An innovative application of Bayesian disease mapping methods to patient safety research: a Canadian adverse medical event study

Abstract

Recently developed disease mapping and ecological regression methods have become important techniques in studies of disease epidemiology and in health services research. This increase in importance is partially a result of the development of Bayesian statistical methodologies that make it possible to study associations between health problems and risk factors at an aggregate (i.e. areal) level while taking into account such matters as unmeasured confounding and spatial relationships. In this paper we present a demonstration of the joint use of empirical Bayes (EB) and full Bayesian inferential techniques in a small area study of adverse medical events (also known as 'iatrogenic injury') in British Columbia, Canada. In particular, we illustrate a unified Bayesian hierarchical spatial modelling framework that enables simultaneous examinations of potential associations between adverse medical event occurrence and regional characteristics, age effects, residual variation and spatial autocorrelation. We propose an analytic strategy for complementary use of EB and FB inferential techniques for risk assessment and model selection, presenting an EB-FB combined approach that draws on the strengths of each method while minimizing inherent weaknesses. The work was motivated by the need to explore relatively efficient ways to analyse regional variations of health services outcomes and resource utilization when a considerable amount of statistical modelling and inference are required.

Regression B‐spline smoothing in Bayesian disease mapping: with an application to patient safety surveillance

“…The parameter could also be called a weight parameter, as the precision matrix CAR interpolates between that of the intrinsic CAR (ICAR) model of Besag et al [2] (when = 1) and that of an independence model (when = 0) [27]. While various definitions of neighbourhood have been discussed in the literature, area 4460 Y. C. MACNAB AND P. GUSTAFSON adjacency (areas that share a common border) has been the most commonly studied neighbourhood structure in disease mapping [1][2][3].…”

“…Hospitalized individuals were aggregated to 84 LHAs based on patient place of residence and to 1991-2000 annually based on date of admission [27]. This analysis is drawn from an ongoing study entitled British Columbia Iatrogenic Injury Study.…”

“…There is a growing literature on hierarchical models for spatiotemporal disease surveillance or registry data [1,[23][24][25]. Applications of recently developed Bayesian disease mapping methodology, computing technology, and geographic information system (GIS) techniques enable us to explore the potential of model-and GIS-based spatiotemporal disease and injury mapping as part of a contemporary space-time disease and injury surveillance and prevention process [7,[23][24][25][26][27]. In this study, we present such an application.…”

Spline smoothing in Bayesian disease mapping

“…, N for N geographical subregions. The subsequent analysis is often pursued by Bayesian hierarchical modeling of the count data via a spatial prior specification for a 'spatially varying' (log) relative risks ensemble (Lawson et al, 1999;Leroux et al, 1999, Ugarte, et al, 2006MacNab et al, 2006; to name a few).…”

Are adverse effects incorporated in economic models? An initial review of current practice

“…The remaining four all included an explicit statement to say that adverse effects had been captured in the utility valuation 25 or through the use of withdrawals. [26][27][28] Therefore, all 13 were considered to have explicitly included adverse effects in the model.…”