keeled components in a biomechanical model. De Wilde and colleagues¹⁵ recently reported on a series of uncemented, central fluted peg glenoids implanted in 34 patients followed clinically and with computed tomography for a minimum of 24 months. The investigators found bony ingrowth into the central peg in 27 (79%) of 34 patients and no signs of loosening in 30 (88%) of 34 patients. Incomplete lucencies around 1 or 2 peripheral pegs were found in 2 (6%) of 34 patients, and complete lucencies around 2 or more peripheral pegs were found in 2 (6%) of 34 patients. However, there was no statistical difference in clinical outcome between patients with and without loosening.

With this type of implant, initial fixation that provides stability while minimizing micromotion under biomechanical loading likely is crucial for attaining bony growth within the central peg flanges. To our knowledge, this is the first biomechanical study to compare micromotion using 3 different fixation methods with a central fluted peg glenoid component design. Of all these fixation methods, fully cemented fixation yielded the most stable glenoid throughout testing with respect to the evaluated parameters. However, this method is not necessarily clinically applicable, as a fully cemented glenoid would inhibit any bony growth within the central flange, which is necessary for long-term biological fixation. Our data showed that, though glenoid distraction was significantly lower with hybrid cement fixation, this fixation method exhibited significantly higher glenoid compression. In addition, there were no significant differences between glenoid components with hybrid fixation and glenoid components with interference-fit fixation with respect to component translation in the superior-inferior direction. These findings may indicate that initial fixation is not significantly improved by the addition of cement to the peripheral pegs in a glenoid component with a central fluted peg design.

The interference fit of the central peg is primarily responsible for conferring long-term implant stability,^13,27 which is ultimately achieved by bony formation between the flutes of the peg. Other authors have reported that, for bony ingrowth to occur, micromotion between the bone–implant interface must not exceed 20 to 150 µm.^28-30 Other than for interference-fit distraction at more than 1000 cycles and hybrid cement fixation compression at each time point throughout testing, our data fall within the reported upper limits of micromotion to support bony ingrowth. Increased micromotion in the interference-fit fixation group is seen at later time points and may be caused by the inability to simulate the potential fixation gained from bony ingrowth allowed with this surgical technique. Research is needed to further explain this increase in distraction.

Results from this study must be interpreted with caution because of limitations of the in vitro testing methodology. This biomechanical model using bone substitute characterizes glenoid fixation at time zero, directly after implantation, followed by repetitive cyclic loading simulating 5 years of implant service. This differs from the clinical scenario in which the shoulder undergoes postoperative immobilization or protected motion during which the early phases of bony remodeling are likely occurring. Furthermore, simulation of 5 years of implant service may not be necessary for an implant that is expected to achieve ultimate fixation by bony ingrowth within the first several months after implantation. Use of this implant without cement is classified off-label, and surgeons should take this into consideration during implantation. Last, this study could not simulate the effect of bony ingrowth on fixation, though our experimental technique of cementing the central peg may be a gross approximation of a fully ingrown central peg and its expected rigid fixation.

Fully cemented fixation of a polyethylene glenoid is superior to hybrid cement fixation and interference-fit fixation with respect to early glenoid micromotion. However, the long-term stability of a fully cemented polyethylene glenoid component remains a clinical concern, as fixation is achieved by bony ingrowth around the central fluted peg of the implant. In this study, interference-fit and hybrid fixation had equivocal component micromotion in biomechanical testing. Our findings suggest that cementation of the peripheral pegs confers no additional initial stability over an uncemented interference-fit technique in a biomechanical model. More research is needed to further evaluate interference-fit fixation as a viable option for implantation of a central fluted, all-polyethylene glenoid component.

Since Neer and colleagues¹ first reported in 1982, glenoid loosening persists as a common cause of anatomic total shoulder arthroplasty (TSA) failure.^1-4 Currently, cemented, all-polyethylene glenoid components are the gold standard, and minimum clinical survival of 10 to 15 years is expected.^3,5 Several clinical studies^5-9 and in vitro biomechanical studies¹⁰ have suggested an advantage of pegged over keeled glenoid components, but glenoid component loosening remains a frequent complication,¹¹ with the cement–implant interface suggested as the weak link of fixation.^10,12 In addition to mechanical loosening, poor cement penetration and heat-induced necrosis have been postulated as contributing to glenoid component loosening.^13,14

Because of these potential complications, there is a growing consideration to minimize or abandon cement fixation and rely on biological fixation to polyethylene for long-term component stability.¹⁵ A newer pegged glenoid component design consists of traditional, peripherally located pegs designed for cement fixation as well as a central, uncemented, fluted, interference-fit peg that allows for bony ingrowth. Short-term clinical studies have shown that bony ingrowth into the space between the flutes can be achieved with a hybrid cementation technique and that, when that occurs, excellent outcomes are likely.^13,16-19 The immediate in vivo stability of this implant design upon initial implantation, before the cement has cured, has prompted some surgeons to consider implanting the device without cement. In a recent series in which this implant design was used without cement, clinical and radiographic results were promising.¹⁵

Despite the widespread clinical use, little biomechanical work has been done to characterize initial fixation of all-polyethylene pegged glenoid implants. We conducted a study to compare glenoid micromotion in an all-polyethylene, centrally fluted pegged glenoid component as a function of 3 fixation techniques: cementless interference-fit fixation, hybrid partial cementation based on manufacturer recommendations, and full cementation to simulate a gold-standard, traditional, cemented, pegged design.

Materials and MethodsBiomechanical Testing

The biomechanical testing methodology used in this study was based on previous studies20-23 and on ASTM standard F2028-1224 using polyurethane bone substitute 0.24 g/cm³ (Pacific Research Laboratories) with ultimate strength of 4.9 MPa and compressive modulus of 123 MPa for component implantation. This material was selected because its mechanical properties are similar to those of cancellous glenoid bone in primary shoulder arthroplasty,25 and it minimizes variability with use of cadaveric specimens. Components were mounted on an MTS 858 Mini-Bionix II materials testing frame (Figure 1). A static compressive load of 756 N (170 lb) was applied via a mass-pulley system simulating the joint compressive force the shoulder is likely to experience during higher load activities.^24,26 The glenoid component was positioned on a linear bearing to allow for joint compression.

Test Groups and Cement Fixation Techniques

All-polyethylene pegged glenoid components (Anchor Peg Glenoid, size 44; DePuy Orthopaedics) were used for biomechanical testing (Figure 1). Polyurethane blocks were reamed with a size 44 reamer until the superior-inferior distance reached 33 mm, ensuring complete seating of implant. Three fixation-technique groups were formed: interference-fit, hybrid cement, and fully cemented. Interference-fit fixation was done without polymethylmethacrylate (PMMA) cement. In hybrid fixation, 2 cm³ of PMMA (SpeedSet Cement, Stryker Orthopaedics) was injected (using a catheter tip syringe) into the peripheral peg holes and manually pressurized; the central peg was press-fit into polyurethane bone substitute. In the fully cemented group, both peripheral and central peg holes received PMMA; the peripheral peg holes were cemented as in hybrid fixation, and the central peg hole was injected with 3 cm³ of PMMA, which was then manually pressurized. The humeral head component (Global Advantage, 44×18 mm; DePuy Synthes) was mounted on the test frame actuator and centrally located within the glenoid at the start of the test.

Determination of Humeral Head Translation via Subluxation Testing

Humeral head subluxation distance, simulating a humeral head rim loading event, was calculated on the basis of preliminary tests outlined in the ASTM standard.²⁴ Three glenoids (1 per fixation technique) were mounted on the test frame with a humeral head positioned centrally within the glenoid. After the joint compressive force was applied, the humeral head was translated along the true superior axis of the glenoid at a rate of 50 mm/min. Testing software was used to record humeral head displacement and load data at a frequency of 100 Hz. Humeral head subluxation displacement was determined at the end of the linear region of the force versus displacement response. This distance, averaged from the 3 subluxation tests, was used as the subluxation distance during cyclic testing.

Determination of Glenoid Component Motion via Cyclic Testing

After subluxation displacement was determined, glenoid components were mounted on the test frame (5 per fixation technique) and subjected to 50,000 cycles of humeral head translation at a frequency of 2 Hz. Amplitude of the humeral head displacement against the glenoid component followed a sinusoidal pattern with maxima and minima represented by the subluxation displacement (positive and negative, respectively). Glenoid edge compression/distraction of the superior edge and glenoid inferior/superior translation were monitored with 2 variable resistance reluctance transducers (Microminiature DVRT; 4.5-µm resolution; MicroStrain) secured to the glenoid component and testing fixture.

Microminiature DVRT measurements of glenoid motion were taken for 5 consecutive cycles at cycles 1, 20, 100, 500, 1000, 5000, 10,000, 15,000, 20,000, 30,000, 40,000, and 50,000. Distraction-compression displacement and superior-inferior translation measurements were recorded relative to the glenoid position with the humeral head at the neutral position at a given cycle. Final glenoid micromotion data were calculated from the mean of consecutive cycles at each cycle time point.

Statistical Analysis

Glenoid motion results are reported as means and standard deviations. Comparisons with 2 factors of fixation technique and number of cycles for glenoid distraction, glenoid compression, and absolute glenoid translation were characterized with 2-way analysis of variance (SigmaPlot Version 11.0; Systat Software), with the Holm-Šídák test used for post hoc determination of significant relationships.

Results

Under subluxation testing, the humeral head translation distance at the end of the linear region was determined to be 0.50 mm. Subsequently for cyclic testing, the humeral head was then translated 0.50 mm from the neutral position of the humeral head along both the superior and inferior axes of the glenoid. All glenoids successfully completed the entire 50,000 cycles of testing. For the glenoid component, Figure 2 depicts distraction and compression, and Figure 3 depicts superior-inferior translation.

Glenoid Component Distraction

Overall, mean (SD) glenoid distraction was significantly higher for interference-fit fixation, 0.21 (0.10) mm, than for hybrid cement fixation, 0.16 (0.05) mm (P < .001), and fully cemented fixation, 0.09 (0.07) mm (P < .001). It was also significantly higher for hybrid fixation than fully cemented fixation (P < .001). From cycle 1000 to cycle 50,000, distraction was significantly higher for interference-fit fixation than for fully cemented fixation at each time point (P < .05).

Glenoid Component Compression

Mean (SD) compression was significantly higher for hybrid cement fixation, 0.31 (0.13) mm, than for interference-fit fixation, 0.17 (0.07) mm (P < .001), and fully cemented fixation, 0.17 (0.08) mm (P < .001). No significant difference was found between interference-fit and fully cemented fixation (P = .793) (Figure 2). At cycles 1, 20, 100, and 500, compression was significantly higher for hybrid fixation than for fully cemented fixation (P < .05). In addition, at cycle 500, it was significantly higher for hybrid fixation than for interference-fit fixation (P < .05).

Glenoid Component Translation

Mean (SD) glenoid translation was significantly lower for fully cemented fixation, 0.10 (0.04) mm, than for interference-fit fixation, 0.13 (0.04) mm (P < .001), and hybrid cement fixation, 0.13 (0.03) mm (P < .001), with all time points considered. There was no significant difference between interference-fit and hybrid fixation (P = .343). Initial translation at cycle 1 was significantly higher for interference-fit and hybid fixation than for fully cemented fixation.

Discussion

Despite advances in glenoid component design, glenoid loosening remains the most common cause of anatomical TSA failure. Recent implants have been designed to take advantage of an all-polyethylene component while providing long-lasting fixation through bony ingrowth into a central peg. In a study of the hybrid cementation technique drescribed here, Groh¹⁷ found no glenoid loosening or radiolucent lines but discovered fingerlike projections of bone between the flanges of the implant in 24 (29%) of 83 cases. Churchill and colleagues¹⁶ also reported bony ingrowth into the central peg in 15 (75%) of 20 patients. Furthermore, Arnold and colleagues13 reported complete bony ingrowth (6/6 inter-fin compartments) in 23 (71%) of 35 shoulders at a mean of 43 months. Wirth and colleagues¹⁹ reported increased radiodensity between the flanges of the central peg in 30 of 44 cases (68%) and osteolysis around the central peg in 3 of 44 cases (7%) at 3 years.

There are also reports of successful bony ingrowth associated with all-polyethylene components implanted without cement. In a canine study using an early ingrowth implant design, Wirth and colleagues²⁷ showed that, though initial fixation was superior with a cemented, keeled implant, pullout strength of the uncemented, pegged implant improved over time and eventually far surpassed that of the cemented, keeled implant owing to both the loosening of the cemented component and the bony ingrowth into the central peg component. Furthermore, Anglin and colleagues¹⁰ confirmed that component micromotion was lower with pegged glenoid components than with keeled components in a biomechanical model. De Wilde and colleagues¹⁵ recently reported on a series of uncemented, central fluted peg glenoids implanted in 34 patients followed clinically and with computed tomography for a minimum of 24 months. The investigators found bony ingrowth into the central peg in 27 (79%) of 34 patients and no signs of loosening in 30 (88%) of 34 patients. Incomplete lucencies around 1 or 2 peripheral pegs were found in 2 (6%) of 34 patients, and complete lucencies around 2 or more peripheral pegs were found in 2 (6%) of 34 patients. However, there was no statistical difference in clinical outcome between patients with and without loosening.

With this type of implant, initial fixation that provides stability while minimizing micromotion under biomechanical loading likely is crucial for attaining bony growth within the central peg flanges. To our knowledge, this is the first biomechanical study to compare micromotion using 3 different fixation methods with a central fluted peg glenoid component design. Of all these fixation methods, fully cemented fixation yielded the most stable glenoid throughout testing with respect to the evaluated parameters. However, this method is not necessarily clinically applicable, as a fully cemented glenoid would inhibit any bony growth within the central flange, which is necessary for long-term biological fixation. Our data showed that, though glenoid distraction was significantly lower with hybrid cement fixation, this fixation method exhibited significantly higher glenoid compression. In addition, there were no significant differences between glenoid components with hybrid fixation and glenoid components with interference-fit fixation with respect to component translation in the superior-inferior direction. These findings may indicate that initial fixation is not significantly improved by the addition of cement to the peripheral pegs in a glenoid component with a central fluted peg design.

The interference fit of the central peg is primarily responsible for conferring long-term implant stability,^13,27 which is ultimately achieved by bony formation between the flutes of the peg. Other authors have reported that, for bony ingrowth to occur, micromotion between the bone–implant interface must not exceed 20 to 150 µm.^28-30 Other than for interference-fit distraction at more than 1000 cycles and hybrid cement fixation compression at each time point throughout testing, our data fall within the reported upper limits of micromotion to support bony ingrowth. Increased micromotion in the interference-fit fixation group is seen at later time points and may be caused by the inability to simulate the potential fixation gained from bony ingrowth allowed with this surgical technique. Research is needed to further explain this increase in distraction.

Results from this study must be interpreted with caution because of limitations of the in vitro testing methodology. This biomechanical model using bone substitute characterizes glenoid fixation at time zero, directly after implantation, followed by repetitive cyclic loading simulating 5 years of implant service. This differs from the clinical scenario in which the shoulder undergoes postoperative immobilization or protected motion during which the early phases of bony remodeling are likely occurring. Furthermore, simulation of 5 years of implant service may not be necessary for an implant that is expected to achieve ultimate fixation by bony ingrowth within the first several months after implantation. Use of this implant without cement is classified off-label, and surgeons should take this into consideration during implantation. Last, this study could not simulate the effect of bony ingrowth on fixation, though our experimental technique of cementing the central peg may be a gross approximation of a fully ingrown central peg and its expected rigid fixation.

Fully cemented fixation of a polyethylene glenoid is superior to hybrid cement fixation and interference-fit fixation with respect to early glenoid micromotion. However, the long-term stability of a fully cemented polyethylene glenoid component remains a clinical concern, as fixation is achieved by bony ingrowth around the central fluted peg of the implant. In this study, interference-fit and hybrid fixation had equivocal component micromotion in biomechanical testing. Our findings suggest that cementation of the peripheral pegs confers no additional initial stability over an uncemented interference-fit technique in a biomechanical model. More research is needed to further evaluate interference-fit fixation as a viable option for implantation of a central fluted, all-polyethylene glenoid component.

Shoulder arthroplasty (SA), including total SA (TSA) and reverse TSA, is an effective surgical treatment for fracture and primary or secondary degenerative disease of the shoulder.¹ Over the past few decades, use of SA has increased dramatically, from about 5000 cases in 1990 to 7000 in 2000 and more than 26,000 in 2008.^1,2

Complications associated with SA generally are classified as perioperative (occurring during the operative index) or long-term (postdischarge).³ Long-term complications include implant loosening, instability, revision, infection, rotator cuff tear, neural injury, and deltoid detachment.^1,4,5 Perioperative complications, which are less commonly reported, include intraoperative fracture, infection, neural injury, venous thromboembolic events (VTEs, including pulmonary embolism [PE] and deep vein thrombosis [DVT]), transfusion, and death.^3,6-10

SA is an attractive treatment option for patients with rheumatoid arthritis (RA), as the effects of pain on these patients are greater in the shoulder joint than in any other joint.¹¹ Patients with RA pose unique orthopedic surgical challenges, including any combination of decreased bone mineralization, poor capsular tissue integrity, and osteonecrosis.^3,12 In addition, RA patients may be taking immunosuppressive medications that have severe side effects, and they may require multiple surgeries.^12,13 These factors predispose patients with RA to complications that include infection and wound dehiscence.^3,5,12-14

The complex nature of RA has prompted investigators to examine outcome measures in this patient group. Hambright and colleagues³ used the Nationwide Inpatient Sample (NIS) to examine perioperative outcomes in RA patients who underwent TSA between 1988 and 2005.³ They found that TSA patients with RA had shorter and less costly hospital stays and were more likely to have a routine discharge.³ Using the same patient population drawn from the period 2006–2011, we conducted a study to determine if this unexpected trend persists as the number of TSAs and quality of postoperative care continue to increase. Given the potential for anemia of chronic disease and the systemic inflammatory nature of RA, we hypothesized that the perioperative complication profile of RA patients would be worse than that of non-RA patients.

Materials and Methods

NIS data were acquired for the period 2006–2011. The NIS is the largest publicly available all-payer inpatient database, with a random 20% sample of about 1000 US hospitals accounting for 7 to 8 million inpatient stays. The database supplies weights used to estimate national totals, at about 35 million inpatient visits per year. NIS inpatient data are limited to the operative index. Postdischarge information is not available. The NIS is managed by the Healthcare Cost and Utilization Project, which is sponsored by the Agency for Healthcare Research and Quality. The quality of NIS data is assessed and validated by an independent contractor. NIS data have been widely used to examine perioperative outcomes.^15-17

NIS data cover patient and hospital demographics, hospital length of stay (LOS), discharge status, payer information, charges, and perioperative outcomes and procedure/diagnosis codes (ICD-9; International Classification of Diseases, Ninth Revision¹⁸).

As our Institutional Review Board (IRB) reviewed the database and determined the project was not human subject research, IRB involvement was not required. This study paralleled successful efforts with similar RA and non-RA patients who had shoulder and elbow surgery.^3,19 SA patients were identified by ICD-9 procedure code 81.80, but this code does not specify whether the prosthesis was unconstrained, semiconstrained, or constrained. ICD-9 coding also does not specify whether the TSA was traditional or reverse. Patients with RA were identified by ICD-9 diagnosis codes 714.0, 714.1, and 714.2. Patients without one of these codes were placed in the non-RA cohort. Patients with codes associated with pathologic fractures secondary to metastatic cancer or bone malignant neoplasm as a secondary or primary diagnosis and patients who had revision surgery indicated by code 81.83 were excluded, as they have a disproportionately higher comorbidity burden.

After each cohort was defined, demographic data (age, sex, race, income quartile based on ZIP postal code) were compared, as were data on primary payer, hospital demographics, LOS (≤5 days, defined as perioperative index), discharge type, inflation-adjusted charges in 2014 dollars based on the Consumer Price Indexes (http://www.bls.gov/cpi/), and mortality. Perioperative complications—respiratory, gastrointestinal, genitourinary, accidental puncture/laceration, central nervous system, wound dehiscence, device-related (including embolism, fibrosis, hemorrhage, pain, stenosis, or thrombus caused by any device, implant, or graft), cardiac, hematoma/seroma, acute respiratory distress syndrome, postoperative shock, VTE, postoperative infection complications, and intraoperative transfusions—were considered using ICD-9 codes (996.X-999.X and 99.X, respectively).²⁰ Although commonly used to determine perioperative comorbidity burden using ICD-9 coding, the modified Charlson index was not considered because RA is a component of the index and would therefore bias the variable.^3,21

Statistical analyses, including χ² tests and 2-sample t tests, were performed for categorical and continuous variables, respectively. P < .05 was considered significant. Fisher exact test was used for cohorts with fewer than 5 occurrences. Multivariate logistic regression models were then calculated to determine the effect of RA on different outcomes and complications, with age, race, sex, hospital region, hospital type, number of hospital beds, primary payer, and hospital ownership as covariates. Statistical analyses were performed using the R statistical programming language.22

Results

Of the 34,970 patients who underwent SA between 2006 and 2011, 1674 (4.8%) had a diagnosis of RA and 33,296 (95.2%) did not. On average, patients with RA tended to be younger than patients without RA (66.4 vs 69.1 years; P < .001), and a larger percentage of RA patients were female (75.5% vs 54.4%; P < .001). Compared with non-RA patients, RA patients comprised a different ethnic group and had a different expected primary payer (P < .001). SA patients with and without RA did not differ in income quartile based on ZIP code, total number of hospital beds, hospital region, or hospital teaching status (P = .34, .78, .59, and .82, respectively) (Table 1).

LOS was significantly (P < .001) statistically longer for RA patients (2.196 days) than for non-RA patients (2.085 days). RA patients were significantly less likely to be discharged home (63.0% vs 67.6%; P < .001). (Routine discharge was defined as discharge home, whereas nonroutine discharge was defined as discharge to a short-term hospital, skilled nursing facility, intermediate care, another type of facility, home health care, against medical advice, or death.) In addition, inflation-adjusted charges associated with SA were significantly higher (P = .018) for RA patients ($54,284) than for non-RA patients ($52,663) (Table 1).

Regarding the rates of complications that occurred during the perioperative index, there were no significant differences between RA and non-RA cohorts. These complications included respiratory, gastrointestinal, genitourinary, accidental puncture/laceration, central nervous system, wound dehiscence, device-related, cardiac, hematoma/seroma, acute respiratory distress syndrome, postoperative shock, VTE, and postoperative infection (Table 2). In addition, there was no significant difference in mortality between the groups (P = .48).

In TSA, blood transfusions were more likely (P < .001) to be given to RA patients (9.00%) than to non-RA patients (6.16%). Multivariate regression analyses were performed with age, race, sex, hospital region, hospital type, number of hospital beds, primary payer, and hospital ownership as covariates. These analyses revealed that transfusion (P < .001), discharge type (P = .002), total inflation-adjusted charges (P < .001), and LOS (P = .047) remained significant (Table 3).

Discussion

Large national databases like NIS allow study of uncommon medical occurrences and help delineate risks and trends that otherwise might be indeterminable. Although it has been suggested that patients with RA may have poorer long-term outcomes after SA, the perioperative risk profile indicates that TSA is well tolerated in RA patients during the operative index.^3,23-25

The data on this study’s 34,970 patients, drawn from the period 2006–2011, demonstrated no significant differences in safety profile with respect to the 14 perioperative complications and outcomes examined, except blood transfusion rate. Rates of postoperative infection (RA, 0.24%; non-RA, 0.14%; P = .303), VTE (RA, 0.30%; non-RA, 0.25%; P = .905), and transfusion (RA, 9.00%; non-RA, 6.16%; P < .001) are of particular interest because of the severity of these situations.

Postoperative infection is a potentially serious complication and often occurs secondary to diabetes, RA, lupus erythematosus, prior surgery, or a nosocomial or remote source.¹ The often costly treatment options include antibiotic suppression, irrigation and debridement with implant retention, 1-stage exchange with antibiotic-impregnated cement fixation, staged reimplantation, resection arthroplasty, arthrodesis, and amputation.¹ The overall 0.14% infection rate determined in this study is lower than the 0.7% reported for SA patients in the literature.¹ Given the nature of the NIS database, this rate underestimates the true postoperative infection rate, as any infection that occurred after the perioperative period is not captured.²⁶ The present study’s perioperative infection rates (RA, 0.24%; non-RA, 0.14%) for the period 2006–2011 are comparable to the rates (RA, 0.17%; non-RA, 0.24%) reported by Hambright and colleagues3 for the same patient population over the preceding, 18-year period (1988–2005) and similarly do not significantly differ between groups. Although infection is uncommon in the immediate perioperative period, the ICD-9 codes used refer specifically to infection resulting from surgery and do not represent concomitant infection.

VTEs, which include PEs and DVTs, are rare but potentially life-threatening surgical complications.^27,28 Mechanical prophylaxis and chemical prophylaxis have been recommended for major orthopedic surgery, particularly lower extremity surgery, such as total hip arthroplasty (THA) and total knee arthroplasty (TKA).^28,29 In the present study, VTE rates were low, 0.30% (RA) and 0.25% (non-RA), and not significantly different in bivariate or multivariate analyses. These rates are comparable to those found in other national-database SA studies.²⁸ VTEs that occur outside the index hospital admission are not captured in this database. Therefore, the rates in the present study may be lower than the true incidence after SA. Mortality secondary to VTE usually occurs within 24 hours but may occur up to 90 days after surgery. DVT rates, on the other hand, are difficult to evaluate because of differences in screening practices.^27,28,30,31

That RA patients were more likely than non-RA patients to receive perioperative blood transfusions supports prior findings that SA patients with RA were more likely than SA patients with osteoarthritis (OA) to receive perioperative blood transfusions.⁸ RA patients have been shown to have high rates of anemia of chronic disease, ranging from 22% to 77%.³² During joint replacement, these patients often require transfusions.^32,33 However, these findings differ from prior findings of no differences between RA and non-RA patients in the same patient population during the period 1988–2005.³ This difference may be a product of the constantly changing transfusion guidelines and increased use; transfusion rates increased 140% between 1997 and 2007, making transfusions the fastest growing common procedure in the United States during that time.³⁴ There was no difference between RA and non-RA patients in household income (as determined by ZIP code analysis), number of hospital beds, hospital region, or hospital teaching status. Compared with non-RA patients, RA patients were more likely to be younger, female, and of a difference race and to have a different expected primary payer (P < .001).These findings are consistent with previous findings in the literature.³ In the present SA study, however, RA patients were more likely than non-RA patients to have longer LOS, higher inflation-adjusted hospital charges, and nonroutine discharge. These findings deviate from those of the study covering the preceding 18 years (1988–2005).³ Despite the findings of a changing environment of care for RA patients, by Hambright and colleagues³ and Weiss and colleagues,³⁵ the trend appears to have shifted. Both groups had shorter average LOS than either group from the preceding 18 years.³ Although statistically significant in bivariate analysis, the difference in LOS between the 2 groups differed by an average of 0.11 day (2 hours 24 minutes) and was not clinically relevant.

In addition, the higher charges for patients with RA represent a deviation from the preceding 18 years.³ Other studies have also shown that RA is associated with increased cost in TSA.³⁶ Patients with RA often have rotator cuff pathology, indicating reverse SA may be used more frequently.^37,38 The increased implant cost associated with reverse SA may account for the increased costs in RA patients.³⁹ As mentioned, TSA type is not captured in the NIS database. In addition, that RA patients were less likely than non-RA patients to have routine discharge may indicate RA cases are more complex because of their complications.^1,5,14,40 A recent study of complications in RA patients (1163 who underwent THA, 2692 who underwent TKA) found that THA patients with RA were significantly more likely than THA patients with OA to dislocate, and TKA patients with RA were significantly more likely than TKA patients with OA to develop an infection after surgery.⁴¹ Postoperative dislocation has been shown to increase hospital costs in other orthopedic procedures.⁴² Also, during TSA, patients with RA are more likely than patients with OA to receive intraoperative blood transfusions.⁸ These complications—combined with the fact that RA is a chronic, progressive, systemic inflammatory disease that can affect soft tissue and blood vessel wall healing and is associated with medications having potential side effects—could contribute to the apparent increased hospital charges and LOS.^3,12,13,43 Factors that include surgeon preference, impact of primary payer, and hospital practice may also affect final charges. Total charges in the NIS database include administrative fees, hospital costs, device-related costs, operating room costs, and ancillary staff costs. Total charges do not include professional fees and differ from the total cost that represents the amount reimbursed by the payer. Charges tend to correlate with but overestimate the total costs.⁴⁴

This study had several important limitations. As mentioned, only events that occur during the operative admission are captured in the NIS database, and thus postoperative complications or serious adverse events that lead to readmission cannot be identified. In addition, outpatient TSAs are not captured in the NIS database, and thus inclusion of only inpatient procedures yields higher average LOS and total charges.⁴⁵ Given the limited granularity of ICD-9 coding, this study could not determine RA severity, estimated blood loss, length of surgery, complication severity, type of TSA procedure/prosthesis, or cause of death. Although commonly used to determine comorbidity burden, the modified Charlson index could not be used, and therefore could not be entered as a covariate in multivariate analysis. Furthermore, the NIS database does not include imaging or patient-reported outcomes information, such as improvements in pain or function, which are of crucial importance in considering surgery.

Conclusion

Our findings corroborated findings that the demographics and the perioperative safety profile for TSA were similar for patients with and without RA. The risk for complications or death in the perioperative period was low. Compared with non-RA patients, RA patients had significantly higher charges and longer LOS and were less likely to be discharged home after surgery. The 0.11-day difference in LOS, though statistically significant, was not clinically relevant. These findings differ from those for the preceding, 18-year period (1988–2005). Future research should focus on the causes of these changes.

Shoulder arthroplasty (SA), including total SA (TSA) and reverse TSA, is an effective surgical treatment for fracture and primary or secondary degenerative disease of the shoulder.¹ Over the past few decades, use of SA has increased dramatically, from about 5000 cases in 1990 to 7000 in 2000 and more than 26,000 in 2008.^1,2

Complications associated with SA generally are classified as perioperative (occurring during the operative index) or long-term (postdischarge).³ Long-term complications include implant loosening, instability, revision, infection, rotator cuff tear, neural injury, and deltoid detachment.^1,4,5 Perioperative complications, which are less commonly reported, include intraoperative fracture, infection, neural injury, venous thromboembolic events (VTEs, including pulmonary embolism [PE] and deep vein thrombosis [DVT]), transfusion, and death.^3,6-10

SA is an attractive treatment option for patients with rheumatoid arthritis (RA), as the effects of pain on these patients are greater in the shoulder joint than in any other joint.¹¹ Patients with RA pose unique orthopedic surgical challenges, including any combination of decreased bone mineralization, poor capsular tissue integrity, and osteonecrosis.^3,12 In addition, RA patients may be taking immunosuppressive medications that have severe side effects, and they may require multiple surgeries.^12,13 These factors predispose patients with RA to complications that include infection and wound dehiscence.^3,5,12-14

The complex nature of RA has prompted investigators to examine outcome measures in this patient group. Hambright and colleagues³ used the Nationwide Inpatient Sample (NIS) to examine perioperative outcomes in RA patients who underwent TSA between 1988 and 2005.³ They found that TSA patients with RA had shorter and less costly hospital stays and were more likely to have a routine discharge.³ Using the same patient population drawn from the period 2006–2011, we conducted a study to determine if this unexpected trend persists as the number of TSAs and quality of postoperative care continue to increase. Given the potential for anemia of chronic disease and the systemic inflammatory nature of RA, we hypothesized that the perioperative complication profile of RA patients would be worse than that of non-RA patients.

Materials and Methods

NIS data were acquired for the period 2006–2011. The NIS is the largest publicly available all-payer inpatient database, with a random 20% sample of about 1000 US hospitals accounting for 7 to 8 million inpatient stays. The database supplies weights used to estimate national totals, at about 35 million inpatient visits per year. NIS inpatient data are limited to the operative index. Postdischarge information is not available. The NIS is managed by the Healthcare Cost and Utilization Project, which is sponsored by the Agency for Healthcare Research and Quality. The quality of NIS data is assessed and validated by an independent contractor. NIS data have been widely used to examine perioperative outcomes.^15-17

NIS data cover patient and hospital demographics, hospital length of stay (LOS), discharge status, payer information, charges, and perioperative outcomes and procedure/diagnosis codes (ICD-9; International Classification of Diseases, Ninth Revision¹⁸).

As our Institutional Review Board (IRB) reviewed the database and determined the project was not human subject research, IRB involvement was not required. This study paralleled successful efforts with similar RA and non-RA patients who had shoulder and elbow surgery.^3,19 SA patients were identified by ICD-9 procedure code 81.80, but this code does not specify whether the prosthesis was unconstrained, semiconstrained, or constrained. ICD-9 coding also does not specify whether the TSA was traditional or reverse. Patients with RA were identified by ICD-9 diagnosis codes 714.0, 714.1, and 714.2. Patients without one of these codes were placed in the non-RA cohort. Patients with codes associated with pathologic fractures secondary to metastatic cancer or bone malignant neoplasm as a secondary or primary diagnosis and patients who had revision surgery indicated by code 81.83 were excluded, as they have a disproportionately higher comorbidity burden.

After each cohort was defined, demographic data (age, sex, race, income quartile based on ZIP postal code) were compared, as were data on primary payer, hospital demographics, LOS (≤5 days, defined as perioperative index), discharge type, inflation-adjusted charges in 2014 dollars based on the Consumer Price Indexes (http://www.bls.gov/cpi/), and mortality. Perioperative complications—respiratory, gastrointestinal, genitourinary, accidental puncture/laceration, central nervous system, wound dehiscence, device-related (including embolism, fibrosis, hemorrhage, pain, stenosis, or thrombus caused by any device, implant, or graft), cardiac, hematoma/seroma, acute respiratory distress syndrome, postoperative shock, VTE, postoperative infection complications, and intraoperative transfusions—were considered using ICD-9 codes (996.X-999.X and 99.X, respectively).²⁰ Although commonly used to determine perioperative comorbidity burden using ICD-9 coding, the modified Charlson index was not considered because RA is a component of the index and would therefore bias the variable.^3,21

Statistical analyses, including χ² tests and 2-sample t tests, were performed for categorical and continuous variables, respectively. P < .05 was considered significant. Fisher exact test was used for cohorts with fewer than 5 occurrences. Multivariate logistic regression models were then calculated to determine the effect of RA on different outcomes and complications, with age, race, sex, hospital region, hospital type, number of hospital beds, primary payer, and hospital ownership as covariates. Statistical analyses were performed using the R statistical programming language.22

Results

Of the 34,970 patients who underwent SA between 2006 and 2011, 1674 (4.8%) had a diagnosis of RA and 33,296 (95.2%) did not. On average, patients with RA tended to be younger than patients without RA (66.4 vs 69.1 years; P < .001), and a larger percentage of RA patients were female (75.5% vs 54.4%; P < .001). Compared with non-RA patients, RA patients comprised a different ethnic group and had a different expected primary payer (P < .001). SA patients with and without RA did not differ in income quartile based on ZIP code, total number of hospital beds, hospital region, or hospital teaching status (P = .34, .78, .59, and .82, respectively) (Table 1).

LOS was significantly (P < .001) statistically longer for RA patients (2.196 days) than for non-RA patients (2.085 days). RA patients were significantly less likely to be discharged home (63.0% vs 67.6%; P < .001). (Routine discharge was defined as discharge home, whereas nonroutine discharge was defined as discharge to a short-term hospital, skilled nursing facility, intermediate care, another type of facility, home health care, against medical advice, or death.) In addition, inflation-adjusted charges associated with SA were significantly higher (P = .018) for RA patients ($54,284) than for non-RA patients ($52,663) (Table 1).

Regarding the rates of complications that occurred during the perioperative index, there were no significant differences between RA and non-RA cohorts. These complications included respiratory, gastrointestinal, genitourinary, accidental puncture/laceration, central nervous system, wound dehiscence, device-related, cardiac, hematoma/seroma, acute respiratory distress syndrome, postoperative shock, VTE, and postoperative infection (Table 2). In addition, there was no significant difference in mortality between the groups (P = .48).

In TSA, blood transfusions were more likely (P < .001) to be given to RA patients (9.00%) than to non-RA patients (6.16%). Multivariate regression analyses were performed with age, race, sex, hospital region, hospital type, number of hospital beds, primary payer, and hospital ownership as covariates. These analyses revealed that transfusion (P < .001), discharge type (P = .002), total inflation-adjusted charges (P < .001), and LOS (P = .047) remained significant (Table 3).

Discussion

Large national databases like NIS allow study of uncommon medical occurrences and help delineate risks and trends that otherwise might be indeterminable. Although it has been suggested that patients with RA may have poorer long-term outcomes after SA, the perioperative risk profile indicates that TSA is well tolerated in RA patients during the operative index.^3,23-25

The data on this study’s 34,970 patients, drawn from the period 2006–2011, demonstrated no significant differences in safety profile with respect to the 14 perioperative complications and outcomes examined, except blood transfusion rate. Rates of postoperative infection (RA, 0.24%; non-RA, 0.14%; P = .303), VTE (RA, 0.30%; non-RA, 0.25%; P = .905), and transfusion (RA, 9.00%; non-RA, 6.16%; P < .001) are of particular interest because of the severity of these situations.

Postoperative infection is a potentially serious complication and often occurs secondary to diabetes, RA, lupus erythematosus, prior surgery, or a nosocomial or remote source.¹ The often costly treatment options include antibiotic suppression, irrigation and debridement with implant retention, 1-stage exchange with antibiotic-impregnated cement fixation, staged reimplantation, resection arthroplasty, arthrodesis, and amputation.¹ The overall 0.14% infection rate determined in this study is lower than the 0.7% reported for SA patients in the literature.¹ Given the nature of the NIS database, this rate underestimates the true postoperative infection rate, as any infection that occurred after the perioperative period is not captured.²⁶ The present study’s perioperative infection rates (RA, 0.24%; non-RA, 0.14%) for the period 2006–2011 are comparable to the rates (RA, 0.17%; non-RA, 0.24%) reported by Hambright and colleagues3 for the same patient population over the preceding, 18-year period (1988–2005) and similarly do not significantly differ between groups. Although infection is uncommon in the immediate perioperative period, the ICD-9 codes used refer specifically to infection resulting from surgery and do not represent concomitant infection.

VTEs, which include PEs and DVTs, are rare but potentially life-threatening surgical complications.^27,28 Mechanical prophylaxis and chemical prophylaxis have been recommended for major orthopedic surgery, particularly lower extremity surgery, such as total hip arthroplasty (THA) and total knee arthroplasty (TKA).^28,29 In the present study, VTE rates were low, 0.30% (RA) and 0.25% (non-RA), and not significantly different in bivariate or multivariate analyses. These rates are comparable to those found in other national-database SA studies.²⁸ VTEs that occur outside the index hospital admission are not captured in this database. Therefore, the rates in the present study may be lower than the true incidence after SA. Mortality secondary to VTE usually occurs within 24 hours but may occur up to 90 days after surgery. DVT rates, on the other hand, are difficult to evaluate because of differences in screening practices.^27,28,30,31

That RA patients were more likely than non-RA patients to receive perioperative blood transfusions supports prior findings that SA patients with RA were more likely than SA patients with osteoarthritis (OA) to receive perioperative blood transfusions.⁸ RA patients have been shown to have high rates of anemia of chronic disease, ranging from 22% to 77%.³² During joint replacement, these patients often require transfusions.^32,33 However, these findings differ from prior findings of no differences between RA and non-RA patients in the same patient population during the period 1988–2005.³ This difference may be a product of the constantly changing transfusion guidelines and increased use; transfusion rates increased 140% between 1997 and 2007, making transfusions the fastest growing common procedure in the United States during that time.³⁴ There was no difference between RA and non-RA patients in household income (as determined by ZIP code analysis), number of hospital beds, hospital region, or hospital teaching status. Compared with non-RA patients, RA patients were more likely to be younger, female, and of a difference race and to have a different expected primary payer (P < .001).These findings are consistent with previous findings in the literature.³ In the present SA study, however, RA patients were more likely than non-RA patients to have longer LOS, higher inflation-adjusted hospital charges, and nonroutine discharge. These findings deviate from those of the study covering the preceding 18 years (1988–2005).³ Despite the findings of a changing environment of care for RA patients, by Hambright and colleagues³ and Weiss and colleagues,³⁵ the trend appears to have shifted. Both groups had shorter average LOS than either group from the preceding 18 years.³ Although statistically significant in bivariate analysis, the difference in LOS between the 2 groups differed by an average of 0.11 day (2 hours 24 minutes) and was not clinically relevant.

In addition, the higher charges for patients with RA represent a deviation from the preceding 18 years.³ Other studies have also shown that RA is associated with increased cost in TSA.³⁶ Patients with RA often have rotator cuff pathology, indicating reverse SA may be used more frequently.^37,38 The increased implant cost associated with reverse SA may account for the increased costs in RA patients.³⁹ As mentioned, TSA type is not captured in the NIS database. In addition, that RA patients were less likely than non-RA patients to have routine discharge may indicate RA cases are more complex because of their complications.^1,5,14,40 A recent study of complications in RA patients (1163 who underwent THA, 2692 who underwent TKA) found that THA patients with RA were significantly more likely than THA patients with OA to dislocate, and TKA patients with RA were significantly more likely than TKA patients with OA to develop an infection after surgery.⁴¹ Postoperative dislocation has been shown to increase hospital costs in other orthopedic procedures.⁴² Also, during TSA, patients with RA are more likely than patients with OA to receive intraoperative blood transfusions.⁸ These complications—combined with the fact that RA is a chronic, progressive, systemic inflammatory disease that can affect soft tissue and blood vessel wall healing and is associated with medications having potential side effects—could contribute to the apparent increased hospital charges and LOS.^3,12,13,43 Factors that include surgeon preference, impact of primary payer, and hospital practice may also affect final charges. Total charges in the NIS database include administrative fees, hospital costs, device-related costs, operating room costs, and ancillary staff costs. Total charges do not include professional fees and differ from the total cost that represents the amount reimbursed by the payer. Charges tend to correlate with but overestimate the total costs.⁴⁴

This study had several important limitations. As mentioned, only events that occur during the operative admission are captured in the NIS database, and thus postoperative complications or serious adverse events that lead to readmission cannot be identified. In addition, outpatient TSAs are not captured in the NIS database, and thus inclusion of only inpatient procedures yields higher average LOS and total charges.⁴⁵ Given the limited granularity of ICD-9 coding, this study could not determine RA severity, estimated blood loss, length of surgery, complication severity, type of TSA procedure/prosthesis, or cause of death. Although commonly used to determine comorbidity burden, the modified Charlson index could not be used, and therefore could not be entered as a covariate in multivariate analysis. Furthermore, the NIS database does not include imaging or patient-reported outcomes information, such as improvements in pain or function, which are of crucial importance in considering surgery.

Conclusion

Our findings corroborated findings that the demographics and the perioperative safety profile for TSA were similar for patients with and without RA. The risk for complications or death in the perioperative period was low. Compared with non-RA patients, RA patients had significantly higher charges and longer LOS and were less likely to be discharged home after surgery. The 0.11-day difference in LOS, though statistically significant, was not clinically relevant. These findings differ from those for the preceding, 18-year period (1988–2005). Future research should focus on the causes of these changes.

Shoulder arthroplasty (SA), including total SA (TSA) and reverse TSA, is an effective surgical treatment for fracture and primary or secondary degenerative disease of the shoulder.¹ Over the past few decades, use of SA has increased dramatically, from about 5000 cases in 1990 to 7000 in 2000 and more than 26,000 in 2008.^1,2

Complications associated with SA generally are classified as perioperative (occurring during the operative index) or long-term (postdischarge).³ Long-term complications include implant loosening, instability, revision, infection, rotator cuff tear, neural injury, and deltoid detachment.^1,4,5 Perioperative complications, which are less commonly reported, include intraoperative fracture, infection, neural injury, venous thromboembolic events (VTEs, including pulmonary embolism [PE] and deep vein thrombosis [DVT]), transfusion, and death.^3,6-10

SA is an attractive treatment option for patients with rheumatoid arthritis (RA), as the effects of pain on these patients are greater in the shoulder joint than in any other joint.¹¹ Patients with RA pose unique orthopedic surgical challenges, including any combination of decreased bone mineralization, poor capsular tissue integrity, and osteonecrosis.^3,12 In addition, RA patients may be taking immunosuppressive medications that have severe side effects, and they may require multiple surgeries.^12,13 These factors predispose patients with RA to complications that include infection and wound dehiscence.^3,5,12-14

The complex nature of RA has prompted investigators to examine outcome measures in this patient group. Hambright and colleagues³ used the Nationwide Inpatient Sample (NIS) to examine perioperative outcomes in RA patients who underwent TSA between 1988 and 2005.³ They found that TSA patients with RA had shorter and less costly hospital stays and were more likely to have a routine discharge.³ Using the same patient population drawn from the period 2006–2011, we conducted a study to determine if this unexpected trend persists as the number of TSAs and quality of postoperative care continue to increase. Given the potential for anemia of chronic disease and the systemic inflammatory nature of RA, we hypothesized that the perioperative complication profile of RA patients would be worse than that of non-RA patients.

Materials and Methods

NIS data were acquired for the period 2006–2011. The NIS is the largest publicly available all-payer inpatient database, with a random 20% sample of about 1000 US hospitals accounting for 7 to 8 million inpatient stays. The database supplies weights used to estimate national totals, at about 35 million inpatient visits per year. NIS inpatient data are limited to the operative index. Postdischarge information is not available. The NIS is managed by the Healthcare Cost and Utilization Project, which is sponsored by the Agency for Healthcare Research and Quality. The quality of NIS data is assessed and validated by an independent contractor. NIS data have been widely used to examine perioperative outcomes.^15-17

NIS data cover patient and hospital demographics, hospital length of stay (LOS), discharge status, payer information, charges, and perioperative outcomes and procedure/diagnosis codes (ICD-9; International Classification of Diseases, Ninth Revision¹⁸).

As our Institutional Review Board (IRB) reviewed the database and determined the project was not human subject research, IRB involvement was not required. This study paralleled successful efforts with similar RA and non-RA patients who had shoulder and elbow surgery.^3,19 SA patients were identified by ICD-9 procedure code 81.80, but this code does not specify whether the prosthesis was unconstrained, semiconstrained, or constrained. ICD-9 coding also does not specify whether the TSA was traditional or reverse. Patients with RA were identified by ICD-9 diagnosis codes 714.0, 714.1, and 714.2. Patients without one of these codes were placed in the non-RA cohort. Patients with codes associated with pathologic fractures secondary to metastatic cancer or bone malignant neoplasm as a secondary or primary diagnosis and patients who had revision surgery indicated by code 81.83 were excluded, as they have a disproportionately higher comorbidity burden.

After each cohort was defined, demographic data (age, sex, race, income quartile based on ZIP postal code) were compared, as were data on primary payer, hospital demographics, LOS (≤5 days, defined as perioperative index), discharge type, inflation-adjusted charges in 2014 dollars based on the Consumer Price Indexes (http://www.bls.gov/cpi/), and mortality. Perioperative complications—respiratory, gastrointestinal, genitourinary, accidental puncture/laceration, central nervous system, wound dehiscence, device-related (including embolism, fibrosis, hemorrhage, pain, stenosis, or thrombus caused by any device, implant, or graft), cardiac, hematoma/seroma, acute respiratory distress syndrome, postoperative shock, VTE, postoperative infection complications, and intraoperative transfusions—were considered using ICD-9 codes (996.X-999.X and 99.X, respectively).²⁰ Although commonly used to determine perioperative comorbidity burden using ICD-9 coding, the modified Charlson index was not considered because RA is a component of the index and would therefore bias the variable.^3,21

Statistical analyses, including χ² tests and 2-sample t tests, were performed for categorical and continuous variables, respectively. P < .05 was considered significant. Fisher exact test was used for cohorts with fewer than 5 occurrences. Multivariate logistic regression models were then calculated to determine the effect of RA on different outcomes and complications, with age, race, sex, hospital region, hospital type, number of hospital beds, primary payer, and hospital ownership as covariates. Statistical analyses were performed using the R statistical programming language.22

Results

Of the 34,970 patients who underwent SA between 2006 and 2011, 1674 (4.8%) had a diagnosis of RA and 33,296 (95.2%) did not. On average, patients with RA tended to be younger than patients without RA (66.4 vs 69.1 years; P < .001), and a larger percentage of RA patients were female (75.5% vs 54.4%; P < .001). Compared with non-RA patients, RA patients comprised a different ethnic group and had a different expected primary payer (P < .001). SA patients with and without RA did not differ in income quartile based on ZIP code, total number of hospital beds, hospital region, or hospital teaching status (P = .34, .78, .59, and .82, respectively) (Table 1).

LOS was significantly (P < .001) statistically longer for RA patients (2.196 days) than for non-RA patients (2.085 days). RA patients were significantly less likely to be discharged home (63.0% vs 67.6%; P < .001). (Routine discharge was defined as discharge home, whereas nonroutine discharge was defined as discharge to a short-term hospital, skilled nursing facility, intermediate care, another type of facility, home health care, against medical advice, or death.) In addition, inflation-adjusted charges associated with SA were significantly higher (P = .018) for RA patients ($54,284) than for non-RA patients ($52,663) (Table 1).

Regarding the rates of complications that occurred during the perioperative index, there were no significant differences between RA and non-RA cohorts. These complications included respiratory, gastrointestinal, genitourinary, accidental puncture/laceration, central nervous system, wound dehiscence, device-related, cardiac, hematoma/seroma, acute respiratory distress syndrome, postoperative shock, VTE, and postoperative infection (Table 2). In addition, there was no significant difference in mortality between the groups (P = .48).

In TSA, blood transfusions were more likely (P < .001) to be given to RA patients (9.00%) than to non-RA patients (6.16%). Multivariate regression analyses were performed with age, race, sex, hospital region, hospital type, number of hospital beds, primary payer, and hospital ownership as covariates. These analyses revealed that transfusion (P < .001), discharge type (P = .002), total inflation-adjusted charges (P < .001), and LOS (P = .047) remained significant (Table 3).

Discussion

Large national databases like NIS allow study of uncommon medical occurrences and help delineate risks and trends that otherwise might be indeterminable. Although it has been suggested that patients with RA may have poorer long-term outcomes after SA, the perioperative risk profile indicates that TSA is well tolerated in RA patients during the operative index.^3,23-25

The data on this study’s 34,970 patients, drawn from the period 2006–2011, demonstrated no significant differences in safety profile with respect to the 14 perioperative complications and outcomes examined, except blood transfusion rate. Rates of postoperative infection (RA, 0.24%; non-RA, 0.14%; P = .303), VTE (RA, 0.30%; non-RA, 0.25%; P = .905), and transfusion (RA, 9.00%; non-RA, 6.16%; P < .001) are of particular interest because of the severity of these situations.

Postoperative infection is a potentially serious complication and often occurs secondary to diabetes, RA, lupus erythematosus, prior surgery, or a nosocomial or remote source.¹ The often costly treatment options include antibiotic suppression, irrigation and debridement with implant retention, 1-stage exchange with antibiotic-impregnated cement fixation, staged reimplantation, resection arthroplasty, arthrodesis, and amputation.¹ The overall 0.14% infection rate determined in this study is lower than the 0.7% reported for SA patients in the literature.¹ Given the nature of the NIS database, this rate underestimates the true postoperative infection rate, as any infection that occurred after the perioperative period is not captured.²⁶ The present study’s perioperative infection rates (RA, 0.24%; non-RA, 0.14%) for the period 2006–2011 are comparable to the rates (RA, 0.17%; non-RA, 0.24%) reported by Hambright and colleagues3 for the same patient population over the preceding, 18-year period (1988–2005) and similarly do not significantly differ between groups. Although infection is uncommon in the immediate perioperative period, the ICD-9 codes used refer specifically to infection resulting from surgery and do not represent concomitant infection.

VTEs, which include PEs and DVTs, are rare but potentially life-threatening surgical complications.^27,28 Mechanical prophylaxis and chemical prophylaxis have been recommended for major orthopedic surgery, particularly lower extremity surgery, such as total hip arthroplasty (THA) and total knee arthroplasty (TKA).^28,29 In the present study, VTE rates were low, 0.30% (RA) and 0.25% (non-RA), and not significantly different in bivariate or multivariate analyses. These rates are comparable to those found in other national-database SA studies.²⁸ VTEs that occur outside the index hospital admission are not captured in this database. Therefore, the rates in the present study may be lower than the true incidence after SA. Mortality secondary to VTE usually occurs within 24 hours but may occur up to 90 days after surgery. DVT rates, on the other hand, are difficult to evaluate because of differences in screening practices.^27,28,30,31

That RA patients were more likely than non-RA patients to receive perioperative blood transfusions supports prior findings that SA patients with RA were more likely than SA patients with osteoarthritis (OA) to receive perioperative blood transfusions.⁸ RA patients have been shown to have high rates of anemia of chronic disease, ranging from 22% to 77%.³² During joint replacement, these patients often require transfusions.^32,33 However, these findings differ from prior findings of no differences between RA and non-RA patients in the same patient population during the period 1988–2005.³ This difference may be a product of the constantly changing transfusion guidelines and increased use; transfusion rates increased 140% between 1997 and 2007, making transfusions the fastest growing common procedure in the United States during that time.³⁴ There was no difference between RA and non-RA patients in household income (as determined by ZIP code analysis), number of hospital beds, hospital region, or hospital teaching status. Compared with non-RA patients, RA patients were more likely to be younger, female, and of a difference race and to have a different expected primary payer (P < .001).These findings are consistent with previous findings in the literature.³ In the present SA study, however, RA patients were more likely than non-RA patients to have longer LOS, higher inflation-adjusted hospital charges, and nonroutine discharge. These findings deviate from those of the study covering the preceding 18 years (1988–2005).³ Despite the findings of a changing environment of care for RA patients, by Hambright and colleagues³ and Weiss and colleagues,³⁵ the trend appears to have shifted. Both groups had shorter average LOS than either group from the preceding 18 years.³ Although statistically significant in bivariate analysis, the difference in LOS between the 2 groups differed by an average of 0.11 day (2 hours 24 minutes) and was not clinically relevant.

In addition, the higher charges for patients with RA represent a deviation from the preceding 18 years.³ Other studies have also shown that RA is associated with increased cost in TSA.³⁶ Patients with RA often have rotator cuff pathology, indicating reverse SA may be used more frequently.^37,38 The increased implant cost associated with reverse SA may account for the increased costs in RA patients.³⁹ As mentioned, TSA type is not captured in the NIS database. In addition, that RA patients were less likely than non-RA patients to have routine discharge may indicate RA cases are more complex because of their complications.^1,5,14,40 A recent study of complications in RA patients (1163 who underwent THA, 2692 who underwent TKA) found that THA patients with RA were significantly more likely than THA patients with OA to dislocate, and TKA patients with RA were significantly more likely than TKA patients with OA to develop an infection after surgery.⁴¹ Postoperative dislocation has been shown to increase hospital costs in other orthopedic procedures.⁴² Also, during TSA, patients with RA are more likely than patients with OA to receive intraoperative blood transfusions.⁸ These complications—combined with the fact that RA is a chronic, progressive, systemic inflammatory disease that can affect soft tissue and blood vessel wall healing and is associated with medications having potential side effects—could contribute to the apparent increased hospital charges and LOS.^3,12,13,43 Factors that include surgeon preference, impact of primary payer, and hospital practice may also affect final charges. Total charges in the NIS database include administrative fees, hospital costs, device-related costs, operating room costs, and ancillary staff costs. Total charges do not include professional fees and differ from the total cost that represents the amount reimbursed by the payer. Charges tend to correlate with but overestimate the total costs.⁴⁴

This study had several important limitations. As mentioned, only events that occur during the operative admission are captured in the NIS database, and thus postoperative complications or serious adverse events that lead to readmission cannot be identified. In addition, outpatient TSAs are not captured in the NIS database, and thus inclusion of only inpatient procedures yields higher average LOS and total charges.⁴⁵ Given the limited granularity of ICD-9 coding, this study could not determine RA severity, estimated blood loss, length of surgery, complication severity, type of TSA procedure/prosthesis, or cause of death. Although commonly used to determine comorbidity burden, the modified Charlson index could not be used, and therefore could not be entered as a covariate in multivariate analysis. Furthermore, the NIS database does not include imaging or patient-reported outcomes information, such as improvements in pain or function, which are of crucial importance in considering surgery.

Conclusion

Our findings corroborated findings that the demographics and the perioperative safety profile for TSA were similar for patients with and without RA. The risk for complications or death in the perioperative period was low. Compared with non-RA patients, RA patients had significantly higher charges and longer LOS and were less likely to be discharged home after surgery. The 0.11-day difference in LOS, though statistically significant, was not clinically relevant. These findings differ from those for the preceding, 18-year period (1988–2005). Future research should focus on the causes of these changes.