Table 2.
Univariate and multivariate analyses of toxigenic C. difficile carriage identified by PCR
| Characteristic | PCR positive Mean ± standard deviation |
PCR negative Mean ± standard deviation |
P value (univariate analysis) (*significant) |
Odds ratio (multivariate analysis) (95% confidence interval) |
|---|---|---|---|---|
| Age | 80.49 ± 8.93 | 78.54 ± 7.20 | 0.123 | |
| Day of stool sample (day after admission) |
2.58 ± 1.68 | 2.42 ± 1.64 | 0.568 | |
| Length of stay in days | 17.14 ± 8.25 | 16.29 ± 6.41 | 0.454 | |
| Number hospital stays past 6 months |
2.07 ± 1.22 | 1.32 ± 1.00 | <0.001* | 1.6 times per hospital stay (1.1–2.2) |
| Length hospital stays past 6 months |
39.88 ± 26.43 | 22.47 ± 26.39 | <0.001* | a |
| Mini Mental State Examination | 25.19 ± 4.99 | 25.10 ± 5.06 | 0.920 | |
| Barthel Index | 47.38 ± 23.48 | 50.86 ± 23.95 | 0.390 | |
| Timed Up & Go | 27.81 ± 11.57 | 23,38 ± 10.83 | 0.084 | |
| Charlson CM Index | 3.91 ± 2.71 | 3.12 ± 2.25 | 0.044* | b |
| Handgrip Strength | 36.95 ± 13.68 | 41.24 ± 16.62 | 0.123 | |
| BMI | 24.65 ± 6.20 | 26.22 ± 7.98 | 0.224 | |
| Characteristic | PCR positive Women/Men |
PCR negative Women/Men |
P value (univariate analysis) (*significant) |
Odds ratio (multivariate analysis) (95% confidence interval) |
| Sex | 65.1%/34.9% | 62.3%/37.7% | 0.726 | |
| Characteristic | PCR positive Yes/No |
PCR negative Yes/No |
P value (univariate analysis) (*significant) |
Odds ratio (multivariate analysis) (95% confidence interval) |
| Immunosuppression | 14.0%/86.0% | 15.6%/84.4% | 0.790 | |
| Residence care facility | 18.6%/81.4% | 6.6%/93.4% | 0.010* | b |
| Previous treatment with antibiotics past 6 months | 93.0%/7.0% | 60.8%/39.2% | <0.001* | 3.7 times (1.0–13.9) |
| Previous episodes of CDI | 11.6%/88.4% | 0.9%/99.1% | <0.001* | 12.3 times (1.9–81.6) |
| PPI premedication | 76.7%/23.3% | 62.3%/37.7% | 0.071 | |
| “Post–Sugery” hospital stay | 53.5%/46.5% | 21.2%/78.8% | <0.001* | 3.5 times (1.4–8.8) |
alength of hospital stay correlated by R = 0.58 to the variable “number of hospital stays” and was therefore excluded from the final multivariate model
bnot statistically significant in the multivariate model