Patients' preferences for technical versus interpersonal quality when selecting a primary care physician

Health Services Research, August, 2005 by Constance H. Fung, Marc N. Elliott, Ron D. Hays, Katherine L. Kahn, David E. Kanouse, Elizabeth A. McGlynn, Mark D. Spranca, Paul G. Shekelle

Wilcoxon's signed-rank test results indicate that the median trust in expert review of medical records is significantly higher than for patient reports (p<.001), with the differences being most apparent at the highest levels of support (35 percent of participants trusting medical records "a lot," as compared with 19 percent trusting patient reports "a lot"). Trust in expert review of medical records was correlated with trust in patient reports (r = 0.34, p<.001), but neither of these variables was correlated with our main dependent variable, the number of times the participant selected the physician who excelled in technical quality for the tradeoff pairs (trust in expert review of medical records r = - 0.04, p = .52; trust in patient reports r = 0.08, p = .16).

Age, gender, race/ethnicity, caregiver status, number of physician visits in past year, chronic disease, subject understanding of hypothetical report cards, computer use, income, and education were not significant predictors of preference for technical quality at the [alpha] = 0.05 level in our ordered logit model.

DISCUSSION

The principal findings from our study are that participants use both technical and interpersonal quality ratings when selecting a PCP and that a majority clearly favors technical quality of care, but not to the exclusion of interpersonal quality. These findings provide insight into the values people place on technical and interpersonal quality when selecting a primary care physician. The results help us understand the choices people might make if they had comparative information about the technical and interpersonal quality of care of primary care physicians in their area.

Our results are directly relevant to the movement to develop and disseminate physician-specific report cards. Although when forced to make tradeoffs, the majority of our participants selected the physician who was higher in technical quality, approximately one-third of the participants chose the physician who was higher in interpersonal quality. In particular, half of participants chose a physician dominant in interpersonal quality when the alternative was a physician with poor interpersonal quality. These findings suggest that future report cards on primary care physicians should include measures of both technical and interpersonal quality and should display these measures in a clear manner. Existing health plan and hospital-level report cards do not consistently provide both types of measures or focus primarily on one type of measure. For example, NCQA's widely used online Health Plan Report Card, which contains the Health Plan Employer Data and Information Set, consists primarily of technical process measures derived from review of records or claims data and interviews of health plan staff. Patient reports on interpersonal quality are included in only one domain, "Staying Healthy," and are aggregated with technical measures. U.S. News and World Report's "Best Hospitals" report card also focuses primarily on technical outcome and volume measures such as mortality ratio and number of discharges rather than interpersonal measures (U.S. News and World Report 2003).