2.1 Background

In general, when estimating the incidence and prevalence of disease or state of health in the population, epidemiologists draw on mortality data, registration data (cancer), data from surveillance and notification, utilisation data, and on results of large-scale population-based studies. Some of these approaches are relevant only to particular diseases. In the case of eye disease in the UK, there has been heavy reliance on registration and utilisation data to the detriment of more valid epidemiological approaches such as population based cross-sectional and cohort studies. Epidemiological studies also address the issues of causation, risk, and the effectiveness of preventive, screening and treatment programmes. The funding of epidemiological studies in eye research has been at a modest level compared with the experience in North America, where there has been substantial investment in pertinent approaches. Nevertheless, there have been a number of important studies in the last decade in the UK and Ireland that have provided useful data. These include The RNIB Survey (1991), the OPCS Causes of Blindness Study (1995), The Irish Glaucoma Survey (1992), The North London Eye Study (NLES) (1998), and The National Cataract Surgery Survey (2000). The sources of data utilised for this report and some of the most important study references are listed at the end of this section (2.4)

  2.1.1 Notes on Methodology

In estimating the magnitude of visual impairment and the major eye disorders in the population, we have collated the prevalence (and incidence) data specific to subgroups defined by age, sex, and (where relevant ethnic groupings), obtained from the most germane recent studies that have investigated 'unbiased' samples from defined populations. These group-specific 'rates' from samples have been applied to the corresponding strata in the older population (60 or older) of England and Wales, to compute the magnitude estimates for the whole older population and for various age groups within it. This preferred methodological approach has not been possible for some of the less frequent eye disorders for which there are no satisfactory prevalence or incidence data. In the absence of longitudinal cohort studies, most of the incidence rates used here have been derived from age-specific prevalence data according to a widely employed statistical procedure. Thus the incidence estimates are not based on direct measures and should be interpreted with due caution. For glaucoma, we have also used predictive equations based on the collective data from several cross-sectional studies in Europe, Australia, and North America. The utility of these predictive equations are addressed in reference 3 of source 2 (2.4). For cataract, some of the findings from an epidemiological model of the population dynamics of the disorder are also reported (source 2, reference 2).

   2.2 The Magnitude of the Problem and the Population Need