Where are we now with paracetamol?

Will the treatment work for my clinical problem?

Paracetamol is recommended in guidelines produced by agencies such as the UK National Institute for Health and Care Excellence. It is an effective mild analgesic but may not work for all types of pain. Paracetamol is effective for postoperative dental pain.3 For headache, it is superior to placebo but less effective than other analgesics.4 For the common cold, the evidence is insufficient to draw conclusions.5 Randomised trial data report that paracetamol is no more effective than placebo for back pain.2 For hip and knee pain, meta-analysis suggests that paracetamol has a small benefit that may not be clinically relevant.6 Further efficacy trials are needed, but, until these are available, paracetamol will remain a first line analgesic in part because of the belief that it is safe in therapeutic doses.

Are there side effects?

Paracetamol does not make people feel unwell but there are recurrent concerns about increased risk of major harm. Liver injury is a well established consequence of paracetamol overdose, and there are also concerns about therapeutic doses. Studies have shown that around 25-40% of healthy volunteers will develop a small and often transient increase in alanine transaminase activity, a biomarker for liver injury, after ingesting therapeutic doses for one to two weeks.7 Among patients with osteoarthritis, those taking paracetamol are reported to be four times more likely than others to have a modest rise in alanine transaminase.6 What does this mean in practice? Decades of clinical experience with paracetamol suggest that at therapeutic doses serious liver injury is unlikely. For example, among 673 patients admitted to a UK liver transplantation unit with paracetamol induced liver injury only four reported taking 4 g or less (although there were still four, which does not provide complete reassurance).8 Furthermore, there were no reported cases of fulminant hepatic injury in a systematic review of 30 865 patients taking therapeutic doses of paracetamol in clinical trials.9

Concerns about the cardiovascular safety of non-steroidal anti-inflammatories and cyclo-oxygenase-2 (COX 2) inhibitors may increase the use of paracetamol in patients with raised cardiovascular risk. However, some studies suggest that it has an adverse cardiovascular safety profile.1 10 Paracetamol has been shown to inhibit COX 2, which has the potential to increase blood pressure and promote thrombosis. Given that paracetamol is widely used and hypertension is common it is surprising that this interaction is not more clearly defined. Human mechanistic studies are needed.11 Nevertheless, a recent study of over 24 000 patients from the UK did not show any association between paracetamol and myocardial infarction or stroke.12 While observational data such as these may be confounded, they provide some reassurance, and the study should be repeated across international datasets.

Pregnant women often take paracetamol because non-steroidal anti-inflammatories have been associated with adverse fetal outcomes. Recent data from rodents suggest that paracetamol for one week, at equivalent doses to those taken by humans, adversely affects testosterone production from fetal testes.13 Although this finding may explain a reported increased risk of cryptorchidism in infants with more than two weeks’ exposure in utero,14 the increase was of marginal significance in other observational studies.15 16 Furthermore, in the rodent model, a short course of paracetamol had no adverse effect. Current guidance advising pregnant women to take paracetamol at the lowest effective dose for the shortest time is appropriate pending further data.

How many tablets do I take?

Paracetamol has a relatively narrow therapeutic index, and it is important that patients understand the implications. Following review in 2012, UK guidance was issued that small therapeutic overdoses should be considered for treatment with the antidote acetylcysteine “as per clinician judgement.” The UK position differs from international guidance (the United States and Australia use higher ingestion thresholds to trigger treatment 17 18) and may need revisiting because a substantial increase in the number of patients admitted to hospital for treatment has been reported since the guidance changed.19 A review could identify new measures to reduce the incidence of therapeutic overdose and reduce the number of low risk patients being unnecessarily treated with a time consuming antidote that commonly produces adverse reactions and results in substantial hospital bed occupancy.

Although the toxicology of paracetamol overdose is well understood, pharmacology at therapeutic doses, and in various clinical indications, is less well characterised. Overall, paracetamol can be less effective than patients and doctors expect. We need further studies to define efficacy in specific settings because the argument that paracetamol is first line treatment because it is safe does not hold if it is ineffective. Cardiovascular safety needs to be established and the management of therapeutic excess needs revisiting. The Medicine and Healthcare Products Regulatory Agency is reviewing all over the counter analgesia, which will hopefully go some way to answering patients’ questions. For now, prescribers should establish whether patients are getting symptom relief from paracetamol to avoid long term exposure without benefit.