Steroid shots are commonly used for back pain, but evidence that they work no better than placebos is mounting.

In a review published Monday in Annals of Internal Medicine, researchers combined data from 30 placebo-controlled studies of epidural steroid injections for radiculopathy (back pain that radiates to the legs) and eight studies of spinal stenosis (back or neck pain caused by narrowing of the spinal canal).

The study showed that for radiculopathy, injections provided some short-term pain relief, but over time were no more likely to be helpful than placebos, and they did not reduce the need for later surgery.

The pooled data showed similar results with injections for spinal stenosis — some moderate temporary pain relief, but no differences between treatment and placebo in pain intensity or functional ability lasting six weeks or longer after the shot.

The authors acknowledge that most of the trials had methodological shortcomings, and that some analyses were based on small numbers of patients.

Dr. Roger Chou, the lead author and a professor of medicine at Oregon Health & Science University, said there was probably some financial motivation for continuing to use the procedure despite the lack of evidence for its effectiveness. “The professional societies are concerned because they worry about the implications for insurance coverage,” he said. “We don’t say anything about that in the study.”

But Dr. Chou also mentioned other, less selfish impulses. “One of the reasons that people advocating for this are doing it is because they see people in horrible pain and are able to see that they do have some relief afterward,” he said. “There is a placebo effect, so there’s a contrast there with what they see anecdotally and what’s in the studies.”