Self-Reported Complications after Tonsillectomy: Comparison of Responders and Nonresponders to a Mailed Questionnaire

Abstract

Some studies of tonsillectomy outcomes have low response rates to mailed quality control questionnaires. This study evaluated the effect of nonresponders to mailed questionnaires about posttonsillectomy complications by determining whether mail responders and nonresponders differ. Questionnaires were mailed to patients 3–6 weeks after tonsillectomy to assess postoperative complications, defined as contact with a private practitioner and/or hospital readmission related to postsurgical bleeding, pain, or infection. Nonresponders to the mailed questionnaire were interviewed by telephone 7–11 weeks postoperatively, and responses of mail and telephone responders were compared. Of 818 patients undergoing tonsillectomy during the study period, 66.3% responded by mail, and 29.5% were interviewed by telephone, for a total response rate of 95.7%. The mail response rate was significantly higher among parents of pediatric patients than among adult patients (71.4% versus 58.7%, p<0.001). In the pediatric group, overall complication rates were 65% higher among mail responders than telephone responders (20.9% versus 12.7%, p=0.049), likely due to their higher rates of both visits to private practitioners and infection, as there were no differences in rates of pediatric readmission, bleeding, or pain between the responder groups. Among adult patients, mail and telephone responders did not differ with respect to their overall complication rate (40.9% versus 34.1%, p=0.226) or their rates of readmission or bleeding. However, similar to the pediatric group, visits to a private practitioner were slightly more common among adult mail responders than telephone responders (30.6% versus 21.1%, p=0.065), as were reports of pain (p=0.001) and infection (p=0.006). Studies relying on mailed questionnaires with low response rates likely overestimate the rate of minor complications handled outside the hospital, but rates of major complications involving readmission are unlikely to be seriously biased by low response rates. Supplementing mailed questionnaires with telephone interviews may increase the validity of surgical outcome studies.