Cox multivariate regression analysis with backward directed stepwise procedure about the effect of new postprandial hyperglycaemia on the risk of late death after pneumonia
| Adjusted HR | 95% CI | p Value | |
|---|---|---|---|
| New postprandial hyperglycaemia | 2.56 | 1.04 to 6.32 | 0.041 |
| Karnofsky equal or less than 80% | 3.26 | 1.12 to 9.47 | 0.030 |
| Age, years | 1.97* | 1.14 to 3.39 | 0.015 |
The included confounders were age, Karnofsky score equal to or less than 80%, elevated plasma NT-proBNP and plasma urea concentration. Only factors with at least suggestive independent association (p<0.10) with the risk of late death are presented. N=131, only the patients without diabetes were included.*HR is expressed per quartile.