mccount

Overview

mccount provides tools to estimate the mean cumulative count (MCC) of recurrent events in the presence of competing risks. Unlike traditional cumulative incidence methods that only consider the first occurrence of an event, MCC accounts for multiple occurrences of the same event type per individual, providing a more informative measure of the total burden of recurrent events in a population.

Available MCC methods in this package include the Dong-Yasui or the sum of cumulative incidences estimators, based on the work of Dong, et al.¹ Users can also estimate the MCC using the weighted Dong-Yasui estimator.²

Installation

You can install the development version of mccount like so:

# install.packages("pak")
pak::pkg_install("KennethATaylor/mccount")

Why Mean Cumulative Count?

In the context of clinical and epidemiological studies, subjects may experience multiple recurrent events over a given follow-up period. Traditional methods like Kaplan-Meier or cumulative incidence only account for the first occurrence of an event and ignore subsequent recurrent events, which can lead to underestimation of the true disease burden.

The MCC:

• Summarizes all events that occur in the population by a given time
• Accounts for competing risks that may terminate follow-up
• Provides a more complete picture of the recurrent event burden
• Is interpretable as the expected number of events per subject by a specific time point

Citation

If you use mccount in your research, please cite the package in addition to the relevant original methodology paper(s).

citation("mccount")

Footnotes

1. Dong H, Robison LL, Leisenring WM, Martin LJ, Armstrong GT, Yasui Y. Estimating the burden of recurrent events in the presence of competing risks: the method of mean cumulative count. Am J Epidemiol. 2015 Apr 1;181(7):532-40. doi: 10.1093/aje/kwu289 ↩

2. Gaber CE, Edwards JK, Lund JL, Peery AF, Richardson DB, Kinlaw AC. Inverse Probability Weighting to Estimate Exposure Effects on the Burden of Recurrent Outcomes in the Presence of Competing Events. Am J Epidemiol. 2023;192(5):830-839. doi: 10.1093/aje/kwad031 ↩