THE ENIGMA OF PRACTICE SIZE
Imagine key stakeholders in primary care — managers, politicians, policy advisors, and patients — have gathered to draft an ideal primary care structure. Some argue for large group practices, others say smaller practices are better, others that team work or salary payment is the key. Is there evidence that would help them? Moreover, is there an optimum sized general practice in terms of delivering safe, quality care that is also highly rated by patients? Or is general practice, by its very nature, best served by practices of different sizes and contexts?
The review in this Journal on ‘whether practice size matters for the quality of care in primary care’ does not resolve these questions.1 Many researchers have tried to find a relevant relationship between attributes of quality and practice size and it is interesting and sobering to see what is left after sifting the evidence. It isn’t a great deal. Defining the optimal size of a practice is a complex decision.
There is conflicting evidence on the ability of small practices to deliver high-quality care and the views of physicians, patients, and health service managers can be at variance.2–5
Moreover, while bigger (and increasing) practice teams are the norm in some countries, such as the UK, smaller practices and indeed single-handed practitioners are prevalent throughout much of continental Europe with comparable or better6 systems health outcomes. The review in this Journal1 suggests that when patients are asked, smaller practices have an advantage; whereas when chronic care and preventive services are the outcome, larger practices can do better. The review therefore condenses the evidence and our primary care planners would have some information for making choices for their desired primary care structure. Yet, we may have to conclude that to look for ‘evidence that group practices or larger practices are providing better care’ is, perhaps needlessly, looking for the Holy Grail.
VARIATION IN CARE
There is considerable variation between practices in quality of care and it can hardly be explained by size. There are perhaps more relevant structural aspects of the practice organisation associated with better quality of care. For example, there is evidence that in the UK longer booking intervals, good teamwork, and support for preventive care all show a correlation with better quality of care.5
In the Netherlands a strong, nearly linear, relation was found, for example, between the ‘overall time the GP spends on the patient’ (including secondary activities, continuing medical education, and being on call) and quality of care.7 With each extra patient over 2000 patients per GP however, quality of care declined. There was no such relation with quality of care and the ‘number of consultations per patient’ or the ‘time the GP spent in the practice’. Other structural keys for better quality of care are ’listed/registered patients’ and ‘continuity of care’.8
MORE PROMISING CLUES
So what do we have to tell our team of primary care planners? Well, based on the actual evidence, very little. We have little information on how much leadership, autonomy and peer pressure we would want GPs and staff in practices to receive in order to foster better quality of care. Should we support practices to offer a broader scope of therapeutic and diagnostic services in primary care? Is an egalitarian team structure better than having some hierarchy? What degree of part time job/presence is required? Is it good to focus on the variation between GPs and invest in the process of analysing the variation with peers or with the team? Which other professionals (pharmacist, social worker, physiotherapist, and so on) should be integrated in the practice to provide better quality of care? Should all GPs or staff be paid a salary or be private care providers? Decisions on these matters are based on conventional wisdom. Research in this field is often inconclusive; yet important for the future design of our premises, teams, and services.
A FUTURE STUDY DESIGN
A possible study design would require the collaboration of thousands of health centres and practices. If these centres could be tempted to collect data (for example, as part of a practice visit scheme) on all relevant structural, process, and outcome data available using the same format, it could shed light on a lot of these research questions. Most practices already collect such data but not in a coordinated way and not using comparable indicators; especially across geographical and country borders. The review shows that analysis of variation in quality of care in a multivariate analysis against organisational aspects is feasible. The EPA-project was such a study, using a collaboration of nine European countries that agreed on a common set of indicators for structure and process.9 Although the project got The European Health Award for its benefit to population health, funding stopped.10
A few other studies have shown the potential of analysing such a database. For example, training practices are associated consistently with a broader scope of services, better patient scores, and less workload and job stress for the GPs.11 We also found that while larger practices do not deliver better chronic care, the workload of GPs in larger practices is less with more tasks delegated to practice nurses.12 Such a database could help address some of the answers raised above.
Most quality assessment/improvement is focused on population-level measures and this helps to provide safe quality care and should apply to all practices, but must not be at the expense of the precedence of the relationship between individual patients and health practitioners.13 The same applies in seeking to identify the optimum practice size. Any search for an ideal practice size must not be at the expense of the focus on the individual patient or practitioner.
Notes
Provenance
Commissioned; not externally peer reviewed.
- © British Journal of General Practice 2013