Brian T. Easton, MD

ABSTRACT

BACKGROUND: Although suturing is the most common method of wound closure, it requires injection of an anesthetic, is time-consuming, carries the risk of a needle stick to the clinician, and requires a return visit for suture removal. Tissue adhesives applied to skin hold wound edges together and usually slough off within 5 to 10 days without requiring removal. This study compared an easy-to-use, topical adhesive, octylcyanoacrylate to standard wound closure methods (sutures, staples, or adhesive tapes).

POPULATION STUDIED: This multicenter study enrolled 814 patients (924 wounds) from emergency departments, an urgent care clinic, outpatient surgery centers, and private practices. Patients had to be in good health (without insulin-dependent diabetes, peripheral vascular disease, bleeding diathesis, multiple trauma, or possibility of keloid formation) and older than 1 year. Wounds that could be repaired with 5-0 or smaller suture were eligible. Bites, puncture wounds, infected wounds, decubitus ulcers, stellate lacerations, wounds located on the vermilion border of the lip or mucosa, and wounds over flexor or extensor surfaces or near the eye were excluded. Seventy percent of the subjects were white and the mean wound length was approximately 2 cm.

STUDY DESIGN AND VALIDITY: Eligible subjects were randomized using computer-generated random numbers and opaque envelopes (concealed allocation assignment) to receive either octylcyanoacrylate or standard wound closure. Although blinding patients or physicians to treatment was not possible, personnel assessing the outcomes at 3 months were blind to treatment group assignment. No baseline differences were noted between treatment groups. Patients were analyzed in the groups to which they were randomized (intention-to-treat analysis) and follow-up was complete (96% at 1 week and 94% at 3 months). This study had adequate power (90%) to detect a difference (10%) in the proportion of patients with optimal wound appearance, but was underpowered to detect a small difference in infection or dehiscence rates.

OUTCOMES MEASURED: The primary outcome measured was cosmetic wound appearance. Other outcomes included wound closure time, infection rates, and dehiscence rates.

RESULTS: At 3 months no difference was noted in the percentage of wounds with optimal appearance (octylcyanoacrylate, 82% vs standard wound closure, 83%; P = .67). Although wound closure with octylcyanoacrylate was faster than with standard wound closure (mean 2.9 vs 5.2 minutes, P < .001), this small difference is probably not clinically significant. Infection rates were similar (octylcyanoacrylate, 2.1% vs standard wound closure, 0.7%; P = .09), as were dehiscence rates (octylcyanoacrylate, 1.6% vs standard wound closure, 0.9%; P = .67).

ABSTRACT

BACKGROUND: Although suturing is the most common method of wound closure, it requires injection of an anesthetic, is time-consuming, carries the risk of a needle stick to the clinician, and requires a return visit for suture removal. Tissue adhesives applied to skin hold wound edges together and usually slough off within 5 to 10 days without requiring removal. This study compared an easy-to-use, topical adhesive, octylcyanoacrylate to standard wound closure methods (sutures, staples, or adhesive tapes).

POPULATION STUDIED: This multicenter study enrolled 814 patients (924 wounds) from emergency departments, an urgent care clinic, outpatient surgery centers, and private practices. Patients had to be in good health (without insulin-dependent diabetes, peripheral vascular disease, bleeding diathesis, multiple trauma, or possibility of keloid formation) and older than 1 year. Wounds that could be repaired with 5-0 or smaller suture were eligible. Bites, puncture wounds, infected wounds, decubitus ulcers, stellate lacerations, wounds located on the vermilion border of the lip or mucosa, and wounds over flexor or extensor surfaces or near the eye were excluded. Seventy percent of the subjects were white and the mean wound length was approximately 2 cm.

STUDY DESIGN AND VALIDITY: Eligible subjects were randomized using computer-generated random numbers and opaque envelopes (concealed allocation assignment) to receive either octylcyanoacrylate or standard wound closure. Although blinding patients or physicians to treatment was not possible, personnel assessing the outcomes at 3 months were blind to treatment group assignment. No baseline differences were noted between treatment groups. Patients were analyzed in the groups to which they were randomized (intention-to-treat analysis) and follow-up was complete (96% at 1 week and 94% at 3 months). This study had adequate power (90%) to detect a difference (10%) in the proportion of patients with optimal wound appearance, but was underpowered to detect a small difference in infection or dehiscence rates.

OUTCOMES MEASURED: The primary outcome measured was cosmetic wound appearance. Other outcomes included wound closure time, infection rates, and dehiscence rates.

RESULTS: At 3 months no difference was noted in the percentage of wounds with optimal appearance (octylcyanoacrylate, 82% vs standard wound closure, 83%; P = .67). Although wound closure with octylcyanoacrylate was faster than with standard wound closure (mean 2.9 vs 5.2 minutes, P < .001), this small difference is probably not clinically significant. Infection rates were similar (octylcyanoacrylate, 2.1% vs standard wound closure, 0.7%; P = .09), as were dehiscence rates (octylcyanoacrylate, 1.6% vs standard wound closure, 0.9%; P = .67).

ABSTRACT

BACKGROUND: Although suturing is the most common method of wound closure, it requires injection of an anesthetic, is time-consuming, carries the risk of a needle stick to the clinician, and requires a return visit for suture removal. Tissue adhesives applied to skin hold wound edges together and usually slough off within 5 to 10 days without requiring removal. This study compared an easy-to-use, topical adhesive, octylcyanoacrylate to standard wound closure methods (sutures, staples, or adhesive tapes).

POPULATION STUDIED: This multicenter study enrolled 814 patients (924 wounds) from emergency departments, an urgent care clinic, outpatient surgery centers, and private practices. Patients had to be in good health (without insulin-dependent diabetes, peripheral vascular disease, bleeding diathesis, multiple trauma, or possibility of keloid formation) and older than 1 year. Wounds that could be repaired with 5-0 or smaller suture were eligible. Bites, puncture wounds, infected wounds, decubitus ulcers, stellate lacerations, wounds located on the vermilion border of the lip or mucosa, and wounds over flexor or extensor surfaces or near the eye were excluded. Seventy percent of the subjects were white and the mean wound length was approximately 2 cm.

STUDY DESIGN AND VALIDITY: Eligible subjects were randomized using computer-generated random numbers and opaque envelopes (concealed allocation assignment) to receive either octylcyanoacrylate or standard wound closure. Although blinding patients or physicians to treatment was not possible, personnel assessing the outcomes at 3 months were blind to treatment group assignment. No baseline differences were noted between treatment groups. Patients were analyzed in the groups to which they were randomized (intention-to-treat analysis) and follow-up was complete (96% at 1 week and 94% at 3 months). This study had adequate power (90%) to detect a difference (10%) in the proportion of patients with optimal wound appearance, but was underpowered to detect a small difference in infection or dehiscence rates.

OUTCOMES MEASURED: The primary outcome measured was cosmetic wound appearance. Other outcomes included wound closure time, infection rates, and dehiscence rates.

RESULTS: At 3 months no difference was noted in the percentage of wounds with optimal appearance (octylcyanoacrylate, 82% vs standard wound closure, 83%; P = .67). Although wound closure with octylcyanoacrylate was faster than with standard wound closure (mean 2.9 vs 5.2 minutes, P < .001), this small difference is probably not clinically significant. Infection rates were similar (octylcyanoacrylate, 2.1% vs standard wound closure, 0.7%; P = .09), as were dehiscence rates (octylcyanoacrylate, 1.6% vs standard wound closure, 0.9%; P = .67).

Osteoarthritis (OA) is a chronic and progressive disease in which damage is done to the joint and surrounding tissue. In the United States, OA is found in 6% of people older than 30 years and in 50% of those older than 60 years.^1,2 It is the most common cause of disability in the United States and leads to considerable costs due to medical and surgical interventions and frequent absences from work.^3-5 OA accounts for 2% of all visits to family physicians and is the 10th leading diagnosis encountered.⁶ Thus, a family physician can expect to have 2 or 3 patient encounters per week in which OA is one of the diagnoses. With the growing elderly population, this burden is likely to increase.

Pathophysiology

In OA, the smooth surface of hyaline cartilage develops irregularities because of alterations at the cellular level and gross mechanical forces.² The role of inflammation has been debated in recent years, and its exact role is unknown. As OA progresses, the nearby bone remodels and forms further joint irregularities and osteophytes. These changes lead to narrowing of the joint space, and in some cases, chronic synovitis. Clinically, this causes pain, restricted movement, and periarticular muscle wasting. Treatment is aimed at these symptoms and structural abnormalities. The joints most commonly affected include the knees, hips, cervical and lumbosacral spine, distal interphalangeal (DIP) joints (producing Heberden nodes), proximal interphalangeal (PIP) joints (producing Bouchard nodes), and the first carpometacarpal joints of the hand. It is not known why these changes occur in some people and not in others. Epidemiologic studies and a recent sibling study raise the question of genetic influences in OA of the hip.^7,8

Diagnosis

The diagnosis of OA is made on the basis of clinical and radiographic features. Few studies have compared a diagnostic test or strategy with a gold standard. Unfortunately, we do not have any good data about the usefulness of individual history and physical examination elements for diagnosing OA. Radiographs suggesting the diagnosis of OA (joint space narrowing, presence of osteophytes, irregular joint surfaces, sclerosis of subchondral bone, or bony cysts) must be closely correlated with clinical symptoms. According to epidemiologic surveys, only one half of patients with radiographic changes of OA of the knee complain of persistent pain.⁹ Classification criteria for OA of the knee, hip, and hands have been developed; they are outlined in Tables 1 Table 2 through Table 3.^10,13

Although these criteria have limitations, they are becoming the standard for defining these types of OA and have been adapted by the American College of Rheumatology.^11-13 Currently, there are no criteria for diagnosing OA of the back. When evaluating a patient with joint pain, other diagnoses must be considered (ie, rheumatoid arthritis, gout, pseudogout, septic arthritis, bursitis, and tendonitis).

Treatment

Most of the treatments of OA address the symptoms rather than the cause of the disease. The short-term goal is to decrease pain; long-term goals are to improve functioning and slow progression of disease. Table 4 provides a summary of treatment options.

Exercise

A systematic review of 12 randomized controlled trials (RCTs) showed beneficial effects of exercise therapy in patients with mild to moderate OA of the knee and, to a lesser extent, the hip.¹⁴ Benefits included improvements in pain, self-reported disability, walking performance, and the patient’s global assessment of symptoms. Insufficient evidence was available to recommend one type of exercise program over another. Exercise interventions included aerobic exercises, strength training, range of motion exercises, and fitness walking. Exercise programs were conducted as individuals or groups, supervised or home-based. Although this review was limited by the small number of good studies, recommending exercise for OA may improve patients’ symptoms and add other health benefits.

Physical Therapy

A recent RCT compared 4 weeks of manual physical therapy plus a supervised knee exercise program with sham ultrasound for treatment of OA of the knee.¹⁵ By 8 weeks, 6-minute walk distances had improved by 13% in the treatment group compared with no change in the placebo group; osteoarthritis index scores had improved 56% over baseline compared with 15% in the placebo group. One year following therapy 5% of the treated group had undergone knee arthroplasty compared with 20% of the untreated group (number needed to treat = 7). Given these clinically important improvements, physical therapy should be an early choice in the treatment of OA.

A systematic review of transcutaneous electrical nerve stimulation (TENS) for treatment of OA of the knee found that this noninvasive modality offered significant pain relief.¹⁶ Both high-frequency and strong burst mode TENS showed significant improvement in pain relief when used for 4 weeks or more. Also, the acupuncture-like TENS improved pain relief, stiffness, and walking time in a 2-week placebo-controlled trial.

Although spa therapy and low-level laser therapy may show some benefit, the true effectiveness cannot be determined, since the trials were small and of poor quality.^17,18 No benefit was found in the one well-done RCT for ultrasound.¹⁹

Patient Education

Group patient education programs teach patients how to manage their disease. A 4-year longitudinal study using the Arthritis Self-Management Program (6 2-hour sessions), suggests that health education may decrease pain and visits to physicians.^20,21 Estimated 4-year savings were $189 per patient.²¹ Telephone-based interventions have also been studied. A randomized trial of telephone contact for patients with OA found statistically significant differences in pain reduction according to the Arthritis Impact Measurement Scales at a very low cost.^22,23 However, these differences may not be clinically significant as the groups differed by less than 1 point on a 10-point scale. Interestingly, educating patients during regularly scheduled appointments had no benefit and even resulted in worsening physical functioning.²² A systematic review of patient education interventions found a nonsignificant trend toward benefit using patient education compared with nonsteroidal anti-inflammatory drugs (NSAIDs).²⁴ Despite the lack of convincing evidence from randomized trials, the possible benefits of educating our patients while using other therapies may outweigh the cost of the time involved, particularly if education can be done in a group setting.

Medication

Acetaminophen (4 g per day) has been shown in randomized trials to reduce pain in OA of the knee by approximately 30%.^25,26 This improvement was seen at 4 weeks and at 2 years. Another study found the combination of codeine plus acetaminophen to be significantly better in reducing pain in patients with OA of the hip than acetaminophen alone, although one third of the patients receiving the combination discontinued therapy because of side effects.²⁷ Narcotic analgesics alone have been shown in a randomized trial to be more effective in reducing pain than placebo but have not been adequately tested against acetaminophen or NSAIDs.²⁸

NSAIDs also produce a 30% reduction in pain caused by OA.^25,26 Two systematic reviews of OA concluded that there is no reliable evidence suggesting that any NSAID is more efficacious than the others in treating OA of the hip and knee.^29,30 A separate meta-analysis found ibuprofen to have the lowest risk of side effects among the nonselective NSAIDs, a finding supported by the Committee on Safety in Medicine.³¹ For this reason, ibuprofen is recommended as the first-line nonselective NSAID.

Acetaminophen and NSAIDs have been found to be equally efficacious when compared in randomized controlled trials.^25,26 According to a meta-analysis, the quality-of-life measures were similar between the 2 groups, despite a greater improvement in both pain at rest and pain on motion in the NSAID-treated patients.³¹ Interestingly, 2 recent surveys showed that patients with OA prefer NSAIDs to acetaminophen.²⁷ However, based on the results of clinical trials, safety profiles, and cost issues acetaminophen should continue to be used as the first-line agent.²⁷

The selective cyclooxygenase-2 (COX-2) inhibitors are effective in relieving the pain caused by OA.^32-35 The COX-2 inhibitors have been shown to minimize fecal blood loss and produce fewer endoscopically proven gastrointestinal erosions/ulcerations than the nonselective inhibitors.^36-39 This, however, is disease-oriented evidence and does not help us know about the more important clinical outcomes.

Using the STEPS approach (Safety, Tolerability, Efficacy, Price, Simplicity) to compare COX-2 inhibitors and nonselective NSAIDs is one way to help guide the choice of NSAIDs in the management of OA. A systematic review of a COX-2 inhibitor found the differences in adverse gastrointestinal events including perforations, symptomatic ulcers, and clinically significant bleeding episodes to be small when compared with the older NSAIDs (incidence = 1.3% vs 1.8%).^36-39 Thus, 200 patients would have to receive a COX-2 inhibitor instead of a nonselective NSAID for 1 year to prevent 1 clinically significant bleeding episode (number needed to harm = 200). These safety differences seem small. However, the safety benefits of the COX-2 inhibitors increase when patients have multiple risk factors for adverse gastrointestinal events (age Ž65, history of peptic ulcer disease or gastrointestinal bleeding, use of oral glucocorticoids or anticoagulants, and comorbid medical conditions).²⁷ Tolerability can be measured by looking at dropout rates. A meta-analysis found that the percentage of patients withdrawing from studies because of gastrointestinal adverse events was lower with the COX-2 inhibitor than with the nonselective NSAIDs (odds ratio = 0.59; 95% confidence interval, 0.52-0.67).³⁷ Tolerability favors the COX-2 agents. Meta-analyses of trials comparing the efficacy of COX-2 inhibitors and nonselective NSAIDs in the treatment of OA show no differences between them.^32-35 The average wholesale price of a month’s supply of a COX-2 inhibitor is approximately $73, and a month’s supply of a generic nonselective NSAID is $15. Nonselective NSAIDs are favored when looking at cost. While the nonselective NSAIDs require multiple dosing throughout the day, the COX-2 inhibitors can be given once daily, thus making the COX-2 inhibitors easier to use for patients. Overall, because of equal efficacy, minimal safety differences, and low cost, the older NSAIDs are recommended for low-risk patients, while the COX-2 inhibitors are recommended for high-risk patients because of increased safety and tolerability. (Of note, the COX-2 inhibitors have not been compared with acetaminophen, which is equally as efficacious as the nonselective NSAIDs and also has minimal side effects.)

Topical NSAIDs are more effective than placebo in treating the pain of OA, according to a systematic review.²⁷ Also, few side effects were noted. A small randomized trial (n=70) of 0.025% capsaicin cream applied topically 4 times daily for 4 weeks reduced pain by 8% compared with placebo in patients with OA of the knee over a 4-week period.¹⁹ Further studies are needed to confirm this finding.

Complementary and Alternative Medicine

A recent meta-analysis found that glucosamine produces a modest to large improvement in pain relief and functional outcomes compared with placebo.⁴⁰ A systematic review of controlled trials found that glucosamine performed equally well or better than NSAIDs.⁴¹ Most trials used a dose of 1500 mg glucosamine per day and addressed OA of the knee. The benefits may be overestimated because of the presence of publication bias. Because glucosamine is not regulated by the United States Food and Drug Administration, preparations and effectiveness may vary. However, it is very well tolerated and costs approximately $20 per month. Based on these results, glucosamine should be considered in the treatment of OA.

Although chondroitin has not been as well studied as glucosamine, a meta-analysis found evidence of effectiveness.⁴⁰ Most studies used 800 to 1200 mg of chondroitin daily for OA of the knee. An added benefit of using both together has not been proved.

Another therapy supported by good-quality evidence is the use of avocado/soybean unsaponifiables. A systematic review of herbal therapy concluded that 300 mg daily of avocado/soybean unsaponifiables can provide long-term symptomatic relief, particularly for patients with chronic stable OA of the hip.⁴² It may also help patients reduce their intake of NSAIDs.

The most recent systematic review of acupuncture for OA of the knee concluded that real acupuncture is better than sham acupuncture in treating pain but not function.⁴³ Four of the studies in the meta-analysis were of high quality, although none assessed the patient’s rating of global improvement. Acupuncture should be considered in the treatment of OA of the knee.

Less convincing therapies include therapeutic touch and electromagnetic fields. Therapeutic touch aims at manipulating a person’s energy system to bring the body system back into balance. A well-done single-blinded randomized trial compared therapeutic touch with mock therapeutic touch for OA of the knee.⁴⁴ The treatment group had a statistically significant decrease in pain and improved function compared with the placebo group in most of the outcomes measured. However, the differences were small (0.5- to 1.4-point difference on a 10-point scale) and may not be clinically important. Further studies are needed to demonstrate a clinical impact and to help us know how to incorporate this into our practices. A small study of 27 patients compared pulsed electromagnetic fields (PEMF) and placebo for the treatment of OA of the knee. Although, this double-blind randomized trial showed a 31% reduction in overall pain compared with sham PEMF, larger studies are needed before widespread use of this therapy is warranted.¹⁹

Injections

Intra-articular steroids for OA of the knee have been recommended by the American College of Rheumatology when an effusion and local signs of inflammation are present.⁴⁵ In 3 studies comparing steroid injections to placebo, 2 found that steroids were effective over 1 to 2 weeks, and 1 trial showed no difference. None of the trials showed any long-term benefit of steroid injections.⁴⁶

A systematic review of injectable biologic agents (glycosaminoglycans, mostly hyaluronic acid) found that in 6 of the 8 randomized trials for OA of the knee they were superior to placebo.⁴⁶ Treatment regimens varied, making it difficult to determine optimal duration of treatment and route of administration (intramuscular vs intra-articular). A comparison of hyaluronic acid injections with naproxen found that they were equally effective in improving pain relief and function, although the injections had fewer side effects.⁴⁷ Dextrose in a 10% solution is another biologic agent used intra-articularly; however, a well-designed study found that it was no more effective than placebo for OA of the knee.⁴⁸ Injectable glycosminoglycans, such as hyaluronic acid, are recommended in the treatment of OA of the knee, although the cost must be considered. The most effective regimen has not been determined.

Surgery

Surgery is reserved for patients with severe disease of the hip or knee that is not controlled by less invasive measures. Total hip arthroplasty has been shown to improve quality of life in patients with advanced hip OA.⁴⁹ Also, a cost-effectiveness analysis showed that total hip arthroplasty can be cost-effective in improving quality-adjusted life expectancy in the short term and long term.⁵⁰ Total joint arthroplasty for OA of the knee has shown similar benefits in pain relief and functional improvement.^45,46 It is not surprising that OA of the hip and knee is the most common indication for elective total hip and knee arthroplasty in the United States.⁵¹

Other Recommendations

On the basis of lower-quality studies, the American College of Rheumatology also recommends the following for individual patients: a weight-loss program, an occupational therapy evaluation, bracing and footwear, joint lavage of the knee, arthroscopic debridement of the knee, and osteotomy of the knee or hip.^45,52 The benefits from these interventions have not been well studied. The costs and benefits of these interventions must be weighed before recommending them to individual patients. Most trials have been conducted in patients with OA of the knee or hip. We do not know if applying treatment modalities to other joints (back and hands) will produce similar results. Also, most trials have been placebo controlled, and we do not know whether some treatments would add benefit to the first-line therapy of acetaminophen and NSAIDs.

Prognosis

OA is a chronic progressive disease of the joints that leads to increased pain and decreased functioning. Two longitudinal studies examined the natural course of OA of the knee in 191 patients. They showed that 56% of patients had no change, and 44% worsened over 15 years.^53-55 A study of OA of the hip showed similar trends except that a small group (7%) experienced improvement over 10 years.⁵⁶ We do not know whether current treatment options can improve the long-term prognosis of OA. Epidemiologic studies suggest that increasing age, obesity, family history, occupation, and joint injury are risk factors for OA.^6,53 Whether modifying these risk factors will decrease symptoms or slow the course of the disease is not known. The progressive nature of OA makes treatment a challenge for the patient and clinician.

Osteoarthritis (OA) is a chronic and progressive disease in which damage is done to the joint and surrounding tissue. In the United States, OA is found in 6% of people older than 30 years and in 50% of those older than 60 years.^1,2 It is the most common cause of disability in the United States and leads to considerable costs due to medical and surgical interventions and frequent absences from work.^3-5 OA accounts for 2% of all visits to family physicians and is the 10th leading diagnosis encountered.⁶ Thus, a family physician can expect to have 2 or 3 patient encounters per week in which OA is one of the diagnoses. With the growing elderly population, this burden is likely to increase.

Pathophysiology

In OA, the smooth surface of hyaline cartilage develops irregularities because of alterations at the cellular level and gross mechanical forces.² The role of inflammation has been debated in recent years, and its exact role is unknown. As OA progresses, the nearby bone remodels and forms further joint irregularities and osteophytes. These changes lead to narrowing of the joint space, and in some cases, chronic synovitis. Clinically, this causes pain, restricted movement, and periarticular muscle wasting. Treatment is aimed at these symptoms and structural abnormalities. The joints most commonly affected include the knees, hips, cervical and lumbosacral spine, distal interphalangeal (DIP) joints (producing Heberden nodes), proximal interphalangeal (PIP) joints (producing Bouchard nodes), and the first carpometacarpal joints of the hand. It is not known why these changes occur in some people and not in others. Epidemiologic studies and a recent sibling study raise the question of genetic influences in OA of the hip.^7,8

Diagnosis

The diagnosis of OA is made on the basis of clinical and radiographic features. Few studies have compared a diagnostic test or strategy with a gold standard. Unfortunately, we do not have any good data about the usefulness of individual history and physical examination elements for diagnosing OA. Radiographs suggesting the diagnosis of OA (joint space narrowing, presence of osteophytes, irregular joint surfaces, sclerosis of subchondral bone, or bony cysts) must be closely correlated with clinical symptoms. According to epidemiologic surveys, only one half of patients with radiographic changes of OA of the knee complain of persistent pain.⁹ Classification criteria for OA of the knee, hip, and hands have been developed; they are outlined in Tables 1 Table 2 through Table 3.^10,13

Although these criteria have limitations, they are becoming the standard for defining these types of OA and have been adapted by the American College of Rheumatology.^11-13 Currently, there are no criteria for diagnosing OA of the back. When evaluating a patient with joint pain, other diagnoses must be considered (ie, rheumatoid arthritis, gout, pseudogout, septic arthritis, bursitis, and tendonitis).

Treatment

Most of the treatments of OA address the symptoms rather than the cause of the disease. The short-term goal is to decrease pain; long-term goals are to improve functioning and slow progression of disease. Table 4 provides a summary of treatment options.

Exercise

A systematic review of 12 randomized controlled trials (RCTs) showed beneficial effects of exercise therapy in patients with mild to moderate OA of the knee and, to a lesser extent, the hip.¹⁴ Benefits included improvements in pain, self-reported disability, walking performance, and the patient’s global assessment of symptoms. Insufficient evidence was available to recommend one type of exercise program over another. Exercise interventions included aerobic exercises, strength training, range of motion exercises, and fitness walking. Exercise programs were conducted as individuals or groups, supervised or home-based. Although this review was limited by the small number of good studies, recommending exercise for OA may improve patients’ symptoms and add other health benefits.

Physical Therapy

A recent RCT compared 4 weeks of manual physical therapy plus a supervised knee exercise program with sham ultrasound for treatment of OA of the knee.¹⁵ By 8 weeks, 6-minute walk distances had improved by 13% in the treatment group compared with no change in the placebo group; osteoarthritis index scores had improved 56% over baseline compared with 15% in the placebo group. One year following therapy 5% of the treated group had undergone knee arthroplasty compared with 20% of the untreated group (number needed to treat = 7). Given these clinically important improvements, physical therapy should be an early choice in the treatment of OA.

A systematic review of transcutaneous electrical nerve stimulation (TENS) for treatment of OA of the knee found that this noninvasive modality offered significant pain relief.¹⁶ Both high-frequency and strong burst mode TENS showed significant improvement in pain relief when used for 4 weeks or more. Also, the acupuncture-like TENS improved pain relief, stiffness, and walking time in a 2-week placebo-controlled trial.

Although spa therapy and low-level laser therapy may show some benefit, the true effectiveness cannot be determined, since the trials were small and of poor quality.^17,18 No benefit was found in the one well-done RCT for ultrasound.¹⁹

Patient Education

Group patient education programs teach patients how to manage their disease. A 4-year longitudinal study using the Arthritis Self-Management Program (6 2-hour sessions), suggests that health education may decrease pain and visits to physicians.^20,21 Estimated 4-year savings were $189 per patient.²¹ Telephone-based interventions have also been studied. A randomized trial of telephone contact for patients with OA found statistically significant differences in pain reduction according to the Arthritis Impact Measurement Scales at a very low cost.^22,23 However, these differences may not be clinically significant as the groups differed by less than 1 point on a 10-point scale. Interestingly, educating patients during regularly scheduled appointments had no benefit and even resulted in worsening physical functioning.²² A systematic review of patient education interventions found a nonsignificant trend toward benefit using patient education compared with nonsteroidal anti-inflammatory drugs (NSAIDs).²⁴ Despite the lack of convincing evidence from randomized trials, the possible benefits of educating our patients while using other therapies may outweigh the cost of the time involved, particularly if education can be done in a group setting.

Medication

Acetaminophen (4 g per day) has been shown in randomized trials to reduce pain in OA of the knee by approximately 30%.^25,26 This improvement was seen at 4 weeks and at 2 years. Another study found the combination of codeine plus acetaminophen to be significantly better in reducing pain in patients with OA of the hip than acetaminophen alone, although one third of the patients receiving the combination discontinued therapy because of side effects.²⁷ Narcotic analgesics alone have been shown in a randomized trial to be more effective in reducing pain than placebo but have not been adequately tested against acetaminophen or NSAIDs.²⁸

NSAIDs also produce a 30% reduction in pain caused by OA.^25,26 Two systematic reviews of OA concluded that there is no reliable evidence suggesting that any NSAID is more efficacious than the others in treating OA of the hip and knee.^29,30 A separate meta-analysis found ibuprofen to have the lowest risk of side effects among the nonselective NSAIDs, a finding supported by the Committee on Safety in Medicine.³¹ For this reason, ibuprofen is recommended as the first-line nonselective NSAID.

Acetaminophen and NSAIDs have been found to be equally efficacious when compared in randomized controlled trials.^25,26 According to a meta-analysis, the quality-of-life measures were similar between the 2 groups, despite a greater improvement in both pain at rest and pain on motion in the NSAID-treated patients.³¹ Interestingly, 2 recent surveys showed that patients with OA prefer NSAIDs to acetaminophen.²⁷ However, based on the results of clinical trials, safety profiles, and cost issues acetaminophen should continue to be used as the first-line agent.²⁷

The selective cyclooxygenase-2 (COX-2) inhibitors are effective in relieving the pain caused by OA.^32-35 The COX-2 inhibitors have been shown to minimize fecal blood loss and produce fewer endoscopically proven gastrointestinal erosions/ulcerations than the nonselective inhibitors.^36-39 This, however, is disease-oriented evidence and does not help us know about the more important clinical outcomes.

Using the STEPS approach (Safety, Tolerability, Efficacy, Price, Simplicity) to compare COX-2 inhibitors and nonselective NSAIDs is one way to help guide the choice of NSAIDs in the management of OA. A systematic review of a COX-2 inhibitor found the differences in adverse gastrointestinal events including perforations, symptomatic ulcers, and clinically significant bleeding episodes to be small when compared with the older NSAIDs (incidence = 1.3% vs 1.8%).^36-39 Thus, 200 patients would have to receive a COX-2 inhibitor instead of a nonselective NSAID for 1 year to prevent 1 clinically significant bleeding episode (number needed to harm = 200). These safety differences seem small. However, the safety benefits of the COX-2 inhibitors increase when patients have multiple risk factors for adverse gastrointestinal events (age Ž65, history of peptic ulcer disease or gastrointestinal bleeding, use of oral glucocorticoids or anticoagulants, and comorbid medical conditions).²⁷ Tolerability can be measured by looking at dropout rates. A meta-analysis found that the percentage of patients withdrawing from studies because of gastrointestinal adverse events was lower with the COX-2 inhibitor than with the nonselective NSAIDs (odds ratio = 0.59; 95% confidence interval, 0.52-0.67).³⁷ Tolerability favors the COX-2 agents. Meta-analyses of trials comparing the efficacy of COX-2 inhibitors and nonselective NSAIDs in the treatment of OA show no differences between them.^32-35 The average wholesale price of a month’s supply of a COX-2 inhibitor is approximately $73, and a month’s supply of a generic nonselective NSAID is $15. Nonselective NSAIDs are favored when looking at cost. While the nonselective NSAIDs require multiple dosing throughout the day, the COX-2 inhibitors can be given once daily, thus making the COX-2 inhibitors easier to use for patients. Overall, because of equal efficacy, minimal safety differences, and low cost, the older NSAIDs are recommended for low-risk patients, while the COX-2 inhibitors are recommended for high-risk patients because of increased safety and tolerability. (Of note, the COX-2 inhibitors have not been compared with acetaminophen, which is equally as efficacious as the nonselective NSAIDs and also has minimal side effects.)

Topical NSAIDs are more effective than placebo in treating the pain of OA, according to a systematic review.²⁷ Also, few side effects were noted. A small randomized trial (n=70) of 0.025% capsaicin cream applied topically 4 times daily for 4 weeks reduced pain by 8% compared with placebo in patients with OA of the knee over a 4-week period.¹⁹ Further studies are needed to confirm this finding.

Complementary and Alternative Medicine

A recent meta-analysis found that glucosamine produces a modest to large improvement in pain relief and functional outcomes compared with placebo.⁴⁰ A systematic review of controlled trials found that glucosamine performed equally well or better than NSAIDs.⁴¹ Most trials used a dose of 1500 mg glucosamine per day and addressed OA of the knee. The benefits may be overestimated because of the presence of publication bias. Because glucosamine is not regulated by the United States Food and Drug Administration, preparations and effectiveness may vary. However, it is very well tolerated and costs approximately $20 per month. Based on these results, glucosamine should be considered in the treatment of OA.

Although chondroitin has not been as well studied as glucosamine, a meta-analysis found evidence of effectiveness.⁴⁰ Most studies used 800 to 1200 mg of chondroitin daily for OA of the knee. An added benefit of using both together has not been proved.

Another therapy supported by good-quality evidence is the use of avocado/soybean unsaponifiables. A systematic review of herbal therapy concluded that 300 mg daily of avocado/soybean unsaponifiables can provide long-term symptomatic relief, particularly for patients with chronic stable OA of the hip.⁴² It may also help patients reduce their intake of NSAIDs.

The most recent systematic review of acupuncture for OA of the knee concluded that real acupuncture is better than sham acupuncture in treating pain but not function.⁴³ Four of the studies in the meta-analysis were of high quality, although none assessed the patient’s rating of global improvement. Acupuncture should be considered in the treatment of OA of the knee.

Less convincing therapies include therapeutic touch and electromagnetic fields. Therapeutic touch aims at manipulating a person’s energy system to bring the body system back into balance. A well-done single-blinded randomized trial compared therapeutic touch with mock therapeutic touch for OA of the knee.⁴⁴ The treatment group had a statistically significant decrease in pain and improved function compared with the placebo group in most of the outcomes measured. However, the differences were small (0.5- to 1.4-point difference on a 10-point scale) and may not be clinically important. Further studies are needed to demonstrate a clinical impact and to help us know how to incorporate this into our practices. A small study of 27 patients compared pulsed electromagnetic fields (PEMF) and placebo for the treatment of OA of the knee. Although, this double-blind randomized trial showed a 31% reduction in overall pain compared with sham PEMF, larger studies are needed before widespread use of this therapy is warranted.¹⁹

Injections

Intra-articular steroids for OA of the knee have been recommended by the American College of Rheumatology when an effusion and local signs of inflammation are present.⁴⁵ In 3 studies comparing steroid injections to placebo, 2 found that steroids were effective over 1 to 2 weeks, and 1 trial showed no difference. None of the trials showed any long-term benefit of steroid injections.⁴⁶

A systematic review of injectable biologic agents (glycosaminoglycans, mostly hyaluronic acid) found that in 6 of the 8 randomized trials for OA of the knee they were superior to placebo.⁴⁶ Treatment regimens varied, making it difficult to determine optimal duration of treatment and route of administration (intramuscular vs intra-articular). A comparison of hyaluronic acid injections with naproxen found that they were equally effective in improving pain relief and function, although the injections had fewer side effects.⁴⁷ Dextrose in a 10% solution is another biologic agent used intra-articularly; however, a well-designed study found that it was no more effective than placebo for OA of the knee.⁴⁸ Injectable glycosminoglycans, such as hyaluronic acid, are recommended in the treatment of OA of the knee, although the cost must be considered. The most effective regimen has not been determined.

Surgery

Surgery is reserved for patients with severe disease of the hip or knee that is not controlled by less invasive measures. Total hip arthroplasty has been shown to improve quality of life in patients with advanced hip OA.⁴⁹ Also, a cost-effectiveness analysis showed that total hip arthroplasty can be cost-effective in improving quality-adjusted life expectancy in the short term and long term.⁵⁰ Total joint arthroplasty for OA of the knee has shown similar benefits in pain relief and functional improvement.^45,46 It is not surprising that OA of the hip and knee is the most common indication for elective total hip and knee arthroplasty in the United States.⁵¹

Other Recommendations

On the basis of lower-quality studies, the American College of Rheumatology also recommends the following for individual patients: a weight-loss program, an occupational therapy evaluation, bracing and footwear, joint lavage of the knee, arthroscopic debridement of the knee, and osteotomy of the knee or hip.^45,52 The benefits from these interventions have not been well studied. The costs and benefits of these interventions must be weighed before recommending them to individual patients. Most trials have been conducted in patients with OA of the knee or hip. We do not know if applying treatment modalities to other joints (back and hands) will produce similar results. Also, most trials have been placebo controlled, and we do not know whether some treatments would add benefit to the first-line therapy of acetaminophen and NSAIDs.

Prognosis

OA is a chronic progressive disease of the joints that leads to increased pain and decreased functioning. Two longitudinal studies examined the natural course of OA of the knee in 191 patients. They showed that 56% of patients had no change, and 44% worsened over 15 years.^53-55 A study of OA of the hip showed similar trends except that a small group (7%) experienced improvement over 10 years.⁵⁶ We do not know whether current treatment options can improve the long-term prognosis of OA. Epidemiologic studies suggest that increasing age, obesity, family history, occupation, and joint injury are risk factors for OA.^6,53 Whether modifying these risk factors will decrease symptoms or slow the course of the disease is not known. The progressive nature of OA makes treatment a challenge for the patient and clinician.

Osteoarthritis (OA) is a chronic and progressive disease in which damage is done to the joint and surrounding tissue. In the United States, OA is found in 6% of people older than 30 years and in 50% of those older than 60 years.^1,2 It is the most common cause of disability in the United States and leads to considerable costs due to medical and surgical interventions and frequent absences from work.^3-5 OA accounts for 2% of all visits to family physicians and is the 10th leading diagnosis encountered.⁶ Thus, a family physician can expect to have 2 or 3 patient encounters per week in which OA is one of the diagnoses. With the growing elderly population, this burden is likely to increase.

Pathophysiology

In OA, the smooth surface of hyaline cartilage develops irregularities because of alterations at the cellular level and gross mechanical forces.² The role of inflammation has been debated in recent years, and its exact role is unknown. As OA progresses, the nearby bone remodels and forms further joint irregularities and osteophytes. These changes lead to narrowing of the joint space, and in some cases, chronic synovitis. Clinically, this causes pain, restricted movement, and periarticular muscle wasting. Treatment is aimed at these symptoms and structural abnormalities. The joints most commonly affected include the knees, hips, cervical and lumbosacral spine, distal interphalangeal (DIP) joints (producing Heberden nodes), proximal interphalangeal (PIP) joints (producing Bouchard nodes), and the first carpometacarpal joints of the hand. It is not known why these changes occur in some people and not in others. Epidemiologic studies and a recent sibling study raise the question of genetic influences in OA of the hip.^7,8

Diagnosis

The diagnosis of OA is made on the basis of clinical and radiographic features. Few studies have compared a diagnostic test or strategy with a gold standard. Unfortunately, we do not have any good data about the usefulness of individual history and physical examination elements for diagnosing OA. Radiographs suggesting the diagnosis of OA (joint space narrowing, presence of osteophytes, irregular joint surfaces, sclerosis of subchondral bone, or bony cysts) must be closely correlated with clinical symptoms. According to epidemiologic surveys, only one half of patients with radiographic changes of OA of the knee complain of persistent pain.⁹ Classification criteria for OA of the knee, hip, and hands have been developed; they are outlined in Tables 1 Table 2 through Table 3.^10,13

Although these criteria have limitations, they are becoming the standard for defining these types of OA and have been adapted by the American College of Rheumatology.^11-13 Currently, there are no criteria for diagnosing OA of the back. When evaluating a patient with joint pain, other diagnoses must be considered (ie, rheumatoid arthritis, gout, pseudogout, septic arthritis, bursitis, and tendonitis).

Treatment

Most of the treatments of OA address the symptoms rather than the cause of the disease. The short-term goal is to decrease pain; long-term goals are to improve functioning and slow progression of disease. Table 4 provides a summary of treatment options.

Exercise

A systematic review of 12 randomized controlled trials (RCTs) showed beneficial effects of exercise therapy in patients with mild to moderate OA of the knee and, to a lesser extent, the hip.¹⁴ Benefits included improvements in pain, self-reported disability, walking performance, and the patient’s global assessment of symptoms. Insufficient evidence was available to recommend one type of exercise program over another. Exercise interventions included aerobic exercises, strength training, range of motion exercises, and fitness walking. Exercise programs were conducted as individuals or groups, supervised or home-based. Although this review was limited by the small number of good studies, recommending exercise for OA may improve patients’ symptoms and add other health benefits.

Physical Therapy

A recent RCT compared 4 weeks of manual physical therapy plus a supervised knee exercise program with sham ultrasound for treatment of OA of the knee.¹⁵ By 8 weeks, 6-minute walk distances had improved by 13% in the treatment group compared with no change in the placebo group; osteoarthritis index scores had improved 56% over baseline compared with 15% in the placebo group. One year following therapy 5% of the treated group had undergone knee arthroplasty compared with 20% of the untreated group (number needed to treat = 7). Given these clinically important improvements, physical therapy should be an early choice in the treatment of OA.

A systematic review of transcutaneous electrical nerve stimulation (TENS) for treatment of OA of the knee found that this noninvasive modality offered significant pain relief.¹⁶ Both high-frequency and strong burst mode TENS showed significant improvement in pain relief when used for 4 weeks or more. Also, the acupuncture-like TENS improved pain relief, stiffness, and walking time in a 2-week placebo-controlled trial.

Although spa therapy and low-level laser therapy may show some benefit, the true effectiveness cannot be determined, since the trials were small and of poor quality.^17,18 No benefit was found in the one well-done RCT for ultrasound.¹⁹

Patient Education

Group patient education programs teach patients how to manage their disease. A 4-year longitudinal study using the Arthritis Self-Management Program (6 2-hour sessions), suggests that health education may decrease pain and visits to physicians.^20,21 Estimated 4-year savings were $189 per patient.²¹ Telephone-based interventions have also been studied. A randomized trial of telephone contact for patients with OA found statistically significant differences in pain reduction according to the Arthritis Impact Measurement Scales at a very low cost.^22,23 However, these differences may not be clinically significant as the groups differed by less than 1 point on a 10-point scale. Interestingly, educating patients during regularly scheduled appointments had no benefit and even resulted in worsening physical functioning.²² A systematic review of patient education interventions found a nonsignificant trend toward benefit using patient education compared with nonsteroidal anti-inflammatory drugs (NSAIDs).²⁴ Despite the lack of convincing evidence from randomized trials, the possible benefits of educating our patients while using other therapies may outweigh the cost of the time involved, particularly if education can be done in a group setting.

Medication

Acetaminophen (4 g per day) has been shown in randomized trials to reduce pain in OA of the knee by approximately 30%.^25,26 This improvement was seen at 4 weeks and at 2 years. Another study found the combination of codeine plus acetaminophen to be significantly better in reducing pain in patients with OA of the hip than acetaminophen alone, although one third of the patients receiving the combination discontinued therapy because of side effects.²⁷ Narcotic analgesics alone have been shown in a randomized trial to be more effective in reducing pain than placebo but have not been adequately tested against acetaminophen or NSAIDs.²⁸

NSAIDs also produce a 30% reduction in pain caused by OA.^25,26 Two systematic reviews of OA concluded that there is no reliable evidence suggesting that any NSAID is more efficacious than the others in treating OA of the hip and knee.^29,30 A separate meta-analysis found ibuprofen to have the lowest risk of side effects among the nonselective NSAIDs, a finding supported by the Committee on Safety in Medicine.³¹ For this reason, ibuprofen is recommended as the first-line nonselective NSAID.

Acetaminophen and NSAIDs have been found to be equally efficacious when compared in randomized controlled trials.^25,26 According to a meta-analysis, the quality-of-life measures were similar between the 2 groups, despite a greater improvement in both pain at rest and pain on motion in the NSAID-treated patients.³¹ Interestingly, 2 recent surveys showed that patients with OA prefer NSAIDs to acetaminophen.²⁷ However, based on the results of clinical trials, safety profiles, and cost issues acetaminophen should continue to be used as the first-line agent.²⁷

The selective cyclooxygenase-2 (COX-2) inhibitors are effective in relieving the pain caused by OA.^32-35 The COX-2 inhibitors have been shown to minimize fecal blood loss and produce fewer endoscopically proven gastrointestinal erosions/ulcerations than the nonselective inhibitors.^36-39 This, however, is disease-oriented evidence and does not help us know about the more important clinical outcomes.

Using the STEPS approach (Safety, Tolerability, Efficacy, Price, Simplicity) to compare COX-2 inhibitors and nonselective NSAIDs is one way to help guide the choice of NSAIDs in the management of OA. A systematic review of a COX-2 inhibitor found the differences in adverse gastrointestinal events including perforations, symptomatic ulcers, and clinically significant bleeding episodes to be small when compared with the older NSAIDs (incidence = 1.3% vs 1.8%).^36-39 Thus, 200 patients would have to receive a COX-2 inhibitor instead of a nonselective NSAID for 1 year to prevent 1 clinically significant bleeding episode (number needed to harm = 200). These safety differences seem small. However, the safety benefits of the COX-2 inhibitors increase when patients have multiple risk factors for adverse gastrointestinal events (age Ž65, history of peptic ulcer disease or gastrointestinal bleeding, use of oral glucocorticoids or anticoagulants, and comorbid medical conditions).²⁷ Tolerability can be measured by looking at dropout rates. A meta-analysis found that the percentage of patients withdrawing from studies because of gastrointestinal adverse events was lower with the COX-2 inhibitor than with the nonselective NSAIDs (odds ratio = 0.59; 95% confidence interval, 0.52-0.67).³⁷ Tolerability favors the COX-2 agents. Meta-analyses of trials comparing the efficacy of COX-2 inhibitors and nonselective NSAIDs in the treatment of OA show no differences between them.^32-35 The average wholesale price of a month’s supply of a COX-2 inhibitor is approximately $73, and a month’s supply of a generic nonselective NSAID is $15. Nonselective NSAIDs are favored when looking at cost. While the nonselective NSAIDs require multiple dosing throughout the day, the COX-2 inhibitors can be given once daily, thus making the COX-2 inhibitors easier to use for patients. Overall, because of equal efficacy, minimal safety differences, and low cost, the older NSAIDs are recommended for low-risk patients, while the COX-2 inhibitors are recommended for high-risk patients because of increased safety and tolerability. (Of note, the COX-2 inhibitors have not been compared with acetaminophen, which is equally as efficacious as the nonselective NSAIDs and also has minimal side effects.)

Topical NSAIDs are more effective than placebo in treating the pain of OA, according to a systematic review.²⁷ Also, few side effects were noted. A small randomized trial (n=70) of 0.025% capsaicin cream applied topically 4 times daily for 4 weeks reduced pain by 8% compared with placebo in patients with OA of the knee over a 4-week period.¹⁹ Further studies are needed to confirm this finding.

Complementary and Alternative Medicine

A recent meta-analysis found that glucosamine produces a modest to large improvement in pain relief and functional outcomes compared with placebo.⁴⁰ A systematic review of controlled trials found that glucosamine performed equally well or better than NSAIDs.⁴¹ Most trials used a dose of 1500 mg glucosamine per day and addressed OA of the knee. The benefits may be overestimated because of the presence of publication bias. Because glucosamine is not regulated by the United States Food and Drug Administration, preparations and effectiveness may vary. However, it is very well tolerated and costs approximately $20 per month. Based on these results, glucosamine should be considered in the treatment of OA.

Although chondroitin has not been as well studied as glucosamine, a meta-analysis found evidence of effectiveness.⁴⁰ Most studies used 800 to 1200 mg of chondroitin daily for OA of the knee. An added benefit of using both together has not been proved.

Another therapy supported by good-quality evidence is the use of avocado/soybean unsaponifiables. A systematic review of herbal therapy concluded that 300 mg daily of avocado/soybean unsaponifiables can provide long-term symptomatic relief, particularly for patients with chronic stable OA of the hip.⁴² It may also help patients reduce their intake of NSAIDs.

The most recent systematic review of acupuncture for OA of the knee concluded that real acupuncture is better than sham acupuncture in treating pain but not function.⁴³ Four of the studies in the meta-analysis were of high quality, although none assessed the patient’s rating of global improvement. Acupuncture should be considered in the treatment of OA of the knee.

Less convincing therapies include therapeutic touch and electromagnetic fields. Therapeutic touch aims at manipulating a person’s energy system to bring the body system back into balance. A well-done single-blinded randomized trial compared therapeutic touch with mock therapeutic touch for OA of the knee.⁴⁴ The treatment group had a statistically significant decrease in pain and improved function compared with the placebo group in most of the outcomes measured. However, the differences were small (0.5- to 1.4-point difference on a 10-point scale) and may not be clinically important. Further studies are needed to demonstrate a clinical impact and to help us know how to incorporate this into our practices. A small study of 27 patients compared pulsed electromagnetic fields (PEMF) and placebo for the treatment of OA of the knee. Although, this double-blind randomized trial showed a 31% reduction in overall pain compared with sham PEMF, larger studies are needed before widespread use of this therapy is warranted.¹⁹

Injections

Intra-articular steroids for OA of the knee have been recommended by the American College of Rheumatology when an effusion and local signs of inflammation are present.⁴⁵ In 3 studies comparing steroid injections to placebo, 2 found that steroids were effective over 1 to 2 weeks, and 1 trial showed no difference. None of the trials showed any long-term benefit of steroid injections.⁴⁶

A systematic review of injectable biologic agents (glycosaminoglycans, mostly hyaluronic acid) found that in 6 of the 8 randomized trials for OA of the knee they were superior to placebo.⁴⁶ Treatment regimens varied, making it difficult to determine optimal duration of treatment and route of administration (intramuscular vs intra-articular). A comparison of hyaluronic acid injections with naproxen found that they were equally effective in improving pain relief and function, although the injections had fewer side effects.⁴⁷ Dextrose in a 10% solution is another biologic agent used intra-articularly; however, a well-designed study found that it was no more effective than placebo for OA of the knee.⁴⁸ Injectable glycosminoglycans, such as hyaluronic acid, are recommended in the treatment of OA of the knee, although the cost must be considered. The most effective regimen has not been determined.

Surgery

Surgery is reserved for patients with severe disease of the hip or knee that is not controlled by less invasive measures. Total hip arthroplasty has been shown to improve quality of life in patients with advanced hip OA.⁴⁹ Also, a cost-effectiveness analysis showed that total hip arthroplasty can be cost-effective in improving quality-adjusted life expectancy in the short term and long term.⁵⁰ Total joint arthroplasty for OA of the knee has shown similar benefits in pain relief and functional improvement.^45,46 It is not surprising that OA of the hip and knee is the most common indication for elective total hip and knee arthroplasty in the United States.⁵¹

Other Recommendations

On the basis of lower-quality studies, the American College of Rheumatology also recommends the following for individual patients: a weight-loss program, an occupational therapy evaluation, bracing and footwear, joint lavage of the knee, arthroscopic debridement of the knee, and osteotomy of the knee or hip.^45,52 The benefits from these interventions have not been well studied. The costs and benefits of these interventions must be weighed before recommending them to individual patients. Most trials have been conducted in patients with OA of the knee or hip. We do not know if applying treatment modalities to other joints (back and hands) will produce similar results. Also, most trials have been placebo controlled, and we do not know whether some treatments would add benefit to the first-line therapy of acetaminophen and NSAIDs.

Prognosis

OA is a chronic progressive disease of the joints that leads to increased pain and decreased functioning. Two longitudinal studies examined the natural course of OA of the knee in 191 patients. They showed that 56% of patients had no change, and 44% worsened over 15 years.^53-55 A study of OA of the hip showed similar trends except that a small group (7%) experienced improvement over 10 years.⁵⁶ We do not know whether current treatment options can improve the long-term prognosis of OA. Epidemiologic studies suggest that increasing age, obesity, family history, occupation, and joint injury are risk factors for OA.^6,53 Whether modifying these risk factors will decrease symptoms or slow the course of the disease is not known. The progressive nature of OA makes treatment a challenge for the patient and clinician.

BACKGROUND: Estrogen is known to affect the urethral mucosa and smooth muscle and has been part of the medical management of urinary incontinence in postmenopausal women. The clinical effectiveness of this therapy is controversial.

POPULATION STUDIED: Study participants included a total of 1525 postmenopausal women younger than 80 years with coronary heart disease who had not had a hysterectomy. All eligible women reported at least weekly urinary incontinence at the beginning of the study period.

STUDY DESIGN AND VALIDITY: The original Heart and Estrogen/progestin Replacement Study (HERS) enrolled 2763 postmenopausal women with established coronary heart disease to evaluate the role of hormone replacement therapy in the prevention of coronary heart disease. Of these women, 1525 participants reported at least weekly urinary incontinence and were included in this analysis. Patients were assigned by computer-generated random numbers to receive daily oral conjugated estrogen (0.625 mg) and medroxyprogesterone acetate (2.5 mg) in one tablet (Prempro; Wyeth Ayerst Research, Radnor, Pa) or identical placebo (concealed allocation assignment). Both participants and study personnel were blind to treatment status. Participants completed questionnaires concerning voiding habits and incontinence.

OUTCOMES MEASURED: The main outcome studied was the severity of incontinence, defined as the number of incontinent episodes per week. Outcomes were measured with a questionnaire conducted by blinded investigators at 4 months and then annually for 4 years.

RESULTS: Follow-up at 4 years was 98% complete in both the treatment and control groups. Urinary incontinence improved in 21% of the women taking hormones compared with 26% assigned to the placebo group, while 39% of the women in the hormone group reported worse incontinence compared with 27% of women in the placebo group (P=.001; number needed to harm=8). These results represented the mean of all follow-up visits and a significant difference was seen as early as 4 months. The actual number of incontinent episodes per week increased an average of 0.7 in the hormone group and decreased by 0.1 in the placebo group (P #060;.001).

BACKGROUND: Estrogen is known to affect the urethral mucosa and smooth muscle and has been part of the medical management of urinary incontinence in postmenopausal women. The clinical effectiveness of this therapy is controversial.

POPULATION STUDIED: Study participants included a total of 1525 postmenopausal women younger than 80 years with coronary heart disease who had not had a hysterectomy. All eligible women reported at least weekly urinary incontinence at the beginning of the study period.

STUDY DESIGN AND VALIDITY: The original Heart and Estrogen/progestin Replacement Study (HERS) enrolled 2763 postmenopausal women with established coronary heart disease to evaluate the role of hormone replacement therapy in the prevention of coronary heart disease. Of these women, 1525 participants reported at least weekly urinary incontinence and were included in this analysis. Patients were assigned by computer-generated random numbers to receive daily oral conjugated estrogen (0.625 mg) and medroxyprogesterone acetate (2.5 mg) in one tablet (Prempro; Wyeth Ayerst Research, Radnor, Pa) or identical placebo (concealed allocation assignment). Both participants and study personnel were blind to treatment status. Participants completed questionnaires concerning voiding habits and incontinence.

OUTCOMES MEASURED: The main outcome studied was the severity of incontinence, defined as the number of incontinent episodes per week. Outcomes were measured with a questionnaire conducted by blinded investigators at 4 months and then annually for 4 years.

RESULTS: Follow-up at 4 years was 98% complete in both the treatment and control groups. Urinary incontinence improved in 21% of the women taking hormones compared with 26% assigned to the placebo group, while 39% of the women in the hormone group reported worse incontinence compared with 27% of women in the placebo group (P=.001; number needed to harm=8). These results represented the mean of all follow-up visits and a significant difference was seen as early as 4 months. The actual number of incontinent episodes per week increased an average of 0.7 in the hormone group and decreased by 0.1 in the placebo group (P #060;.001).

BACKGROUND: Estrogen is known to affect the urethral mucosa and smooth muscle and has been part of the medical management of urinary incontinence in postmenopausal women. The clinical effectiveness of this therapy is controversial.

POPULATION STUDIED: Study participants included a total of 1525 postmenopausal women younger than 80 years with coronary heart disease who had not had a hysterectomy. All eligible women reported at least weekly urinary incontinence at the beginning of the study period.

STUDY DESIGN AND VALIDITY: The original Heart and Estrogen/progestin Replacement Study (HERS) enrolled 2763 postmenopausal women with established coronary heart disease to evaluate the role of hormone replacement therapy in the prevention of coronary heart disease. Of these women, 1525 participants reported at least weekly urinary incontinence and were included in this analysis. Patients were assigned by computer-generated random numbers to receive daily oral conjugated estrogen (0.625 mg) and medroxyprogesterone acetate (2.5 mg) in one tablet (Prempro; Wyeth Ayerst Research, Radnor, Pa) or identical placebo (concealed allocation assignment). Both participants and study personnel were blind to treatment status. Participants completed questionnaires concerning voiding habits and incontinence.

OUTCOMES MEASURED: The main outcome studied was the severity of incontinence, defined as the number of incontinent episodes per week. Outcomes were measured with a questionnaire conducted by blinded investigators at 4 months and then annually for 4 years.

RESULTS: Follow-up at 4 years was 98% complete in both the treatment and control groups. Urinary incontinence improved in 21% of the women taking hormones compared with 26% assigned to the placebo group, while 39% of the women in the hormone group reported worse incontinence compared with 27% of women in the placebo group (P=.001; number needed to harm=8). These results represented the mean of all follow-up visits and a significant difference was seen as early as 4 months. The actual number of incontinent episodes per week increased an average of 0.7 in the hormone group and decreased by 0.1 in the placebo group (P #060;.001).