X-ray scans for nonspecific low back pain: a nonspecific pain?

Evidence • A meta-analysis1 (N = 1804) examined 2 magnetic resonance imaging (MRI) and computed tomography (CT) trials and 4 x-ray trials; 0% to 44% of patients had sciatica. -The trials were of relatively good quality but there was a lot of heterogeneity (except with pain). -Short-term and long-term outcomes of pain, function, quality of life, mental health, and patient satisfaction did not differ significantly. —Pain at 3 months approached significantly worse with x-ray scan (standard mean difference 0.19, 95% CI -0.01 to 0.39). • A UK RCT2 of 421 general practice patients with LBP for 6 weeks or longer found -at 3 months statistically significant differences in —the proportion of patients still in pain (74% in x-ray group vs 65%, number needed to harm [NNH] = 12), —the proportion of patients requiring follow-up doctor visits (53% in x-ray group vs 30%, NNH = 5), and —self-rated health status (lower in x-ray group); -after 6 more months the differences were borderline but not statistically significant. -However, 80% of both groups wanted x-ray scans. —Those who received x-ray scans were more satisfied. —X-ray scan and clinical findings were not correlated.