Loss to follow-up in longitudinal psychiatric research

Epidemiol Psichiatr Soc. 2008 Apr-Jun;17(2):138-47.

Abstract

Aims: To analyse factors that differentiate patients who attend follow-up assessments versus those who do not, and to identify predictors for drop-out within the context of the European Day Hospital Evaluation Study (EDEN-Study).

Methods: The EDEN-Study, a multi-center RCT comparing acute psychiatric day care with inpatient care, required re-assessment of patients at discharge, 3 and 12 months after discharge. Follow-up rates varied between 54.0% and 99.5%. Socio-demographic and clinical characteristics of patients who did and did not attend follow-up were analysed using uni- and multivariate statistical methods.

Results: Univariate analyses showed differences between patients regarding study site, treatment setting, living situation, employment, age, psychopathological symptoms and treatment satisfaction. They were not confirmed in multivariate analyses thus meaningful predictors of drop-out could not be identified.

Conclusions: Results emphasize the general need to compare patients re-assessed and not re-assessed in terms of their most relevant socio-demographic and clinical variables to assess the generalizability of results.

MeSH terms

  • Adult
  • Female
  • Follow-Up Studies
  • Humans
  • Male
  • Mental Disorders / therapy*
  • Patient Dropouts*
  • Psychiatry*
  • Research