In the period from 2009–2015, 594 hospitalized patients developed enterococcal urinary tract infections. The average number of infected patients in a year was 84.9 patients, with 2012 showing the lowest number of infected patients (n = 65) and 2015 showing the highest number (n = 110). In addition, the number of infected patients exhibited a yearly increasing trend (Fig. 1). There were more female patients than male patients in our sample; male-to-female ratio was 0.7 (range, 0.67–0.84) (Fig. 2). The mean age of infected patients was 72.3 ± 16.1 years (Fig. 3).
There were more infected patients in the GW than in the ICU. In the GW, the number of patients with enterococcal infections was lowest in 2012 (51 patients) and highest in 2015 (68 patients). In the ICU, the number of patients with enterococcal infections was lowest in 2012 (14 patients) and highest in 2015 (42 years) (Table 1, Fig. 4).
Table 1
Number of enterococcal nosocomial infections staying in different ward and ratio in 2009–2015
Ward | bacterial \year | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | Total |
GW | VRE | 11 | 12 | 12 | 13 | 12 | 22 | 26 | 108 |
VSE | 51 | 51 | 54 | 38 | 44 | 35 | 42 | 315 |
VRE percentage | 17.7% | 19.0% | 18.2% | 25.5% | 21.4% | 38.6% | 38.2% | 25.5% |
ICU | VRE | 4 | 10 | 12 | 7 | 11 | 14 | 26 | 84 |
VSE | 17 | 14 | 12 | 7 | 11 | 10 | 16 | 87 |
VRE percentage | 19.0% | 41.7% | 50.0% | 50.0% | 50.0% | 58.3% | 61.9% | 49.1% |
GW: General ward; ICU: Intensive care units; VRE: Vancomycin-resistant Enterococci; VSE: Vancomycin-susceptible Enterococci |
VRE percentage = VRE/ (VRE + VSE) |
The percentage of VRE infections was higher in the ICU than in the GW every year. In the GW, VRE accounted for 17.7% of all enterococcal urinary tract infections in 2009, which was the lowest during the study period. This proportion showed a yearly increasing trend, reaching the highest level at 38.6% in 2014, followed by 38.2% in 2015. In the ICU, VRE accounted for 19.0% of all enterococcal urinary tract infections in 2009, which was the lowest during the study period. Subsequently, this proportion showed a yearly increasing trend, reaching the highest level at 61.9% in 2015. In 2011, the proportion of VRE infections out of all enterococcal urinary tract infections was 50%, meaning half of all ICU enterococcal urinary tract infections were caused by VRE (Table 1, Fig. 4).
Patients with VRE infections were older than those with VSE infections by a mean of 2.2 years (73.8 ± 15.6 vs. 71.6 ± 16.4, P = .127), and the mean ages of both groups were greater than 70. Female sex distribution was higher for both groups (111 vs. 81 and 238 vs. 164, respectively). The proportions of VRE and VSE infections among women were 57.8% and 59.2%; the difference was not statistically significant (P = .747) (Table 2).
Table 2
Risk factors of the vancomycin-resistant Enterococci urinary tract infections
Risk factors | Total | VRE | VSE | p value |
n(%) | 594 | 192(32.3) | 402(67.7) |
Patients' characteristic | | | | |
Age | 72.3(± 16.10)b | 73.8(± 15.58)b | 71.6(± 16.35)b | 0.127a |
Age group | | | | 0.35 |
21-61.9 | 148 | 39(20.3) | 109(27.1) | |
62-75.9 | 151 | 53(27.6) | 98(24.4) | |
76-84.6 | 148 | 50(26.0) | 98(24.4) | |
84.7-100.6 | 147 | 50(26.0) | 97(24.1) | |
Gender | | | | 0.747 |
Male | 245 | 81(42.2) | 164(40.8) | |
Female | 349 | 111(57.8) | 238(59.2) | |
Environment | | | | |
Ward | | | | < 0.001* |
ICU | 171 | 84(43.8) | 87(21.6) | |
GW | 423 | 108(56.3) | 315(78.4) | |
Enterococcal infection days | 32.3(± 38.1)b | 34.7(± 29.6)b | 31.2(± 42.1)b | < 0.001a* |
Enterococcal infection days(group) | | | | < 0.001* |
3–12 days | 156 | 32(16.7) | 124(30.8) | |
13–21 days | 141 | 48(25.0) | 93(23.1) | |
22–40 days | 154 | 51(26.6) | 103(25.6) | |
41–263 days | 143 | 61(31.8) | 82(20.4) | |
Medical history | | | | |
Malignant tumor | 166 | 45(23.4) | 121(30.1) | 0.091 |
Cardiovascular disease | 145 | 49(25.5) | 96(23.9) | 0.663 |
Chronic obstructive pulmonary disease | 42 | 11(5.7) | 31(7.7) | 0.378 |
Diabetes | 190 | 70(36.5) | 120(29.9) | 0.106 |
Cirrhosis | 20 | 8(4.2) | 12(3.0) | 0.455 |
Statin usage | 46 | 16(8.3) | 30(7.5) | 0.71 |
Long-term bedridden | 123 | 52(27.1) | 71(17.7) | 0.008* |
Invasive procedures | | | | |
Peripheral venous catheter | 430 | 132(68.8) | 298(74.1) | 0.17 |
Central venous catheter | 132 | 63(32.8) | 69(17.2) | < 0.001* |
Total parenteral nutrition | 13 | 8(4.2) | 5(1.2) | 0.033*c |
Arterial catheter | 141 | 65(33.9) | 76(18.9) | < 0.001* |
Arteriovenous fistula | 10 | 6(3.1) | 4(1.0) | 0.084c |
Double-lumen catheter | 59 | 34(17.7) | 25(6.2) | < 0.001* |
Urinary catheter | 453 | 157(81.8) | 296(73.6) | 0.029* |
Endotracheal tube | 119 | 64(33.3) | 55(13.7) | < 0.001* |
Tracheotomy tube | 155 | 60(31.3) | 95(23.6) | 0.048* |
Ventilator usage | 228 | 111(57.8) | 117(29.1) | < 0.001* |
Nasogastric tube insertion | 295 | 128(66.7) | 167(41.5) | < 0.001* |
GW: General ward; ICU: Intensive care units; VRE: Vancomycin-resistant Enterococci; |
VSE: Vancomycin-susceptible Enterococci |
a: Mann -Whitney U |
b: Standard Deviation, SD |
c: Fisher exact test |
*: p < 0.05 |
More patients were infected with VRE and VSE in the GW. The proportions of VRE infections in the ICU and GW were similar (43.8% and 56.3%, respectively) as opposed to the proportions of VSE infections (21.6% and 78.4%, respectively), indicting statistically significant differences (P < .001). Patients in the ICU had a higher risk of VRE urinary tract infection (odds ratio [OR] = 1.84, 95% confidence interval [CI]: 0.95–3.58, P = .071) than those in the GW (Table 2, Table 3).
Table 3
Logistic regression of vancomycin-resistant enterococci urinary tract infections
Risk factors | Univariate regression | Multivariate regression a |
| OR | p value | OR | 95% CI | p value |
Patients' characteristic | | | | | |
Age | 1.009 | 0.109 | | | |
Age group | | | | | |
21-61.9 | | | | | |
62-75.9 | 1.512 | 0.102 | | | |
76-84.6 | 1.426 | 0.164 | | | |
84.7-100.6 | 1.441 | 0.153 | | | |
Gender | | | | | |
Male | 1.059 | 0.747 | | | |
Female | | | | | |
Environment | | | | | |
Ward | | | | | |
ICU | 2.816 | < 0.001* | 1.842 | 0.949–3.575 | 0.071 |
GW | | | | | |
Enterococcal infection days | 1.002 | 0.309 | | | |
Enterococcal infection days(group) | | | |
3–12 days | | | | | |
13–21 days | 2 | 0.009* | 1.651 | 0.939–2.903 | 0.081 |
22–40 days | 1.919 | 0.013* | 1.573 | 0.887–2.790 | 0.121 |
41–263 days | 2.883 | < 0.001* | 2.716 | 1.494–4.939 | 0.001* |
Medical history | | | | | |
Malignant tumor | 0.711 | 0.091 | | | |
Cardiovascular disease | 1.092 | 0.663 | | | |
Chronic obstructive pulmonary disease | 0.727 | 0.38 | | | |
Diabetes | 1.348 | 0.107 | | | |
Cirrhosis | 1.413 | 0.457 | | | |
Statin usage | 0.127 | 0.711 | | | |
Long-term bedridden | 1.732 | 0.008* | 1.272 | 0.781–2.071 | 0.333 |
Invasive procedures | | | | | |
Peripheral venous catheter | 0.768 | 0.171 | | | |
Central venous catheter | 2.357 | < 0.001* | 1.051 | 0.635–1.740 | 0.847 |
Total parenteral nutrition | 3.452 | 0.032* | 2.795 | 0.776–10.060 | 0.116 |
Arterial catheter | 2.195 | < 0.001* | 0.702 | | 0.289 |
Arteriovenous fistula | 3.21 | 0.073 | | | |
Double-lumen catheter | 3.245 | < 0.001* | 1.926 | 1.035–3.586 | 0.039* |
Urinary catheter | 1.606 | 0.030* | 1.094 | 0.675–1.774 | 0.714 |
Endotracheal tube | 3.155 | < 0.001* | 1.281 | 0.556–2.952 | 0.561 |
Tracheotomy tube | 1.469 | 0.049* | 0.822 | 0.401–1.684 | 0.593 |
Ventilator usage | 3.338 | < 0.001* | 1.884 | 0.915–3.878 | 0.086 |
Nasogastric tube insertion | 2.814 | < 0.001* | 1.938 | 1.301–2.885 | 0.001* |
GW: General ward; ICU: Intensive care units; VRE: Vancomycin-resistant Enterococci; |
VSE: Vancomycin-susceptible Enterococci |
a : Hosmer and Lemeshow = 0.289 |
*: p < 0.05 |
The mean numbers of days taken to develop VRE and VSE infections after hospitalization were 34.7 ± 29.6 days and 31.2 ± 42.1 days, respectively (P < .001) (Table 2). The numbers of days taken to develop an enterococcal infection were divided into four groups (3–12 days, 13–21 days, 22–40 days, and > 40 days) for the chi-square test. The VRE and VSE infection groups showed differences in the number of days to enterococcal infection (P = .001). After controlling for other significant variables, length of hospitalization of more than 40 days led to a higher risk of VRE urinary tract infection than a length of 3–12 days (OR = 2.72, 95% CI: 1.49–4.94) (Table 3).
Medical history or conditions before infection were divided into seven variables (malignant tumour, cardiovascular disease, chronic obstructive pulmonary disease, diabetes, cirrhosis, statin usage, and long-term bedridden status) for analysis. Differences in the aforementioned diseases/conditions were analysed between VRE-infected and VSE-infected patients using a chi-square test; only the “long-term bedridden status” variable showed a statistically significant difference (P = .008). When multivariate logistic regression analysis was used for analysis, no significant difference was observed after controlling for other variables (Table 2).
To analyse the situation 48 hours before infection, we included the following invasive procedures in patients with enterococcal infections as relevant variables: peripheral venous catheter, central venous catheter, total parenteral nutrition, arterial catheter, arteriovenous fistula, double-lumen catheter, placement of urinary catheter, endotracheal tube, tracheotomy tube, ventilator usage, and nasogastric tube insertion. In the group with enterococcal infections, the proportion of patients with peripheral venous catheter placement and urinary catheter placement were higher. About 65% of infected patients had a peripheral venous catheter or urinary catheter. In VRE-infected group, 81.8% of the patients had a urinary catheter and the VSE-infected group also had a high proportion of urinary catheter placement (73.6%). With regard to nasogastric tube, the proportion of VRE infected patients with a nasogastric tube was higher than that of VRE infected patients without the tube (66.7% vs. 41.5%) (Table 2).
When the chi-square test was carried out for VRE and VSE infections, using the aforementioned 11 variables, significant differences were observed for central catheter, total parenteral nutrition, arterial catheter, double-lumen catheter, placement of urinary catheter, endotracheal tube, tracheotomy tube, usage of ventilator, and nasogastric tube placement (p < 0.05) (Table 2). After adjusting for other significant variables, significant differences were observed between the presence and absence of a double-lumen catheter or nasogastric tube placement (OR = 1.93, 95% CI: 1.04–3.59 vs. OR = 1.94, 95% CI: 1.30–2.89, respectively) (Table 3).