Adaptation to life after surgical removal of the bladder-an application of graphical Markov models for analysing longitudinal data.

Abstract

Graphical Markov models have been developed particularly for the analysis of observational data. They allow the control of various background variables when analysing theoretically relevant associations. This paper demonstrates the application and some advantages of graphical Markov models in comparison to conventional statistical analyses. The aim of the study was to identify patients at risk for developing decreased health-related quality of life (QoL) after cystectomy and to explore the influence of coping on QoL in this situation. Therefore, the method was applied to analyse the data of a prospective study, in which 81 patients with bladder cancer were interviewed pre-operatively and in a 1-year follow-up. QoL was assessed both times, and two basic coping strategies (active and depressive) were measured preoperatively. The explanatory variables of theoretical interest were active and depressive coping strategies. As a result of the analysis, relevant proportions of variance in the development of QoL could be explained by the suggested model (60 per cent in mental component, 40 per cent in physical component of QoL). Active coping was positively related to QoL, depressive coping negatively. These effects were linear in the physical component of QoL, moderated by working status and the type of urinary diversion in the mental component of QoL.