-
Introduction
-
Methods
- Study cohort
- Primary outcome and secondary outcomes
- Statistical methods
- The doubly robust estimation
- The gradient boosted model (GBM)
- inverse probabilities weighting (IPW) model
- Sensitivity analysis
- Covariates
- Comorbidities
- Vital signs
- Interventions
- Laboratory results
-
Results
- fig.1 流程图
- table 1 Comparison of the basic demographics cohort and the adjusted (weighted) cohort
- fig.2 The contributions of individual covariates to the final propensity score are illustrated in Fig. 2.
- Doubly robust analysis
- table 1 Comparison of the basic demographics cohort and the adjusted (weighted) cohort
- Primary outcome and sensitivity studies
- table 2 Primary outcome analysis with five different models
- (1) doubly robust model with unbalanced covariates
- (2) doubly robust model with all covariates,
- (3) propensityscore IPW model
- (4) propensity score matching model,
- (5)multivariate logistic regression model
- table 2 Primary outcome analysis with five different models
- Secondary outcomes studies with propensity score
- table 3 Secondary outcome analysis with propensity score matched cohorts matching
-
Discussion
-
Conclusions