We read with great interest the paper by Bismark et al1 about identification of doctors at risk of recurrent complaints in Australia. The authors examined a sample of 18 000 formal complaints lodged within an 11-year period. In brief, 3% of doctors accounted for 49% of all complaints. Doctors who were complained about were at significantly increased risk of a re-complaint. The authors concluded that it is feasible to predict which doctors are at high risk of recurring complaints.

Acknowledging that this article contributes to the current debate on this issue, the oversimplification is such that, unfortunately, it has the potential to do more harm than good.

Throughout the article, there is the notion that it is predominantly the doctors’ behaviour which results in complaints. The authors even recommend interventions, such as limiting doctors’ practices to address their bad behaviour. This idea is a rather naive, implausible and unrealistic view of the doctor–patient relationship. While it may be that in some instances doctors’ behaviour translates into complaints, it is extremely unlikely that this is typically the case. In the authors’ own sample, only 23% of complaints related to miscommunication, including attitude or manner.

Current research2 suggests that a breakdown in communication between the doctor and the patient is the greatest risk factor for a complaint. While the doctor's behaviour may contribute to a breakdown in communication, typically the likelihood of a complaint increases if the patient experiences an adverse event and a satisfactory explanation for the poor outcome is not provided or is not available. In the absence of an appropriate explanation for an adverse event, patients or relatives may suspect a cover-up and the easiest way (for patients) to obtain details about the causes of the adverse event (ie, what really happened) is through a formal complaint to health authorities. In addition, the financial cost of lodging formal complaints has decreased over the past few years due to advertising campaigns promising ‘No win, no fee.’3

How is it that 3% of doctors account for 49% of complaints? Male doctors, doctors over the age of 45 years and doctors in surgical specialities seem to be at greatest risk. Doctors in this demographic typically see a greater number of patients—a key factor not considered at all in the multivariate model. Logically, the higher a doctor's surgical volume, the higher the risk that some patients may experience an adverse event. Rather complex surgical cases with a high risk of surgical complications are referred to a low number of surgical subspecialists at tertiary centres. Due to the ageing population, surgeons increasingly push the boundaries, operating on patients with severe comorbidities or after multiple previous treatments. In this paper, more meaningful and better-structured models (including doctors without prior complaint) could have been developed rather than simply models with over 35 parameters, which demonstrate little more than statistical significance. Lacking key denominators throws doubt on the value of the conclusions reached.