Bayesian consensus clustering for multivariate longitudinal data

In clinical and epidemiological studies, there is a growing interest in studying the heterogeneity among patients based on longitudinal characteristics to identify subtypes of the study population. Compared to clustering a single longitudinal marker, simultaneously clustering multiple longitudinal markers allow additional information to be incorporated into the clustering process, which reveals co-existing longitudinal patterns and generates deeper biological insight. In the current study, we propose a Bayesian consensus clustering (BCC) model for multivariate longitudinal data. Instead of arriving at a single overall clustering, the proposed model allows each marker to follow marker-specific local clustering and these local clusterings are aggregated to find a global (consensus) clustering. To estimate the posterior distribution of model parameters, a Gibbs sampling algorithm is proposed. We apply our proposed model to the primary biliary cirrhosis study to identify patient subtypes that may be associated with their prognosis. We also perform simulation studies to compare the clustering performance between the proposed model and existing models under several scenarios. The results demonstrate that the proposed BCC model serves as a useful tool for clustering multivariate longitudinal data.     

A Tool to Assess Patient and Surrogate Knowledge About the POLST (Physician Orders for Life-Sustaining Treatment) Program

ContextIt is especially important that patients are well informed when making high-stakes, preference-sensitive decisions like those on the Physician Orders for Life-Sustaining Treatment (POLST) form. However, there is currently no way to easily evaluate whether patients understand key concepts when making these important decisions.ObjectivesTo develop a POLST knowledge survey.MethodsExpert (n = 62) ratings of key POLST facts were used to select items for a POLST knowledge survey. The survey was administered to nursing facility residents (n = 97) and surrogate decision-makers (n = 112). A subset (n = 135) were re-administered the survey after a standardized advance care planning discussion to assess the scale's responsiveness to change.ResultsThe 19-item survey demonstrated adequate reliability (α = 0.72.). Residents' scores (x = 11.4, standard deviation 3.3) were significantly lower than surrogate scores (x = 14.7, standard deviation 2.5) (P < 0.001). Scores for both groups increased significantly after administration of a standardized advance care planning discussion (P < 0.001). Although being a surrogate, age, race, education, cognitive functioning, and health literacy were significantly associated with higher POLST Knowledge Survey scores in univariate analyses, only being a surrogate (P < 0.001) and being white (P = 0.028) remained significantly associated with higher scores in multivariate analyses.ConclusionThe 19-item POLST Knowledge Survey demonstrated adequate reliability and responsiveness to change. Findings suggest the survey could be used to identify knowledge deficits and provide targeted education to ensure adequate understanding of key clinical decisions when completing POLST.