Intended for healthcare professionals

Editor's Choice

Rebalancing risks in favour of benefits

BMJ 2024; 384 doi: (Published 28 March 2024) Cite this as: BMJ 2024;384:q758
  1. Kamran Abbasi, editor in chief
  1. The BMJ
  1. kabbasi{at}
    Follow Kamran on Twitter @KamranAbbasi

Balancing risks is life, the universe, and everything. One classic challenge is applying population based data to individuals. A related row over use of statins—the most prescribed drug class in England—shifted in favour of even greater prescription of statins last year, when NICE issued updated guidelines that lowered cardiovascular risk thresholds. Yet the benefits of statins may not outweigh the harms in people at low risk, and population risk thresholds may oversimplify matters. Guidelines, argue Sam Finnikin and colleagues, should be the starting point for medical practice—not strict rules to abide by (doi:10.1136/bmj-2023-076774).1

Three research papers this week attempt to quantify or predict risk. A retrospective cohort study finds that people with an episode of delirium have three times the risk of developing dementia (doi:10.1136/bmj-2023-077634).2 Additional episodes add to that risk by 20%. A second paper validates a simple risk prediction score for patients on intravenous cisplatin developing acute kidney injury (doi:10.1136/bmj-2023-077169).3 Acute kidney injury is strongly associated with death. Our third paper examines the use of progestogens and risk of intracranial meningioma, based on data from a French national database (doi:10.1136/bmj-2023-078078).4 Of the various progestogens, the risk associated with injectable formulations, used widely for contraception, was of particular concern.

Interpretation of these risks can be difficult. The research enterprise itself is fraught with risk, as demonstrated by large scale retractions at for-profit, open access publisher, Frontiers (doi:10.1136/bmj.q659).5 Threats to research integrity, through research fraud and misconduct, are the biggest challenges facing all publishers of science—they always were, but the scale of the threat is greater than ever and requires a strong response from editors and publishers.

But a conversation on risk becomes even more heated outside the research environment. What are the implications of poor health policies (doi:10.1136/bmj.q602)?6 If a government fails to implement patient safety recommendations (doi:10.1136/bmj.q740),7 or delays roll out of a vaccine programme to support tax cuts (doi:10.1136/bmj.q706),8 how should it be held accountable for the additional risk passed on to patients? What if it erodes prison healthcare to the point of perpetual crisis (doi:10.1136/bmj.q562)?9 Or keeps ignoring the unequivocal data on the health harms of widening inequalities and social determinants of health, leaving the solutions, such as Marmot Places, to local government (doi:10.1136/bmj.q654)?10

There is, unfortunately, no respite from risk concerns. The ongoing row over physician associates, which has damaged the Royal College of Physicians, is all about risks (doi:10.1136/bmj.q732, doi:10.1136/bmj.q737).1112 The risks associated with the GMC’s newly proposed changes to medical education are causing alarm (doi:10.1136/bmj.q728, doi:10.1136/bmj.q708).1314 The risks from industrial action (doi:10.1136/bmj.q717)15 are at the heart of a dispute between NHS England and the BMA. Some trusts claim that the BMA’s refusal to allow several derogation requests during strikes compromised patient safety. The BMA disagrees that any requests were inappropriately refused (doi:10.1136/bmj.q740).16 What’s clear is that, just as a good cricket umpire has a good relationship with the players (doi:10.1136/bmj.q399),17 to minimise risks in healthcare we need better working relationships between professions and professional organisations, regulators, and government.