Intended for healthcare professionals

Practice Therapeutics

Non-steroidal anti-inflammatory drugs (NSAIDs) for musculoskeletal pain

BMJ 2021; 372 doi: (Published 29 January 2021) Cite this as: BMJ 2021;372:n104
  1. Gustavo C Machado, research fellow1 2,
  2. Christina Abdel-Shaheed, research fellow1 2,
  3. Martin Underwood, professor3 4,
  4. Richard O Day, professor5 6
  1. 1Institute for Musculoskeletal Health, University of Sydney and Sydney Local Health District, Sydney, Australia
  2. 2Sydney School of Public Health, Faculty of Medicine and Health, University of Sydney, Sydney, Australia
  3. 3Warwick Clinical Trials Unit, University of Warwick, Coventry, UK
  4. 4University Hospitals Coventry and Warwickshire, Coventry, UK
  5. 5Department of Clinical Pharmacology and Toxicology, St Vincent’s Hospital, Sydney, Australia
  6. 6St Vincent’s Clinical School, University of New South Wales, Sydney, Australia
  1. Correspondence to: G C Machado gustavo.machado{at}

What you need to know

  • Oral non-steroidal anti-inflammatory drugs (NSAIDs) can reduce musculoskeletal pain but increase the risk of gastrointestinal (perforation, ulcers, bleeding), cardiovascular (myocardial infarction, heart failure, hypertension), and renal adverse events

  • Topical NSAIDs are also effective for osteoarthritis, with fewer adverse events than oral formulations

  • Opioids do not provide greater pain relief than NSAIDs for musculoskeletal pain and may cause serious harms such as dependence

A 65 year old patient has been troubled with intermittent low back pain for years, for which you prescribed ibuprofen 400 mg up to three times daily as needed. For the past four months, persistent knee pain, not associated with stiffness, has limited his mobility. He is overweight, physically inactive, and has recently been diagnosed with hypertension. Paracetamol has had little benefit. He asks if taking ibuprofen regularly might ease his knee pain.

What are non-steroidal anti-inflammatory drugs (NSAIDs)?

NSAIDs are commonly used for pain management. They reduce inflammation and pain by reducing the activity of cyclo-oxygenase (or COX) enzymes and inhibiting prostaglandin synthesis.1

  • Older, non-selective NSAIDs such as ibuprofen, diclofenac and naproxen inhibit both COX-1 and COX-2 enzymes

  • Newer, selective NSAIDs (COX-2 inhibitors, coxibs) such as celecoxib and etoricoxib selectively inhibit COX-2, which plays a greater role in prostaglandin mediated pain and inflammation.

The National Institute for Health and Care Excellence (NICE) guidelines recommend NSAIDs as first line analgesics for low back pain and sciatica2 and osteoarthritis.3 They are also used for musculoskeletal pain from acute injury. In England there were about 11.5 million prescriptions for oral NSAIDs in 2018.4 In the US around 60% of patients with osteoarthritis or chronic low back pain are prescribed NSAIDs, based on analysis of records in a claims database in 2008.5

Search strategy

We searched PubMed (NLM database) and the Cochrane Library with the terms “non-steroidal anti-inflammatory drugs” OR “NSAIDs” AND “spinal pain” OR “osteoarthritis” OR “musculoskeletal pain” for systematic reviews of randomised trials published from 2000 to June 2020. We prioritised systematic reviews and studies on the benefits, harms, and cost effectiveness of NSAIDs for musculoskeletal pain in primary care. We also searched clinical guidelines, our personal archives of references, and screened the reference lists of studies retrieved by the searches.

How well do NSAIDs work for musculoskeletal pain?


Spinal pain

NSAIDs are slightly more effective than placebo at reducing pain intensity and disability in people with acute and chronic low back pain,67 as per Cochrane systematic reviews. The differences are small and not likely to be clinically relevant; about 7 points on a 0–100 pain scale. There is no difference in effects of selective and non-selective NSAIDs.

Overall, five people (95% CI 4–6) with spinal pain (that is, low back or neck pain with or without radicular pain) need to be treated with NSAIDs, rather than placebo, for one additional person to achieve clinically important pain relief (>10 points on a 0–100 scale) over two weeks.8


NSAIDs marginally improve osteoarthritis pain compared with placebo over 12 weeks, based on a network meta-analysis (76 randomised controlled trials, 58 451 participants).9 More notable benefits (>10 points on a 0–100 scale) are seen with high daily doses of diclofenac (150 mg), naproxen (1000 mg), ibuprofen (2400 mg) and etoricoxib (60 mg).9 Table 1 lists effect sizes for oral and topical NSAIDs compared with placebo.

Table 1

Effects of non-steroidal anti-inflammatory drugs (NSAIDs) compared with placebo on musculoskeletal pain from systematic reviews of randomised trials

View this table:

Other arthritic and musculoskeletal conditions

NSAIDs are effective in reducing pain from axial spondyloarthritis,11 rotator cuff tendinopathy (shoulder pain),12 and acute ankle sprain,14 as per systematic reviews. There is scant and conflicting evidence for lateral elbow pain.13 Recent draft NICE guidance advises against the use of NSAIDs for some types of chronic pain, such as fibromyalgia and chronic pelvic pain.16

Topical NSAIDs


Topical NSAIDs are superior to placebo in relieving pain from osteoarthritis, as per a network meta-analysis,10 and a Cochrane review.17 Diclofenac patches were most effective for pain, and piroxicam for functional improvement. Only diclofenac patch and ibuprofen gel/cream provide clinically important pain relief (>1.2 points on a 0–10 scale).10 Pain relief is similar to oral NSAIDs in knee osteoarthritis with fewer side effects, as per two randomised controlled trials.1819 A randomised controlled trial (775 participants) found no additional benefit of combining topical and oral NSAIDs over either formulation alone for knee osteoarthritis.20 Topical NSAIDs do not increase the risk of gastrointestinal adverse events (odds ratio 1.0, 95% CI 0.7 to 1.3) or cardiovascular events (odds ratio 1.2, 95% CI 0.6 to 2.4) compared with placebo.21 European guidelines advise topical NSAIDs as first line treatment for hand osteoarthritis.22 The Cochrane review found no evidence for their use in other chronic musculoskeletal pain, such as low back pain.17

Acute injuries

Topical NSAIDs provide good levels of pain relief for musculoskeletal pain from acute injuries such as a sprained ankle or a muscle pull, as per high quality evidence from a Cochrane review (61 studies, 8386 participants). Gel formulations of diclofenac, ibuprofen, and ketoprofen and diclofenac patches are the most effective.15 Topical NSAIDs may improve lateral elbow pain (or tennis elbow), but the evidence is of low quality.13

How do NSAIDs compare with other drugs?

Paracetamol23 and opioids2425 are ineffective for low back pain and may provide modest short term pain relief in osteoarthritis, as per systematic reviews.

For acute low back pain, there is no difference between NSAIDs and paracetamol on pain intensity (mean difference −0.7, 95% CI −2.5 to 1.1), as per a Cochrane review.6 Small trials show comparable pain relief between NSAIDs and paracetamol-codeine, tramadol, or meptazinol—these effects were not clinically important.6 Most systematic reviews comparing NSAIDs with other drugs focus on pain intensity. Data on function or quality of life are often limited. Most studies consider a between-group difference of 10 points (on a 0–100 scale) as clinically important.

A Cochrane review on chronic low back pain found no difference between NSAIDs and paracetamol and pregabalin in pain intensity, based on data from small trials.7 A single trial reported greater global improvement with celecoxib compared with tramadol (risk ratio 1.3, 95% CI 1.2 to 1.4).7 There is insufficient evidence on the comparative effectiveness of NSAIDs versus other drugs for sciatica or neck pain.

For knee or hip osteoarthritis, NSAIDs are slightly more effective than paracetamol (standardised mean difference −0.3, 95% CI −0.4 to −0.2) as per a systematic review (15 randomised controlled trials).26 This small effect is equivalent to 7.3 points on a 0–100 scale. A trial with 892 participants found that combining ibuprofen with paracetamol reduces knee pain in the short term more than paracetamol but not ibuprofen alone.27 A systematic review (17 studies) found comparable reduction in pain from knee osteoarthritis with NSAIDs (−18 points), weak opioids such as tramadol (−18 points), and strong opioids such as hydromorphone and oxycodone (−19 points).28 Withdrawal rates due to toxicity were lower with NSAIDs. There was no difference between duloxetine, an antidepressant, and oral NSAIDs in osteoarthritis in a systematic review.29

What are the harms?

Safety is a concern for all NSAIDs, especially in older people, those with multi-system disorders, and persons at higher risk of cardiovascular and renal adverse events. The risk of adverse events with NSAIDs seems to be dose dependent. There is insufficient evidence on long term risks. Table 2 provides relative risks of common NSAIDs compared with placebo.

Table 2

Risk ratios (95% confidence intervals) for adverse events of oral NSAIDs compared with placebo from randomised trials30

View this table:

Gastrointestinal events

Taking any NSAID increases the risk of upper gastrointestinal adverse events (perforations, ulcers, bleeding) by two to four times compared with non-use, as per a systematic review (639 randomised trials).30 Coxibs and diclofenac yield the lowest increase in risk. Dyspepsia (indigestion) affects 8-12% of people taking high dose non-selective NSAIDs.31 The risk is 12% lower with coxibs and 66% lower with a combination of a non-selective NSAID plus a proton pump inhibitor.32

The increase in gastrointestinal risk can be up to sevenfold greater with use of high versus low/medium daily doses of NSAIDs.3033 For example, diclofenac <75 mg/day does not increase risk of gastrointestinal bleeds, but daily doses of 75-149 mg or ≥150 mg increase the risk by three and 12 times respectively compared with non-use.33 The risk seems to be higher in the first week, reduces thereafter with continuing use, and drops one week after stopping use.33 Taking two or more NSAIDs concurrently substantially increases gastrointestinal risk.33

In older adults, combining NSAIDs with aspirin or anticoagulants increases the risk of gastrointestinal bleeding by 13 times compared with non-use of either drug.34

Cardiovascular events

Coxibs and diclofenac increase the risk of myocardial infarction by about 70% compared with placebo,30 while ibuprofen 2400 mg daily doubles the risk.30 Naproxen seems to be safer, but observational data show it increases the odds of myocardial infarction by 53%.35 A trial with 24 081 participants found that celecoxib was equivalent to ibuprofen and diclofenac for cardiovascular risk, but the trial had significant methodological concerns such as high drop-out rate and low average dose of celecoxib taken.36 Observational studies find NSAIDs increase the risk of haemorrhagic stroke by a third, mostly related to diclofenac or meloxicam use.37

NSAIDs double the risk of heart failure compared with placebo in randomised trials,30 and observational data confirm this.38 NSAIDs can worsen heart failure, the risk being greater in people with renal impairment, particularly if they are taking diuretics.39 The risk of hospitalisation for heart failure while taking NSAIDs is higher in patients with established cardiovascular disease than in those without; odds ratio 10.5 (95% CI 2.5 to 44.9) versus 1.6 (95% CI 0.7 to 3.7).40

No safe NSAID treatment window exists in people with cardiovascular disease. Risk starts to increase within a week of starting NSAID treatment35 and increases further with longer use. Stopping NSAID use drops the risk level similar to that of non-users after three months.41 Diclofenac was associated with the highest risk of death and recurrent myocardial infarction immediately after the start of treatment (hazard ratio 3.3, 95% CI 2.6 to 3.9, at day 1-7 of treatment) and throughout the treatment course in a Danish nationwide cohort study.42 The increased cardiovascular risk seems to be dose dependent for celecoxib, ibuprofen, and naproxen, whereas both low and high doses of diclofenac increase risk by similar amounts.43

Renovascular events and hypertension

NSAIDs can impair renal function because COX enzymes are also involved in the synthesis of renal prostaglandins and prostacyclin. Risk ratios range from 1.3 to 2.2 among individual NSAIDs compared with non-use.44 Coxibs have a similar risk of renal failure. The risk seems to be dose dependent, regardless of the type of NSAID used—relative risk of renal failure in NSAID users (compared with non-users) is 2.5 with low or medium daily doses and 3.4 with high daily doses.45 High daily doses of NSAIDs also increase the risk of renal events in active young and middle aged adults.46

Most NSAIDs increase blood pressure, which could be a mechanism for increased cardiovascular risk. NSAID use can increase mean blood pressure by 5.0 mm Hg (95% CI 1.2 to 8.7).47 Ibuprofen, compared with celecoxib or naproxen, is associated with increased systolic blood pressure and a higher incidence of new onset hypertension.48


In the first trimester NSAIDs can increase the risk of miscarriage.49 In the third trimester, NSAIDs pose fetal risks, such as preterm birth50 and premature closure of the ductus arteriosus.51 NSAID use in the second trimester is reasonably safe, but there have been reports of ductus arteriosus constriction and oligohydramnios (deficient volume of amniotic fluid).52 Fetal monitoring from 20 weeks of pregnancy onwards is recommended in women with long term NSAID use.

Respiratory events

The prevalence of NSAID-exacerbated respiratory disease is about 9% in adults with chronic asthma.53 Coxibs are less likely to be a problem for asthmatic patients, although there have been case reports.54 NSAID use is associated with a complicated course of respiratory tract infections, as per a recent review of case-control studies.55

Despite recent concerns, there is currently no evidence that the acute use of NSAIDs causes an increased risk of developing covid-19 or of developing more severe covid-19 mortality.56 However, there is only limited data for those taking NSAIDs long term.

How are NSAIDs given and monitored?

Managing chronic musculoskeletal pain is complex. Given the potential harms of drugs, it is important to discuss the benefits and risks with your patient, and their expectations for symptomatic relief. Box 1 lists useful pointers to discuss with your patient. It is important to consider non-pharmacological therapies as well.

Box 1

Tips for patients about use of NSAIDs

  • NSAIDs provide small to moderate pain relief for back pain, osteoarthritis, and musculoskeletal pain from overuse or injury

  • Topical preparations are just as good as, and safer than, oral NSAIDs for knee or hand osteoarthritis

  • NSAIDs should be used for the shortest possible time and at the lowest effective dose (if any beneficial effect), and stopped if no benefit is experienced

  • There is no added benefit from taking two different oral NSAIDs at once, nor combining oral with topical NSAIDs, and this can be harmful

  • If you have asthma, renal disease, or a history of stomach ulcers or are taking blood pressure lowering medication or blood clotting inhibitors such as aspirin, taking NSAIDs can increase your risk of serious adverse events. Check with your doctor first before taking any NSAID

  • Bleeding from the stomach is uncommon, but taking NSAIDs can increase the risk by up to four times. Speak to your doctor or pharmacist about how this could be avoided

  • Oral NSAIDs may double the risk of developing heart disease, may increase blood pressure, and may triple the risk of kidney disease if you have impaired kidney function

  • Look out for signs and symptoms of NSAID related adverse events, such as blood in stools (black, tarry stools) or any unusual changes, including difficulty breathing, nausea, fatigue, reduced urine output, fainting, ankle or leg swelling, and chest or abdominal pain. Report these to your doctor immediately


Assess your patient’s risk profile before prescribing NSAIDs—such as age, history of gastrointestinal or cardiovascular events, hypertension, asthma, renal insufficiency, or bleeding disorders. Lower doses, and intermittent or as needed use may be safer than regular use, particularly for gastrointestinal safety. NSAIDs often have a ceiling dose, above which additional benefit is unlikely but risk of harm is more likely—for example, the analgesic ceiling dose of ibuprofen is 400 mg/dose,57 with a maximum dose of 2400 mg/day. Box 2 lists tips for safer prescribing.

Box 2

Tips for safer prescribing of NSAIDs

  • Use with caution—that is, for the shortest time and lowest effective dose possible

  • Review existing medications, such as concurrent use of anticoagulants or antiplatelet medicines, before prescribing NSAIDs

  • For knee or hand osteoarthritis, topical preparations are preferred to oral NSAIDs as they are associated with similar benefits and less risk of harms

  • Ibuprofen (≤1200 mg daily) and naproxen have the lowest cardiovascular risk but do increase risk of gastrointestinal adverse events58

  • Avoid high dose ibuprofen (2400 mg daily), diclofenac, and coxibs in people with cardiovascular conditions58

  • Moderate doses of celecoxib may be non-inferior to naproxen and ibuprofen in terms of cardiovascular safety

  • Avoid NSAIDs in people with chronic kidney disease. Regularly monitor renal function in people at risk of renal failure, including obtaining a baseline serum creatinine level when starting NSAID therapy

  • NICE guidelines advise co-prescription of a proton pump inhibitor with non-selective NSAIDs in people with musculoskeletal pain aged over 45 years3 and in those with a history of dyspepsia or past ulcer or taking antithrombotic medicines58

  • For people who have had an NSAID-related upper gastrointestinal bleed, but who must take NSAIDs, co-prescription of a proton pump inhibitor, misoprostol, or double-dose histamine blockers has been shown to prevent future events59

  • Avoid NSAIDs in pregnancy, particularly in the first and third trimester


Drug interactions are important. NSAIDs can block the antiplatelet effect of low dose aspirin used for prevention of cardiovascular disease, so combined use is discouraged; aspirin should not be stopped however. Coxibs such as celecoxib do not seem to do this as aspirin affects platelets via COX-1,60 but they can nevertheless increase cardiovascular risk.6162 The antihypertensive effect of diuretics may be negated with NSAID use because of blockage of prostaglandins responsible for natriuresis.

How cost effective are they?

NSAIDs that are available by over the counter (such as ibuprofen and naproxen) or on prescription (such as celecoxib) are relatively cheap. In England the average cost of NSAIDs per prescription item in 2018 was £5.38.4 Cost effectiveness of long term oral NSAID therapy has been evaluated for osteoarthritis in modelling studies. A modelling study found that co-prescribing a proton pump inhibitor with any NSAID is cost effective in osteoarthritis, even in those at low risk of gastrointestinal events.63 In a simulation model, naproxen or ibuprofen with a proton pump inhibitor are more cost-effective in managing osteoarthritis pain than opioids or celecoxib.64 In a decision-tree modelling, perceived optimal osteoarthritic pain relief from paracetamol, ibuprofen, or celecoxib seems to depend on the willingness-to-pay threshold (that is, the maximum amount a patient might be prepared to pay for the health benefit),65 but another modelling found celecoxib more cost-effective than paracetamol and non-selective NSAIDs.66

Education into practice

  • In which situations would you consider prescribing oral NSAIDs for musculoskeletal pain relief?

  • What are the implications of prescribing NSAIDs for musculoskeletal pain in patients taking aspirin?

  • How would you discuss the risks and benefits with your patient when prescribing NSAIDs for chronic musculoskeletal pain?

How patients were involved in the creation of this article

Two patients from a community pharmacy in Sydney reviewed this paper and had an informal interview with one of the authors. The patients suggested it was important to clarify whether taking two different oral NSAIDs at once or combining oral with topical NSAIDs result in additional benefits. They also suggested recommendations for identifying adverse events associated with NSAIDs, such as stomach bleeds, and examples of health conditions and medications where it may be best to check with a doctor or a pharmacist before taking NSAIDs. Based on their feedback, we searched for evidence and revised the manuscript and the content for the section “Tips for patients” accordingly. We thank the patients for their comments.


  • This is one of a series of occasional articles on therapeutics for common or serious conditions, covering new drugs and old drugs with important new indications or concerns. The series advisers are Robin Ferner, honorary professor of clinical pharmacology, University of Birmingham and Birmingham City Hospital, and Patricia McGettigan, clinical senior lecturer in clinical pharmacology, Queen Mary's University, London. To suggest a topic, please email us at

  • Contributors: GCM wrote the first draft of the manuscript. All authors complemented, revised, and finalised subsequent versions and approved the final version of the manuscript. GCM is guarantor.

  • Funding: None.

  • Competing interests: We have read and understood the BMJ Group policy on declaration of interests and declare the following interests: GCM holds an Australian National Health and Medical Research Council (NHMRC) Early Career Fellowship (APP1141272). Flexeze provided heat wraps at no cost for the SHaPED trial that GCM is an investigator on. CAS is an academic fellow at the University of Sydney. Flexeze provided heat wraps at no cost for the HEAT pilot trial that CAS is lead investigator on. MU is chief investigator or co-investigator on multiple previous and current research grants from the UK NIHR, Arthritis Research UK, and is a co-investigator on grants funded by the Australian NHMRC. He is an NIHR senior investigator. He has received travel expenses for speaking at conferences from the professional organisations hosting the conferences. He is a director and shareholder of Clinvivo, which provides electronic data collection for health services research. He is part of an academic partnership with Serco related to return to work initiatives. He is a co-investigator on two NIHR funded studies that receive additional support from Stryker. He has accepted honoraria for teaching/lecturing from a consortium for advanced research training in Africa. Until March 2020 he was an editor of the NIHR journal series, and a member of the NIHR Journal Editors Group, for which he received a fee. ROD has been involved with consultancies for Reckitt Benkiser on advice on ibuprofen; all payments were made to an audited trust fund at St Vincent’s Hospital. Payment to ROD was made for expert testimony regarding a veterinary anti-arthritic product unrelated to NSAIDs.

  • Transparency: The lead author (GCM) affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; no important aspects of the study have been omitted.

  • Provenance and peer review: Commissioned, based on an idea from the author; externally peer reviewed.


View Abstract