It has been recognised for quite some time now that considerable differences occur in the treatments offered by doctors to apparently similar patients, both across countries and within countries.1-5 Reasons for studying these variations are based upon the wish to improve practice, to help to develop innovative aspects of care and also, very importantly, to monitor patient uptake and access to care. Variations in clinical practice occur despite the fact that healthcare professionals in their chosen clinical specialty basically share the same body of knowledge as a basis for their activities. Cataract surgery is apparently no exception, as shown in the paper by Norregaard et al, in this issue of theBJO (p1107), who have studied and described international variations in the indications for surgery.

Within the ophthalmic healthcare service there is an approach which we might adopt in contributing to the study of variations in ophthalmic care. We could:

(1)

Lay more emphasis on the collection and use of basic clinical indicators such as visual acuity at the time of patient presentation, operation, and discharge

(2)

Encourage interested colleagues to document the process of professional judgment involved in clinical decision making surrounding the use of clinical indicators

(3)

Gradually broaden these indicators to include items on social and economic factors

(4)

Contribute to the work of understanding the reasons for clinical variation at the level of profession, institution, and health system, nationally and internationally.

 International comparisons of variations are of great interest. Our understanding of the underlying causes of variations, however, is problematic. This is particularly so at international level because of diversity of background and also because of issues surrounding definitions and standardisation of the methods that are used. Let us start in the UK by looking more closely at the variations in our own practice and relate these outwards to the experience of other countries. In doing so, the methods and information sources we develop, and on which we agree, will need to be carefully considered if we are to make our own contribution to this interesting area.

The study by Norregaard et al is in keeping with other reported studies (both national and international) on variations in clinical practice.2-5 Their research on variation in indications for cataract surgery was, for the most part, focused on clinical factors relating to preoperative visual status and the coexistence of more than one eye disease. They conclude, however, that social, demographic, and societal factors were likely to be major contributory factors to the variations in practice that they observed. Yet very few of the economic and social factors that are put forward in the discussion section of this (and similar) articles are subject to the same level of investigation as the clinical factors within these studies.

As we move forward in developing methods for investigating variations in clinical practice in ophthalmology we can broaden the scope of the content at the same time. Work on differences in practice can either confine itself to reporting on those clinical factors that have been studied, or move forward in a truly multidisciplinary perspective in the study of clinical practice variations. Social and clinical epidemiological methods have a similar basis but if they are to be fused in clinical settings their different requirements for investigation have to be considered. An example here would be that of sample size, the importance of which we recognise yet do not necessarily find easy to calculate. We must now take account of the observations of social scientists and health economists that sample size estimates for clinical purposes may not be adequate or appropriate for investigating social variables.

If, as suggested by Norregaard et al, that contributory factors for variations in indications for surgery may be related to “patient demand”, then broad socioeconomic profiling would help to investigate this issue. The authors also suggest that “access to care and practice style” may also be important contributory factors. Investigation of income levels and national charging policies could help to shed some light on these aspects.

While we can develop the methods for investigation, the ways and means of data acquisition are likely to be constrained by the budget available for information collection generally in the health sector. For the present therefore, and for some time ahead, we must maximise our use of available material. Our main sources of information must be those we can obtain from large routine datasets, supplemented by a few key socioeconomic and health policy factors. These data could be complemented by smaller in depth studies on specific issues—for example, outcomes assessment and operational research on clinical decision making surrounding choice of patient for surgery. For the latter to be a comprehensive assessment it must include monitoring of indicators, such as visual acuity, for those patients who present with self reported diminishing vision but are discharged to return at some distant time in the future or not at all.

Variations in thresholds for surgery raise important public health issues for investigation on equity, quality, and outcomes of care to which we as ophthalmologists can contribute with the help of epidemiology and the social sciences. In doing so, national variations could be quantified and monitored and used to inform allocation of resources for ophthalmic services that are both appropriate and equitable.