Background:

Hospitals are increasingly interested in patient satisfaction as an outcome measure. The Press‐Ganey (PG) patient satisfaction survey contains items related to satisfaction with individual physician's communication. On receipt of the completed survey these results are matched to the discharging attending. However, the survey does not specify which physician, among the many a patient might see, should be rated. We sought to compare the physician‐specific results of PG surveys to a communication survey conducted during hospitalization that asked the patient to verify the hospitalist's identity.

Methods:

The PG patient satisfaction survey was mailed to a random sampling of 40% of hospitalized patients after discharge, and results are summarized based on a typical response rate of 30%. The physician section of the survey consists of 5 questions related to the patient's satisfaction with her or his physician's communication and skill. Items are answered on a 5‐point Likert scale (very poor = 1, poor, fair, good, very good = 5). Returned surveys are matched via a reference number to the discharging hospitalist named in the patient's account. The Communication Assessment Tool (CAT) is a 14‐item psychometricaily validated instrument for patient assessment of physician communication skills. Items are answered on a 5‐point Likert scale (poor = 1, fair, good, very good, excellent = 5), Results for each physician are reported as the percentage of ”5” responses. Trained research assistants verbally administered the CAT to patients admitted to the hospital medicine or teaching service. Patients who could not identify their hospitalist by name or photo were excluded from the analysis. Pearson's R correlation was used to compare the PG survey results to the CAT survey results for each individual hospitalist. Twenty CAT surveys were collected per hospitalist, and the number of PG surveys ranged between 20 and 41 (mean = 27.5). Pearson's R was weighted against an n of 20. The content of 2 items on the PG survey match closely to CAT items: PG item “Time spent with you” versus CAT item “Spent the right amount of time with me,” and PG item “How well physician Kept you informed” versus CAT item “Gave me as much information as I wanted.”

Results:

CAT survey scores did not correlate with PG survey scores regardless of whether percentage “5” responses (r = 0.20, P = ns) or average score (r = 0.32. P = ns) was used. A correlation was found between the PG and CAT information question scores when comparing the PG average score for that question to the CAT score (R = 0.47. P < 0.05), and a trend for the PG score reported as percent 5s (r = 0.42, P = 0.06).

Conclusions:

There was no correlation found between the PG physician‐specific ratings and the CAT ratings for hospitalists. The questions related to information did show correlation. Results indicale that Press‐Ganey physician‐specific survey results may not be accurately tied to the discharging attending.

Author Disclosure:

D. Ferranti, none; J. Lee, none; G. Gavron, none; V. Forth, none; M. Williams, none.