The content of this website is intended for healthcare professionals only

Diagnosis targets unethical says GP

Targets based on error-prone prevalence estimates and could lead to overdiagnosis

Louise Prime

Wednesday, 03 December 2014

Target rates for diagnosis in primary care are unethical as well as misleading, writes* a GP today on bmj.com. Dr Martin Brunet, from Godalming in Surrey, argues in his Personal View that targets in healthcare always undermine the doctor-patient relationship, and may be based on flawed assumptions – so he is dismayed that the Government is setting diagnosis targets for six conditions in addition to that for dementia. It last week had to back down on its highly controversial scheme to offer GPs a £55 payment-per-diagnosis for dementia.

Recent updates to NHS England’s General Practice Outcome Standards and the Primary Care Web Tool include practice-level data on 29 separate indicators across seven clinical areas, and each practice in England has a set diagnosis rate for each of the seven conditions – diabetes, atrial fibrillation, coronary heart disease, asthma, chronic obstructive pulmonary disease, dementia, and depression – with the intention of exerting pressure on GPs to “improve” diagnosis rates.

In theory, the targets set are estimated from practice data and “expected” prevalence. But Dr Brunet questions the robustness of the “error-prone national estimates of prevalence … usually presented as indisputable fact” on which these targets are based.

He points out that significant differences between regions in terms of their demographics – for example rurality, ethnicity, and deprivation – lead inevitably to true differences between practices in the prevalence of many health problems. So, he writes, basing targets for diagnosis on a nationwide prevalence figure for these conditions – even when there’s been an attempt to account for these variations – could incentivise potentially harmful overdiagnosis as some practices try to “improve” their apparently low diagnosis rates that are in fact already better than average.

He is concerned that setting target diagnosis rates could lead practices to “inadvertently introduce screening by the back door, even though the UK National Screening Committee does not recommend screening in any of the clinical areas in question”. He warns: “Such unofficial, ad hoc screening could do more harm than good through overdiagnosis, misdiagnosis, and the diversion of resources away from people with symptoms.”

Dr Brunet writes that patients must be able to trust their doctor to act solely in their best interests, unencumbered by competing interests. He argues that healthcare targets will always undermine trust in the doctor-patient relationship, although he adds: “Mechanisms such as exception reporting in the Quality and Outcomes Framework mitigate this risk, because they enable the doctor to exempt individual patients from a health target on the grounds of patient choice.”

He insists that the ethical principles are the same whether pressure to diagnose is for direct financial gain – as with the £55 incentive to diagnose dementia – or arises from concern for a practice’s diagnosis rates.

He concludes: “NHS England needs to hear a clear message from doctors and patients that setting targets for diagnosis is problematic, unscientific, and unethical. Instead, it needs to trust doctors and their patients to know when to seek a diagnosis.”

* Martin Brunet. Target diagnosis rates in primary care are misleading and unethical. BMJ 2014;349:g7235. doi: http://dx.doi.org/10.1136/bmj.g7235

Registered in England and Wales. Reg No. 2530185. c/o Wilmington plc, 5th Floor, 10 Whitechapel High Street, London E1 8QS. Reg No. 30158470