Background Most low back pain trials have limited applicability to the emergency department (ED) because they provide treatment and measure outcomes after discharge from the ED. We investigated the efficacy and safety of pharmacological and non-pharmacological interventions delivered in the ED to patients with non-specific low back pain and/or sciatica on patient-relevant outcomes measured during the emergency visit.

Methods Literature searches were performed in MEDLINE, EMBASE and CINAHL from inception to week 1 February 2020. We included all randomised controlled trials investigating adult patients (≥18 years) with non-specific low back pain and/or sciatica presenting to ED. The primary outcome of interest was pain intensity. Two reviewers independently screened the full texts, extracted the data and assessed risk of bias of each trial using the Physiotherapy Evidence Database (PEDro) scale. The overall quality of evidence, or certainty, provided by a set of trials evaluating the same treatment was evaluated using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach, which considers imprecision, inconsistency, indirectness and bias in the evidence.

Results Fifteen trials (1802 participants) were included with 12 of 15 at low risk of bias (ie, PEDro score >6). Based on results from individual trials and moderate quality evidence, ketoprofen gel was more effective than placebo for non-specific low back pain at 30 min (mean difference (MD) −15.0, 95% confidence interval (CI) −21.0 to −9.0). For those with sciatica (moderate quality evidence), intravenous paracetamol (acetaminophen) (MD −15.7, 95% CI −19.8 to −11.6) and intravenous morphine (MD −11.4, 95% CI −21.6 to −1.2) were both superior to placebo at 30 min. Based on moderate quality of evidence, corticosteroids showed no benefits against placebo at emergency discharge for non-specific low back pain (MD 9.0, 95% CI −0.71 to 18.7) or sciatica (MD −6.8, 95% CI −24.2 to 10.6). There were conflicting results from trials comparing different pharmacological options (moderate quality evidence) or investigating non-pharmacological treatments (low quality evidence).

Conclusion Ketoprofen gel for non-specific low back pain and intravenous paracetamol or morphine for sciatica were superior to placebo, whereas corticosteroids were ineffective for both conditions. There was conflicting evidence for comparisons of different pharmacological options and those involving non-pharmacological treatments. Additional trials measuring important patient-related outcomes to EDs are needed.