In urgent and primary care settings, when a patient requests medication for acute low back pain without radicular symptoms, I typically prescribe naproxen and cyclobenzaprine, adding oxycodone/acetaminophen if the pain seems especially severe. But two recent articles in American Family Physician will likely lead me to change my practice.
The first article is a Medicine by the Numbers that reviewed the number needed to treat (NNT) and number needed to harm (NNH) from a pooled analysis of trials evaluating cyclobenzaprine for low back pain. Compared to placebo, cyclobenzaprine was more likely to lead to global symptom improvement by day 10 of treatment, with an impressive NNT of 3. Unfortunately, it was also much more likely to cause dizziness, nausea, drowsiness, and dry mouth, with a NNH of 4 for any adverse effect. In other words, participants were almost as likely to feel worse on the drug as they were to feel better. Further, most trials did not use intention-to-treat analysis or had other important flaws in quality, making even this marginal benefit uncertain.
The second article is a POEM that summarized a randomized trial comparing functional outcomes in adults with acute, nontraumatic, nonradicular low back pain who received naproxen plus placebo, naproxen plus cyclobenzaprine, or naproxen plus oxycodone/acetaminophen for 10 days. Research associates blinded to treatment arm assignment assessed participants for pain and functional outcomes in telephone interviews conducted at 7 days and 3 months of follow-up. There were no statistical differences between groups in either outcome at either time point. However, the NNH for adverse effects was 7.8 for cyclobenzaprine and 5.3 for oxycodone/acetaminophen.
Based on this information, I plan to prescribe naproxen alone for most patients with acute low back pain and no contraindications to nonsteroidal anti-inflammatory drugs (NSAIDs); reserve cyclobenzaprine for patients who can't use NSAIDs; and prescribe oxycodone/acetaminophen only in patients who can't tolerate NSAIDs or cyclobenzaprine.
This post first appeared on the AFP Community Blog.