Odette Wegwarth1, Robert W Day, Gerd Gigerenzer. 1. Max Planck Institute for Human Development, Berlin, Germany and Humboldt University, Berlin, Germany. wegwarth@mpib-berlin.mpg.de
Abstract
BACKGROUND: With the increase in molecular genetic understanding of disease, diagnostic test development and availability are growing rapidly. This study investigated oncologists' decision making on using pharmacogenomic tests for cancer treatment and examined cross-cultural differences between the USA and Germany. METHODS: Pilot studies revealed that the following cues play a role in decisions on pharmacogemonic tests: stage of cancer, availability of treatment options, cost of the treatment options, severity of side effects of the treatments, therapeutic consequence of the test, cost of the test and guideline recommendation specifying use of the test. All cues were used for designing the main study comprising nine scenarios, for each of which oncologists were asked to decide whether they would order a pharmacogenomic test. RESULTS: On average, US oncologists opted for the test in 6.5 out of the nine scenarios (SE = 0.2), and German oncologists in 5.4 scenarios (SE = 0.2). The majority of oncologists' decisions in both the USA (76.1%) and Germany (64%) were best explained by a simple sequential model (heuristical strategy). In the USA, the information about cost of the test was most influential on the decisions; in Germany it was the guideline recommendation of the test. When the side effects of therapy B were described as being more severe within the scenarios, choices in favour of a non-recommended test increased by about 20% within both samples. CONCLUSION: Both US and German oncologists were highly inclined to use pharmacogenomic tests, but differed in what information influenced their decisions - a difference possibly explained by the differences in the health insurance systems. Although many oncologists' heuristical decisions were based on the valid cue of a test's guideline recommendation, an alarming number abandoned it when a therapy had potentially severe side effects.
BACKGROUND: With the increase in molecular genetic understanding of disease, diagnostic test development and availability are growing rapidly. This study investigated oncologists' decision making on using pharmacogenomic tests for cancer treatment and examined cross-cultural differences between the USA and Germany. METHODS: Pilot studies revealed that the following cues play a role in decisions on pharmacogemonic tests: stage of cancer, availability of treatment options, cost of the treatment options, severity of side effects of the treatments, therapeutic consequence of the test, cost of the test and guideline recommendation specifying use of the test. All cues were used for designing the main study comprising nine scenarios, for each of which oncologists were asked to decide whether they would order a pharmacogenomic test. RESULTS: On average, US oncologists opted for the test in 6.5 out of the nine scenarios (SE = 0.2), and German oncologists in 5.4 scenarios (SE = 0.2). The majority of oncologists' decisions in both the USA (76.1%) and Germany (64%) were best explained by a simple sequential model (heuristical strategy). In the USA, the information about cost of the test was most influential on the decisions; in Germany it was the guideline recommendation of the test. When the side effects of therapy B were described as being more severe within the scenarios, choices in favour of a non-recommended test increased by about 20% within both samples. CONCLUSION: Both US and German oncologists were highly inclined to use pharmacogenomic tests, but differed in what information influenced their decisions - a difference possibly explained by the differences in the health insurance systems. Although many oncologists' heuristical decisions were based on the valid cue of a test's guideline recommendation, an alarming number abandoned it when a therapy had potentially severe side effects.
Authors: Nicholas J Keeling; Tyler J Dunn; John P Bentley; Sujith Ramachandran; James M Hoffman; Meagen Rosenthal Journal: Genet Med Date: 2021-04-29 Impact factor: 8.822