Purpose: The aims of this study were: (i) to create a structural simulation model capable of predicting the future need and cost of eyecare services in Finland; and (ii) to test and rank different policy alternatives for access to care and the required physician workforce.
Methods: Using the system dynamics approach, the number and cost of patients with cataract, glaucoma, diabetic retinopathy and age-related macular degeneration (AMD) were described with causal-loop diagrams and were then translated into a set of mathematical equations to build a computer simulation model. Mathematically, the problem was formulated as a set of differential equations that were solved numerically with specialized software. The validity of the model was tested against prevalence and administrative historical data. The costs covered by the public sector in Finland were obtained from 2003 from the Finnish Hospital Discharge Register (including outpatient care), the Finnish Social Insurance Institution and a survey of hospital price lists. Different levels of access to public care were then simulated in four eye diseases, for which the model estimated the need for services and resources and their costs in the years 2005-2040.
Results: The model forecasted that the adoption of the 2005 national 'access to care' criteria for cataract surgery would shorten waiting lists. If the workload of Finnish ophthalmologists were kept at the 2003 level, the graduation rate of new ophthalmologists would have to increase by 75% from the current level. If all glaucoma patients were followed in the public sector in future, even this increase in training would not meet the demand for physician workforce. The current model indicated that the screening frequency of diabetes can be increased without large sacrifices in terms of costs. AMD therapy has a significant role in the allocation of future resources in eyecare. The modelling study predicted that ageing alone will increase the costs of eyecare during the next four decades in Finland by about 1% per year in real terms (undiscounted and without inflation of unit costs). The increases in total yearly costs were on average 8.6% between 2001 and 2003.
Conclusions: The results of this modelling study indicate that policy initiatives, such as defining criteria for access to care, can have substantial implications on the demand for care and waiting times whereas the effect of ageing alone was relatively small. Measures to control several other factors - such as the adoption and price level of new technologies, treatments and practice patterns - will be at least equally important in order to restrain healthcare costs effectively.