gms | German Medical Science

Purpose: The upcoming pharmacodiagnostic tests will offer the opportunity to better tailor cancer treatments to the needs of individual patients and allow oncologists a more individualized treatment decision for their patients. However, if used inappropriately to guide the decision, or if the test itself is of questionable quality, test use may be more misleading than helpful. Yet little is known about how health care professionals deal with these issues about the test’s quality and what influences them to employ these tests in their daily treatment decisions. This study is a first attempt to examine the determinants of oncologists’ decisions and the decision-making strategies applied. Since pathologists perform such tests for oncologists, and patients' treatment is affected by test results, these groups are examined, too.

Method: For the pre-study, focused interviews were used to identify relevant determinants of decision-making as well as decision-making strategies for oncologists, pathologists and patients in both Germany and the US. A sample of 19 oncologists, 15 pathologists and 17 patients (total) was interviewed about their past decision-making behavior regarding such tests in both countries. As most patients lacked experience with such tests, they were mainly asked about their decision-making regarding cancer treatments. Discrete choice experiments (DCE) are currently being carried out for each group to test the identified decision-making strategies and the relative impact of each of the named main determinants.

Results pre-study: For oncologists, target setting behavior was found. The attributes they took into account were the tests’ level of evidence (target variable), cost of the treatment and test, severity of side effects of the treatment as well as utility of the test result for the decision (trade off variables). For pathologists, compensatory decision-making was identified. They reported trade offs between several factors of cost effectiveness, and accuracy of the test system. Patients mainly reported to trust in the recommendations given by their oncologists (heuristic). This strategy would shift to compensatory decision-making if they were to be charged for the test. No obvious differences between the German and the US samples were found. With the discrete choice experiments, these results are currently being tested on large samples for each group. Final results are expected by the end of December 2005.