Development and validation of a geriatrics attitudes scale for primary care residents

OBJECTIVE: To develop and validate an instrument measuring attitudes toward older persons and caring for older patients. DESIGN: Cross-sectional and longitudinal studies. SETTING: An academic medical center. PARTICIPANTS: Initial Study: 121 primary care residents (n = 96), fellows (n = 14), and faculty (n = 11) participated in instrument development in 1995. Longitudinal Study: 95 residents (n = 87) and fellows (n = 8) of the initial cohort participated in the 1996 follow-up study, and 61 of the initial cohort (57 residents and 4 fellows) participated in the 1997 follow-up study. Cross Validation Study: 96 first-year residents (n = 78) and fellows (n = 18) participated in this study. MEASUREMENTS: A 14-item geriatrics attitudes scale was developed. The items were selected from a pool of 37 items administered to the 121 participants in the initial study. RESULTS: The instrument demonstrated high reliability (Cronbach's alpha = .76) and known-groups and construct validity. Attitudes were progressively more positive with more medical training (P < .001), and residents with greater career interest in geriatrics scored higher than those less interested (P = .007). Cross validation results supported the reliability and validity of the instrument. Longitudinal data showed significantly different trends of attitude changes among groups of residents and fellows over a 2-year period. CONCLUSIONS: The 14-item geriatrics attitudes scale developed in this study shows sound reliability, validity, and sensitivity to change among primary care residents. The performance of other groups of medical trainees and the relationship of attitude changes to specific medical training warrant further investigation.