Policy evaluation in diabetes prevention and treatment using a population-based macro simulation model: the MICADO model.

AIMS: To test a simulation model, the MICADO model, for estimating the long-term effects of interventions in people with and without diabetes. METHODS: The MICADO model includes micro- and macrovascular diseases in relation to their risk factors. The strengths of this model are its population scope and the possibility to assess parameter uncertainty using probabilistic sensitivity analyses. Outcomes include incidence and prevalence of complications, quality of life, costs and cost-effectiveness. We externally validated MICADO's estimates of micro- and macrovascular complications in a Dutch cohort with diabetes (n = 498,400) by comparing these estimates with national and international empirical data. RESULTS: For the annual number of people undergoing amputations, MICADO's estimate was 592 (95% interquantile range 291-842), which compared well with the registered number of people with diabetes-related amputations in the Netherlands (728). The incidence of end-stage renal disease estimated using the MICADO model was 247 people (95% interquartile range 120-363), which was also similar to the registered incidence in the Netherlands (277 people). MICADO performed well in the validation of macrovascular outcomes of population-based cohorts, while it had more difficulty in reflecting a highly selected trial population. CONCLUSIONS: Validation by comparison with independent empirical data showed that the MICADO model simulates the natural course of diabetes and its micro- and macrovascular complications well. As a population-based model, MICADO can be applied for projections as well as scenario analyses to evaluate the long-term (cost-)effectiveness of population-level interventions targeting diabetes and its complications in the Netherlands or similar countries.

Estimating net transition probabilities from cross-sectional data with application to risk factors in chronic disease modeling.

A problem occurring in chronic disease modeling is the estimation of transition probabilities of moving from one state of a categorical risk factor to another. Transitions could be obtained from a cohort study, but often such data may not be available. However, under the assumption that transitions remain stable over time, age specific cross-sectional prevalence data could be used instead. Problems that then arise are parameter identifiability and the fact that age dependent cross-sectional data are often noisy or are given in age intervals. In this paper we propose a method to estimate so-called net annual transition probabilities from cross-sectional data, including their uncertainties. Net transitions only describe the net inflow or outflow into a certain risk factor state at a certain age. Our approach consists of two steps: first, smooth the data using multinomial P-splines, second, from these data estimate net transition probabilities. This second step can be formulated as a transportation problem, which is solved using the simplex algorithm from linear programming theory. A sensible specification of the cost matrix is crucial to get meaningful results. Uncertainties are assessed by parametric bootstrapping. We illustrate our method using data on body mass index. We conclude that this method provides a flexible way of estimating net transitions and that the use of net transitions has implications for model dynamics, for example when modeling interventions.

Case fatality of COPD exacerbations: a meta-analysis and statistical modelling approach.

The aim of our study was to estimate the case fatality of a severe exacerbation from long-term survival data presented in the literature. A literature search identified studies reporting ≥1.5 yr survival after a severe chronic obstructive pulmonary disease (COPD) exacerbation resulting in hospitalisation. The survival curve of each study was divided into a critical and a stable period. Mortality during the stable period was then estimated by extrapolating the survival curve during the stable period back to the time of exacerbation onset. Case fatality was defined as the excess mortality that results from an exacerbation and was calculated as 1 minus the (backwardly) extrapolated survival during the stable period at the time of exacerbation onset. The 95% confidence intervals (CI) of the estimated case fatalities were obtained by bootstrapping. A random effect model was used to combine all estimates into a weighted average with 95% CI. The meta-analysis based on six studies that fulfilled the inclusion criteria resulted in a weighted average case-fatality rate of 15.6% (95% CI 10.9-20.3), ranging from 11.4% to 19.0% for the individual studies. A severe COPD exacerbation requiring hospitalisation not only results in higher mortality risks during hospitalisation, but also in the time-period after discharge and contributes substantially to total COPD mortality.

Cost-effectiveness analyses of health promotion programs: a case study of smoking prevention and cessation among Dutch students.

Little research has been done to connect health promotion programs to outcomes in terms of life expectancy, health care costs and cost-effectiveness. For a policy maker, economic evaluation may be an important tool to support decisions on how to allocate the health care budget. The aim of this paper was to determine the cost-effectiveness of a Dutch school-based smoking education program. The incremental cost-effectiveness ratio of the school program was estimated at euro19 900 per quality adjusted life year gained. For a complete analysis, not only intervention costs but also savings for smoking-related diseases and differences in total health care costs should be taken into account. As several assumptions had to be made in order to estimate cost-effectiveness, the study outcomes should be interpreted with caution. Main problem in estimating the cost-effectiveness was the lack of proper effectiveness data on daily smokers among adolescents. Absence of specific effectiveness data often is an obstacle in the economic evaluation of public health interventions. While some problems may be the result of insufficient sample size or follow-up, another possible explanation might be the different basic principles of analysis of health promoters and economists.

A dynamic population model of disease progression in COPD.

To contribute to evidence-based policy making, a dynamic Dutch population model of chronic obstructive pulmonary disease (COPD) progression was developed. The model projects incidence, prevalence, mortality, progression and costs of diagnosed COPD by the Global Initiative for Chronic Obstructive Lung Disease-severity stage for 2000-2025, taking into account population dynamics and changes in smoking prevalence over time. It was estimated that of all diagnosed COPD patients in 2000, 27% had mild, 55% moderate, 15% severe and 3% very severe COPD. The severity distribution of COPD incidence was computed to be 40% mild, 55% moderate, 4% severe and 0.1% very severe COPD. Disease progression was modelled as decline in forced expiratory volume in one second (FEV1) % predicted depending on sex, age, smoking and FEV1 % pred. The relative mortality risk of a 10-unit decrease in FEV1 % pred was estimated at 1.2. Projections of current practice were compared with projections assuming that each year 25% of all COPD patients receive either minimal smoking cessation counselling or intensive counselling plus bupropion. In the projections of current practice, prevalence rates between 2000-2025 changed from 5.1 to 11 per 1,000 inhabitants for mild, 11 to 14 per 1,000 for moderate, 3.0 to 3.9 per 1,000 for severe and from 0.5 to 1.3 per 1,000 for very severe COPD. Costs per inhabitant increased from 1.40 Euro to 3.10 for mild, 6.50 Euro to 9.00 for moderate, 6.20 Euro to 8.50 for severe and from 3.40 Euro to 9.40 for very severe COPD (price level 2000). Both smoking cessation scenarios were cost-effective with minimal counselling generating net savings. In conclusion, the chronic obstructive pulmonary disease progression model is a useful instrument to give detailed information about the future burden of chronic obstructive pulmonary disease and to assess the long-term impact of interventions on this burden.

The impact of aging and smoking on the future burden of chronic obstructive pulmonary disease: a model analysis in the Netherlands.

Chronic obstructive pulmonary disease (COPD) causes extensive disability, primarily among the elderly. On the World Health Organization ranking list of disability-adjusted life years (DALYs), COPD rises from the twelfth to the fifth place from 1990 to 2020. The purpose of this study is to single out the impact of changes in demography and in smoking behavior on COPD morbidity, mortality, and health care costs. A dynamic multistate life table model was used to compute projections for the Netherlands. Changes in the size and composition of the population cause COPD prevalence to increase from 21/1,000 in 1994 to 33/1,000 in 2015 for men, and from 10/ 1,000 to 23/1,000 for women. Changes in smoking behavior reduce the projected prevalence to 29/1,000 for men, but increase it to 25/ 1,000 for women. Total life years lost increase more than 60%, and DALYs lost increase 75%. Costs rise 90%; smokers cause approximately 90% of these costs. The model demonstrates the unavoidable increase in the burden of COPD, an increase that is larger for women than for men. The major causes of this increase are past smoking behavior and the aging of the population; changes in smoking behavior will have only a small effect in the nearby future.

The impact of smoking on future pancreatic cancer: a computer simulation.

BACKGROUND: We studied the impact of several smoking cessation-based scenarios on future pancreatic cancer incidence in the European Union by means of computer simulation. MATERIAL AND METHODS: Among other data, published data on pancreatic cancer incidence rate and smoking prevalence in ten member states of the European Union, and on the relative risk of smoking were entered into a simulation model. Four different scenarios were simulated: one reference scenario, one based on theoretically maximal smoking reduction and two feasible scenarios based on WHO's Health for All targets. In each scenario, pancreatic cancer incidence was computed from 1994 up to the year 2020. Results were extrapolated to the European Union as a whole. RESULTS: When the percentage of smokers remains unchanged, 627,000 and 588,000 newly diagnosed pancreatic cancer cases among males and females respectively will arise in the European Union up to 2020. Theoretically, if all smokers would give up smoking instantly, this number can be reduced by 133,000 cases among men and 43,000 cases among women. In more feasible scenarios up to 35,500 male and 32,500 female pancreatic cancer cases can be prevented. CONCLUSION: Giving up smoking substantially reduces future burden of pancreatic cancer up to almost 68,000 patients in the European Union up to the year 2020.

Reaching homosexual men for HIV surveillance through a gay magazine.

HIV surveillance in homosexual men is poor in most countries, as this risk group is difficult to sample. The aim of this study is to test the feasibility of reaching homosexual men for national HIV surveillance using gay community media. In 1989, a questionnaire on general gay issues, with a section on AIDS, was included in a widely sold gay magazine in the Netherlands. Among 17,700 sold copies, 1134 responses were obtained from males (6%). Of these, 669 men (59%) gave their address, of which in turn 84% responded to a questionnaire on risk behaviour in 1990. In 1991/1992, the 669 men were asked to participate in an HIV serosurvey, in which eventually 308 participated with a blood test (46%) and 147 without (total 68%). Participation in the serosurvey with blood test was associated with reporting multiple partners in 1989. Twenty participants were infected (6.5%). In logistic regression analysis, risk factors for infection were recent unprotected receptive anal intercourse with multiple partners (odds ratio (OR): 10.7; 95% confidence interval (CI): (2.18-52.2); one partner 1.17 (0.31-4.48); none 1) and living in Amsterdam (OR: 3.92; 95% CI: (0.99-15.5); urbanised western Netherlands 2.15 (0.57-8.03); elsewhere 1), while a high educational level was protective (OR: 0.29 (0.08-0.96); middle 0.41 (0.11-1.54); low 1). Among those who participated in 1991/1992, risk behaviour increased between 1989 and 1991/1992 (reporting multiple casual partners rose from 55% to 64%; reporting inconsistent condom use with receptive anal sex from 58% to 71%). Using a predictive model which included self-reported serostatus in 1989, the HIV prevalence rate in 1991/1992 among all male responders to the 1989 questionnaire was estimated to be 5.3% (95% CI: 3.1 7.7%). In conclusion, unless initial response is improved, recruitment through a gay magazine may not allow reliable estimates of HIV prevalence in homosexual men. However, it can be useful at the national level for monitoring changes in prevalence and risk behaviour over time, geographical differences and risk factors for infection.

Forecasting the number of diabetic patients in The Netherlands in 2005.

OBJECTIVES: There is evidence from past decades that the number of diabetic patients has increased independently of changes in demography. A static model that takes into account only demographic changes is therefore unable to forecast the expected number of diabetic patients correctly. METHODS: We developed a dynamic model in which actual incidence, prevalence, and life expectancy data are used and alternative assumptions about future trends in these parameters can be incorporated. RESULTS: This dynamic model forecasts higher numbers of diabetic patients than the less sophisticated static model. According to the dynamic model, a 46% increase in the number of diabetic patients in The Netherlands can be expected, from 244,000 in 1990 to 355,000 in 2005 (about 2.5% annually). The static model forecasts a 22% increase. CONCLUSIONS: Diabetes mellitus will become a more serious public health problem than can be expected from demographic changes only. In planning future health care, monitoring of trends in incidence, prevalence, remission, and mortality or life expectancy is a necessary prerequisite.