Table 4

Possible risk factors for incident as compared with a denominator analysis during calendar weeks 11 and 12

ItemPatient groupOR (95% CI)
(crude)
OR (95% CI)
(adjusted)
DenominatorIncidents
Number of observations26 852197//
Age (mean±SD), years46.7±27.569.3±20.81.004 (1.001 to 1.006)§* 1.001 (0.996 to 1.005)§
Gender
Male
Female
47.0
53.0
37.6
62.4
1
1.048 (0.916 to 1.197)
1
1.055 (0.878 to 1.196)
Care-dependency, number (%)‡ †
None
Yes, by proxies
Yes, by community nurse
 Yes, by institution
16 335 (85.5%)
954 (5.0%)
723 (3.8%)
1099 (5.8%)
96 (51.6%)
12 (6.5%)
22 (11.8%)
56 (30.1%)
1
1.121 (0.789 to 1.594)
1.458 (1.025 to 2.073)*
1.802 (1.399 to 2.323)***
1
0.979 (0.674 to 1.423)
1.201 (0.821 to 1.758)
1.528 (1.141 to 2.046)*
Number of conditions (median, IQR)2 (0 to 4)5 (3 to 7)1.052 (1.029 to 1.075)§*** 1.030 (0.994 to 1.067)§
Number of chronic active treatments (median, IQR)1 (0 to 4)6 (3 to 9)1.052 (1.030 to 1.074)§*** 1.030 (0.995 to 1.067)§
Evans’ Index (median, IQR)3 (0 to 8)11 (6 to 17)1.009 (1.005 to 1.013)§*** n.a.¶
Thurgau Morbidity Index value (%)† 
0
1
2
3
4
5
6
missing values
8463 (31.5%)
3611 (13.4%)
4102 (15.3%)
3877 (14.4%)
2119 (7.9%)
1539 (5.7%)
709 (2.6%)
18
24 (12.2%)
8 (4.1%)
23 (11.7%)
39 (19.8%)
39 (19.8%)
38 (19.3%)
26 (13.2%)
0
1
0.989 (0.787 to 1.242)
1.049 (0.847 to 1.300)
1.131 (0.914 to 1.399)
1.292 (1.004 to 1.662)*
1.420 (1.078 to 1.868)**
1.680 (1.178 to 2.396)***
1
0.908 (0.694 to 1.190)
0.898 (0.685 to 1.169)
0.830 (0.611 to 1.127)
0.901 (0.643 to 1.265)
0.823 (0.547 to 1.239)
0.866 (0.523 to 1.436)
  • *p<0.05, **p<0.01, ***p<0.001—significance levels.

  • †Because GENLINMIXED procedure was not able to process ordinal scaled variables, correlations between study group and TMI or care-dependency were tested with Spearman's rho: the correlation coefficients were +0.075 or +0.094, respectively, p<0.001.

  • ‡Because of question ambiguity this analysis was restricted to adult patients (age >19 years); this led to 19 812 valid observations in the denominator and 183 in the incident groups.

  • §Per one conditions, medication, year or index point.

  • ¶Because Evans’ Index is a composite of condition and medication numbers, it was not included in the multiple regression analysis.

  • n.a, not available.