Measuring diabetes care quality in general practice: A whole bundle of trouble?

The place of general practice in Western health systems has changed profoundly in the last 20 years, moving out from the shadows of the hospitals to centre stage. This has been particularly pertinent in the management of one condition – type 2 diabetes. The growing prevalence of diabetes, the increasing age of the general population and the rising pressure on hospital services has meant that the majority of care for “low-risk” people with diabetes now takes place almost exclusively in general practice.

Against this background, evidence is accumulating of steady progress in the management of chronic conditions in the NHS. For example, Oluwatowoju et al (2010) have shown improving care of people with type 2 diabetes. The pervasive influence of the Quality and Outcomes Framework (QOF) and NICE guidance has had an important effect on clinicians in general practice. However, there is still a debate about the quality of care, how it should be measured, the standards that are appropriate and achievable, and how equity of access to high-quality care can be ensured for all.

In some quarters, the term “scandal” has been applied to the state of care for people with diabetes in general practice. This is something that we responded to in a recent article in the British Journal of General Practice (Pereira Gray et al, 2014). This contrasted the approach of the QOF with that of the National Diabetes Audit (NDA) and highlighted, via work from our own practice, a number of data inconsistencies within the NDA. These systematically underestimated the quality of diabetes care in general practice. The NDA has consistently reported that a sizeable proportion of patients do not receive all nine processes of care as recommended by NICE (e.g. 46% in 2011; Gadsby and Young, 2013). Gadsby and Young (2014), in their response to our letter, stressed the formative rather than summative nature of the NDA – yet in many circles the NDA is seen as a summative assessment of diabetes care in general practice.

The use of these data in a negative manner to denigrate general practice care also has the potential for demoralisation of GPs and their teams, increasing work-related stress and hence possibly causing patient care to suffer further.

Subsequently, NICE has agreed in principle to bundle the current diabetes QOF indicators into a single “all or nothing” diabetes indicator. However, the introduction of such a bundled indicator is in itself fraught with potential difficulties and will undoubtedly lead to an increase in exception reporting.

Another potential consequence is the threat of de-motivation of practitioners and reduction in quality of care, once GPs realise that comprehensive coverage of all nine indicators is impossible (for example, people who have undergone bilateral lower-limb amputation cannot have foot pulses recorded). We suggested that “levels of achievement should be reported separately for every target and the interactivity convention played down” (Pereira Gray et al, 2014). It is also important that best practice is shared and that the maximum achievable standard for each indicator is made public to calibrate underperforming practices and encourage the provision of ever higher standards. 

British general practice is seriously short of GPs, if measured in whole-time equivalents, and faces increasing workload pressures. It is essential that the appropriate measurement of quality of care in diabetes and how best to improve performance are awarded the highest priority. The measurement and management of other chronic diseases can then follow in the “footsteps of diabetes”.