Acute low back pain: what the new 2026 guideline says about imaging and opioids

Most low back pain guidance in general circulation, including our own broader guide to low back pain, leans on the 2017 American College of Physicians guideline and NICE’s NG59, both built mostly on evidence for chronic and subacute pain. This new guideline is narrower and more current: a multidisciplinary panel, developed in partnership with the American Academy of Pain Medicine (AAPM), built it specifically for acute low back pain, pain lasting less than six weeks, using GRADE methodology applied to a purpose-built systematic review of the literature through March 2025, with a bridge search in May 2026.^[1][2]

The guideline itself was published online in Pain Medicine on September 2, 2026, with its companion systematic review following later that month.^[1][2] Trade press coverage followed in early October. The timing matters less than the scope: this is the first major guideline update specifically for acute low back pain management since the 2017 ACP guideline, and it leans further against routine imaging and opioid prescribing than that document did.

The guideline’s clinician-facing summary is direct on this point: do not routinely obtain imaging for atraumatic acute low back pain lasting less than six weeks, reserving it for cases with red-flag symptoms or suspected serious spine pathology.^[3]That recommendation rests on the systematic review’s own findings, not just precedent. Looking specifically at randomized trials of early imaging versus no imaging in patients without signs of serious spine pathology, the reviewers found no difference in pain or function outcomes between the two approaches.^[2]

This echoes a point already well established in other guidelines: structural findings on a scan (disc bulges, degeneration, and the like) are extremely common in people with no back pain at all, which is why a scan this early rarely changes management and can lead to overtreatment of incidental findings. Imaging still belongs in the picture when red flags are present, after significant trauma, or when pain has not improved after four to six weeks.

The clinician summary is equally direct here: do not routinely use opioids as initial treatment for acute low back pain.^[3] The systematic review behind that advice looked at opioids versus placebo and found they generally produced similar pain and function improvement, while being associated with an increased risk of adverse events.^[2]The review’s own confidence ratings are worth stating plainly: low confidence that opioids outperform placebo for pain or function, and moderate confidence that they carry a higher risk of harm.^[2]

The guideline does not ban opioids outright. It leaves room for a short course of a short-acting opioid, at the lowest effective dose for the shortest time, for severe and disabling pain that has not responded to other options, or when other medications cannot safely be used.^[3] It also states explicitly that it should not be used to restrict or deny coverage of a medication a clinician has legitimately prescribed.^[3] The change is in the default starting point, not an absolute rule.

Several options cleared a higher evidence bar in the review. Across trials pooled for oral medications, NSAIDs and muscle relaxants improved function and/or pain versus placebo in the first few weeks, while acetaminophen did not.^[2] For nonpharmacologic options, the review found that manipulation and mobilization, acupuncture, and simple advice to stay active all improved function and/or pain versus sham treatment or usual care, and early referral to physical therapy (which included similar exercise content) improved pain and function versus usual care.^[2]

The reviewers were careful to flag two limits worth repeating here. First, results from exercise-specific trials (as distinct from the broader physical therapy referral pathway) were mixed, so no single exercise protocol stood out as clearly superior.^[2] Second, improvement across most of these interventions, where it occurred, showed up within the first few weeks and the effect sizes were generally small, which is consistent with the natural course of acute low back pain: most episodes improve substantially on their own regardless of which of these options is added.

For the subset of patients with radiculopathy (pain radiating down a leg from nerve-root irritation), the review found that systemic steroids and epidural steroid injections both improved function versus placebo or usual care, with epidural injections also improving back pain specifically.^[2] The review graded its confidence in these benefits as low, and the clinician summary reflects that caution, advising that epidural steroid injections be reserved for cases with significant symptoms after appropriate clinical consultation, rather than used early and routinely.^[2][3]

For a fuller picture of leg-pain-specific guidance, including when imaging and specialist referral are actually warranted, see our separate guide to sciatica, which this new guideline’s radiculopathy findings are broadly consistent with.

Beyond opioids and routine imaging, the clinician summary specifically advises against carisoprodol, citing its abuse potential, and does not include benzodiazepines among the treatment options it covers at all.^[3]Acetaminophen is kept as a fallback for patients who cannot safely take an NSAID, rather than a routine first choice, reflecting the review’s finding that it did not improve outcomes in the trials assessed.^[2]

None of this is a reversal of prior advice; it is a sharpening of it, built on a dedicated acute-pain evidence base rather than one built primarily for chronic pain. The broad shape (stay active, be skeptical of early imaging, lead with lower-risk options, escalate if things are not improving after a couple of weeks) matches what our existing low back pain guide and neck pain guide already describe for the broader spine-pain picture. What this guideline adds is a harder line specifically on opioids, direct evidence that early imaging does not improve outcomes in low-risk acute presentations, and explicit confidence grading for each option rather than a general consensus statement.

It is also worth separating this guideline from the PACBACK trial we covered in September, which asked a different question: whether early, clinician-supported self-management can prevent acute pain from becoming chronic. This guideline is about what to do for acute pain in the moment; PACBACK is about reducing the odds that today’s acute episode becomes next year’s chronic problem. The two are complementary, not competing.

The guideline’s case against routine imaging applies specifically to people without warning signs. A smaller set of symptoms should prompt assessment rather than a wait-and-see approach.

Outside of these patterns, the guideline’s central message holds: most acute low back pain improves with time and activity, and the lower-risk options it favors are a reasonable place to start before escalating to stronger medications or imaging.

This guideline does not say back pain does not matter or that nothing helps. It says the evidence, now reviewed specifically for the acute phase, supports starting with education, staying active, heat, and short-term NSAIDs over reflexive imaging or opioids, and it grades its confidence in each recommendation honestly rather than presenting all of them as equally certain. Options like physical therapy, manipulation, and acupuncture have moderate-confidence support if symptoms persist past one to two weeks; steroid injections for nerve-root pain have a thinner evidence base and a narrower role. See how we grade evidence strength across topics in our editorial and evidence standards, and browse the rest of our Mobility & Pain coverage.

None of this replaces an individual assessment from a clinician who can rule out the red-flag presentations above and account for your specific history. It reflects what a purpose-built systematic review and the guideline panel built from it actually found, not a universal protocol for every case of back pain.