Premium
A survival analysis for recurrent events in psychiatric research
Author(s) -
Baethge Christopher,
Schlattmann Peter
Publication year - 2004
Publication title -
bipolar disorders
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.285
H-Index - 129
eISSN - 1399-5618
pISSN - 1398-5647
DOI - 10.1111/j.1399-5618.2004.00104.x
Subject(s) - survival analysis , carbamazepine , lithium (medication) , psychiatry , event (particle physics) , statistical analysis , proportional hazards model , medicine , psychology , clinical psychology , statistics , mathematics , epilepsy , physics , quantum mechanics
Objectives: Time to first recurrence, as analyzed by the Kaplan–Meier (KM) survival analysis, is a commonly applied statistical method in psychiatric research. However, many psychiatric disorders are characterized not by a single event, but rather by recurrent events, such as multiple affective episodes. This study aims to demonstrate a method of survival analysis that takes multiple recurrences into account. Methods: We examined data on sex differences in a sample of 181 patients undergoing prophylactic treatment with lithium or carbamazepine (serum level assayed) for bipolar disorder (ICD‐10). The classical KM method was compared with an approach developed by Peña, Strawderman and Hollander (PSH) that uses recurrent event data to estimate survival function. Results: The results obtained with the multiple events method differed considerably from those acquired using the standard KM analysis. When taking recurrent event data into account, the probability of remaining well was lower and survival times were longer. In addition, whereas the standard KM analysis indicated that male patients had a higher likelihood of remaining well, the alternative method revealed that both sexes were similarly likely to remain well. Conclusions: Survival analysis techniques that take recurrent events into account are potentially important instruments for the study of psychiatric conditions characterized by multiple recurrences. In many cases, the standard KM analysis appears to provide only a rough approximation of the course of illness.