I’ve always been perplexed about why doctors would give a not-risk-free drug (steroids) for a problem that is not clearly ‘inflammatory’.  Giving steroids for acute radiculopathy – lower extremity nerve pain caused by a herniated disc – is, as far as I know, not recommended by any professional society and there is no evidence to support its use for this indication.  And yet it’s apparently very commonly done.  My own research supports this – I found a lot of prescriptions for steroids apparently for low back pain in the VA system.

So this study (behind a paywall; sorry) in JAMA is somewhat valuable.  Like many other studies of injectable and oral steroids for other indications, it suggests that there may be a tiny bit of value in relieving pain up front – though not much in this case: only a 0.3 point difference on the Oswetry Disability index, a measure of spinal function and pain, and not statistically significantly different from zero when compared to placebo – but no difference in long term pain and function.

I hope this will convince some physicians that this is a useless practice, but I’m not holding my breath.  Getting physicians to stop doing something is really hard.