Determinants-viral-clearance

Statistical analysis of an individual patient data meta-analysis for temporal changes in viral clearance kinetics and optimal design of the PLATCOV trial. The PLATCOV trial is registered at clinicaltrials.gov number NCT05041907.

This work is licensed under a Creative Commons Attribution 4.0 International License.

Overview

This github repo provides the data and code for the statistical analysis of the article "Temporal changes in SARS-CoV-2 clearance kinetics and the optimal design of antiviral pharmacodynamic studies: an individual patient data meta-analysis of a randomised, controlled, adaptive platform study (PLATCOV)" published in The Lancet Infectious Diseases.

The analysis in this repo included:

1. The meta-analysis for individual patient data in unblinded patients of the PLATCOV trials between September, 2021 and October, 2023 to describe temporal changes in SARS-CoV-2 clearance kinetics. Temporal trends of baseline viral densities and viral clearance rates are described using a penalized B-spline fit.

   • The analysis is described in 01_temporal_splines_analysis.qmd and 01_run_local_temporal_splines_analysis.R.
   • Data used in this analysis is Unblinded_meta_analysis.csv.

2. The boostrapping analysis for the impacts follow-up duration on the z scores.

   • The analysis is described in 02_bootstrapping_analysis.qmd and 02_run_local_bootstraps_analysis.R.
   • Data used in this analysis are:
     • Ineffective_analysis.csv
     • Paxlovid_Molnupiravir_analysis.csv
     • Paxlovid_recent_analysis.csv
     • REGN_analysis.csv
     • Remdesivir_analysis.csv

3. The boostrapping analysis for the impacts sampling frequencies on the z scores.

   • The analysis is described in 03_bootstrapping_analysis_sampling_fq.qmd and 03_run_local_bootstraps_analysis_sampling_fq.R.
   • Data used in this analysis are:
     • Ineffective_analysis.csv
     • Paxlovid_Molnupiravir_analysis.csv
     • Paxlovid_recent_analysis.csv
     • REGN_analysis.csv
     • Remdesivir_analysis.csv

Stan models used in this analysis are in the Stan_models folder:

• Temporal_spline_additive.stan
• Temporal_spline_proportional.stan: This is the main model for this analysis.

Any questions or comments or if any bugs spotted drop me a message at [email protected] or [email protected].