Many endoscopy units administer patient satisfaction surveys. We hypothesized that the survey collection method would affect satisfaction scores.

To compare satisfaction scores obtained by using on-site (OS) surveys versus mail-back (MB) surveys.

Quasi-randomization based on alternating weeks.

Teaching hospital.

Patients undergoing elective routine outpatient colonoscopy or upper endoscopy.

Every patient was given an 11-question survey that asked about the patient’s satisfaction with the nurses and the physician, wait times, the bowel-preparation process, patient education, procedural comfort, and sedation. Survey collection methods alternated weekly between an OS versus an MB method.

Satisfaction scores on a Likert scale ranged from 1 (worst) to 7 (best).

A total of 1698 subjects were included. The response rate was higher for the OS group (95%) than the MB group (62%). OS scores were significantly higher than MB scores for 5 of 11 questions, which concerned nurse satisfaction, physician satisfaction, bowel-preparation comfort, postprocedure education, and overall satisfaction (Bonferroni adjusted P < .05 for all). Younger patients gave lower scores than older patients for all questions, whereas women gave significantly lower scores than men for bowel-preparation satisfaction.

Lack of true randomization and formal validation of the satisfaction survey.

Survey collection methods may bias not only response rates but also satisfaction scores. OS survey collection methods tend to result in higher satisfaction scores than MB methods. This bias should be noted when comparing scores among studies that used different survey collection methods.