The Phase2.1NeuroRPackage has been updated with the following changes:
- Stratify all analyses by adult and pediatric populations
- Modify survival analyses approach:
- Perform Cox-Proportional hazard models for time to mortality and time to discharge
- Remove severity survival analysis
- Incorporate censor parameter (30, 60, and 90 days)
- Implement models with different methods of handling comorbidities: (1) Elixhauser comorbidities as independent predictors, (2) Elixhauser summary score as a predictor, (3) the top 10 components from a Logistic Principal Component Analysis (LPCA)
- Kaplan-Meier curves are computed by calculating the average covariate adjusted survival time among all patients
- Include event/risk/censored tables
- Implement logistic regression based approach to evaluate association between neurological status and COVID-19 severity and evaluate with Pointwise Mutual Information (PMI)
- Computer a new table for demographic counts stratified by individual comorbidities
- Include the number of patients discharged in our demographics tables
- Include data dictionary of .rda output
- Only include patients who were admitted up to 11-01-2021 in order to preclude the inclusion of potential Omicron patients