Case-mix adjustments for patient reported experience and outcome measures in primary care: an empirical approach to identify patient characteristics as case-mix adjusters based on a secondary analysis of an international survey among patients and their general practitioners in 34 countries

Journal of Patient-Reported Outcomes

Table 1 Summary of the analyses for the dependent variable Communication (PREM), multilevel linear regression

Potential case-mix variable	Fixed effect significant (y/n)	Slope effect GP level important* (y/n)	Slope effect Country level important (y/n)	Case-mix adjustment (y/n)
Self-reported general health	Yes Worse self-reported health → experienced communication worse	No	No	Yes
Longstanding disease	Yes Longstanding disease → experienced communication worse	No	No	Yes
Patient’s age	Yes, Older patients → experienced communication worse	No	No	Yes
Patient’s sex	No	No	No	No
Education	Yes Higher education → experienced communication better	Yes	No	No Importantly more variation at GP level with lower educated
Income	Yes Higher household income → experienced communication better	No	No	Yes
Migrant status	Yes 1st generation migrant → experienced communication worse	Yes	No	No Importantly more variation at GP level with 1st generation migrants
Place of living	No	Yes	No	No No significant fixed effect, but importantly more variation in small towns compared to suburbs and rural areas

*Important means that the difference in variance between categories is more that 0.25*variance in the model with fixed effect and random intercept