Diazepam Is No Better Than Placebo When Added to Naproxen for Acute Low Back Pain

Abstract

Study objective: Low back pain causes more than 2.5 million visits to US emergency departments (EDs) annually. Low back pain patients are often treated with nonsteroidal anti-inflammatory drugs and benzodiazepines. The former is an evidence-based intervention, whereas the efficacy of the latter has not been established. We compare pain and functional outcomes 1 week and 3 months after ED discharge among patients randomized to a 1-week course of naproxen+diazepam versus naproxen+placebo.

Methods: This was a randomized, double-blind, comparative efficacy clinical trial conducted in an urban health care system. Patients presenting with acute, nontraumatic, nonradicular low back pain of no more than a duration of 2 weeks were eligible for enrollment immediately before discharge from an ED if they had a score greater than 5 on the Roland-Morris Disability Questionnaire, a validated 24-item inventory of functional impairment caused by low back pain. Higher scores on the questionnaire indicate greater functional disability. The primary outcome in the trial was improvement in the score between ED discharge and 1 week later. Secondary outcomes included pain intensity 1 week and 3 months after ED discharge, as measured on a 4-point descriptive scale (severe, moderate, mild, and none). All patients were given 20 tablets of naproxen 500 mg, to be taken twice a day as needed for low back pain. Additionally, patients were randomized to receive either 28 tablets of diazepam 5 mg or identical placebo, to be received as 1 or 2 tablets every 12 hours as needed for low back pain. All patients received a standardized 10-minute low back pain educational session before discharge. Using a between-group mean difference of 5 Roland-Morris Disability Questionnaire points, a previously validated threshold for clinical significance, we calculated the need for at least 100 patients with primary outcome data.

Results: Enrollment began in June 2015 and continued for 9 months. Five hundred forty-five patients were screened for eligibility. One hundred fourteen patients met selection criteria and were randomized. Baseline demographic characteristics were not substantially different between the 2 groups. One hundred twelve patients (98%) provided 1-week outcome data. The mean Roland-Morris Disability Questionnaire score of patients randomized to naproxen+diazepam improved by 11 (95% confidence interval [CI] 9 to 13), as did the mean score of patients randomized to naproxen+placebo (11; 95% CI 8 to 13). At 1-week follow-up, 18 of 57 diazepam patients (32%; 95% CI 21% to 45%) reported moderate or severe low back pain versus 12 of 55 placebo patients (22%; 95% CI 13% to 35%). At 3-month follow-up, 6 of 50 diazepam patients (12%; 95% CI 5% to 24%) reported moderate or severe low back pain versus 5 of 53 placebo patients (9%; 95% CI 4% to 21%). Adverse events were reported by 12 of 57 diazepam patients (21%; 95% CI 12% to 33%) and 8 of 55 placebo patients (15%; 95% CI 7% to 26%).

Conclusion: Among ED patients with acute, nontraumatic, nonradicular low back pain, naproxen+diazepam did not improve functional outcomes or pain compared with naproxen+placebo 1 week and 3 months after ED discharge.

Figure 1. CONSORT flow diagram

* 4 patients were excluded for more than one reason