Table 2.
Main regression results
| Outcome variables | (1) | (2) |
|---|---|---|
| Ignoring time-dependent CDI exposure | Using time-to-exposure stratification | |
| Costs | 1.08*** [1.00;1.15] | 0.31*** [0.25;0.37] |
| Marginal effect | 23,313*** [20,772;25,854] | 9016*** [7152;10,880] |
| Reimbursement | 0.94*** [0.87;1.01] | 0.29*** [0.23;0.352 |
| Marginal effect | 18,678*** [16,580;20,776] | 7838*** [6072;9605] |
| Extra length of stay | 0.96*** [0.87;1.02] | 0.22*** [0.18;0.26] |
| Marginal effect | 21.6*** [19.6;23.6] | 6.4*** [5.1;7.8] |
| N | 2709 | 2500 |
This table summarizes a total of six GLM regression analyses, each using a gamma distribution and a log link function. Each cell gives the effect of CDI infection on the respective outcome on the left from different GLM regressions. All regressions use a within-main-diagnosis estimation and control for comorbidities according to Resch (2008), for age, age2, and for sex. Regressions in the second column additionally apply a time-to-exposure stratification. 95% confidence intervals are in brackets
* p < 0.1, ** p < 0.05, *** p < 0.01