This randomised methodological study sought to test the reliability of an Internet questionnaire and investigate the differences in response rates between traditional pen-and-paper questionnaires and Internet questionnaires for measuring patient-reported outcome after total hip arthroplasty (THA) surgery.

From the Swedish Hip Arthroplasty Register, 2 400 patients were chosen at random but stratified by age, sex and diagnosis for inclusion in a four-year follow-up using the health-related quality of life (HRQoL) tool EQ-5D and visual analogue scales for pain and satisfaction. The patients were randomized to answer the follow-up model protocol either via a password-protected Internet questionnaire or via a mailed pen-and-paper questionnaire.

A reliability test for the Internet follow-up instrument showed adequate correlation. However, the Internet group and the pen-and-paper group differed significantly (p<0.001) with a 92% response rate in the latter and 49% in the former. Adjusted to the normal age distribution of the THA population, the Internet response rate was 34%.

The patient-administered Internet questionnaire alone does not give a sufficient response rate in the THA population to replace the pen-and-paper questionnaire. However, the system is reliable and could be used for measuring patient- reported outcome if supplemented with traditional pen-and-paper questionnaires for Internet non-respondents. It is expected that this answer procedure will soon predominate in view of the general development of Internet functions. Register work may then become less resource-consuming and the results may be analysed in real time.