Statistics from Altmetric.com
Seeing the broader picture
Riad and colleagues have presented us, in this issue of BJO (p 493), with an interesting and thorough review of primary care especially as it applies to ophthalmology in the United Kingdom. In their far ranging article the authors consider topics such as the history and definitions of primary care, the attributes of good primary care, and the primary-secondary interface. The review accurately portrays the complexity of primary healthcare provision and implicitly raises the question of how such a complex system can be planned, managed, or evaluated. The review will be particularly helpful when health service researchers need to view ophthalmic primary care as a whole because it is not sufficient for them merely to concentrate on one specific aspect.
Most health service research concentrates on the qualities of some small part of healthcare provision, perhaps suggesting how that particular aspect could be improved. While important, this approach sometimes ignores the knock-on effect that changing one part of a service will have on other provision. The coverage of diabetic retinopathy screening schemes in England and Wales was recently audited with the aim of describing the characteristics of services that are best at reaching people with diabetes.1 One striking feature of the audit was the way that general practitioners tried to compensate for the absence of an organised scheme by performing more eye checks themselves. Introducing an organised diabetic retinopathy screen will therefore have wider repercussions. Not only will tests given by specialists tend to be more reliable, but general practitioners will become less involved in giving eye checks, allowing them more time for other aspects of their work; this might be positive, but the withdrawal from eye screening could reduce their contact with their diabetic patients and so impact on the total package of care that they provide. Whenever change is advocated for one part of primary care, the impact on the service as a whole also needs to be considered, because changes tend to ripple through. Although obvious, this is frequently ignored when research findings are reported and implemented. Riad et al‘s review may help researchers who are looking into a particular aspect of ophthalmic primary care to place their work in the broader context.
Those of us who study particular aspects of primary care provision must guard against missing the impact on the service as a whole
As part of the audit of coverage the process of diabetic retinopathy screening was divided into its component parts; identifying people with diabetes, coverage of the screen, the quality of the test, and treatment.2 Each of those components has been the subject of much separate research and by bringing together the information on the various components it might be possible to suggest the “ideal” scheme. Unfortunately, reality is more complex than that and the various components may interact with one another. Perhaps increasing the coverage would put added pressure on the screening services so that the sensitivity or specificity of the test would decline. In such circumstances a simpler screening method that is less sensitive to the pressure of numbers of patients might be preferable to a “high tech” solution, even if research demonstrates the theoretical benefits of the latter. Although a lot can be learned from studying parts of the service in isolation, at some stage the programme as a whole needs to be considered. Eventually, more studies will be conducted that specifically relate to the interactions between parts of primary care so that models of the whole system can be developed, but in the meantime the review of Riad et al will help us think about these issues informally.
The difficulties experienced by those of us who are interested in healthcare evaluation are mirrored in the extreme difficulty that is currently being experienced by those trying to manage the NHS. Most ophthalmologists will be familiar with the way that management targets, genuinely aimed at improving one part of the service, end up causing unanticipated repercussions elsewhere. Thus, moving resources to control waiting times for first appointments affects the service‘s ability to provide follow up care. A centrally planned healthcare system has all the problems of a centrally planned economy. Ironically, the NHS used to be lightly managed with care provided by semiautonomous ethically motivated consultants and general practitioners; with proper financing this might actually be one of the best ways of organising a non-fee based service. Management was introduced to control costs but resulted in the more centralised and bureaucratic system that government is now trying to unravel. Primary care is seen as leading the new NHS and this strategic position makes evaluation even more complex, as the primary care services might need to do things to drive the NHS that do not necessarily improve the quality of primary care itself.
Just as scientists who study disease at a cellular level need to guard against missing the impact on the person, so those of us who study particular aspects of primary care provision must guard against missing the impact on the service as a whole. Riad et al‘s review may help us all in that.
Note in Proof
Seeing the broader picture