Does sponsorship matter in patient satisfaction surveys? A randomized trial

The purpose of this study was to assess the impact of sponsorship of a mailed satisfaction survey on response patterns and patient satisfaction ratings, and to establish whether satisfaction ratings depend on the timing of response. The study was part of a patient satisfaction survey conducted in a medical group practice in Geneva, Switzerland. All persons who had received an outpatient visit at the medical practice in the last week of September 1993 were assigned randomly to receiving the survey package either on university letterhead or on the medical practice letterhead. Participation rates were 80.5% and 80.4% in the two groups, respectively. The sponsorship of the survey had no detectable impact on response speed or response completeness. Six out of seven satisfaction scores were higher in the "medical practice" group, but the differences were not statistically significant. As data collection progressed, the cumulative satisfaction rating decreased gradually. This association was statistically significant but moderate. Whether a satisfaction survey is sponsored by a university or a health care provider does not seem to affect survey results. However, low survey response rates may moderately bias satisfaction estimates toward higher values.