Validity of residents' self-reported cardiovascular disease prevention activities: The Preventive Medicine Attitudes and Activities Questionnaire

Background. This article describes the development, reliability, and validity of three cardiovascular disease (CVD) prevention subscales - CVD prevention behaviors, perceived importance, and perceived effectiveness - of the Preventive Medicine Attitudes and Activities Questionnaire (PMAAQ). Methods. The PMAAQ was administered three times to University of Minnesota family practice residents (178) over 2 years (91% response rate). Stability measures were calculated, and validity was demonstrated in four ways: content validity through an expert panel; calculation of internal consistency reliabilities; demonstration of divergent validity; and external validation via a separate chart review. Results. High internal consistency reliabilities among the subscales were seen (Cronbach's α = 0.77 to 0.92). Divergent validity was verified by low intercorrelations among the subscales (r = -0.23 to 0.27). Two-month test-retest scores ranged from Cronbach's α = 0.47 to 0.64. Significant correlations were seen between the chart review scale and both the CVD behaviors subscale and the PMAAQ smoking scale (r = 0.25 and 0.36, respectively). Conclusions. Results indicate that the PMAAQ can validly and reliably measure residents' CVD prevention behaviors and provide insight into their preventive health care attitudes. Further, the independence among the subscales suggests that importance and effectiveness by themselves do not affect behavior and that other factors are likely to be important in influencing physician behavior change. (C) 2000 American Health Foundation and Academic Press.