Patient-centered clinical decisions and their impact on physician adherence to clinical guidelines.

BACKGROUND: This study was undertaken to assess the impact of traditionally unmeasured patient-centered factors on primary care physicians' decisions to adhere to an evidence-based clinical practice guideline for heart failure.
METHODS: Experimental and control scenarios were developed to test three patient-centered factors hypothesized to influence physician nonadherence to a heart failure guideline: patient concerns about finances, quality of life, and location of care. Each factor represented an implicit patient goal potentially in conflict with a goal of the guideline recommendations. A control scenario for one factor and an experimental scenario for a second were placed within a cross-sectional survey and questionnaires were mailed by random assignment to 978 Upstate New York family physicians. Experimental and control responses were compared by chi square.
RESULTS: The response rate was 47% (n = 456). Each hypothetical patient-centered factor resulted in significant reductions in physicians' predicted adherence. Reductions in reported pharmaceutical usage and testing of left ventricular (LV) function were associated with patient financial difficulties (P < .01). The poor quality-of-life scenario was associated with reduced testing for LV function but increased discussion of advance directives (P < .01). The clinical scenario limiting access to services for a rural patient was associated with decreases in physician choice of LV function tests and cardiology referrals (P < .05).
CONCLUSIONS: Patient-specific factors are associated with physician decisions to comply with guideline recommendations. These findings suggest that performance profiles measuring physician adherence to guidelines should be interpreted with caution, and that current case-mix methodologies may not adequately control for patient-centered factors that may influence health care quality.