National Surveillance of Stroke Quality of Care and Outcomes by Applying Post-Stratification Survey Weights on the Get With The Guidelines-Stroke Patient Registry

Abstract

Background: The U.S. lacks a stroke surveillance system. This study develops a method to transform an existing registry into a nationally representative database to evaluate acute ischemic stroke care quality.Methods: Two statistical approaches were used to develop post-stratification weights for the Get With The Guidelines-Stroke registry by anchoring population estimates to the National Inpatient Sample. Post-stratification survey weights were estimated using a raking procedure and Bayesian interpolation methods. Weighting methods were adjusted to limit the dispersion of weights and make reasonable epidemiologic estimates of patient characteristics, quality of hospital care, and clinical outcomes. Standardized differences in national estimates were reported between the two post-stratification methods for anchored and non-anchored patient characteristics to evaluate estimation quality. Primary measures evaluated were patient and hospital characteristics, stroke severity, vital and laboratory measures, disposition, and clinical outcomes at discharge. Results: A total of 1,388,296 acute ischemic strokes occurred between 2012 and 2014. Raking and Bayesian estimates of clinical data not recorded in administrative databases were estimated within 5% to 10% of the margins of expected values. Median weights for the raking method were 1.386 and the weights at the 99th percentile were 6.881 with a maximum weight of 30.775. Median Bayesian weights were 1.329 and the 99th percentile weights were 11.201 with a maximum weight of 515.689. Conclusions: Leveraging existing databases with patient registries to develop post-stratification weights is a reliable approach to estimate acute ischemic stroke epidemiology and monitoring for stroke quality of care nationally. These methods may be applied to other diseases or settings to better monitor population health.

US Surveillance of Acute Ischemic Stroke Patient Characteristics, Care Quality, and Outcomes for 2019

“…Although the GWTG program contains many small, rural, and nonacademic hospitals, these hospital types are under-represented compared with the overall US hospitalized population. 8,11 Therefore, the sampling strategy does not directly estimate national AIS clinical characteristics as currently structured.…”

“…The derived 2019 NIS population counts were used to generate poststratification weights for 2019 GWTG-Stroke observations using Bayesian population interpolation method previously validated. 11 GWTG-Stroke observations (Bayesian prior) are fit to the marginal distributions of the 2019 NIS anchoring counts to estimate poststratification weights for each hospitalization. Weighted GWTG-Stroke data is used to estimate national AIS clinical characteristics, laboratory values, and quality metrics.…”

“…8,9 Nonrepresentative samples may be transformed into a representative ones using statistical methods such as poststratification weights that rebalance over and under-represented segments of the target population of interest. 10,11 A few community cohort and case-control studies are currently featured in the annual American Heart Association statistical update on heart disease and stroke statistics but are not nationally representative and inadequate to measure AIS burden and quality of care nationally. 12–14 For this study, we use Bayesian interpolation to estimate post-stratification survey weights for the GWTG-Stroke registry to quantify the 2019 AIS burden, hospital quality of care, and clinical outcomes.…”