Each week, we will be publishing layman’s abstracts of new articles from our prestigious portfolio of journals in statistics. The aim is to highlight the latest research to a broader audience in an accessible format.
The article featured today is from Statistics in Medicine, with the full article now available to read here.
Doubly robust estimation and causal inference for recurrent event data. Statistics in Medicine. 2020; 39: 2324– 2338. https://doi.org/10.1002/sim.8541
, , .Recurrent event data commonly arise in studies from industry, public health and medicine where each individual may experience multiple episodes of the same type of event during the observed follow-up time. In a recurrent injury study for artists, for example, it is of interest to understand whether acrobat and non-acrobat groups have the same rates of recurrent injuries over time if all artists were in the acrobat group and in the non-acrobat group, respectively. However, it is not valid to compare these two groups when artists are not randomly assigned to either the acrobat or non-acrobat group, but instead the assignment depends on the artist-specific covariates such as age and gender.
This article proposes a novel statistical method for reducing confounding effects induced by covariates when analyzing recurrent event data. Using our approach, researchers can achieve three goals: 1) explore risk factors for the recurrent event of interest, 2) properly investigate the impact of risk factors for the group assignment, 3) estimate and compare the cumulative number of events for each group such as treated and untreated group at a given time point. The new methodology is illustrated with an application to a database of circus artist injuries.