Potentially modifiable influences on cancer family history-taking among primary care physicians in Vermont and Massachusetts

Abstract

6067 Background. Taking a complete family history is key in identifying individuals at increased risk for cancer. The objectives of this study were to test hypothesized relationships between potentially modifiable factors and family history-taking by primary care physicians. Methods. Primary care physicians (family medicine, internal medicine, gynecology) were surveyed by mail in Vermont, urban and suburban Boston (n=880) using comprehensive rosters from employers or state-wide networks, with a response rate of 70%. Measures of family history-taking quality and factors that may influence this behavior were developed systematically and validated using factor analysis methods. The primary outcome variable was either extent of history-taking or age of cancer diagnoses. These dependent variables were regressed on scales measuring physician perceptions of the advantages and disadvantages of history-taking, confidence in taking cancer family histories, supportive resources, and knowledge of management for patients at high risk for breast and colon cancer; relevant factors controlled in these multiple regressions included type and location of practice. Results. Most physician perception factors hypothesized to influence quality of history taking were found to be significantly associated in the analysis. The strongest predictor of extent of family history-taking was perceived disadvantages of taking a cancer family history (p<.001). Reports of supportive resources ((p<.01); perceived advantages (p=.02); and confidence (p=.02) were also associated with this behavior. Gathering age of cancer diagnoses was significantly associated with perceived advantages, supportive resources and confidence (all p<.0001). Knowledge of high risk management was not associated with quality of family history-taking in either analysis. Conclusions. Potentially modifiable resource and perception factors were significantly associated with quality of family history in a large and diverse sample of primary care physicians. Improving family history quality for identification of high risk individuals among primary care providers will need to address multiple factors. No significant financial relationships to disclose.