Objective: The role of psychological factors in adult survival after bone marrow transplant (BMT) has not been adequately established. BMT survival rates have been shown to depend largely on disease and demographic variables. With sample sizes no larger than 123, the majority of psychosocial studies found little definitive statistical evidence that mood, marital/cohabiting status, coping styles, or smoking affect survival. The purpose of this study was to comprehensively examine the relative contribution of psychological variables to survival after BMT in a unified model, controlling for medical and demographic variables.
Methods: Pre-BMT, 305 consecutive patients were psychologically assessed with interview and Minnesota Multiphasic Personality Inventory (MMPI). Transplant Evaluation Rating Scale (TERS) ratings were assigned retrospectively by two raters (interrater reliability r=.89). We employed a fully parametric, accelerated failure time regression model (Weibull), which provides richer extrapolation and interpretation than the semiparametric Cox proportional hazards regression model.
Results: Overall our results indicate that for a given medical condition, the type of BMT patient likely to survive longest was a young, married, educated, European-American, nonsmoker who was more defiant, better adjusted, and less depressed. Moreover, the longer a patient survived, the smaller the probability of dying in the short run.
Conclusions: This study, the first large-scale statistical analysis using a fully parametric model (Weibull), provides evidence that select psychosocial variables can affect BMT survival. Future investigations could explore possible mediating variables, and whether identifying high-risk individuals pretreatment could enhance resource allocation, psychological intervention, and possibly even survival itself.