Term: High-Dimensional Propensity Score (HDPS)
Last Updated: 2015-05-12
A High-Dimensional Propensity Score (HDPS) is a probability score between zero and one typically obtained from a logistic regression model that reflects the likelihood of the outcome regressed, be it the likelihood of belonging to a group or having a disease. An HDPS is different from a propensity score in that hundreds of covariates are included in the regression model.
In Martens et al. (2015). covariates from six data dimensions were included in the HDPS regression model: diagnoses and physician tariff codes from medical claims data, diagnoses and interventions from hospital abstract data, Anatomical Therapeutic Chemical (ATC) codes from prescription drug data and survey questions from survey data.
- Martens P, Nickel N, Forget E, Lix L, Turner D, Prior H, Walld R, Soodeen RA, Rajotte L, Ekuma O. The Cost of Smoking: A Manitoba Study. Winnipeg, MB: Manitoba Centre for Health Policy, 2015. [Report] [Summary] [Updates and Errata] [Additional Materials] (View)
Manitoba Centre for Health Policy
Community Health Sciences, Max Rady College of Medicine,
Rady Faculty of Health Sciences,
Room 408-727 McDermot Ave.
University of Manitoba
Winnipeg, MB R3E 3P5 Canada