Use of machine learning to improve the measurement of health system performance

To improve the efficiency of health system performance measurement, we applied unsupervised learning methods to assess the performance of a currently used facility readiness index recommended by the World Health Organization, and compared it to empirically derived indexes using machine learning approaches. With colleagues Hannah Leslie, Xin Zhou, and Margaret Kruk of HSPH, we published the manuscript describing this work:

Liao X, Zhou X, Wang M, Hart JE, Laden F, Spiegelman D. Survival analysis with functions of mismeasured covariate histories: the case of chronic air pollution exposure in relation to mortality in the nurses' health study. J R Stat Soc Ser C Appl Stat. 2018 Feb;67(2):307-327. doi: 10.1111/rssc.12229. Epub 2017 Jun 15. PubMed PMID: 29430064; PubMed Central PMCID: PMC5801717.