Presenting comparative information on the quality of health care to consumers.
Damman, O.C., Hendriks, M., Rademakers, J.J.D.J.M., Delnoij, D.M.J., Groenewegen, P.P. Presenting comparative information on the quality of health care to consumers.: , 2008.
Objective: To assess which presentation approaches contribute most to consumers’ interpretation and effective use of information on the quality of health care. Methods: We used fictitious Consumer Quality Index (CQ-index) data on the quality of home care providers to construct 15 different presentation formats of comparative information. CQ-index data regard the quality of health care from the patient’s perspective. In the Netherlands, this family of standardised questionnaires is used to generate comparative information, which should enable consumers to make informed choices in health care. For this study, one quality aspect was selected, namely ‘provider-client interaction’. An experimental design using conjoint analysis methodology was applied to assess the effects of five presentation features: 1) the use of bar charts and / or star ratings; 2) the number of stars (three or five); 3) the type of stars (based on absolute versus relative performance); 4) the ordering of providers (alphabetical or in order of performance); and 5) the addition of a global rating for the home care provider. A total of 438 online access panel members responded to 4 different formats. We performed multilevel regression analysis to investigate the influence of the presentation features. Results: Both person characteristics (age and education) and presentation features significantly affected consumers’ responses. The significant regression coefficients (standard errors) of the person characteristics varied from 0.372 (0.176) to -1.170 (0.468). For the presentation features, the significant regression coefficients (standard errors) varied from 0.431 (0.153) to 1.808 (0.142). In general, approaches using a combination of bar chart and stars, an alphabetical ordering of performances, three stars, and no inclusion of a global rating supported consumers. However, the effects of the presentation features differed across the different outcome variables. Conclusions: Our findings show that certain features of the presentation approach of health care quality information support consumers. This provides us with important suggestions for presenting such public information on the Internet. The effects differ across the different outcomes: interpretation and use of the information. In our opinion, health care policy makers should reflect whether publishing comparative health care information is intended to help consumers interpret the information, or to make an informed choice. Further research should investigate consumers’ decision strategies and their interest in information that is adjusted to their needs and interests. (aut. ref.)
Objective: To assess which presentation approaches contribute most to consumers’ interpretation and effective use of information on the quality of health care. Methods: We used fictitious Consumer Quality Index (CQ-index) data on the quality of home care providers to construct 15 different presentation formats of comparative information. CQ-index data regard the quality of health care from the patient’s perspective. In the Netherlands, this family of standardised questionnaires is used to generate comparative information, which should enable consumers to make informed choices in health care. For this study, one quality aspect was selected, namely ‘provider-client interaction’. An experimental design using conjoint analysis methodology was applied to assess the effects of five presentation features: 1) the use of bar charts and / or star ratings; 2) the number of stars (three or five); 3) the type of stars (based on absolute versus relative performance); 4) the ordering of providers (alphabetical or in order of performance); and 5) the addition of a global rating for the home care provider. A total of 438 online access panel members responded to 4 different formats. We performed multilevel regression analysis to investigate the influence of the presentation features. Results: Both person characteristics (age and education) and presentation features significantly affected consumers’ responses. The significant regression coefficients (standard errors) of the person characteristics varied from 0.372 (0.176) to -1.170 (0.468). For the presentation features, the significant regression coefficients (standard errors) varied from 0.431 (0.153) to 1.808 (0.142). In general, approaches using a combination of bar chart and stars, an alphabetical ordering of performances, three stars, and no inclusion of a global rating supported consumers. However, the effects of the presentation features differed across the different outcome variables. Conclusions: Our findings show that certain features of the presentation approach of health care quality information support consumers. This provides us with important suggestions for presenting such public information on the Internet. The effects differ across the different outcomes: interpretation and use of the information. In our opinion, health care policy makers should reflect whether publishing comparative health care information is intended to help consumers interpret the information, or to make an informed choice. Further research should investigate consumers’ decision strategies and their interest in information that is adjusted to their needs and interests. (aut. ref.)
