For new back pain, guidance starts with heat, not scans or opioids

You bend to lift a laundry basket and something low in your back locks. By evening, standing up from a chair takes planning. The instinct, often shared by whoever you call for help, is to escalate: a scan to find the damage, a pill stronger than anything in the cabinet, maybe the name of a spine surgeon.

The American College of Physicians, the national organization for internal medicine doctors, went the other way in 2017. Its first-line advice for new low back pain was superficial heat, such as a heating pad, or massage, acupuncture or spinal manipulation. Acetaminophen does not appear in any of its recommendations.

The placebo-controlled trials are blunter still. In a 2023 trial, people with new back or neck pain who took a short opioid course did no better than people taking dummy pills, and on average reported slightly more pain at six weeks. In an earlier trial, acetaminophen users recovered on almost exactly the same day as placebo users.

American practice had been heading the other way despite earlier guidelines. Between 1999 and 2010, the share of back and neck pain visits involving a narcotic prescription rose from about 19% to 29%. Here is what the guidelines and trials say, and the short list of warning signs that do call for fast action.

What the physicians’ guideline puts first

The American College of Physicians (ACP) published its low back pain guideline, limited to noninvasive treatments, in the Annals of Internal Medicine in February 2017, drawing on trials through April 2015 with searches updated to November 2016. It gives separate advice for newer pain, which it calls acute or subacute, and for chronic pain. The National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) counts back pain as chronic once it passes 12 weeks.

For newer pain, the guideline starts from its own observation that most people improve over time whatever treatment they get. It then puts superficial heat first, on moderate-quality evidence, with massage, acupuncture and spinal manipulation as alternatives on low-quality evidence. If a drug is wanted, it names nonsteroidal anti-inflammatory drugs (NSAIDs, the ibuprofen and naproxen family) or skeletal muscle relaxants, on moderate-quality evidence. All of that sits under one strong recommendation.

For chronic pain, the first step is again something other than a pill, chosen from a long menu. Grouped by type, it includes:

• Movement: general exercise, motor control exercise, tai chi and yoga.
• Mind and behavior: mindfulness-based stress reduction, cognitive behavioral therapy, progressive relaxation, operant therapy (a behavioral approach that reinforces activity rather than pain behaviors) and electromyography biofeedback.
• Hands-on or device-based: acupuncture, spinal manipulation and low-level laser therapy.
• Team-based: multidisciplinary rehabilitation.

Moderate-quality evidence backs only four of them: mindfulness-based stress reduction, acupuncture, exercise and multidisciplinary rehabilitation. The rest are rated low-quality. The hands-on and movement options each get a closer look in the library’s articles on manipulation, yoga, tai chi, massage, acupuncture and osteopathic care, and the mind-focused ones in its review of psychological pain treatments.

Drugs come in only after that menu has failed. NSAIDs are first, then tramadol or duloxetine, a drug from the antidepressant family. Opioids come last, only for people for whom those options have failed, only if benefits are judged to outweigh risks for that person, and only after a frank conversation about both. This whole drug step is a weak recommendation on moderate-quality evidence, unlike the strong ones for the nondrug menus.

Why a strong recommendation is not strong proof, and what is missing

The guideline prints two ratings side by side, and they often disagree. Heat rests on moderate-quality evidence, the three hands-on options on low-quality evidence, yet one strong recommendation covers all four. The strength label says how firmly the panel wants these tried first; the evidence label says how sure anyone can be about the size of the benefit.

None of the three recommendations names acetaminophen, and none sets rules for X-rays or MRI, which come from radiology and family medicine groups instead. The acetaminophen gap matters, because the pill was long described as the standard first painkiller for back pain.

The drug trials, where placebo kept pace

Acetaminophen: tested and not supported for back pain

The PACE trial, published in The Lancet in 2014, enrolled 1,652 people with acute low back pain at 235 primary care centers in Sydney, Australia. Its authors noted that acetaminophen was then the recommended first-line painkiller for this problem without high-quality evidence behind it. Participants took regular acetaminophen, acetaminophen as needed, or placebo for up to four weeks, alongside evidence-based advice.

Median time to recovery was 17 days, 17 days and 16 days. Side effects were reported by about 18.5% of each group. Australia’s national medical research council and a pharmaceutical company funded the trial.

A 2015 review in The BMJ pooled 13 placebo-controlled trials in spinal pain and hip or knee osteoarthritis. For low back pain, it rated the evidence high quality that acetaminophen did not reduce pain or disability in the short term. Across the trials, people taking it were nearly four times as likely to have abnormal liver test results, though the reviewers said the clinical importance of that was uncertain.

NSAIDs and muscle relaxants: named in the guideline, small in the trials

Start with the documented harm. In July 2015 the Food and Drug Administration toughened its heart attack and stroke warning for non-aspirin NSAIDs, noting a greater likelihood in people who already have heart disease or its risk factors. The fuller list of label cautions, from stomach bleeding to pregnancy, appears in our article on inflammation.

On benefit, a 2020 Cochrane review of 32 trials and 5,356 people found NSAIDs beat placebo for short-term pain by about 7 points on a 100-point scale, and for disability by about 2 points on a 24-point questionnaire. The reviewers called both effects small and probably not clinically relevant. Almost half the trials were industry-funded, and the review could not judge longer-term or rare harms because the trials were short.

Muscle relaxants look similar. A 2021 review in The BMJ covering 49 trials found that the main group, called non-benzodiazepine antispasmodics, cut acute back pain by roughly 8 points out of 100 within two weeks, on very low-certainty evidence, with no measurable effect on disability and possibly more side effects. The authors described the pain reduction as small and not clinically important.

That is not necessarily a contradiction: a guideline can name a drug whose benefit is real but small, for people who want a medicine. In the library’s terms, both drug classes are recommended by name in a guideline, with effects measured but below the threshold that matters. Weighing that risk against a few points of pain relief, for a particular person and medicine list, is a job for the prescriber or pharmacist.

The opioid trial

OPAL is the trial that tested the instinct most people have about severe back pain. Researchers in Sydney recruited 347 adults at 157 primary care and emergency department sites between 2016 and 2022. Everyone had 12 weeks or less of at least moderately severe low back pain, neck pain or both.

All participants received the care guidelines recommend. Half also got a deliberately cautious, short course of the opioid oxycodone combined with naloxone, for up to six weeks. The other half got identical placebo tablets, and the trial was triple-blinded, so patients did not know which they had.

At six weeks, average pain on a 10-point scale was 2.78 in the opioid group and 2.25 in the placebo group. That gap narrowly missed statistical significance, so the fair reading is that the opioid brought no benefit, with the direction favoring the dummy pill. Overall side effects were similar, 35% against 30%, while constipation ran 7.5% against 3.5%. The authors, publishing in The Lancet in 2023, wrote: “Opioids should not be recommended for acute non-specific low back pain or neck pain…”

One Australian trial of new pain does not settle every case, but its result lines up with the Centers for Disease Control and Prevention’s 2022 opioid prescribing guideline. That document lists low back pain and neck pain among common acute conditions where nonopioid treatments work at least as well as opioids, and says opioids are not recommended as first-line therapy for them. It also cites an observational study in which people who got an early opioid prescription for acute low back pain had about twice the odds of long-term opioid use, with the odds rising for larger prescriptions. That is an association, not proof the prescription caused it.

For anyone already taking an opioid, any decision about continuing or changing it belongs with the prescriber.

How often backs get scanned, and why it rarely helps

The American Academy of Family Physicians, through the Choosing Wisely campaign, advises against imaging for low back pain in the first six weeks unless red flags are present, because early imaging adds cost without improving outcomes. The American College of Radiology’s 2021 criteria go further: “Uncomplicated acute low back pain and/or radiculopathy is a benign, self-limited condition that does not warrant any imaging studies.” Radiculopathy is nerve pain running into the leg.

The strongest evidence comes from trials. A 2009 analysis in The Lancet pooled six randomized trials with 1,804 patients that compared immediate X-ray, CT or MRI against usual care without immediate imaging. Pain and function did not differ in the short term or at six to twelve months. Routine early imaging for back pain without warning signs sits in the tested-and-not-supported tier.

Many guidelines let older adults skip that wait, the authors of a 2015 JAMA study noted. Their study followed 5,239 Americans aged 65 and older with a new primary care visit for back pain; those imaged within six weeks did no better on disability a year later than closely matched people who were not. It was observational, but it points the same way as the trials.

Practice has not caught up. A review in the British Journal of Sports Medicine, published online in 2019 and pooling 45 studies from several countries and more than 19 million consultations, found about one in four primary care patients with low back pain was imaged, and about one in three in emergency departments. Complex imaging, the more advanced scans, rose from 7.4% of patients in 1995 to 11.4% in 2015 despite guideline advice. A 2013 Harvard-led analysis of US national survey data from 1999 to 2010 showed the same drift: CT or MRI use per back or neck pain visit went from 7.2% to 11.3%, and referrals to other physicians doubled, while physical therapy referrals stayed flat at about one in five.

Part of the problem is what a scan finds. Disc wear and bulges are common in people with no pain at all, a mismatch tallied in our piece on why pain doesn’t measure damage. One US trial tried a direct fix: in a secondary analysis across four health systems and nearly 239,000 patients, adding a note to spine imaging reports about how common each finding is in pain-free people did not change how many went on to spinal injections or surgery.

Where you lived shaped whether you had spine surgery

Researchers at Dartmouth mapped lumbar surgery among Medicare enrollees across 306 US hospital referral regions. Fusion, the operation that joins vertebrae together, rose from 0.3 to 1.1 per 1,000 enrollees between 1992 and 2003, and Medicare spending on it grew from $75 million to $482 million. Rates did not track how many spine surgeons a region had. Their 2006 paper in Spine put it starkly: in 2002 and 2003, fusion rates ranged nearly 20-fold among regions, the widest variation they saw for any surgical procedure.

The harm side shows up in a 2010 JAMA analysis of Medicare patients having surgery for spinal stenosis, a narrowing of the canal around the nerves. From 2002 to 2007, complex fusions rose 15-fold, from 1.3 to 19.9 per 100,000 beneficiaries. In 2007, life-threatening complications occurred in 2.3% of people having decompression alone and 5.6% of those having complex fusion, and average hospital charges were about $23,700 against $80,900.

The fair counterpoint comes from a 2019 study of national hospital data. Elective lumbar fusions rose 62% between 2004 and 2015, to about 199,000 a year. The biggest growth was for slipped vertebrae (spondylolisthesis) and spinal curvature (scoliosis), indications the authors say have relatively good evidence, and the share done for less-proven reasons slightly fell. Even so, disc degeneration, herniation and stenosis still accounted for 42% of fusions in 2015.

So the evidence does not say spine surgery is always overused. It says that in the Medicare data of the early 2000s, rates varied enormously by region, and more extensive operations carried more risk. Trials comparing surgery with physical therapy are weighed in the library’s physical therapy article.

The warning signs worth acting on

Serious causes of back pain are rare. In a Sydney study published in 2009, of 1,172 people seeing primary care for acute low back pain, 11 (0.9%) turned out to have a serious underlying condition, most often a fracture. Yet 80% had at least one of the red flags clinicians were asked to screen for. Only three fracture flags helped: prolonged steroid use, age over 70 and significant trauma.

Cauda equina syndrome, compression of the bundle of nerve roots at the base of the spine, is among the serious conditions the radiology criteria say red flags should raise. A 2019 systematic review by physical therapy researchers at Western University in Canada found its classic signs are more specific than sensitive. When they appear they mean something, and the authors said they justify prompt testing. Low sensitivity means many people with the condition lack any given sign, so a clean checklist early on is not a guarantee, and symptoms that appear later count.

The signs below come from the family physicians’ red flags, the radiology criteria and NIAMS. The two tiers are this article’s grouping, not a ranking any of those bodies publishes.

Same-day emergency care:
– New trouble starting or controlling urination, or new loss of bowel control
– Numbness in the saddle area: groin, buttocks or inner thighs
– Leg weakness that is getting worse

Promptly, without waiting it out:
– Back pain with fever, or with unexplained weight loss
– New back pain in someone with a history of cancer
– Pain after a fall or other injury, especially in someone over 70
– Sudden pain with tenderness over the spine, especially with osteoporosis, cancer or long-term steroid use
– Weakness, pain, numbness or tingling in the legs
– Pain that is not improving, or severe pain that does not ease with treatment

Questions to ask before a scan, a prescription or a referral

For an appointment about back pain without those warning signs:

1. What would a scan change about my plan right now? If the answer is nothing, that is the situation the radiology and family medicine guidance describes as not needing a scan.
2. Which first-line options fit my situation, and what will they cost me?
3. For any medicine: how big was the benefit in trials, and what are the risks for someone with my health history? That question applies to NSAIDs as much as to opioids.
4. If you are referring me, what question should the specialist answer?
5. If surgery comes up: what is the evidence for this operation for my diagnosis, and how does its risk compare with a less extensive one?
6. Which changes should bring me back sooner?

Whether any test, drug or referral makes sense is for the clinician who examines you to decide; these questions bring the reasoning into the open.

The size of the problem, and the most useful thing to know

Low back pain is the leading cause of years lived with disability worldwide, and the radiology criteria say the same for the United States. The Global Burden of Disease study estimated 619 million people had it in 2020, a global figure projected to reach 843 million by 2050. For people who have recovered from an episode, a 2024 Australian trial of 701 adults found a physiotherapist-guided walking and education program pushed the median time to the next activity-limiting episode from 112 days to 208, with more lower-limb side effects along the way.

If one idea from this article sticks, make it a reversal of instinct. For new back pain without warning signs, the evidence-backed serious response is not the scan, the stronger pill or the operating room. It is knowing the handful of signs that turn ordinary back pain into a same-day emergency: new bladder or bowel trouble, saddle numbness and worsening leg weakness above all.

Keep exploring: back belts.

Sources

1. Annals of Internal Medicine (American College of Physicians). “Qaseem A, Wilt TJ, McLean RM, Forciea MA, et al. "Noninvasive Treatments for Acute, Subacute, and Chronic Low Back Pain: A Clinical Practice Guideline From the American College of Physicians." Annals of Internal Medicine, 2017. PubMed. DOI: 10.7326/M16-2367.” pubmed.ncbi.nlm.nih.gov
2. American Academy of Family Physicians. “Choosing Wisely: Imaging for Low Back Pain ("Don't do imaging for low back pain within the first six weeks, unless red flags are present.").”
3. Journal of the American College of Radiology (American College of Radiology). “Hutchins TA, Peckham M, Shah LM, et al. "ACR Appropriateness Criteria Low Back Pain: 2021 Update." Journal of the American College of Radiology, 2021. PubMed. DOI: 10.1016/j.jacr.2021.08.002.” pubmed.ncbi.nlm.nih.gov
4. Centers for Disease Control and Prevention. “Dowell D, Ragan KR, Jones CM, Baldwin GT, Chou R. "CDC Clinical Practice Guideline for Prescribing Opioids for Pain, United States, 2022." MMWR Recommendations and Reports 71(3), November 4, 2022.” cdc.gov
5. National Institute of Arthritis and Musculoskeletal and Skin Diseases (National Institutes of Health). “Living With Back Pain: Health Information Basics for You and Your Family.” niams.nih.gov
6. U.S. Food and Drug Administration. “FDA Drug Safety Communication: FDA strengthens warning that non-aspirin nonsteroidal anti-inflammatory drugs (NSAIDs) can cause heart attacks or strokes (July 9, 2015).”
7. The Lancet. “Williams CM, Maher CG, Latimer J, et al. "Efficacy of paracetamol for acute low-back pain: a double-blind, randomised controlled trial." The Lancet, 2014. PubMed. DOI: 10.1016/S0140-6736(14)60805-9.” pubmed.ncbi.nlm.nih.gov
8. The BMJ. “Machado GC, Maher CG, Ferreira PH, et al. "Efficacy and safety of paracetamol for spinal pain and osteoarthritis: systematic review and meta-analysis of randomised placebo controlled trials." BMJ, 2015. PubMed. DOI: 10.1136/bmj.h1225.” pubmed.ncbi.nlm.nih.gov
9. Cochrane Database of Systematic Reviews. “van der Gaag WH, Roelofs PD, Enthoven WT, van Tulder MW, Koes BW. "Non-steroidal anti-inflammatory drugs for acute low back pain." Cochrane Database of Systematic Reviews, 2020. PubMed. DOI: 10.1002/14651858.CD013581.” pubmed.ncbi.nlm.nih.gov
10. The BMJ. “Cashin AG, Folly T, Bagg MK, et al. "Efficacy, acceptability, and safety of muscle relaxants for adults with non-specific low back pain: systematic review and meta-analysis." BMJ, 2021. PubMed. DOI: 10.1136/bmj.n1446.” pubmed.ncbi.nlm.nih.gov
11. The Lancet. “Jones CMP, Day RO, Koes BW, et al. "Opioid analgesia for acute low back pain and neck pain (the OPAL trial): a randomised placebo-controlled trial." The Lancet, 2023. PubMed. DOI: 10.1016/S0140-6736(23)00404-X.” pubmed.ncbi.nlm.nih.gov
12. The Lancet. “Chou R, Fu R, Carrino JA, Deyo RA. "Imaging strategies for low-back pain: systematic review and meta-analysis." The Lancet, 2009. PubMed. DOI: 10.1016/S0140-6736(09)60172-0.” pubmed.ncbi.nlm.nih.gov
13. JAMA. “Jarvik JG, Gold LS, Comstock BA, et al. "Association of early imaging for back pain with clinical outcomes in older adults." JAMA, 2015. PubMed. DOI: 10.1001/jama.2015.1871.” pubmed.ncbi.nlm.nih.gov
14. British Journal of Sports Medicine. “Downie A, Hancock M, Jenkins H, et al. "How common is imaging for low back pain in primary and emergency care? Systematic review and meta-analysis of over 4 million imaging requests across 21 years." British Journal of Sports Medicine, 2020 (online 2019). PubMed. DOI: 10.1136/bjsports-2018-100087.” pubmed.ncbi.nlm.nih.gov
15. JAMA Internal Medicine. “Mafi JN, McCarthy EP, Davis RB, Landon BE. "Worsening trends in the management and treatment of back pain." JAMA Internal Medicine, 2013. PubMed. DOI: 10.1001/jamainternmed.2013.8992.” pubmed.ncbi.nlm.nih.gov
16. Pain Medicine. “Suri P, Meier EN, Gold LS, et al. "Providing Epidemiological Data in Lumbar Spine Imaging Reports Did Not Affect Subsequent Utilization of Spine Procedures: Secondary Outcomes from a Stepped-Wedge Randomized Controlled Trial." Pain Medicine, 2021. PubMed. DOI: 10.1093/pm/pnab065.” pubmed.ncbi.nlm.nih.gov
17. Spine. “Weinstein JN, Lurie JD, Olson PR, Bronner KK, Fisher ES. "United States' trends and regional variations in lumbar spine surgery: 1992-2003." Spine, 2006. PubMed. DOI: 10.1097/01.brs.0000248132.15231.fe.” pubmed.ncbi.nlm.nih.gov
18. JAMA. “Deyo RA, Mirza SK, Martin BI, Kreuter W, Goodman DC, Jarvik JG. "Trends, major medical complications, and charges associated with surgery for lumbar spinal stenosis in older adults." JAMA, 2010. PubMed. DOI: 10.1001/jama.2010.338.” pubmed.ncbi.nlm.nih.gov
19. Spine. “Martin BI, Mirza SK, Spina N, Spiker WR, Lawrence B, Brodke DS. "Trends in Lumbar Fusion Procedure Rates and Associated Hospital Costs for Degenerative Spinal Diseases in the United States, 2004 to 2015." Spine, 2019. PubMed. DOI: 10.1097/BRS.0000000000002822.” pubmed.ncbi.nlm.nih.gov
20. Arthritis and Rheumatism. “Henschke N, Maher CG, Refshauge KM, et al. "Prevalence of and screening for serious spinal pathology in patients presenting to primary care settings with acute low back pain." Arthritis and Rheumatism, 2009. PubMed. DOI: 10.1002/art.24853.” pubmed.ncbi.nlm.nih.gov
21. Musculoskeletal Science and Practice. “Dionne N, Adefolarin A, Kunzelman D, et al. "What is the diagnostic accuracy of red flags related to cauda equina syndrome (CES), when compared to Magnetic Resonance Imaging (MRI)? A systematic review." Musculoskeletal Science and Practice, 2019 (two published corrigenda, 2019 and 2021). PubMed. DOI: 10.1016/j.msksp.2019.05.004.” pubmed.ncbi.nlm.nih.gov
22. The Lancet Rheumatology. “GBD 2021 Low Back Pain Collaborators. "Global, regional, and national burden of low back pain, 1990-2020, its attributable risk factors, and projections to 2050: a systematic analysis of the Global Burden of Disease Study 2021." The Lancet Rheumatology, 2023. PubMed. DOI: 10.1016/S2665-9913(23)00098-X.” pubmed.ncbi.nlm.nih.gov
23. The Lancet. “Pocovi NC, Lin CC, French SD, et al. "Effectiveness and cost-effectiveness of an individualised, progressive walking and education intervention for the prevention of low back pain recurrence in Australia (WalkBack): a randomised controlled trial." The Lancet, 2024. PubMed. DOI: 10.1016/S0140-6736(24)00755-4.” pubmed.ncbi.nlm.nih.gov