Skip to main content
. 2018 Dec 14;8(12):e023853. doi: 10.1136/bmjopen-2018-023853

Table 2.

Predictors of long-term mortality—univariate regression analysis

Risk factor n HR 95% CI P value
Age Under 20 20
  (reference <20 years) 20–29 72 1.6 0.2 to 13.1 0.674
30–39 131 2.9 0.4 to 21.8 0.296
40–49 175 3.9 0.5 to 28.5 0.180
50–59 206 5.0 0.7 to 36.5 0.110
60–69 218 7.9 1.1 to 57.0 0.040
70–79 221 8.7 1.2 to 62.4 0.032
80+ 113 17.3 2.4 to 124.4 0.005
Gender Male 650
 (reference male) Female 506 1.2 1.0 to 1.5 0.049
Charlson score 0 833
 (reference 0) 1 168 2.6 2.0 to 3.4 <0.001
2 112 4.7 3.6 to 6.2 <0.001
3 31 3.8 2.4 to 6.1 <0.001
4 or more 9 5.3 2.4 to 12.0 <0.001
Number of comorbid conditions 0 833
 (reference 0) 1 262 3.2 2.6 to 4.0 <0.001
2 47 4.5 3.0 to 6.7 <0.001
3 11 3.3 1.5 to 7.5 0.004
Length of stay Less than 20 days 1079
 (reference <20 days) Longer than 20 days 77 0.4 0.2 to 0.8 0.006
Level of critical care ICU 251
 (reference ICU) HDU 905 1.2 1.0 to 1.4 0.019

The HR, 95% CI and p value of Wald’s test are presented for each variable found to significantly affect postdischarge survival on univariate regression analysis. P<0.05 was considered significant. The reference category for each variable is appended. Age has been transformed to a categorical variable for the purposes of the analysis.

ccAP, acute pancreatitis requiring critical care admission; HDU, high-dependency unit; ICU, intensive care unit; n, number of patients per category.