Academic journal article Applied Health Economics and Health Policy

Cost Effectiveness of Preventive Screening Programmes for Type 2 Diabetes Mellitus in Germany

Academic journal article Applied Health Economics and Health Policy

Cost Effectiveness of Preventive Screening Programmes for Type 2 Diabetes Mellitus in Germany

Article excerpt


Germany's statutory health insurance (SHI) expenditure is growing in line with the GDP.[1] However, contribution rates are rising due to a slower growth of assessable income and an increasing share of retired SHI insurants.[2] This is problematic due to the impact of SHI contributions on ancillary wage costs and the labour market, especially the labour-intensive service sector. Consequently, rationalizing and rationing of healthcare services are gaining in importance to limit contribution growth. Prioritization of early detection and prevention of chronic diseases is considered to provide an opportunity for a more effective use of resources and to slow down the growth of SHI contribution rates.[3] However, controversy surrounds the costs and effects in terms of medical and financial outcomes of such programmes.[4]

Several recent studies have assessed the relative cost effectiveness of primary prevention and interventions in type 2 diabetes mellitus (T2DM) based on health economic models. However, none was considered able to fully determine the cost effectiveness of diabetes screening in Germany as was intended with this study. Reasons for this include that existing models have not been fully published with respect to methodology;[5] focused on treatment of patients with known T2DM only;[6-9] included preventive treatment of patients with either impaired fasting glucose (IFG) or impaired glucose tolerance (IGT) but neither both groups simultaneously nor the screening programme necessary for identifying these groups;[5,10] or were restricted with regard to the inclusion or projection of all relevant co-morbidities.[7,11-13]

Therefore, we examined the cost effectiveness of screening for T2DM in the setting of the German SHI.


Screening Programme

The screening strategy that was compared with non-intervention encompassed (i) early detection of pre-diabetes or T2DM by oral glucose tolerance test (OGTT); and (ii) prevention of T2DM in subjects diagnosed with pre-diabetes either by intensive lifestyle intervention as defined by the Diabetes Prevention Program (DPP)[14] or with metformin. Non-intervention refers to diagnosis after the occurrence of the first clinical symptoms without preventive screening, and reflects the status quo of routine clinical practice in Germany. Diagnosis by OGTT is considered the gold standard in detecting pre-diabetes and T2DM. Pre-diabetes is defined as either IFG or IGT as defined by the WHO.[15] Yearly screening is offered to individuals aged 35-75 years covered by the German SHI.

The Model Outline

Due to a lack of data on long-term effects of the proposed screening programme, decision analytic modelling in comparison with the status quo was the method of choice to evaluate efficient allocation of scarce resources. Considering the necessarily long period under review, and parameter uncertainties requiring sensitivity analysis, a Markov model was used to reproduce the time-discrete stochastic process. The major limitation of Markov models (the memoryless feature) was avoided by using decision analytic software (TreeAge Pro Suite 2007) for microsimulation of a cohort of individual potential screening participants. Therefore, transition probabilities in the model also accounted for the individual disease history of each patient.

In contrast to a cohort model, this approach also accounts for the different characteristics of the general German population that would be targeted by the screening programme. While, theoretically, results from microsimulation and cohort models are equivalent if designed correctly, the latter need a much higher number of disease states to allow for the same detailed subgroup analysis. For this reason, cost effectiveness can be predicted more realistically with less effort in a microsimulation. Specifically, this allows us to analyse, for example, differences in disease progression in individuals diagnosed with T2DM in routine clinical care or 'no screening' compared with those diagnosed with pre-diabetes by screening, and either progressing to T2DM or not, and those diagnosed with T2DM by screening. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.