BACKGROUND

In 1998 the BJC Health System (St Louis) made the decision to migrate its patient satisfaction measurement system from a mail-out/mail-back method to a phone interview method. Out of concern that results obtained by phone would not be comparable with the 4 years of mail-based data, a controlled study was undertaken to directly compare mail and phone responses and to evaluate response rates, patient sample demographics, and patient satisfaction ratings.

METHODS

Mail and phone responses obtained from parallel random samples selected from inpatient, outpatient test/treatment, outpatient surgery, and emergency service patient populations were compared. Patients were randomly selected to receive a standardized satisfaction survey by either phone or mail 10 to 14 days postdischarge.

RESULTS

Significantly higher response rates were obtained by telephone then via the mail-based method for all four samples. After adjusting for demographic differences, numerous significant differences in mean scores as well as percentages of excellent and fair or poor responses were observed, and more positive ratings were obtained by phone.

DISCUSSION

Crude comparisons of satisfaction scores between organizations using phone and mail-based responses may lead to erroneous conclusions about consumer-perceived quality. Organizations that use mixed-mode surveys should conduct careful side-by-side studies of the methods used on the survey of interest and then establish a correction formula to adjust the results for the measurement biases.