Abstract
Improving the quality of care that patients receive is a major focus of clinical research, particularly in the setting of cardiovascular hospitalization. Quality improvement studies seek to estimate and visualize the degree of variability in dichotomous treatment patterns and outcomes across different providers, whereby naive techniques either over-estimate or under-estimate the actual degree of variation. Various statistical methods have been proposed for similar applications including (1) the Gaussian hierarchical model, (2) the semi-parametric Bayesian hierarchical model with a Dirichlet process prior and (3) the non-parametric empirical Bayes approach of smoothing by roughening. Alternatively, we propose that a recently developed method for density estimation in the presence of measurement error, moment-adjusted imputation, can be adapted for this problem. The methods are compared by an extensive simulation study. In the present context, we find that the Bayesian methods are sensitive to the choice of prior and tuning parameters, whereas moment-adjusted imputation performs well with modest sample size requirements. The alternative approaches are applied to identify disparities in the receipt of early physician follow-up after myocardial infarction across 225 hospitals in the CRUSADE registry.
Highlights
Clinical studies frequently seek to improve the quality of care provided to patients by identifying discrepancies between providers
In accordance with the CRUSADE early follow-up example, our conclusions about variability are meant to apply to hospitals from a broader population, not limited to those participating in the registry
Thomas et al.[18] introduced moment-adjusted imputation (MAI) which replaces mis-measured data, W, with estimators that have asymptotically the same distribution as a latent variable of interest, X, up to some finite number of moments. They show that MAI is related to other measurement error methods but has superior performance in many settings
Summary
Clinical studies frequently seek to improve the quality of care provided to patients by identifying discrepancies between providers. Recent publications emphasize conclusions about the magnitude in variation across providers: ‘‘the degree of variability in clinical practice we observed represents a potential quality improvement opportunity’’;1 ‘‘in-hospital major bleeding rates varied widely across hospitals’’;2 and ‘‘ICU admission rates for heart failure (HF) varied markedly across hospitals’’.3 These data arise hierarchically, with a sample of providers, and within them, a sample of patients who experience a binary treatment or outcome. Hierarchical models and shrinkage estimators are advocated as a strategy to remove the excessive sampling error in estimates for small sites and are widely used in the medical literature.[6,7,8] the objective of making predictions for individual sites differs from the specific goal of estimating the magnitude of variation across sites For the latter objective, estimating potentially hundreds of nuisance parameters (the performance of each hospital) and pulling them back together to form a distribution results in more error than is necessary, in the form of either over-dispersion (exhibited in the raw proportions above) or shrinkage toward the mean. The methods are applied and interpreted in the CRUSADE example of early physician follow-up
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
