Paracetamol, known as acetaminophen in the United States, may have more risks than originally thought, particularly when it is taken at the higher end of standard therapeutic doses, according to a new systematic review.
The authors and an outside expert recommend caution when interpreting the data, as they are observational in nature and are subject to uncontrolled confounders. That said, the authors do note that the dose–response curves seen for each adverse outcome examined suggest “a considerable degree of paracetamol toxicity especially at the upper end of standard analgesic doses.”
Emmert Roberts, from South London and the Maudsley Mental Health Trust, Maudsley Hospital, London, United Kingdom, and colleagues present their findings in an article published online March 1 in BMJ.
“Paracetamol is the most widely used over-the-counter and prescription analgesic worldwide. It is the first step on the [World Health Organization] pain ladder and is currently recommended as first-line pharmacological therapy by a variety of international guidelines for a multitude of acute and chronic painful conditions,” the authors write.
They conducted a systematic literature review to determine the adverse event profile of paracetamol by searching Medline and Embase from the date of inception to May 1, 2013. They identified observational studies written in English that reported mortality, cardiovascular, gastrointestinal, or renal adverse events in adults in the general population who took standard analgesic doses of paracetamol.
Ultimately, they included eight of 1888 studies retrieved. All of the included studies were cohort studies. The researchers assessed study quality using Grading of Recommendations Assessment, Development and Evaluation. They pooled or adjusted summary statistics for each outcome.
Both studies that examined mortality risk among adults who took paracetamol and those who did not found an elevation in overall risk. In one study, the standardized mortality ratio was 1.9 (95% confidence interval [CI], 1.88 – 1.94) for those taking the drug. The other study showed an overall risk of 1.28 (95% CI, 1.26 – 1.30), as well as a dose-response increase in the relative rate of mortality from 0.95 (95% confidence interval [CI], 0.92 – 0.98) at the lowest exposure, compared with nonusers, to 1.63 (95% CI, 1.58 – 1.68) at the highest exposure.
Of four studies that reported cardiovascular adverse events, all found a dose-response, with one study demonstrating an increased risk ratio of all cardiovascular events from 1.19 (95% CI, 0.81 – 1.75) at the lowest exposure to 1.68 (95% CI, 1.10 – 2.57) at the highest.
One study that reported gastrointestinal adverse events found a dose-response with relative rate of gastrointestinal adverse events or bleeding increasing from 1.11 (95% CI, 1.04 – 1.18) to 1.49 (95% CI, 1.34 – 1.66).
Four studies reported adverse events; of those, three found a dose-response, with one that reported an odds ratio of 30% or more decrease in estimated glomerular filtration rate increasing from 1.40 (95% CI, 0.79 – 2.48) to 2.19 (95% CI, 1.4 – 3.43).
“Because this literature review was based on long-term observational data, there are many potential biases that could influence the results, so it cannot be called ‘hard’ data at all,” study author Philip Conaghan, MBBS, PhD, professor of musculoskeletal medicine, University of Leeds; consultant rheumatologist, Leeds Teaching Hospitals National Health Service Trust; National Institute for Health Research senior investigator; and deputy director, National Institute for Health Research Leeds Musculoskeletal Biomedical Research Unit, United Kingdom said.
“For example, one confounder that is impossible to measure is the use of over-the-counter medicines, which are usually not recorded and can include drugs with significant side effects, such as ibuprofen. Of course it’s almost impossible to get long-term data from clinical trials: They usually don’t run for many years, so we are dependent on this sort of imperfect data to explore long-term potential drug side-effects,” Dr Conaghan said.
“I don’t think this study is reproducible because of the softness of the data. [Also], that kind of risk profile is very hard to imagine is meaningful,” Norton M. Hadler, MD, emeritus professor of medicine and microbiology/immunology, University of North Carolina at Chapel Hill, said.
Implications for Clinical Practice
The first thing clinicians should do when reading studies like this is to closely examine the methods of the study and not simply rely on the abstract, Dr Hadler noted.
Moreover, clinicians should ask themselves whether a patient needs medication in the first place, Dr Hadler said. Although over-the-counter medications are generally safe, it makes sense for clinicians and patients to try nonmedication ways of relieving pain first.
Dr Conaghan agrees. “First they should assess if paracetamol is needed for a given patient. It might not add much to people also taking other pain killers such as [nonsteroidal anti-inflammatory drugs] or opioids. Second, they should ask their patients about all their pain killers, including over-the-counter pills, to get a complete picture of analgesic use (note [that nonsteroidal anti-inflammatory drugs] are analgesics too). Thirdly, they should be conscious that people using moderate to high doses of paracetamol over long periods of time may be more prone to certain side effects that they need to look out for,” Dr Conaghan added.
“I am assuming that the common long-term use of paracetamol is for musculoskeletal pain in this response…. [T]here is a massive need for pain control with ageing communities, increased levels of back pain and osteoarthritic joint pain, and lots of people can’t tolerate aspirin and ibuprofen,” he concluded.
He also noted that it is worth reassessing every so often whether the drug is still helping the patient. “That might mean stopping it for a couple of days and seeing if it makes much difference to their pain. Then they have to consider if they are doing the simple things that effectively improve joint pain (if that’s their problem) without side effects; for example, muscle strengthening exercises followed by increased physical activity, and weight loss if needed, all help knee pain. Fitting these things into busy lives is difficult, but ultimately they are more effective and safer than pills,” Dr Conaghan explained.
“[W]e should consider the benefit-risk ratio for particular conditions, and would need to see where paracetamol has demonstrated benefits. A recent study in Lancet suggested paracetamol wasn’t effective for treating acute lower back pain, although its safety was good over the 4-week period of that study,” Dr Conaghan said.