OBJECTIVES: Little is known about doctors' preferences regarding public report card design. Taiwan just announced the first diabetes report card on April 2008. The aim of this study was to investigate the Diabetes Mellitus (DM) providers' preferences towards four report card attributes: update frequency, risk adjustment, content information and display format. METHODS: A discrete choice questionnaire was mailed to hospital and primary care doctors in the northern part of Taiwan, with 221 study targets. The response rate was 29%. RESULTS: Using random effect logistic regression, doctors' preference attribute rankings were risk adjustment for patients (44.7%), content information (25.2%), display format (18.3%) and update frequency (11.8%). One-year update frequency, risk adjustment, detailed scores of technical quality and interpersonal quality and bar chart display were the most important items noted in our survey. CONCLUSIONS: This is the first study to investigate provider's preferences for a diabetes report card. It enables the policy maker to clearly see the implication of trade-offs between different choices when designing a report card that doctors will like. Our findings suggest that doctors do not favour the 'less is more' principle, or the higher frequency of updates that patients may prefer. Rather, our findings suggest that risk adjustment and more information content are the most important factors for doctors. Future studies should use discrete choice experiment on different aspects of report card design, such as vulnerable patients or health care administrations.
OBJECTIVES: Little is known about doctors' preferences regarding public report card design. Taiwan just announced the first diabetes report card on April 2008. The aim of this study was to investigate the Diabetes Mellitus (DM) providers' preferences towards four report card attributes: update frequency, risk adjustment, content information and display format. METHODS: A discrete choice questionnaire was mailed to hospital and primary care doctors in the northern part of Taiwan, with 221 study targets. The response rate was 29%. RESULTS: Using random effect logistic regression, doctors' preference attribute rankings were risk adjustment for patients (44.7%), content information (25.2%), display format (18.3%) and update frequency (11.8%). One-year update frequency, risk adjustment, detailed scores of technical quality and interpersonal quality and bar chart display were the most important items noted in our survey. CONCLUSIONS: This is the first study to investigate provider's preferences for a diabetes report card. It enables the policy maker to clearly see the implication of trade-offs between different choices when designing a report card that doctors will like. Our findings suggest that doctors do not favour the 'less is more' principle, or the higher frequency of updates that patients may prefer. Rather, our findings suggest that risk adjustment and more information content are the most important factors for doctors. Future studies should use discrete choice experiment on different aspects of report card design, such as vulnerable patients or health care administrations.