Analysing the length of care episode after hip fracture: a nonparametric and a parametric Bayesian approach

Abstract

Effective utilisation of limited resources is a challenge for health care providers. Accurate and relevant information extracted from the length of stay distributions is useful for management purposes. Patient care episodes can be reconstructed from the comprehensive health registers, and in this paper we develop a Bayesian approach to analyse the length of care episode after a fractured hip. We model the large scale data with a flexible nonparametric multilayer perceptron network and with a parametric Weibull mixture model. To assess the performances of the models, we estimate expected utilities using predictive density as a utility measure. Since the model parameters cannot be directly compared, we focus on observables, and estimate the relevances of patient explanatory variables in predicting the length of stay. To demonstrate how the use of the nonparametric flexible model is advantageous for this complex health care data, we also study joint effects of variables in predictions, and visualise nonlinearities and interactions found in the data.

Risk of Bleeding With Dabigatran in Atrial Fibrillation

“…We further examined the incidence of bleeding in subgroups stratified by age (<75 or ≥75 years) and among African Americans, users with renal impairment, and patients with at least 7 priority CMS conditions other than AF. 12 Subgroup analyses were performed following the same methods and controlling for all covariates except for the one defining the subgroup. For example, age-stratified analysis controlled for the same covariates except for age.…”

“…The diagnosis of AF was defined as having 1 inpatient or 2 outpatient claims with primary or secondary International Classification of Diseases, Ninth Revision (ICD-9), code 427.31. 12 We also required that individuals in our study sample had filled an outpatient prescription for either dabigatran or warfarin within 2 months of the first diagnosis (N = 9562). Those who filled prescriptions for dabigatran and warfarin during the first 2 months after diagnosis were excluded (N = 158).…”

Phase-Type Survival Trees and Mixed Distribution Survival Trees for Clustering Patients' Hospital Length of Stay

“…However, for a large sample population with few ties, hospital length of stay can still be modelled as a continuous time random variable. In the healthcare literature, many researchers have modelled hospital length of stay as a continuous random variable (Abbi et al, 2008;El-darzi et al, 2009;Fackrell et al, 2009;Marshall et al, 2007;Marshall and McClean, 2004;Marshall and Zenga, 2009;McClean et al, 2007;Riihimäki et al, 2010).…”

Comparing clinical and economic outcomes of biologic and conventional medications in postmenopausal women with osteoporosis

“…We followed each individual from the index date, defined as 180 days after drug initiation, until they switched their osteoporosis treatment, died or at end of the study period (31 December 2011). We set the index date as 180 days after drug initiation for two reasons: (1) to allow treatments to achieve therapeutic effects [10,11]; and (2) to make sure our clinical outcomes did not capture follow‐up visits of fractures that happened before drug initiation [11,12]. We did not consider discontinuation of treatment as a censoring event because of the residual effects of the treatments after discontinuation [13,14].…”