Skip to main content

Table 4 Main findings in this study

From: Dealing with missing data in laboratory test results used as a baseline covariate: results of multi-hospital cohort studies utilizing a database system contributing to MID-NET® in Japan

#

Brief description

1

Point estimates for aHR can be similar for SRI and MI, but there may be differences in 95% CI when adjusted with PS methods (especially PS matching)

2

In the IPW method and imputation methods, aHR may be similar, but differences can occur.

3

In the CC method and the method excluding the laboratory test item, aHR may be similar to that of the MAR-based method, but differences can occur.

4

It can be pointed out that the difference in approaches considering hospital variations may not significantly affect the difference in aHR.

5

It can be pointed out that the difference in confounding adjustment methods can be different for aHR, especially between PS matching and other methods.

  1. Abbreviations: aHR Adjusted hazard ratio, CC Complete case, CI Confidence interval, IPW Inverse probability weighted, MAR Missing at random, MI Multiple imputation, PS Propensity score, SRI Single regression imputation