Top POEMs of 2016 Consistent with the Principles of the Choosing Wisely Campaign: Musculoskeletal Conditions

Top 20 POEMs of 2016 Consistent with the Principles of the Choosing Wisely Campaign

Knee Surgery Does Not Reduce Knee Catching or Locking in Patients with Meniscal Tear (FIDELITY)

Clinical question
Does partial meniscectomy fix mechanical symptoms -- knee catching or locking -- better than sham surgery?

Bottom line
I guess it's time to stop the knee-jerk reaction of sending patients with occasional catches and locking to ortho for meniscal resection. Removing the torn bits of meniscus in middle-aged patients who have intermittent knee catches or locking does not decrease their likelihood of experiencing symptoms in the following year as compared with diagnostic arthroscopy (ie, looking but not touching). In general, meniscectomy does not improve knee pain, regardless of the symptoms (N Engl J Med 2013;369(26):2515-24). (LOE = 1b-)(www.essentialevidenceplus.com)

Reference
Sihvonen R, Englund M, Turkiewicz A, Jarvinen TL, for the Finnish Degenerative Meniscal Lesion Study Group. Mechanical symptoms and arthroscopic partial meniscectomy in patients with degenerative meniscus tear: A secondary analysis of a randomized trial. Ann Intern Med. 2016;164(7):449-455.

Study design: Randomized controlled trial (double-blinded)

Funding source: Foundation

Allocation: Concealed

Setting: Outpatient (specialty)

Synopsis
This report is a substudy of a larger study investigating the effect of arthroscopic surgery on (relatively) young patients with meniscal tear but without signs of osteoarthritis. These Finnish investigators enrolled 146 patients, aged 35 to 65 years, who had knee pain for at least 3 months and evidence of a degenerative meniscal tear but did not respond to conservative treatment. They excluded patients with a verified locked knee (unable to straighten), though they included patients (n = 69) who had symptoms of "catching" or occasional or frequent locking. All patients underwent arthroscopic surgery, though slightly more than half were randomly assigned, using concealed allocation, to a group that did not have the tear addressed (sham surgery). In the surgery group, damaged and loose parts were removed; in the sham surgery group, diagnostic arthroscopy was performed and the surgeon simulated actual surgery (since patients were awake) without removing anything. In the subsequent 12 months, 23 (72%) in the surgery group and 22 (59%) in the sham surgery group with preoperative mechanical symptoms reported symptoms at least once. Only 9 of 32 patients (28%) in the surgery subgroup and 15 of 37 (41%) in the sham surgery subgroup reported complete resolution of their symptoms.

Allen F. Shaughnessy, PharmD, MMedEd
Professor of Family Medicine
Tufts University
Boston, MA

Hyaluronic Acid = Sham Injections in Patients with Knee DJD

Clinical question
Do hyaluronic acid injections in patients with knee degenerative joint disease improve pain and function?

Bottom line
The highest quality studies, which are now fairly plentiful, show that hyaluronic acid injections are only minimally better than sham injections in improving pain and function in patients with knee degenerative joint disease. (LOE = 1a-)(www.essentialevidenceplus.com)

Reference
Jevsevar D, Donnelly P, Brown GA, Cummins DS. Viscosupplementation for osteoarthritis of the knee. J Bone Joint Surg Am 2015;97(24):2047-2060.

Study design: Meta-analysis (randomized controlled trials)

Funding source: Self-funded or unfunded

Setting: Outpatient (specialty)

Synopsis
These authors searched multiple databases and a clinical trial registry to identify randomized trials comparing hyaluronic acid injections with control treatments in patients with knee degenerative joint disease. Studies had to include at least 30 patients in each group, include pain and function scales for which minimal clinically important differences are established, and have at least 4 weeks of follow up. The authors don't describe the process of article inclusion. Ultimately, they included 19 studies with nearly 4500 patients: 14 used sham injections as the comparator; 2 used usual care; 3 used injections combined with some other active treatment. The authors did not find any statistically significant potential for publication bias. In the studies using sham injections, hyaluronic acid was slightly better at improving pain and function, but the improvement was not clinically important. Similarly, the improvements in double-blind trials were also not clinically significant. Among the other studies with designs at higher risk of bias, the magnitude of improvement in pain and function were really impressive. We have seen this story before: The stuff looks really effective until the studies are done correctly. In the meantime, all the placebo responders are writing testimonials about how their "lives were saved."

Henry C. Barry, MD, MS
Professor
Michigan State University
East Lansing, MI

OA Shoes No Better Than Walking Shoes for Knee OA Pain and Function

Clinical question
Are special shoes more effective than conventional walking shoes to decrease the pain of knee osteoarthritis and improve physical function?

Bottom line
A quick search of the Internet will yield many walking shoes targeted at the bad knees market. But specifically designed shoes with modified midsoles and a wedge to unload the medial aspect of the knee are no more effective than typical walking shoes at relieving pain and improving function in patients with documented knee osteoarthritis. In this study, patients in both groups improved, which either might be an artifact of the study or because any type of walking shoe improves pain and function. (LOE = 1b)(www.essentialevidenceplus.com)

Reference
Hinman RS, Wrigley TV, Metcalf BR, et al. Unloading shoes for self-management of knee osteoarthritis. A randomized trial. Ann Intern Med 2016;165(6):381-389.

Study design: Randomized controlled trial (double-blinded)

Funding source: Government

Allocation: Concealed

Setting: Outpatient (any)

Synopsis
Some shoes are specifically designed to "unload" knees through modified midsoles that have variable stiffness and a lateral wedge to decrease the biomechanical load on the medial aspect of the knee. This study, conducted in Australia, enrolled 164 patients recruited through advertising who were at least 50 years old, had knee pain of at least 4 on a scale of 0 to 10 for most days, and radiographic evidence of medial knee osteoarthritis. The participants were randomly assigned, using concealed allocation, to wear either unloading shoes (Asics GEL-Melbourne OA) or neutral walking shoes (Asics GEL-Odyssey) for at least 4 hours a day for 6 months. Shoes for both groups were provided by the researchers. Pain scores changed from an average 5.7 to 6.0 before the start of the study to an average 4.2 in both groups on an 11-point scale (P = NS). Function also improved to a clinically significant degree (> 6 units on the Western Ontario and McMaster Universities Osteoarthritis Index) in both groups (function improved in 44% and 48% of patients in the unloading and neutral shoe groups, respectively).The study had 90% power to find a 2.5-unit difference in pain and a 10.5-unit difference in function. The results might be cause in part by an "acquiescence bias," in which people in both groups felt they needed to report improvement to researchers who gave them free shoes. It would have been great if this study had included a third group who stuck to their usual shoe-wearing habits.

Allen F. Shaughnessy, PharmD, MMedEd
Professor of Family Medicine
Tufts University
Boston, MA

Lumbar Fusion No Better Than Exercise Therapy in the Long Term

Clinical question
Is lumbar fusion effective for patients with chronic low back pain?

Bottom line
This trial is a good example of how to do just about everything wrong in order to get the results you want. The authors did not conceal allocation, did not mask anyone in the study, used an unvalidated and subjective primary outcome, and downplayed the intention-to-treat analysis. Funding for the original study came from industry, and the authors have numerous conflicts of interest. Two other trials in the United Kingdom and Norway found no benefit to lumbar fusion, and the results of this study are consistent with those findings, despite what the authors conclude. (LOE = 2b)(www.essentialevidenceplus.com)

Reference
Hedlund R, Johansson C, Hagg O, et al. The long-term outcome of lumbar fusion in the Swedish lumbar spine study. Spine 2016;16(5):579-587.

Study design: Randomized controlled trial (nonblinded)

Funding source: Industry

Allocation: Unconcealed

Setting: Outpatient (specialty)

Synopsis
This is an important question; one not without controversy. This study reports a mean 12.8 years of follow-up from a trial that randomized 294 persons with severe chronic low back pain in a 3:1 ratio to lumbar fusion or physical therapy. This report provides almost no detail about their methods, but a look at their earlier publication reveals that outcome assessors (and, obviously, patients) were not masked to treatment assignment. The earlier report, after 2 years of follow-up, showed generally favorable results for surgery. Approximately 20% of patients in each group died or were lost to follow-up. In the long-term results, using intention-to-treat analysis, there is no difference between groups for any outcome, including the patient's Global Assessment (GA) of back pain score, the Oswestry Disability index score, a visual analog scale for pain score, pain medication use, pain frequency, or employment status. The authors also report an "as treated" analysis, which counts the 19 of 72 patients who crossed over to surgery as if they had originally been assigned to surgery (they were not!), and they report a per-protocol analysis, which ignores patients who crossed over or were lost to follow-up. Both of these analyses found an improvement in the patient's GA score with surgery, but failed to find improvement in any other outcomes. On the basis of the single outcome of GA score in the more biased analyses, the authors' conclusion is that surgery should be considered effective. An accompanying editorial, which strongly disagrees with the authors, begins with the snide headline: Consensus at last... fusion is no better than nonoperative care in improving pain and disability in chronic low back pain.

Mark H. Ebell, MD, MS
Professor
University of Georgia
Athens, GA

Long-Acting Opioids for Noncancer Pain Associated with Increased Risk of Death from Causes Other Than Overdose

Clinical question
Do long-acting opioids increase the risk of death from causes other than overdose in adults with noncancer chronic pain?

Bottom line
The use of long-acting opioids for chronic noncancer pain as compared with anticonvulsants or cyclic antidepressants is associated with a significantly increased risk of premature mortality--not only from unintentional overdose, but also from other causes. Although it is possible for observational trials like this to be confounded by unadjusted treatment associations, this is likely to be the best evidence we'll have on this question; it would be difficult to perform a more rigorous long-term randomized controlled trial. A recent systematic review (Chaparro LE, et al. Spine 2014;39:556-63) also found minimal evidence for long-term improvement in function from opioid therapy. (LOE = 3b)(www.essentialevidenceplus.com)

Reference
Ray WA, Chung CP, Murray KT, Hall K, Stein CM. Prescription of long-acting opioids and mortality in patients with chronic noncancer pain. JAMA 2016;315(22):2415-2423.

Study design: Cohort (retrospective)

Funding source: Foundation

Setting: Population-based

Synopsis
It is well known that chronic use of long-acting opioids increases the risk of overdose-related deaths. These investigators retrospectively evaluated data obtained from Tennessee Medicaid records of enrollees from 1999 through 2012. Pertinent data included enrollment, pharmacy, hospital, outpatient, and nursing home files; death certificates; and hospital discharge information. Only information on patients without evidence of cancer care, palliative care, or end-of-life care was included. Patients with recorded evidence of drug abuse were excluded. Noncancer-related chronic pain diagnoses included chronic back, other musculoskeletal, abdominal, headache, or other neurologic pain in the past 90 days. Study drugs included long-acting opioids (morphine, oxycodone, methadone, and transdermal fentanyl). Control drugs included anticonvulsants (gabapentin, pregabalin, carbamazepine) or low-dose cyclic antidepressants. Case patients with new episodes of therapy for long-acting opioids were matched to control patients according to demographics; chronic pain diagnoses; and the use of benzodiazepines, other psychotropic medications, or short-acting opioids and other pain medications. Cohort initiation occurred on the date of filling the first study drug prescription and ended after 1 year without filling a study drug prescription. After matching, the cohort included 22,912 long-acting opioid use patient-episodes and an equal number of matched control medication episodes. The mean age of patients was 48 years and 60% were women. More than half (68%) the study patients had a current prescription for short-acting opioids at the beginning of the follow-up period. Patients frequently filled prescriptions for other pain medications, including muscle relaxants, nonsteroidal anti-inflammatory drugs, benzodiazepines, and antidepressants. The median dose of long-acting opioids at study entry was 50 mg morphine equivalents. There were 185 deaths (167.1 per 10,000 person-years) in the long-acting opioid group compared with 87 deaths (107.9 per 10,000 person-years) in the control medication group, for a risk difference of 68.5 excess deaths per 10,000 person-years; of these, more than one-half were cardiovascular deaths. The risk difference for the long-acting opioids cohort compared with the control group was 47.4 per 10,000 person-years for out-of-hospital deaths with a cause of death other than unintentional overdose. The increased mortality risk for the long-acting opioid group was limited to the first 180 days of prescribed therapy and was similar for both low doses and high doses of the study drugs.

David Slawson, MD
Director of Information Sciences
University of Virginia Health System
Charlottesville, VA

POEMs are provided by Essential Evidence Plus, a point-of-care clinical decision support system published by Wiley-Blackwell. For more information, please see http://www.essentialevidenceplus.com(www.essentialevidenceplus.com).

Advertisement