• patient safety
• medical errors
• malpractice
• primary care

“Medications that clean bile and phlegm are a source of danger, and of blame for the person treating”. Hippocrates. Affections 33.¹

The tangled relations between disease, treatment, patient harm, medical fallibility, and physician culpability have been debated since classical times. But it is only historically recently that actions alleging negligence by doctors have become a commonplace feature of the health care landscape.

One hundred years ago an experienced Scottish judge, while hearing a legal case against an Edinburgh general practitioner (see box), commented on its rarity: “This action is certainly one of a particularly unusual character. It is an action of damages against a medical man. In my somewhat long experience I cannot remember having seen a similar case before.”²

Only a century later the medicolegal landscape of health care could hardly be more different. In the year 2000 the UK General Medical Council received 5000 complaints which alleged doctors’ misconduct or poor performance and National Health Service (NHS) hospitals in England faced 23 000 outstanding claims for compensation.^3,4 The annual incidence of NHS written complaints concerning GPs’ behaviour or the organisation of primary health care in 2001 relating to GPs and community dentists amounted to 44 000, an increase of 12% on the previous year and an overall increase of 20% since the current complaints procedure was implemented in 1996.⁵ Although legal cases against GPs remain relatively uncommon in the UK (a high proportion of claims settle before trial), the Medical Protection Society observed a tenfold increase in the number of claims dealt with on behalf of its GP membership between 1989 and 1998, and estimated that 3.6% of its GP members would face legal action as a result of work undertaken in the previous 12 months.⁶

In depicting a pre-NHS patient’s attempt—unsuccessful in the event—to gain redress for harm he believed had been caused by a GP’s failure to arrange reasonable monitoring and follow up, the case of Farquhar v Murray tellingly foreshadowed many aspects of modern day claims. As Phillips and colleagues convincingly show in a study in this issue of QSHC of 26 126 malpractice claims lodged against US primary care physicians, those judged to result from negligence (23% of the total) feature failures of diagnosis (34%) and monitoring (16%), with poor communication—problems with records or inter-doctor communication—being a major contributing factor on 9% of occasions.⁷ Their study of Physician Insurers Association of America malpractice claims data significantly adds to the epidemiological understanding of conditions associated with claims for clinical negligence in primary care, and shows that certain medical conditions are much more likely than others to generate a claim—the diagnosis of appendicitis, for example, being some 25 times more likely on their data to generate a claim grounded in negligence than the diagnosis of breast cancer.

Other US researchers have analysed data sets relevant to malpractice. A study of primary care physicians in Colorado and Oregon compared doctors not facing malpractice claims with those with such claims; those without claims had significantly longer consultations (18.3 v 15 minutes), used more orientating statements (explaining what is likely to happen next), more facilitating statements (asking patient opinions and checking understanding), laughed more, and used more humour in consultations than did physicians with claims. Consultation length and physician affect (particularly laughter and behaviour demonstrative of concern, approval and empathy) were found to predict physician claims status.⁸ This association of facets of communication style with the likelihood of facing or having faced malpractice suits is strengthened by the findings of a more recent study which examined standardised extracts from 114 taped conversations of office visits between community practising surgeons and patients. The conversations were rated for warmth, hostility, dominance, and anxiety based on analysis of tone and content. Controlled for content, ratings of higher dominance identified surgeons with previous claims compared with those who had no claims, whereas greater concern/anxiety in their tone of voice identified those without claims (OR 2.74, 95% CI 1.16 to 6.43 for dominance; OR 0.46 95% CI 0.21 to 1.01 for concern/anxiety).⁹

These studies build on previous analyses of US primary care malpractice data sets: a study in Florida covering a 13 year period of information found primary care doctors with a favourable claims profile to be older but no more likely to have more prestigious professional credentials, to have qualified in the USA or Canada, to be in solo or group practice, or to be involved in research or teaching than doctors with an unfavourable claims profile.¹⁰ Although such data sets cannot always allow for confounding—from unstudied differences in, for example, patient complexity (case mix) or physician interpersonal skills¹¹—such data, when properly analysed, can identify aspects of medical practice, organisation, and interpersonal behaviour likely to lead to malpractice claims. Because malpractice claims are in part the expression of patient expectation and societal and legal arrangements for redress and compensation, US findings cannot easily be generalised to other countries. UK organisations which hold comparable data sets should make them available, with appropriate safeguards, to independent researchers who can learn much from the research questions and approaches taken by their US counterparts.