Why paracetamol (acetaminophen) is a suitable first choice for treating mild to moderate acute pain in adults with liver, kidney or cardiovascular disease, gastrointestinal disorders, asthma, or who are older

Abstract

Acute pain is among the most common reasons that people consult primary care physicians, who must weigh benefits versus risks of analgesics use for each patient. Paracetamol (acetaminophen) is a first-choice analgesic for many adults with mild to moderate acute pain, is generally well tolerated at recommended doses (≤4 g/day) in healthy adults and may be preferable to non-steroidal anti-inflammatory drugs that are associated with undesirable gastrointestinal, renal, and cardiovascular effects. Although paracetamol is widely used, many patients and physicians still have questions about its suitability and dosing, especially for older people or adults with underlying comorbidities, for whom there are limited clinical data or evidence-based guidelines. Inappropriate use may increase the risks of both overdosing and inadequate analgesia. To address knowledge deficits and augment existing guidance in salient areas of uncertainty, we have researched, reviewed, and collated published evidence and expert opinion relevant to the acute use of paracetamol by adults with liver, kidney, or cardiovascular diseases, gastrointestinal disorders, asthma, or/and who are older. A concern is hepatotoxicity, but this is rare among adults who use paracetamol as directed, including people with cirrhotic liver disease. Putative epidemiologic associations of paracetamol use with kidney or cardiovascular disease, hypertension, gastrointestinal disorders, and asthma largely reflect confounding biases and are of doubtful relevance to short-term use (<14 days). Paracetamol is a suitable first-line analgesic for mild to moderate acute pain in many adults with liver, kidney or cardiovascular disease, gastrointestinal disorders, asthma, and/or who are older. No evidence supports routine dose reduction for older people. Rather, dosing for adults who are older and/or have decompensated cirrhosis, advanced kidney failure, or analgesic-induced asthma that is known to be cross-sensitive to paracetamol, should be individualized in consultation with their physician, who may recommend a lower effective dose appropriate to the circumstances.

Keywords:

• Acetaminophen
• acute pain
• older adult (frail)
• liver
• kidney
• gastrointestinal diseases
• asthma
• cardiovascular diseases

Introduction

People worldwide use paracetamol (acetaminophen) more than any other analgesic^1–3; besides being readily available over-the-counter (OTC), its popularity rests on being generally well tolerated and considered safer than other analgesics when taken as directed, especially in high-risk users such as older people or adults with comorbidities^1–6. Accordingly, international guidelines and medical experts consistently recommend paracetamol for first-line treatment of acute mild/moderate pain^6–8.

Paracetamol has antipyretic and analgesic effects (Figure 1), with complex mechanisms that distinguish it from non-steroidal anti-inflammatory drugs (NSAIDs) and opioids^1,2,6,9. Its most common everyday uses in adults include headaches, and musculoskeletal, or menstrual pain⁷; musculoskeletal pain and sprains/injury account for 45% of pain episodes for which adults may use OTC or prescription analgesics (Figure 2)¹⁰. The standard adult dose of paracetamol is 0.5–1.0 g, taken as needed at least 4–6 h apart, to a maximum of 4 g/day; it is recommended to use the lowest dose that relieves symptoms, taken for the shortest possible treatment duration^8,11,i.

Figure 1. Paracetamol (N-acetyl-p-aminophenol) pharmacotherapeutic mechanisms, metabolism, and toxicity. Abbreviations. TRPV, transient receptor potential cation channel subfamily V; CYP2E1, cytochrome P450 2E1; COX, cyclooxygenase.

Figure 2. Pain episodes from 6527 diary entries by adults aged 18–65 years. Data source: Hersch et al.¹⁰.

Paracetamol very rarely has adverse effects in healthy adults who take ≤4 g/day in episodic use^2,12,ii and as directed by the label, but doses exceeding the recommended daily maximum may be harmful due to the accumulation of a toxic metabolite, N-acetyl-p-benzoquinoneimine (NAPQI), that can cause liver failure (Figure 1)^2,4,12. Although some epidemiological studies have also implied that paracetamol use may contribute to kidney or cardiovascular disease (CVD), gastrointestinal (GI) disorders, and asthma, especially among chronic users, these findings remain inconclusive and of unproven relevance to episodic acute use^2,4,13,14. Nevertheless, such observations have raised questions about the suitability of paracetamol for people with underlying comorbidities or who are elderly, and whether dose reduction might be appropriate in certain circumstances^4,14–16.

Such concerns compound widespread lack of awareness, especially among older people, which conduces to inappropriate use that may increase the likelihoods of both accidental overdose and inadequate pain management^3,5,17,18. Many overdoses are unintentional, because people do not know the recommended maximum dose or are unaware of the presence or amount of paracetamol as an ingredient of numerous OTC medications^2,7,17. On the other hand, some adults take doses lower than are recommended⁸ and inadequate pain relief may lead them to use stronger analgesics with well-documented side effects^5,12.

Acute pain is one of the most common reasons that people use primary care services^7,10 and despite available consumer information, many still have questions about paracetamol, especially regarding its safety and dosing for people with specific health concerns³. In a survey of adults with liver disease, 60.5% said they would ask a physician before choosing an OTC analgesic containing paracetamol¹⁷. Physicians recommending any analgesic must be confident that its benefit will outweigh potential harms^2,7,14. This necessitates thorough appraisal of each patient’s circumstances, based on which they can be advised which treatment options may be suitable, and on what dose to take. Healthcare professionals should be best placed to do this but may themselves be uncertain about the appropriate use of paracetamol, especially for patients with special health concerns, for whom there are limited clinical data or specific dosing instructions^3,5,16,17. Pertinently, patient information cautions people with liver or kidney disease, or who have aspirin-sensitive asthma, to consult their physician before using paracetamol, but does not specify whether or how dosing should be adjusted^11,19.

To address knowledge gaps and augment existing guidance, we have reviewed published evidence and expert opinion pertaining to the acute use of paracetamol by adults with liver or kidney disease, CVD, GI disorders, asthma, or who are older. Briefly, independent researchers shortlisted articles retrieved from PubMed searches that combined the terms “paracetamol” or “acetaminophen” with others chosen to retrieve articles relevant to each of the aforementioned areas of potential health concern, which the authors had prespecified. The authors checked that selected articles were within the scope of their review and suggested others based on their specialist knowledge. Reference lists of reviewed articles were perused to identify additional sources. Although this review discusses the use of GlaxoSmithKline products in this context, the authors strove to remain impartial and produce an objective, evidence-based summary that reflects the current consensus of scientific expertise in these areas. We encourage physicians and other healthcare professionals to use this resource to help people make informed choices about pain management.

Liver disease and hepatotoxicity

Many patients and physicians do not know whether paracetamol is suitable for people with liver disease – potential hepatotoxicity is a particular concern^17,20–22. Paracetamol is almost entirely metabolized by the liver (Figure 1): ∼80% is eliminated as inactive glucuronide (40–65%) and sulphate (25–35%) conjugates but 5–10% is oxidized by cytochrome P450 2E1 (CYP2E1) to generate NAPQI, a highly reactive intermediate that is normally rapidly neutralized by conjugation with glutathione, which is constantly replenished^4,12,22–24. However, if NAPQI accumulates faster than it is detoxified (e.g. due to paracetamol overdose or, hypothetically, in glutathione-depleted states) it can cause acute necrotic liver failure^4,6,23,24. Theoretically, pathologies that affect this pathway could increase the risk of paracetamol-induced hepatotoxicity^25,26; for example, hyperglycemia and starvation induce CYP2E1 activity²⁷, and malnutrition may deplete hepatic glutathione, potentially abating NAPQI detoxification^23,25. Expert reviewers, however, maintain that paracetamol at therapeutic doses generates insufficient NAPQI to exhaust glutathione, and that there is no convincing clinical evidence that malnourished individuals who use paracetamol as directed are at increased risk of hepatotoxicity^23,25. Despite the lack of good-quality evidence, some precautionary guidance nevertheless suggests that people weighing <50 kg should take paracetamol at lower doses (contrary to the drug label)²⁵. Studies on how diabetes influences paracetamol metabolism found limited effects on its conjugation and clearance pathways and there has been no suggestion that people with diabetes should use a lower dose^28,29.

Perceived risk of hepatotoxicity may lead patients and physicians either to believe incorrectly that paracetamol is absolutely contraindicated in liver disease and opt for other drugs that may be less appropriate (e.g. NSAIDs), or use too low a dose to relieve pain satisfactorily^{12,17,20–22}. The challenges of pain management in patients with liver disease are compounded by the dearth of prospective studies of analgesic pharmacology and safety in this population and lack of specific evidence-based dosing recommendations. As a result, prescribing and usage practices vary considerably and pain may be undertreated^{12,17,19–22}.

Reappraisal of reported paracetamol poisoning cases has suggested that acute hepatotoxicity is much rarer when used as its label recommends than was commonly supposed⁴; in retrospect, many purported cases can actually be attributed to unrecognized overdosing, which is mostly unintentional and often occurs in the context of alcoholism^{4,12,20,24,30,31}. Nonetheless, there has been conjecture as to whether hepatotoxicity at therapeutic doses of paracetamol might be more likely in people with liver disease^12,20,22,32, especially chronic alcoholics, who may have depleted glutathione reserves and potentiated oxidation to NAPQI^{4,12,20,22,23}. Relevant research over 40 years has consistently shown negligible risk of hepatotoxicity from short-term use of paracetamol at standard doses up to 4 g/day by patients with liver disease, including moderate to heavy drinkers (Table 1)^{12,21–23,32–35}. Although patients with severe liver disease who took single doses of either 1 g or 1.5 g paracetamol had a longer plasma half-life versus controls, there was no evidence of impaired glutathione conjugation or clinical hepatotoxicity, nor of drug accumulation after taking 3 g/day for 5 days^33,36. Likewise, there were neither accumulation nor evidence of hepatotoxicity in patients with stable chronic liver disease who took 4 g/day paracetamol for 5 days consecutively, and no clinically significant changes after 13 days continuous use³². In patients with cirrhosis, including recent alcohol users, episodic doses of ≤3 g/day paracetamol during the previous month were not associated with acute hepatic decompensation (ascites, variceal bleeds, encephalopathy, jaundice)³⁷. In other studies, there was no evident liver damage in newly abstinent chronic alcoholics who took 4 g/day paracetamol for 3 days³⁵, or in moderate drinkers who took 4 g/day for 10 days³⁴.

Table 1. Effects of short-term use of standard oral doses of paracetamol in adults who have liver disease, including alcoholic cirrhosis.

First author, year	Patient population	Study design (analytic cohort)	Paracetamol dosing	Main findings
Benson^32, 1983	Stable chronic liver disease	Open-pilot study in patients with advanced cirrhosis (n = 6)	1.0 g 4/day, 5 days consecutive	Clinical status and lab test results unchanged  – No progressive increase in plasma paracetamol level  – Paracetamol concentration never approached potentially toxic levels
		Double-blind randomized crossover to paracetamol or placebo (n = 20)	1.0 g 4/day, 13 days consecutive	No exacerbation of clinical features nor abnormal lab test results
Andreasen^33, 1979	Hepatic cirrhosis	Compared patients with cirrhosis (n = 11) vs without liver disease (n = 12)	Two 0.5 g tablets simultaneously then…	No clinical or biochemical signs of hepatotoxicity Cirrhotic patients had significantly higher plasma paracetamol levels, longer half-live (3.7 vs 2.1 h) and slower clearance (162 vs 337 mL/min)
		Cirrhosis (n = 4) vs controls (n = 9)	1.0 g 3/day, next 5 days	No paracetamol accumulation in blood
Heard^34, 2007	Moderate alcohol drinkers^a	Double-blind randomized to paracetamol (n = 100) vs placebo (n = 50)	Two 0.5 g capsules simultaneously 4/day, 10 days consecutive	No subjects met criteria for hepatotoxicity or liver failure or developed clinically evident liver injury  – Paracetamol group had increased mean ALT at day 11
Kuffner^35, 2007	Newly abstinent chronic alcoholics^b	Double-blind randomized to paracetamol (n = 258) vs placebo (n = 114)	Two 0.5 g capsules simultaneously 4/day, 3 days consecutive	No subjects had increased ALT, AST, or other markers of liver injury  – No statistical difference between paracetamol vs placebo in subgroup with alcoholic hepatitis
Forrest^36, 1979	Chronic liver disease	Compared groups with severe (n = 7) or mild (n = 8) liver disease, and normal controls (n = 8)	1.5 g single administration	No evidence for increased risk of hepatotoxicity in people with severe liver disease  – Mean plasma half-life significantly longer in severe liver disease group (4.3 h) vs mild liver disease (2.2 h) and normal groups (2.4 h)
Khalid^37, 2009	Hepatic cirrhosis^c	Cases with decompensated cirrhosis (n = 91) compared with compensated cirrhotic (n = 153) and non-cirrhotic (n = 89) controls	Highest doses in two cirrhotic controls: 3.0 g/day for 2 days ∼1.0 g/day for 25 days	Paracetamol at maximal doses of 3.0 g/day for ≤ 2 days/month or 1.0 g/day for ≤ 25 days/month is not associated with acute hepatic decompensation in cirrhotic patients, even those with recent alcohol ingestion

Abbreviations. ALT, alanine aminotransferase; AST, aspartate transaminase.

^aSubjects consuming an average of 1–3 alcoholic drinks/day.

^bIncluding 44 people fulfilling case definition for alcoholic liver disease.

^cIncluding 89 patients with alcoholic cirrhosis.

Collectively, these findings constitute persuasive evidence that short-term use (<1–2 weeks) of up to 4 g/day paracetamol is an appropriate treatment for mild to moderate pain in patients with cirrhotic liver disease, even moderate to heavy drinkers^12,20,22,38. No high-quality evidence supports reducing the maximum recommended dose of paracetamol (4 g/day) for short-term use by patients with compensated cirrhosis¹⁹. It is difficult to make dosing recommendations for people with decompensated alcoholic liver disease due to insufficient data on this group − 3–4 g/day short-term for acute pain may be reasonable¹². Experts also concur that 2–3 g/day may be more prudent if used by alcoholics for more than 14 days, without significantly compromising analgesia^{12,19–22,38}.

The potential risks and benefits of paracetamol for patients with liver disease must be weighed against those of alternative treatment options, such as NSAIDs or opioids^4,21,23. Given its established safety profile and lack of sedating or nephrotoxic effects, paracetamol is the analgesic of choice irrespective of liver disease etiology, whereas NSAIDs are best avoided due to increased risk of GI bleeding, renal failure, and antiplatelet effects^{4,12,17,21,22,38}.

Cardiovascular disorders

People with CVD or cardiovascular (CV) risk factors account for a significant proportion of the primary care caseload^39,40. To relieve musculoskeletal pain in such patients, the American Heart Association has advocated stepwise escalation that starts by using analgesics with the least associated risk of adverse CV effects, usually paracetamol or aspirin at the lowest effective dose, and suggested paracetamol initially for people at risk of GI bleeding⁴¹. If first-line treatment is ineffective or poorly tolerated, a non-selective NSAID or cyclo-oxygenase 2 (COX-2) inhibitor may be considered but should be used cautiously, due to increased risks of serious CV or GI complications^39,41–44. Contrastingly, there is no conclusive evidence associating paracetamol with increased risk of CV events in patients with or at risk of CVD^39,42,45. Many patients with CV comorbidities require polypharmacy, which substantially increases the likelihood of drug-drug interactions; using paracetamol instead of NSAIDs has been recommended to avoid potential interactions that may warrant avoiding a combination or modifying therapy⁴⁰. Accordingly, CVD patients worldwide often use paracetamol for pain relief^43,46.

The pharmacology of paracetamol differs from NSAIDs and has only recently become better understood^2,6,9. Given such widespread use, even slightly increased risk to potentially vulnerable patients would raise concern^13,45,47; however, studies of CV-related outcomes in paracetamol users have produced inconsistent results^43,48. In particular, there are insufficient high-quality data to assess hypothetical CV risks associated with paracetamol in typical episodic use⁴⁹. Further studies, especially randomized controlled trials (RCTs), may help to resolve remaining uncertainties^43,45,49.

Hypertension

Although some observational studies have associated paracetamol use with elevated blood pressure (BP)¹⁴, there is inconclusive evidence of clinically significant hypertensive effects from episodic acute use^49,50. Reported epidemiological associations between chronic/frequent paracetamol use and incident hypertension^51–53, may be attributable in part to confounding factors⁴⁹; for example, the more often people are prescribed analgesics by a physician, the likelier they are to have their BP measured and to have hypertension diagnosed¹³. Other investigators have found no association between paracetamol and hypertension^54,55. A retrospective cohort of adults with hypertension who were first prescribed NSAIDs had a 2 mmHg increase in systolic BP at 1-year follow-up relative to new paracetamol users, who were matched by propensity scores for baseline BP, age, and cardiometabolic comorbidities⁵⁶. Besides clearly increasing BP and the risk of serious CV events in both normotensive and hypertensive individuals, NSAIDs directly interfere with the actions of many antihypertensive and diuretic agents, whereas paracetamol does not appear to^{39,42,49,52,57}; reassuringly, there is no signal that hypertensive paracetamol users have increased risks of myocardial infarction or stroke⁴⁵, which have been associated with BP elevation by just a few mmHg during follow-up studies^39,49. In this context, it is noteworthy that the sodium salts in soluble paracetamol products can significantly raise BP^58,59, which may partly explain apparent associations between paracetamol use and hypertension⁶⁰; healthcare professionals should advise hypertensive patients to use sodium-free formulations⁵⁹.

Small prospective RCTs have also investigated whether paracetamol affects BP^14,49. The earliest found a 4 mmHg rise in systolic BP in 20 hypertensive patients who took 3 g paracetamol per day (two 500 mg tablets, 8-hourly for 4 weeks)⁶¹. However, subsequent RCTs found no significant BP changes in people with mild/moderate hypertension^57,62,63. In another RCT, 33 patients with coronary heart disease who took 1 g paracetamol thrice-daily for 2 weeks had significantly increased ambulatory BP, which raised concern about using paracetamol in such patients^49,64; however, the validity and clinical relevance of this study have been challenged⁴⁷. To augment this limited and conflicting evidence, a recent British RCT investigated the effect of giving 110 people with hypertension paracetamol 1 g four times per day for 14 consecutive days, typifying the dosing schedule used to treat chronic pain⁶⁵. Uninterrupted use of 4 g/day paracetamol increased systolic BP by 4.7 mmHg versus placebo, suggesting that regular long-term use of paracetamol may be inadvisable for people with risk factors for CVD⁶⁵; notably, however, the authors saw no grounds for such concern about short-term episodic use in this setting⁶⁶.

CV endpoints

The CV safety of widely used analgesics has become a major public health concern, especially given evidence accrued since 2011 that NSAIDs which are still often used by patients with CVD have no risk-neutral treatment window; even at low doses for <7 days, non-aspirin NSAIDs have been associated with increased risk of thrombotic CV events⁴⁴. A Swedish national study found heart failure to be associated with NSAIDs but not paracetamol⁶⁷. Among 70,971 United States (US) female nurses without known CVD, occasional paracetamol use (1–21 days/month) was not associated with increased CV risk; although women who took paracetamol frequently (≥22 days/month) were increasingly likely to have CV events compared with non-users, with greatest risk at ≥15 tablets/week⁴⁸, the analytic methodology has been criticized¹³. A 2016 meta-analysis also reported a dose-response relationship between paracetamol and adverse CV events¹⁵ but its findings too have been disputed, particularly concerning inadequate treatment of potential biases^13,68. A 10-year follow-up study of 10,878 British hypertensive patients found no relationship between paracetamol use or exposure and the risk of any CV event or stroke⁴⁵. In a recent cohort of frail nursing home residents with multimorbidity and polypharmacy, neither paracetamol use nor dosing affected CV mortality, although participants with diabetes had increased stroke risk⁶⁹.

Available evidence does not indicate that episodic use of paracetamol at recommended doses is associated with any additional risk of major CV events^43,45; it can therefore be considered the first-choice analgesic for mild to moderate acute pain in most people with CV risk factors, especially those liable to NSAID-related GI bleeding^6,39,41.

Gastrointestinal disorders

Paracetamol is generally accepted to have excellent GI tolerability and is consequently an analgesic of choice for patients at risk for adverse GI events^4,6,70. Unlike aspirin and some other NSAIDs that reportedly increase the risk of GI bleeding at any dose⁷¹, paracetamol appears to have negligible risk of causing or exacerbating GI ulcers or related complications^6,70. Prospective studies have shown that paracetamol does not damage the GI tract⁴, and a meta-analysis found no significant effect of paracetamol at any dose on upper GI bleeding⁷². Epidemiological studies of health records have produced conflicting results; although some found no association between paracetamol use and GI bleeding⁷⁰, others have suggested that the incidence of adverse GI reactions may increase dose-dependently^73–75. However, these positive associations must be interpreted cautiously, as they could reflect inherent biases and confounding factors that may invalidate inferences about causality. A likely explanation is that physicians recommend paracetamol preferentially over NSAIDs to patients at high risk for gastropathy because they consider paracetamol to be safer (confounding by indication), while widespread use of paracetamol means it is often the first-line treatment for GI pain due to a pre-existing condition (protopathic bias)^4,70. For example, observations that paracetamol users sometimes experience dyspepsia⁷⁵ probably reflect that paracetamol is the preferred analgesic for patients with GI risk factors⁷⁰. A recent analysis of the Japanese Adverse Drug Event Report (JADER) pharmacovigilance database affirmed that paracetamol causes less GI injury than NSAIDs¹.

Based on evidence that paracetamol lacks major GI toxicity in short-term use and has GI safety superior to that of many NSAIDs^1,4,6,70, paracetamol can be recommended as a first-line treatment for mild to moderate acute pain in patients at risk of GI ulceration or bleeding^6,70.

Nevertheless, it is important for physicians to be aware of how GI disorders such as gastric dysmotility, which is a common complication of type 2 diabetes⁷⁶, GI ulcers, or other conditions, may affect the absorption and/or pharmacokinetics of paracetamol⁷⁷. Absorption of orally administered paracetamol depends on its dissolution in the stomach and the gastric emptying rate⁷⁸. A rapid-dissolving oral paracetamol formulation that enables faster absorption with less inter/intra user variability, may be more suitable than standard paracetamol tablets for people with delayed gastric emptying^79,80. Sixteen patients with gastric or duodenal ulcers had prolonged and increased absorption of 1 g paracetamol relative to healthy controls and slower clearance, particularly among those with duodenal ulceration, suggesting that reduced paracetamol dosing may be appropriate⁸¹; however, larger studies would be needed to corroborate this provisional evidence. Conversely, people with ulcerative colitis had no difference versus healthy controls in the clearance of 1 g peroral paracetamol or its metabolites⁸², which would support these patients using a standard dose. The jejunum is the major site of paracetamol absorption, and people with short bowels are likely to have impaired uptake of oral doses⁷⁷; a rectal paracetamol suppository may be administered in such cases but will have slower, less complete absorption and a less predictable effect compared with an oral dose^2,6,77.

Kidney disease

Many people with chronic kidney disease (CKD) need pain relief; some 40–60% of those receiving dialysis and 60–70% or more with advanced/end-stage renal disease are reported to suffer pain, predominantly musculoskeletal^83,84. However, high-quality evidence in this setting is limited, and diverse prescription practices suggest uncertainty as to the most expedient pharmacotherapeutic approach⁸⁵. Consequently, pain management is suboptimal in many such patients, which diminishes their quality of life and imposes additional healthcare burdens^83,86.

There has been speculation that paracetamol may be associated with renal impairment^87,88. Historically, realization that classic analgesic nephropathy was caused by phenacetin overuse raised concerns that paracetamol, its primary metabolite, might have similar injurious renal effects^88,89; however, this hypothesis has been dismissed^88–91. Although acute nephrotoxicity has been documented in rodents given paracetamol doses of ≥500 mg/kg and following overdoses in a few clinical cases^87,90,91, there is no evidence that episodic use of paracetamol alone at recommended doses causes renal damage, nor that cumulative lifetime doses exacerbate advanced CKD^{83,87,88,90–92}.

Epidemiologic inferences that paracetamol may increase users’ risk of developing CKD are contentious; contradictory findings of observational studies suggest that putative associations might be due to protopathic bias and/or confounding by indication rather than signifying underlying causation^{4,87,88,90,93}. In particular, first-line use of paracetamol in preference to other analgesics by people with early renal impairment may lead to spurious associations with incident CKD in observational studies^13,87. Conversely, besides increasing the risk of GI bleeding, NSAIDs have proven nephrotoxic class effects and should be avoided where possible in patients with symptoms of renal impairment, as should COX-2 inhibitors which induce similar adverse effects^{83,84,86,90,94}. Unlike COX-2 inhibitors, there have been no reports that therapeutic doses of paracetamol are associated with acute renal failure⁹⁵. Analysis of the large real-world JADER pharmacovigilance database supports the acute renal safety of paracetamol versus NSAIDs¹.

Prevailing guidance from the US Kidney Foundation and other nephrology experts, advocates stepwise escalation of analgesia, starting with non-opioids for mild pain, and recommends paracetamol as the preferred first-line analgesic for episodic treatment of mild pain in patients with renal dysfunction, CKD, and/or requiring dialysis^{83,84,86,90,92,95}. Concordantly, a recent meta-analysis found paracetamol to be the mainstay of treatment for mild to moderate pain in CKD patients⁸⁵. Dose modification is not necessary for patients on dialysis, who may take 1 g orally 3–4 times per day⁸⁶. Among CKD patients on hemodialysis, 500 mg peroral paracetamol has been used to treat headaches, shivering, or fever⁹⁶. However, dose minimization may sometimes be warranted⁸⁴; for example, a maximum of 3 g/day has been recommended for patients with advanced kidney failure⁸³.

Asthma

Several studies over the years have investigated the hypothesis that adulthood asthma might be related to paracetamol use^97–105; however, almost all were non-interventional association studies (Table 2)¹⁰⁶, and only one analyzed whether occasional use of paracetamol affects symptom severity in adults with established asthma¹⁰¹. Consequently, there is a paucity of high-quality clinical evidence on which to base guidance¹⁰⁷.

Table 2. Reported associations between frequency of paracetamol exposure and asthma incidence or prevalence.

First author, year	Design (setting)	Analytic cohort(s) (age range)	Outcome of paracetamol use versus never-use	Association with prior exposure to paracetamol (OR/RR, 95% CI)^a
				← Lower		Higher →
Shaheen^101, 2000	Case/control (UK general practices)	644 asthmatics vs 910 non-asthmatics (16–49 years)	OR for prevalent asthma	Less than monthly: 1.06 (0.77–1.45)	Monthly: 1.22 (0.87–1.72)	Weekly: 1.79 (1.21–2.65)*	Daily: 2.38 (1.22–2.65)*
Barr^97, 2004	Prospective cohort (USA)	346 women with incident asthma (30–55 years)	RR for incident asthma	1–4 days/month: 1.27 (0.96–1.66)	5–14 days/month: 1.43 (0.99–2.07)	15–21 days/month 1.78 (1.04–3.04)*	>22 days/month: 1.53 (0.95–2.46)
				<14 days/month: NR		>14 days/month: 1.63 (1.11–2.39)*
Davey^100, 2005	Prospective cohort (Ethiopia)	78 people with self-reported asthma (5–≥70 years)^b	OR for self-reported asthma	1–3 doses/month: 1.36 (0.93–1.99)		>3 doses/month 1.28 (0.76–2.15)
McKeever^103, 2005	Cross-sectional (USA)	934 people with physician-diagnosed asthma (20–80 years)	OR for prevalent asthma	1–5 days/month: 1.15 (0.99–1.34)	6–29 days/month:1.40 (1.12–1.75)*		≥30 days/month: 1.75 (1.34–2.28)*
Kelkar^99, 2012	Retrospective case/control (USA)	28,892 people with incident asthma vs 86,676 non-asthmatics (≥18 years)	OR for incident asthma	Recent pre-index prescription		Chronic pre-index prescription
				>1/30 days: 0.97 (0.90–1.04)	≥1/7 days: 1.02 (0.93–1.11)	≤30/365 days: 1.07 (1.03–1.11)*	>30/365 days: 1.39 (1.27–1.53)*

Abbreviations. CI, confidence interval; NR, not reported; OR, odds ratio; RR, risk ratio.

^aAdjusted for confounding variables.

^bProbably includes some children, as there was no breakdown of people with asthma by age.

*Bold type indicates statistical significance.

Epidemiological studies have consistently shown that paracetamol is associated with asthma incidence/prevalence, especially with more frequent exposure^{97,99,101–104}. These positive associations do not prove causality and may be partly attributable to confounding factors or biases, particularly people with asthma avoiding NSAIDs and preferentially using paracetamol. Nevertheless, accruing evidence supports a weak association between having asthma and frequent paracetamol use^{14,99,101–104,106,108,109}. Asthma etiology hypothetically involves depletion of glutathione, an antioxidant that abates airway inflammation^{101,102,108,109}, though some doubt that this is significant at standard therapeutic doses of paracetamol²³. Glutathione depletion also suppresses T-helper 1 production, thus favoring allergenic T-helper 2 responses^102,108. Other biologically plausible mechanisms are antigenic effects of paracetamol that stimulate immunoglobulin E responses, and COX inhibition that increases leukotriene levels and reduces prostaglandin E2 production^{14,99,104,108}.

On the other hand, epidemiological studies found no significant associations with asthma among the least frequent paracetamol users (Table 2)^{97,99,101,103}. In one study, increasing asthma severity was associated with weekly or daily paracetamol use, whereas monthly users were no more likely to have asthma compared with never users¹⁰¹. However, in the only RCT to date in adults, 1 g paracetamol taken twice daily for 12 weeks had no statistically significant effect on the severity or control of stable mild to moderate asthma, probably ruling out overt morbid effects with short-term, low-dose use¹⁰⁵.

Analgesic-induced asthma

Aspirin and other NSAIDs provoke bronchospasm and exacerbate asthma in up to ∼20% of asthmatics who are hypersensitive to COX inhibitors; these acute reactions may be severe, even life-threatening^1,107–111. Although most people with asthma tolerate paracetamol well, its weak COX inhibitory action has dose-dependent bronchospastic effects in 20–30% of individuals with aspirin-induced asthma (AIA) who are cross-sensitive to paracetamol^107,109. In one study, 17 of 50 patients with AIA had bronchospastic reactions to paracetamol challenges of 1 g/day or 1.5 g/day¹¹². However, less than 2% of asthmatics overall are hypersensitive to paracetamol, and such reactions are generally mild, occur at high doses, and are not known to have caused fatalities^{107,109,111,112}. Most people with asthma, even those with AIA, have little risk of bronchospasm at paracetamol doses of <650 mg¹¹². The JADER database showed little sign of asthma induction with real-world use of paracetamol compared with NSAIDs¹.

Given the risks of hypersensitivity and GI bleeding associated with NSAIDs, paracetamol may be preferable for short-term treatment of mild to moderate pain in people with asthma^{6,14,99,101,107,110,111}. Although there are no RCTs on the respiratory effects of short-term standard doses of paracetamol in adults with established asthma, available evidence suggests that episodic paracetamol use does not exacerbate respiratory symptoms^{1,14,99,105,109}. However, people with AIA that is known to be cross-sensitive to paracetamol should use the lowest effective single dose, for example, ≤500 mg per dosing occasion^{14,109,110,113}.

Older age

Pain is considerably more common among people older than 60 versus younger adults and affects up to half or more of older community-dwellers, with even higher prevalence in elderly care home residents^114,115. But despite older adults being among the highest users of analgesics, their pain is often inadequately treated and significantly impairs their quality of life^5,114–120. Managing pain in older people is often complicated by comorbidities, potential drug-drug interactions from polypharmacy, and age-related physiological and pharmacokinetic/dynamic changes^5,119–122; however, a “start low and go slow” approach may make physicians over-cautious about up-titrating analgesic dosing as necessary, leading to undertreatment¹¹⁸. In such circumstances, paracetamol may be the first choice for treating mild to moderate acute pain^6,16,69,122, especially musculoskeletal¹¹⁵, and is often prescribed to avoid using NSAIDs associated with increased risks of GI bleeding and renal dysfunction in older people^5,16,118,121.

Drug interactions

Paracetamol may alter the International Normalized Ratio of oral anticoagulants, notably warfarin, potentially increasing the risk of bleeding^25,122–124. There is no convincing evidence that other potential drug interactions with paracetamol are clinically significant (Table 3)^25,124.

Table 3. Potential drug interactions with paracetamol and their clinical relevance.^a

Drug action	Interacting drug(s)	Potential effect of interaction	Clinical evidence & implications
Antioxidant	N-acetylcysteine	Attenuates NAPQI-induced hepatotoxicity	Principal antidote for paracetamol overdose
Ion exchange substrate	Activated charcoal, cholestyramine	Reduced paracetamol absorption	Oral activated charcoal ≤ 1 h after paracetamol ingestion treats an overdose, cholestyramine reduces absorption if given within 1 h
Anticoagulant	Warfarin/coumarin	Potentiated anticoagulation, elevated warfarin INR, risk of bleeding may increase	Assiduous INR monitoring during and 1 week after concomitant paracetamol is prudent
Gastric emptying retardant	Propantheline, anti-cholinergic agents, opioid analgesics	Decrease paracetamol absorption rate	Effects have limited clinical importance and do not cause toxicity in therapeutic use
Accelerate gastric emptying	Metoclopramide	Increase paracetamol absorption rate	Lower peak paracetamol concentration but total clearance unaffected
	Cisapride		No clinically significant interaction
Anticonvulsants	Phenytoin	Increase paracetamol elimination rate	No good evidence for hepatotoxicity and possibly hepatoprotective.
	Carbamazepine^b	May increase NAPQI production but data are equivocal	No good evidence for hepatotoxicity, anecdotal association may reflect confounding factors
	Phenobarbitaone	Suppress glucuronidation	No good evidence for increased risk of hepatotoxicity
Antibiotics	Chloramphenicol	May increase chloramphenicol concentration or clearance	Conflicting evidence precludes firm conclusions
	Rifampicin^b	Increase paracetamol elimination rate	No good evidence of increased hepatotoxicity.
Tuberculosis drugs	Isoniazid^b	CYP2E1 induction, increased NAPQI production after discontinuing isoniazid	Effect too small to significantly potentiate hepatotoxicity
Antiviral (NTRI)	Zidovudine	Inhibited glucuronidation	Studies have not confirmed interactions at therapeutic doses; no good evidence for hepatotoxicity or hepatoprotection
Proton pump inhibitor	Omeprazole	Increased NAPQI production	No good evidence that omeprazole increases hepatotoxicity
H2 histamine receptor antagonists	Ranitidine	Inhibit glucuronidation	No noticeable change in efficacy
	Cimetidine	Decrease NAPQI clearance	No clinically significant interaction reported
Gout treatment	Probenecid	Inhibits glucuronic acid conjugation, reducing paracetamol clearance	Consider reducing paracetamol dose if taken concomitantly
Tyrosine kinase inhibitors	Sorafenib, dasatinib, imatinib	Reduced glucuronidation may increase NAPQI production	No good evidence that tyrosine kinase inhibitors increase paracetamol toxicity

Abbreviations. INR, International Normalized Ratio; CYP2E1, cytochrome 450 2E1; NAPQI, N-acetyl-p-benzoquinoneimine; NTRI, nucleoside reverse transcriptase inhibitor.

^aData sources: Caparrotta et al.²⁵; Sharma and Mehta¹²³; Toes et al.¹²⁴.

^bDocumented hepatotoxicity.

Paracetamol pharmacokinetics and pharmacodynamics in older people

Although some guidance advocates reduced dosing of paracetamol for older people, there is limited supporting evidence, because this population is under-represented in clinical trials and the pharmacodynamics of paracetamol in older people have not been studied systematically^5,118,120. Consequently, questions about whether or how therapeutic doses should be adapted for both frail and robust older people remain unresolved^120,125. Understanding the pharmacology of paracetamol in older people can help to improve pain management¹¹⁸.

Conjecture that gastric emptying slows with advancing age is based largely on indirect findings or small cohorts of aged people with related comorbidities (e.g. Parkinson’s disease, diabetes) and/or medication use^126,127. Healthy older people, however, generally have gastric emptying rates within the normal range, and modestly slower emptying than in younger people is unlikely to be clinically significant^126–128. Paracetamol absorption was not impaired in healthy older study subjects and time to maximum plasma concentration also appeared unaffected by age (Table 4)^{120,128–131,133}; however, exposure to a 1 g intravenous dose increased with age because paracetamol had a longer elimination half-life¹³⁰ and somewhat slower clearance in older people compared with young adults^130–133, with creatinine clearance rather than age determining the clearance rate¹²⁹. Paracetamol users aged 84–95 years who took three 1 g doses per day for 5 days consecutively had no drug accumulation¹³², and although frail inpatients older than 70 who took 3–4 g/day for 5 days did have higher plasma concentrations than younger patients, none had abnormally elevated serum alanine aminotransferase, which suggests that slower clearance of therapeutic paracetamol doses in geriatric patients may not necessarily increase the risk of hepatotoxicity¹¹⁷. A study of 36 inpatients ≥80 years old who received 3 g/day paracetamol (concomitant with other analgesics), found wide inter-individual variations in all pharmacokinetic metrics; however, only seven had an average plasma concentration above the predefined analgesic target of 10 mg/L¹²⁵. Pharmacokinetic analyses in 30 fit, older people given 1 g intravenous paracetamol after knee surgery also found large (unexplained) variability; in population simulations, 1 g/6 h achieved a steady state of 10 mg/L on average, whereas a longer dosing interval of 1 g/8 h resulted in underdosing for most patients¹³⁴.

Table 4. Studies comparing paracetamol absorption, concentration, and elimination in older versus younger adults.

First author, year	Participant population and stratification			Paracetamol dosing	Main findings (reported in older vs younger subjects, respectively)
	All subjects	Analysis groups^a	Age range (mean)
Gainsborough^128, 1993	Healthy  volunteers	Elderly (n = 19)	69–86 (75.0)	Oral tablets at ∼20.0 mg/kg	No significant difference in median T1/2abs (0.12 vs 0.11 h) or Tmax (0.57 vs 0.43 h) Gastric emptying does not become impaired with older age
		Young (n = 19)	20–33 (24.8)
Triggs^129, 1975	Apparently  healthy  individuals	Geriatric (n = 7)	73–91 (80.9)	0.8% solution at 14.3 mg/kg	Non-significant differences in absorption rate, Tmax (0.74 vs 0.67 h), volume distribution (1.05 vs 1.03 L/kg) or clearance (379 vs 477 mL/min/1.73 m^2) Older group had significantly longer T1/2 (130 vs 109 min) Plasma clearance correlated significantly (but not strongly) with creatinine clearance, not with age
		Young (n = 6)	22–30 (24.0)
Liukas^130, 2011	Orthopedic  surgery  patients	Older (n = 10)	60–70	1.0 g intravenous	Significant associations of age with total clearance (age 80–90 vs 20–40: 3.3 vs 4.6 mL/min/kg) and T1/2 (age 80–90 vs 20–40: 3.6 vs 2.7 h), but not with Tmax (age 80–90 vs 20–40: 0.26 vs 0.31 h) Drug exposure increased with age: 36–68% higher in 70–80 and 80–90-age groups vs ≤ 40-year-olds
		Older (n = 10)	70–80
		Oldest (n = 10)	80–90
		Young (n = 10)	20–40
Divoll^131, 1982	Considered  healthy^b	Older (n = 16)	61–78 (67.9)	0.65 g intravenous	Similar T1/2 in older vs young groups (mean 2.7 h) Older group had significantly lower weight-adjusted volume distribution Mean weight-adjusted clearance was lower in older vs young subjects of each sex, but differences had borderline significance
		Young (n = 16)	24–37 (29.0)
Bannwarth^132, 2001	Hospital inpatients	Very old (n = 12)	84–95 (89.0)	Two 0.5 g capsules simultaneously 3/day, 5 days consecutive	No drug accumulation despite mean plasma clearance ∼50–66% of that for healthy young adults given 1 g oral paracetamol (5–6 mL/min/kg) Longer mean T1/2 compared to value usual in young subjects (2.7 vs 2.0 h)
Miners^133, 1988	Healthy young or  older^c men	Elderly (n = 8)	72–92 (79.3)	Two 0.5 g tablets simultaneously	No significant difference in Tmax (0.76 vs. 0.87 h) or Cmax (14.5 vs 13.9 mg/L) No significant difference in apparent oral clearance, T1/2, or clearance to glucuronide or glutathione conjugates
		Young (n = 8)	18–26 (20.8)
Mitchell^117, 2011	Hospital  inpatients	Young (n = 11)	18–55 (35.6)	3–4 g/day, 5 days consecutive	All older vs younger participants had significantly higher mean day 5 paracetamol concentrations, with the highest in frail subjects who took 4 g/day ALT in frail paracetamol users remained within the normal range throughout
		Fit older (n = 12)	≥70 (81.7)
		Frail older (n = 13)	≥70 (85.4)

Abbreviations. T_max, time to maximum drug concentration; C_max, peak drug concentration; T_1/2abs, absorption half-life; T_1/2, elimination half-life; ALT, alanine aminotransferase.

^aAs designated in respective data sources.

^bThree older subjects were taking medications for cardiovascular disease, which was fully compensated and stable.

^cOlder group excluded men with evident renal hepatic or cardiovascular disease.

Despite decreasing volume of distribution and clearance of paracetamol with advancing age^16,118, most older people tolerate 1 g/6 h per day well, with no need for dose reduction provided physicians have carefully considered patient-specific risk factors such as liver disease, heavy alcohol intake, or renal insufficiency^16,119,121. Although liver damage from standard therapeutic doses (0.5–1 g/4–6 h, max ≤4 g/day) rarely occurs in healthy older people¹⁶, it may nevertheless be prudent to consider lower or less frequent doses in those with risk factors for hepatotoxicity, which include weighing <50 kg, renal insufficiency, concomitant medications inducing liver enzymes, hepatitis or chronic alcohol use, or glutathione-depleted states¹⁶; 0.5 g four times per day or 1 g thrice daily may be suitable for such patients¹⁶.

Speculation that paracetamol could increase the risk of hypothermia in vulnerable older adults remains unsubstantiated^135–137. Rare case reports have linked pediatric hypothermia to therapeutic doses of paracetamol¹³⁵ and limited experimental data suggest that paracetamol (20 mg/kg lean body mass) may impair thermogenesis in adult subjects during cold stress, but not in thermoneutral conditions^136,137. However, the mechanisms are not well understood and there is currently no evidence that paracetamol at a standard dose is a factor in adult cold-related hypothermia deaths^135–137.

In summary, there is insufficient evidence to conclude that reduced paracetamol clearance with age is clinically significant, or to support routine dose reduction for older patients^119,120,133. Although some guidelines provide general dosing advice, specific recommendations for managing acute pain in older people are lacking^16,120. Based on pharmacokinetic data, adjusting the dose of paracetamol is generally unnecessary^131,133. A dose of 0.5–1 g every 4–6 h (maximum of 4 g/day) is probably appropriate for healthy older people, while it may be advisable for very elderly or frail people or those otherwise at increased risk for hepatotoxicity to take the lowest effective dose, not exceeding 3 g/day^16,132.

Conclusions and recommendations

Notwithstanding prevalent self-treatment with non-prescription OTC analgesics, acute pain is still among the most common reasons for visits to primary care physicians^7,10. Before recommending any analgesic, physicians must first weigh the potential risks versus benefit for each patient^2,7,14. Although paracetamol is very widely used, many patients and some physicians have misconceptions about its suitability and dosing, especially for people with underlying comorbidities or who are older^3,5,17. Such uncertainties are compounded by under-representation of these patient populations in clinical studies and consequent lack of evidence-based guidance, particularly specific dosing recommendations, which results in inconsistent prescription and usage practices^{3,5,16,17,19,22,85,118}. While it is important to educate people that the maximum recommended dose of paracetamol for healthy individuals is 4 g/day, people should also avoid using too low a dose to relieve pain, which may lead them to use stronger, less safe, analgesics^{2,5,12,17,21,22,83}. Given a lack of large-scale RCTs of paracetamol involving users with health concerns such as liver disease, older age and others reviewed herein, it becomes necessary to make recommendations based on pharmacokinetic/pharmacodynamic data, observational studies, expert opinion, and real-world pharmacovigilance^1,12,19,118. Considering this, it is important to bear in mind that because the perceived safety of paracetamol makes it the first-choice analgesic for many patients with comorbidities, epidemiological associations between paracetamol and hypertension, GI disorders, renal disease, and asthma could be largely attributable to protopathic bias and/or confounding by indication^4,13.

The weight of evidence accrued over many years indicates that adults who take paracetamol episodically (<14 days) at recommended doses have almost negligible risk of serious adverse effects^2,4,6,12; accordingly experts concur that paracetamol should still be considered a first-line analgesic for acute mild to moderate pain in patients with liver or kidney disease, CVD, GI disorders, asthma, and/or who are older^{6,7,14,16,20,21,24,38,39,70,83,84,99,107,110,111,115}. Dose reduction is generally unnecessary but may be warranted in certain circumstances (Table 5); therefore, advice on appropriate dosing for paracetamol users with comorbidities or who are older should be based on careful consideration of relevant individual-specific factors^16,19.

Table 5. First-line analgesia for acute mild to moderate pain in adults with special health concerns.

User group	Recommendations and expert advice for paracetamol use and dosing^a
All adults	Before taking paracetamol, users should carefully read the product information issued in their country/region regarding their specific formulation and pain indication The maximum recommended dose of paracetamol is 4 g/day, dosed as needed at ≥ 4–6-h intervals Use the lowest effective dose, for no longer than is necessary to relieve symptoms People whose symptoms persist or who experience adverse effects should seek medical advice
Liver disease	Use paracetamol in preference to NSAIDs Paracetamol ≤ 4 g/day is suitable for episodic acute^b use by patients with compensated cirrhosis, including moderate to heavy drinkers Paracetamol 3–4 g/day is suitable for patients with decompensated alcoholic liver disease; however, it is prudent to reduce the maximum recommended dose to 2–3 g/day for alcoholics who need to use paracetamol for > 14 days
Cardiovascular disease	Use paracetamol in preference to NSAIDs associated with increased risk of adverse CV events Episodic acute^b use of ≤ 4 g/day paracetamol poses very little, if any, increased risk of adverse CV events for most patients Hypertensive patients should use sodium-free paracetamol formulations
Gastrointestinal disorders	Use paracetamol in preference to NSAIDs in patients with risk factors for GI ulcers or bleeding Short-term use of paracetamol at recommended doses poses negligible risk of causing or exacerbating GI ulcers or related complications Dose reduction is generally unnecessary. Weak evidence that patients with gastric or duodenal ulcers may have increased paracetamol absorption and slower clearance is insufficient to support dose adjustment without corroborating studies.
Kidney disease	Use paracetamol in preference to NSAIDs, which have nephrotoxic class effects Dose modification is not necessary for patients on dialysis, who may take 1 g paracetamol orally as needed ≤ 3–4 times per day Paracetamol is suitable for patients with chronic kidney disease; however, patients with advanced kidney failure are recommended to use a reduced dose of ≤ 3 g/day
Asthma	Use paracetamol in preference to NSAIDs, which may exacerbate asthma symptoms in hypersensitive users Infrequent/short-term^c use of paracetamol at recommended therapeutic doses does not increase asthma severity People with analgesic-induced asthma that is known to be cross-sensitive to paracetamol, should use the lowest effective dose (e.g. ≤0.5 g) per dosing occasion
Older age (>60 years)	Paracetamol 0.5–1 g every 4–6 h (≤4 g/day) is suitable for healthy^c older people A reduced maximum dose of 0.5 g four times per day or 1 g three times per day is appropriate for older people with risk factors for hepatotoxicity (body weight < 50 kg; renal insufficiency; concomitant medications inducing liver enzymes; hepatitis or chronic alcohol use; glutathione depletion)

Abbreviations. NSAIDs, non-steroidal anti-inflammatory drugs; CV, cardiovascular; GI, gastrointestinal.

^aThis summary is intended only to supplement official product information and is not exhaustive.

^bShort-term use for treating intermittent acute pain episodes (e.g. headache, menstrual), as opposed to uninterrupted long-term use to treat chronic pain (e.g. osteoarthritis).

^cPeople without comorbidities or conditions to which product contraindications, warnings, or precautions apply.

Transparency

Declaration of funding

This work was funded by GlaxoSmithKline Consumer Healthcare. GlaxoSmithKline authors conceived the review and were involved in establishing its terms of reference, and selected literature. GlaxoSmithKline employees reviewed manuscript drafts and suggested revisions, but the authors were independent in deciding which content and recommendations to include in their final version.

Declaration of financial/other relationships

John Alchin has received honorarium payments for serving as a member of the GlaxoSmithKline Global Pain Faculty. Arti Dhar and Kamran Siddiqui are employees of GlaxoSmithKline Consumer Healthcare. Paul J. Christo has received honorarium payments for serving as a member of the GlaxoSmithKline Global Pain Faculty, has also been a consultant/advisor to Eli Lilly and Y-mAbs Therapeutics, Inc., and undertakes paid media work via Algiatry, LLC; he receives grant funding from the US National Institute on Aging. Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

AD and KS conceived the review. JA and PC advised on the review scope and literature search parameters. All authors revised manuscript drafts for important intellectual content and approved the final version submitted.

Acknowledgements

The authors thank Yen-May Ong and Huw Williams PhD of MIMS, Singapore, and David Neil PhD, (affiliated to MIMS, Singapore, at time of writing), for providing medical writing support, which was funded by GlaxoSmithKline Consumer Healthcare in accordance with Good Publication Practice (GPP3) guidelines (http://www.ismpp.org/gpp3).

Notes

i Based on package leaflet for Panadol optizorb, 500 mg film-coated tablets Paracetamol. GlaxoSmithKline Dungarvan Ltd., Ireland. Users should read their own package leaflet, as exact recommendations vary between countries and products.

ii This review defines “episodic” as, short-term use as needed to treat intermittent acute pain episodes (e.g. headache, menstrual), as opposed to uninterrupted long-term use to treat chronic pain conditions (e.g. osteoarthritis).