Economic consequences of near-patient test results - the case of tests for the Helicobacter Pylori bacterium in dyspepsia

Abstract Diagnostic tests and in particular laboratory tests are often important in diagnostic work-up and monitoring of patients. Therefore the economic consequences of medical actions based on test results may amount to a substantial proportion of health service costs. Thus, it is of public interest to study the consequences and costs of using laboratory tests. We develop a model for economic evaluation related to the diagnostic accuracy (sensitivity and specificity) of near patient tests. Blood sample based tests to detect the bacterium Helicobacter Pylori (HP) are useful in diagnosing peptic ulcer and suitable to illustrate the model. First, general practitioners’ initial management plans for a dyspeptic patient are elucidated using a paper vignette survey. Based on survey results, and medical literature, a decision tree is constructed to visualize expected costs and outcomes resulting from using three different HP tests in the clinical situation described in the vignette. Tests included are two rapid tests for use in general practice, and one hospital laboratory test for comparison. The tests had different sensitivities and specificities. Then a costeffectiveness analysis is undertaken from a societal perspective. Finally we use sensitivity analyses to model the decision uncertainty. Estimating for a follow-up period of 120 days, the rapid test with lower sensitivity and specificity than the hospital HP test is cost-effective because the laboratory result is available immediately. Further, in general practice, the rapid test with the highest sensitivity is significantly cost effective compared to the test with the highest specificity when the willingness to pay for each dyspepsia-free day exceeds €42.6. When deciding whether a laboratory analysis should be analysed in the office laboratory or not, it is important to consider both the diagnostic accuracy of the tests and the waiting time for the alternative, i.e. a hospital laboratory result.