This article uses a novel instrumental variable to assess whether patients' assessment of the communication skills of their doctor was related to overall satisfaction with hospital care. The authors found that there is a significant and strong relationship between quality of physician communication and overall patient satisfaction, even after adjusting for patient-level factors.
The authors used data from a survey administered to 3,123 patients at a large urban hospital between July 1997 and June 2000. Patients rated their doctors on four communication behaviors and also rated their overall satisfaction with hospital care. The authors performed two analyses, an ordinary least squares regression and a two-stage regression with a novel instrumental variable.
In the ordinary regression, researchers found a significant positive relationship between physician communication skills and overall satisfaction.
The two-stage regression confirmed the significant positive relationship after controlling for patient factors.
By adjusting for patient factors, researchers removed the possibility of confounding results due to the tendency of a patient to give a high or low rating and the possibility that a patient's positive or negative hospital experiences would influence their assessment of physician communication.
The authors' instrumental variable may provide a new method to analyze situations where there is confounding due to patient-level factors.