Priyonto Saha, Master's candidate, Dalla Lana School of Public Health

Supervised by Zahra Shakeri, Dalla Lana School of Public Health, Aya Mitani, Dalla Lana School of Public Health, and Kuan Liu, Dalla Lana School of Public Health

Project Title: A causal inference approach to equitably characterize long COVID across geographic clusters

Project Summary:

Long COVID is a global health crisis inadequately addressed by current medical and public health measures. However, the causal risk factors of PCC for different individuals across Canada’s multicentre healthcare system are still unclear. As such, this project aims to quantify the impact of health disparities on PCC across Canada’s diverse subpopulations. To do so, we propose an innovative algorithm which combines causal inference, survival analysis, and clustered data analysis. Although these concepts and their pairwise combinations have been thoroughly researched and validated in isolation, we are amongst the first to explore all three together. To provide a brief technical outline, our proposed method extends inverse probability of treatment and censoring (IPTC) weighting, a method used in causal survival analysis, to two different clustering perspectives: central clustering and incidental clustering.

Our project consists of three primary components: the development and programming of our proposed algorithm, a simulation study of our causal survival analysis methods under varying multi-level environments and clustering perspectives, and a tutorial paper to assist with the knowledge translation of our new approach for non-technical audiences. All our R code will be shared on GitHub to ensure reproducibility and transparency, and we also hope to develop an R package for our algorithm if time permits.