What not to do with a back that has just gone

WHAT NOT TO DO

Most acute back pain settles. What decides how fast is mostly a list of things not to do, and three of the four are still being recommended.

Start with the baseline, because it changes how you read everything below. In a community inception cohort following people from the first days of an episode, most improved substantially within six weeks. The interventions on offer are competing against a condition that largely resolves on its own, which means the ones that slow resolution down do real harm even when they feel like care.

Do not go to bed

Rest is the most intuitive response and one of the least useful. Guideline reviews are consistent that advice to stay active outperforms rest for acute low back pain, and the gap is not subtle. The mechanism is not mysterious: trunk muscles lose endurance quickly, the segment stiffens, and you get up two days later with less capacity than you went down with.

Relative modification is a different thing and it is sensible. Skip the deadlifts, keep walking. What you do in the first fourteen days decides more than most of what follows.

Do not start on opioids

This is the clearest result in the recent literature and it deserves to be stated plainly. The OPAL trial randomized people with acute back and neck pain to an opioid or to placebo on top of guideline care, and at six weeks the opioid group was not better. They did have more adverse events, and at one year they reported slightly worse pain. Earlier dose-ranging work in chronic low back pain had already shown effects that are small, dose-limited by tolerability, and clinically marginal.

Our position on opioids is not abstinence and never has been. It is that starting one for an acute back is a bad trade on the evidence, and the population that arrives here on high doses usually got there one reasonable-seeming prescription at a time.

Do not get a scan in the first six weeks

Absent a red flag, early imaging does not improve outcomes. The systematic review comparing imaging against no imaging found no benefit in pain, function or quality of life, and it costs more. What it reliably produces is findings — and past forty, nearly every lumbar MRI contains disc desiccation, a bulge, or facet arthrosis that is also present in people with no pain at all.

A scan taken too early does not answer the question. It supplies a plausible culprit that then shapes every conversation afterward. Your scan described a back that belongs to somebody your age.

The findings that do earn an urgent scan are a short list, and it is short on purpose.

Do not buy the belt

Lumbar supports have been studied mostly as occupational prevention, and the evidence does not support them for that. Worn during an episode they do what a collar does to a neck: they take over the work the muscle needs to be doing. If you already own one, use it for a specific heavy task rather than as a garment.

What to do instead, in one paragraph

Keep moving within tolerance, use a simple analgesic if you want one, sleep as well as the pain allows, and give it six weeks before concluding anything. Exercise is the intervention with the most consistent support — the type matters far less than that it happens, which is the most liberating finding in this field. Physical therapy fails most often when nobody established what it was treating.

And if it has not moved at six weeks, the question changes from how to manage it to which structure is generating it. That question is answerable. There is no scan that finds facet pain, which is exactly why so many people are still looking.