Confounding adjustment via a semi-automated high-dimensional propensity score algorithm: an application to electronic medical records

• Published in: Pharmacoepidemiology and drug safety Vol. 20; no. 8; pp. 849 - 857
• Main Authors: Toh, Sengwee, García Rodríguez, Luis A., Hernán, Miguel A.
• Format: Journal Article
• Language: English
• Published: Chichester, UK John Wiley & Sons, Ltd 01-08-2011; Wiley Subscription Services, Inc
• Subjects: Adult; Aged; Aged, 80 and over; Algorithms; Anti-Inflammatory Agents, Non-Steroidal - adverse effects; confounding; Confounding Factors (Epidemiology); Cyclooxygenase 2 Inhibitors - adverse effects; Databases, Factual - statistics & numerical data; Electronic Health Records - statistics & numerical data; Female; Gastrointestinal Hemorrhage - chemically induced; Gastrointestinal Hemorrhage - epidemiology; Gastrointestinal Hemorrhage - etiology; Humans; Male; Middle Aged; pharmacoepidemiology; Pharmacoepidemiology - methods; Propensity Score; propensity score analysis; Risk Factors; THIN; United Kingdom
• ISSN: 1053-8569; 1099-1557
• DOI: 10.1002/pds.2152

ABSTRACT
Purpose
A semi‐automated high‐dimensional propensity score (hd‐PS) algorithm has been proposed to adjust for confounding in claims databases. The feasibility of using this algorithm in other types of healthcare databases is unknown.
Methods
We estimated the comparative safety of traditional non‐steroidal anti‐inflammatory drugs (NSAIDs) and selective COX‐2 inhibitors regarding the risk of upper gastrointestinal bleeding (UGIB) in The Health Improvement Network, an electronic medical record (EMR) database in the UK. We compared the adjusted effect estimates when the confounders were identified using expert knowledge or the semi‐automated hd‐PS algorithm.
Results
Compared with the 411,616 traditional NSAID initiators, the crude odds ratio (OR) of UGIB was 1.50 (95%CI: 0.98, 2.28) for the 43,569 selective COX‐2 inhibitor initiators. The OR dropped to 0.81 (0.52, 1.27) upon adjustment for known risk factors for UGIB that are typically available in both claims and EMR databases. The OR remained similar when further adjusting for covariates—smoking, alcohol consumption, and body mass index—that are not typically recorded in claims databases (OR 0.81; 0.51, 1.26) or adding 500 empirically identified covariates using the hd‐PS algorithm (OR 0.78; 0.49, 1.22). Adjusting for age and sex plus 500 empirically identified covariates produced an OR of 0.87 (0.56, 1.34).
Conclusions
The hd‐PS algorithm can be implemented in pharmacoepidemiologic studies that use primary care EMR databases such as The Health Improvement Network. For the NSAID–UGIB association for which major confounders are well known, further adjustment for covariates selected by the algorithm had little impact on the effect estimate. Copyright © 2011 John Wiley & Sons, Ltd.