Differences between responders and nonresponders in intensive care unit (ICU) related survey studies.

Abstract

In survey research studies, there are often a proportion of non-responders, that is patients or relatives of patients who do not want to fill in a survey. Especially in intensive care medicine, these proportions of non-responders can be relatively high. However, with the increasing incidence of post-intensive care syndrome (PICS), it can be beneficial to know what the predisposing factors are. We performed a sub analysis of a recent survey study we have conducted on relatives of deceased ICU patients. Of the 219 eligible relatives, 139 were willing to fill in the questionnaires and 95 of them actually returned the questionnaire, which raises the question whether there are differences between the 95 relatives who did fill in the questionnaires and the 44 relatives who did not. We collected information from the medical records of deceased patients whose relatives did and did not fill in the questionnaires, that is responders and non-responders. Patient's whose relatives did not fill in the questionnaire had a significantly lower median age compared with patients whose relatives returned the questionnaires (64 [56-75] vs. 71 [65-76] years, respectively, p = .039). The non-responders were two relatives of a deceased patient below 60 years in 34.1% of the cases, whereas this percentage was 14.7% in the responders. The distribution of men and women was comparable between the groups (p = .54) as well as the length of ICU stay (p = .48). Our findings suggest that ICU-related variables did not influence response rate, but the age of the deceased ICU patient does: The non-responders are related to deceased ICU patients with a significantly lower age.