If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Invasive VRE cases have a long hospital stay and higher costs of antibiotic use.
•
Active screening in oncology wards is suggested.
Abstract
Objective
To analyze 48 cases the risk factors of vancomycin-resistant Enterococcus (VRE) infections, the antibiotic costs after infection, and the survival conditions.
Design
1:3 matched case-control study a medical center in the eastern Taiwan area. The case group, patients with VRE bacterial strains detected at the sterile sites, and the control group were randomly selected from invasive vancomycin-sensitive Enterococcus (VSE) infected patients at the nearest time point by taking the occurrence time of each VRE infection case as the reference time. Fisher exact tests were conducted in order to verify the existence of differences between the case and control groups; survival analysis was applied to explore the prognoses of the VRE infection cases.
Results
The mortality rate of the invasive VRE infection cases was 64.6%, which is obviously higher than that of the invasive VSE infection cases (39.4%); the fact of taking chemotherapy during a hospital stay as well as the use of third-generation cephalosporin, glycopeptides, and medicines of the metronidazole category before the infections, are the risk factors of future invasive VRE infections. Moreover, the antibiotic costs after the infections of invasive VRE infection cases are much higher than those of the VSE infection cases (the average daily cost is 3,433 new Taiwan dollars (NTD) vs. 1,742 NTD).
Conclusions
The history of receiving chemotherapy, the use of third-generation cephalosporin, glycopeptides, and medicines of the metronidazole category before the infections are the risk factors of VRE infections. The antibiotic costs after the infections of invasive VRE infection cases are much higher than those of the VSE infection cases.
and have become a major nosocomial pathogen in Taiwan. Taiwan Nosocomial Infections Surveillance (TNIS) shows that the percentage of nosocomial Infections accounted for by VRE in ICUs in medical centers increased from 3% in 2003 to 24% in 2011. The study found cases of VRE infection crude mortality rate of 64% in National Taiwan University Hospital from 1993 to 2000. It demonstrates that hospital-acquired infections caused by VRE increases not only mortality rates but also costs of treatment and care. Nevertheless, there is no study on increased costs and little research on risk factors of VRE infections with large sample size in Taiwan.
2. Methods
From a medical center in Hualien, researchers collected bacterial culture data from 2007 to 2012. After all subjects under 18 years of age were excluded, cases and control groups were selected according to set requirements under a 1:3 matched case-control study.
The variables data, as required by this study, were collected by reviewing the medical records and electronic medical records maintained by the hospital. First, Fisher exact tests were conducted in order to verify the existence of differences between the categorical variables of the risk factors of the VRE infection cases and those of VSE infection cases, as well as the existence of differences between the continuous variables of nonparametric analysis. Then, the prognoses of the VRE infection cases were explored through survival analysis. The statistical analysis all p-values < 0.05 was considered statistically significant.
In the case group, patients with VRE bacterial strains detected at the normally sterile sites were taken as the individual cases, and the bacterial culture time was mainly taken as the infection time. In the control group, the individual cases were collected by taking the occurrence time of each VRE infection case as the reference time in order to search for invasive VSE infected patients at the nearest time point as the individual cases of the control group. The selection was made randomly in the event that there were more than two individual cases suitable for the control group.
3. The collected data included
3.1 Basic variables
They include age, gender, place of residence, infection location (community, ward or intensive care unit (ICU)), medical care related infections or not, hospital discharge status (mortality or survival), and number of days of hospital stay.
3.2 Disease-related factors
The variables included the severity of illness (using APACHE 2 score to calculate the disease severity status), related risk factors (including whether the patient had an operation two weeks before the inspection; whether the patient had a malignant tumor, poor renal function, chronic liver function disease, diabetes, organ transplantation, hemodialysis, or HIV infection; whether the patient used steroid or cancer chemotherapy, was incontinent, had pressure sores or leucopenia during the current stay at hospital); whether the patient had septicemia; and the types of antibiotics used before the infection and the Carlson comorbidity index (CCI), and invasive treatment items (48 hours before specimen collection, or the time of collecting the specimens and the information, such as whether a breathing apparatus, a central venous catheter indwelling, urethral catheter indwelling, or tracheal tube indwelling was used at the time or 48 hours before the time).
The variables included antibiotic cost before VRE infection, VRE treatment, and when the antibiotic treatment for VRE started (VRE treatment refers to the use of quinupristin-dalfopristin, linezolid, daptomycin, and tigecycline;
and when the antibiotic treatment for VRE started is the timeframe between testing of VRE infection and the administration of antibiotic, which are within 24 hours after bacterial culture, between 24 to 48 hours after bacterial culture, and over 72 hours after bacterial culture, and no treatment).
3.4 Prognostic factors
The survival times of each individual case of the two groups were analyzed using the Kaplan-Meier test.
3.5 Definitions of the variables of each category
The invasive VRE infection cases were defined as the individual cases where the VRE bacterial strains were detected in blood, ascetic fluid, intra-abdominal abscess, pleural fluid, and joint fluid. Poor renal function was defined as creatinine clearance <60 mL/min; and poor liver function was defined when at least two of the following items were met: bilirubin concentration was >2.5 mg/dL; the aspartate aminotransferase (AST) or alanine aminotransferase (ALT) exceeded more than two times the normal value; and liver disease was diagnosed. Leucopenia was defined as white blood cell count being <500/mm3.
To determine VRE, BD Phoenix Automated Microbiology System was used to test Enterococcus strains for Vancomycin minimum inhibitory concentration (MIC). VRE was confirmed when MIC result was >32 ug/Ml.
The antibiotic treatment costs were calculated as follows: the types, dosages, frequencies and usage days of all antibiotics used; the types, dosages, and usage days of the antibiotics used before and after the infection, respectively, by taking the infection days as the tangent points; and then calculate the costs of antibiotics. The usage cost data of each antibiotic was collected by taking the unit price of each antibiotic, as stipulated by the National Health Insurance Administration in 2013,
National Health Insurance Administration Ministry of Health and Welfare: National health insurance drug price adjustment paid job.http://www.nhi.gov.tw.Accessed2013/06/01.
as the standard in order to add up the total usage costs of each antibiotic, in order to reduce the cost differences caused by currency inflation and medicine cost fluctuations.
Cases of healthcare-associated infections (HAIs) were determined based on the 2009 revision of HAI Surveillance Definition of Taiwan Center for Diseases Control (CDC).
4. Results
There were a total of 50 invasive VRE infection cases during 2007-2012. Excluding the two cases where the age was under 18 years old (0 and 2 years old, respectively), the study team collected 48 cases for study. The control group collection was made according to the designed study method of 1:3, with 144 cases selected as the control group. However, two cases were not incorporated, as the related data of the two patients’ medical records could not be retrieved in time. In the end, 48 invasive VRE infection cases and 142 invasive VSE infection cases were collected for analysis.
There were 31 males (accounting for 64.6% of the gender ratio) and 17 females (35.4%) in the invasive VRE infection cases; and 87 males (61.3%) and 55 females (38.7%) in the control group, which shows no statistically significant difference (p = 0.682) in gender between the two groups. Regarding the average age, that of the case group was 60.5 ± 18.2 (median: 59.5), and that of the control group was 65.7 ± 15 (median: 67), which shows no statistically significant difference (p = 0.078), although the age of the VSE group was older according to the age medians of the two groups.
In terms of infection occurrence location, the VRE cases mostly occurred in ICUs with 24 people (accounting for 50%); thus, there were 41 (28.9%) VSE cases occurring in the ICUs. Chi-square testing found that the invasive VRE infection cases had a higher probability of infection in ICUs than the VSE infection cases, which reached the statistically significant level (p = 0.027). Regarding hospital discharge, there were 31 (64.6%) mortality cases by the time of hospital discharge in the VRE infection cases, and the number of VSE infection cases was 56 (39.4%). Chi-square testing found that the mortality rate of the VER cases was higher than that of the VSE cases, and reached a statistically significant level (p = 0.003).
We found 52.1% (n = 25) of VRE infection cases used respirators when the infections occurred, which is significantly higher than the 30.3% (n = 43) of VSE infections cases (P = 0.0006). In addition, regarding the aspect of central catheter use, the VRE cases had a higher use rate of central venous catheters at 81.3% (n = 39), which is also higher than the 62.7% (n = 89) of the VSE infection cases (p = 0.018). With regard to having septic shock, the rate of the VRE infection cases was 43.8% (n = 21), which shows a significant difference (p = 0.028) from the 26.8% (n = 38) of VSE infection cases. Regarding the aspect of steroid use before the infection, the VRE infection cases had a higher use rate of 43.8% (n = 21), which is higher than the 24.6% (n = 35) of VSE infection cases, and reached a statistically significant level (p = 0.012). Regarding the CCI index and APACHE2 score, no significant difference was found between the data of the two groups after relevant testing (as shown in Table 1).
Table 1Descriptive statistical analysis of data of the invasive VRE case group and the invasive VSE control group from a medical center in Hualien during 2007 - 2012
VRE
VSE
P-value
(n = 48)
(n = 142)
Gender
Male
31(64.6)
87(61.3)
0.682
Female
17(35.4)
55(38.7)
Age
60.5 ± 18.2 (59.5)
65.7 ± 15.0 (67)
0.057
Infection location
0.027*
Community
5(10.4)
25(17.6)
Ordinary ward
19(39.6)
76(53.5)
ICU
24(50)
41(28.9)
Enterococcus species
< 0.0001*
E.faecium
46(95.8)
40(28.2)
E.faecalis
2(4.2)
102(71.8)
Infection part
0.026*
Blood
26(54.2)
102(71.8)
Ascitic fluid
18(37.5)
30(21.1)
Pleural fluid
3(6.5)
4(2.8)
Spinal fluid
1(2.1)
0(0)
Other
6(4.2)
Medical care related infections
23(47.9)
62(43.7)
0.608
Operation
11(22.9)
43(30.3)
0.328
Malignant tumor
16(33.3)
52(36.6)
0.681
Poor renal function
10(20.8)
35(24.6)
0.591
Liver disease
20(41.7)
49(34.5)
0.373
Diabetes
9(18.8)
42(29.6)
0.143
Mechanical ventilation
25(52.1)
43(30.3)
0.0006*
Central line use
39(81.3)
89(62.7)
0.018*
Urinary catheter
31(64.6)
88(62.0)
0.746
Incontinence
35(72.9)
89(62.7)
0.198
Pressure sore
9(18.8)
25(17.6)
0.858
Leucopenia
5(10.4)
6(4.2)
0.112
Septic shock
21(43.8)
38(26.8)
0.028*
Using steroids
21(43.8)
35(24.6)
0.012*
Hemodialysis
10(20.8)
23(16.2)
0.464
CCI
1.46 ± 1.53 (1.0)
1.71 ± 1.70 (1.0)
0.401
APACHE2
17.8 ± 6.8 (16.0)
17.0 ± 7.2 (16.0)
0.453
Total length of stay, median
55.4 ± 47.7 (46)
38 ± 34 (25)
0.009*
Hospital stay days after infection
32.5 ± 34.3 (21.5)
24.2 ± 25.9 (15.5)
0.082
Hospital discharge status
Mortality
31(64.6)
56(39.4)
0.003*
Survival
17(35.4)
86(60.6)
Note: * p < 0.05; the () in the category variable is the percentage; data presented as continuous variables are the average ± standard deviation (median). VRE: vancomycin-resistant Enterococcus; VSE: vancomycin sensitive Enterococcus; ICU: intensive care unit; CCI: Charlson comorbidity index; APACHE 2: acute physiological and chronic health evaluation 2.
With regard to antibiotic usage types before infection, it was found that the average types of antibiotics used before the infection by the VRE infection cases were 4.2 ± 2.7 types (the median was four kinds), while those by the VSE infection cases were 2.0 ± 2.1 types (the median was one kind), which shows the more the types of antibiotics are used, the bigger the chance of VRE infection (p > 0.0001).
In terms of antibiotic usage costs during a hospital stay, the average daily antibiotic usage cost of VRE infected patients during their stay in the hospital was 2,666 ± 2,077.9 new Taiwan dollars (NTD) (median: 2,183.3 NTD), and that of VSE infected patients was 1,504 ± 1,398 NTD (median: 1,216.3 NTD), showing that antibiotic costs during a hospital stay, in the event that the patient is infected with drug-resistant Enterococcus, are significantly higher than those of the non-drug-resistant Enterococcus cases (p < 0.0001). The same result was found from a comparison of antibiotic usage costs after an infection. The average daily antibiotics usage cost of the VRE cases after an infection was 3,433 ± 2,966 NTD (median: 2,984.6 NTD), while that of the VSE infection cases was 1,742 ± 1,611 NTD (median: 1425.2 NTD). p < 0.0001 was obtained upon testing the data of the two groups, and reacheed a statistically significant level (as shown in Table 2).
Table 2Descriptive statistical analysis of related antibiotic usage status and costs of the invasive VRE case group and the invasive VSE control group from a medical center in Hualien during 2007 – 2012
VRE
VSE
P-value
(n = 48)
(n = 142)
Types of antibiotics used before infection
4.2 ± 2.7 (4)
2.0 ± 2.1 (1)
< 0.0001*
Antibiotic types used after infection
4.3 ± 2.7 (4.5)
3.2 ± 2.3 (3)
0.006*
Average daily antibiotic cost during the hospital stay
2666.1 ± 2077.9 (2183.3)
1503.7 ± 1397.9 (1216.3)
< 0.0001*
Average daily antibiotic cost before infection
1589.9 ± 1458.5 (1136.6)
902.9 ± 1275.8 (472.4)
< 0.0001*
Average daily antibiotic cost after infection
3433.3 ± 2965.5 (2984.6)
1742.1 ± 1611.6 (1425.2)
< 0.0001*
Types of antibiotics used before infection
Aminoglycosides
22(45.8)
9(6.3)
0.611
Antifungals
21(43.8)
13(9.2)
0.058
1st-generation Cephalosporins
43(89.6)
11(7.7)
0.315
2nd-generation Cephalosporins
22(45.8)
5(3.5)
0.375
3rd-generation Cephalosporins
35(72.9)
20(14.1)
0.026*
4th-generation Cephalosporins
4(8.3)
4(2.8)
0.115
Glycopeptides
12(25)
17(12.0)
< 0.0001*
Glycylcyclines
5(10.4)
7(4.9)
0.013*
Macrolides
4(8.3)
2(1.4)
0.645
Oxazolidione
0(0)
2(1.4)
0.063
Penicillins
32(72.9)
25(17.6)
< 0.0001*
Quinolones
25(52.1)
21(14.8)
< 0.0001*
Sulfonamides
3(6.3)
4(2.8)
0.049*
Tetracyclines
0(0)
2(1.4)
0.063
Colistin
5(10.4)
4(2.8)
0.235
Metronidazole
12(25)
18(12.7)
< 0.0001*
Carbapenem
36(75)
28(19.7)
< 0.0001*
beta-lactamase Inhibitors
5(10.4)
10(7.0)
0.001*
Note:* p < 0.05; the () in the category variable is the percentage; data presented as continuous variables are the average ± standard deviation (median). VRE: vancomycin-resistant Enterococcus; VSE: vancomycin sensitive Enterococcus.
Survival was analyzed using a Kaplan-Meier curve on the invasive VRE infection cases and VSE infection cases. The average survival days of the VRE cases was 52 days, while that of VSE infection cases was 61 days, which was not significant (as shown in Fig. 1), and had an overall test result of p = 0.246.
Fig. 1Survival analysis of the invasive VRE infection case group and of the control group of invasive VSE infection cases from a medical center in Hualien during 2007 – 2012. Note: VRE: vancomycin-resistant Enterococcus.
Among the 190 patients in the collected invasive Enterococcus infection cases, 103 (54.2%) survived and 87 (45.8%) did not. Cox regression analysis was performed on the data of the two groups, and the p-value of the overall model was tested to be lower than 0.001, indicating that the overall Cox regression model was significant. The APACHE2 score was higher when the patient was found to have diabetes, leucopenia, or using the 4th-generation cephalosporin; glycopeptides and carbapenem before an infection constituted a risk factor affecting the invasive Enterococcus prognosis. However, the drug–resistant or non-drug–resistant nature of the bacterial strain infected does not constitute a risk factor affecting the prognosis. (p = 0.45) (Table 3).
Table 3Survival and mortality predictor data analysis of the invasive Enterococcus infection cases from a medical center in Hualien during 2007 - 2012
Survival
Mortality
HR
P-value
(n = 103)
(n = 87)
(95%CI)
Gender
1.0(0.56-1.82)
0.984
Male
42(40.8)
30(34.5)
Female
61(59.2)
57(65.5)
Age
63.5 ± 16.3 (66)
65.5 ± 15.6 (66)
1.0(0.98-1.03)
0.597
Infection location
0.236
ICU
18(17.5)
47(54)
Reference
Ordinary ward
67(65)
28(32.2)
0.4(0.17-1.18)
0.103
Community
18(17.5)
12(13.8)
0.64(0.27-1.48)
0.295
Bacterial strain
E.faecium
34(33)
52(59.8)
Reference
E.faecalis
69(67)
35(40.2)
0.8(0.36-1.66)
0.511
Medical care related infections
41(39.8)
44(50.6)
0.7(0.34-1.28)
0.218
Operation
30(29.1)
24(27.6)
1.5(0.74-3.04)
0.262
Malignant tumor
37(35.9)
31(35.6)
1.0(0.46-2.16)
1.000
Poor renal function
22(21.4)
23(26.4)
0.9(0.44-1.95)
0.835
Liver disease
35(34.0)
34(39.1)
1.6(0.76-3.30)
0.220
Diabetes
31(30.1)
20(23.0)
0.3(0.15-0.61)
0.001*
Mechanical ventilation
22(21.4)
46(52.9)
0.8(0.37-1.88)
0.664
Central line use
54(52.4)
74(85.1)
1.8(0.77-4.11)
0.179
Urinary catheter
56(54.4)
63(72.4)
0.6(0.22-1.58)
0.290
Chemotherapy
6(5.8)
6(6.9)
0.5(0.10-2.21)
0.335
Incontinence
53(51.5)
71(81.6)
1.9(0.72-5.06)
0.190
Pressure sore
15(14.6)
19(21.8)
1.1(0.54-2.33)
0.768
Leucopenia
5(4.9)
6(6.9)
3.6(1.92-6.70)
0.000*
Septic shock
11(10.7)
48(55.2)
3.0(0.76-11.9)
0.116
Using steroids
20(19.4)
36(41.4)
1.9(0.91-3.80)
0.091
Hemodialysis
10(9.7)
23(26.4)
1.5(0.69-3.32)
0.307
CCI
1.6 ± 1.7 (1)
1.7 ± 1.6 (1)
1.1(0.84-1.34)
0.603
APACHE2
14.9 ± 6.2 (15)
19.9 ± 7.1 (18)
1.1(1.01-1.10)
0.022*
VRE
17(16.5)
31(35.6)
0.8(0.38-1.53)
0.450
Types of antibiotics used before infection
Aminoglycosides
19(18.4)
12(13.8)
0.8(0.31-1.96)
0.603
Antifungals
11(10.7)
23(26.4)
0.9(0.37-2.04)
0.754
1st-generation Cephalosporins
36(35)
18(20.7)
0.7(0.33-1.34)
0.257
2nd-generation Cephalosporins
15(14.6)
12(13.8)
0.7(0.27-2.00)
0.547
3rd-generation Cephalosporins
24(23.3)
31(35.6)
0.9(0.42-1.83)
0.730
4th-generation Cephalosporins
2(1.9)
6(6.9)
0.2(0.04-0.83)
0.027*
Glycopeptides
11(10.7)
18(20.7)
0.4(0.14-0.92)
0.033*
Glycylcyclines
3(2.9)
9(10.3)
3.3(0.86-12.92)
0.082
Macrolides
2(1.9)
4(4.6)
3.6(0.89-14.64)
0.073
Oxazolidinone
0(0)
2(2.3)
0.8(0.05-12.00)
0.857
Penicillins
23(22.3)
34(39.1)
1.0(0.50-1.99)
0.985
Quinolones
17(16.5)
29(33.3)
0.9(0.45-1.71)
0.700
Sulfonamides
4(3.9)
3(3.4)
1.4(0.25-7.72)
0.705
Tetracyclines
1(1)
1(1.1)
1.5(0.14-16.92)
0.727
Colistin
3(2.9)
6(6.9)
0.6(0.14-2.14)
0.391
Metronidazole
10(9.7)
20(23)
1.8(0.78-4.02)
0.174
Carbapenem
28(27.2)
36(41.4)
0.3(0.13-0.61)
0.001*
betalactamine
6(5.8)
9(10.3)
1.0(0.31-2.96)
0.947
Note: * p < 0.05; the () in the category variable is the percentage; data presented as continuous variables are the average ± standard deviation (median). VRE: vancomycin-resistant Enterococcus; VSE: vancomycin sensitive Enterococcus; ICU: intensive care unit; CCI: Charlson comorbidity index; APACHE 2: acute physiological and chronic health evaluation 2. HR: hazard ratio.
The risk factors of VRE infections were explored by logistic regression analysis, where the variables that may affect the drug-resistant factors were placed into the model (gender, age, infection occurrence location, nosocomial infection or not, use of a central catheter or not during the infection, respirator usage, taking chemotherapy or not during the hospital stay, being diagnosed as having septic shock or not, previous use of steroids, usage of 3rd-generation cephalosporins, glycopeptides, glycylcyclines, penicillins, quinolones, metronidazole, and carbapenem before the infection). The R2 of the overall model was 0.283. The test result of overall fitness was p < 0.001, which reached the statistically significant level (as shown in Table 4).
Table 4Logistic regression model analysis of risk factors affecting drug resistance
VRE
VSE
Regression Test Analysis
(n = 48)
(n = 142)
OR(95%CI)
P-value
Gender
Male
31(64.6)
87(61.3)
Reference
Female
17(35.4)
55(38.7)
1.2(0.48-2.82)
0.734
Age
60.5 ± 18.2 (59.5)
65.7 ± 15.0 (67)
1.0(0.95-1.0)
0.195
Infection location
ICU
24(50)
41(28.9)
Reference
Ordinary ward
19(39.6)
76(53.5)
0.8(0.23-2.86)
0.744
Community
5(10.4)
25(17.6)
3.0(0.61-14.2)
0.177
Medical care related infections
23(47.9)
62(43.7)
0.6(0.22-1.50)
0.258
Mechanical ventilation
25(52.1)
43(30.3)
2.2(0.62-7.60)
0.229
Central line use
39(81.3)
89(62.7)
1.2(0.39-3.62)
0.769
Chemotherapy
6(12.5)
6(4.2)
6.3(1.10-37.0)
0.042*
Septic shock
21(43.8)
38(26.8)
1.3(0.49-3.50)
0.592
Using steroids
21(43.8)
35(24.6)
0.6(0.19-1.81)
0.359
CCI
1.46 ± 1.53 (1.0)
1.71 ± 1.70 (1.0)
0.8(0.60-1.10)
0.174
APACHE2
17.8 ± 6.8 (16.0)
17.0 ± 7.2 (16.0)
1.0(0.91-1.05)
0.503
Types of antibiotics used before infection
3rd-generation Cephalosporins
35(72.9)
20(14.1)
2.7(1.04-7.14)
0.040*
Glycopeptides
12(25)
17(12)
4.2(1.36-12.91)
0.013*
Glycylcyclines
5(10.4)
7(4.9)
0.6(0.12-2.80)
0.492
Penicillins
32(72.9)
25(17.6)
2.5(0.93-6.69)
0.069
Quinolones
25(52.1)
21(14.8)
2.0(0.77-5.44)
0.150
Metronidazole
12(25)
18(12.7)
4.0(1.32-11.96)
0.014*
Carbapenem
36(75)
28(19.7)
2.4(0.92-6.30)
0.075
Note: data presented as continuous variables are the average ± standard deviation (median); the () in the category variable is the percentage; *p < 0.05; CI = confidence interval; OR = odds ratio. VRE: vancomycin-resistant Enterococcus; VSE: vancomycin sensitive Enterococcus; ICU: intensive care unit; CCI: Charlson comorbidity index; APACHE 2: acute physiological and chronic health evaluation 2.
Our results show that the use of 3rd-generation cephalosporins, glycopeptides, and metronidazole-category medicines before the infection of a patient who had cancer chemotherapy during a hospital stay could effectively predict and explain the presence of VRE. The odds ratio of patients infected with the vancomycin-resistant enterococci who received chemotherapy was 6.3, p = 0.042, while that of patients who received third-generation cephalosporin-category medicines before infection was 2.7, p = 0.040; 4.2, p = 0.013 for those receiving glycopeptide-category medicines; and 4.0, p = 0.014 for those receiving metronidazole (as shown in Table 4).
The factors influencing survival of VRE infection cases were analyzed using Cox regression. After gender, age, CCI, and APACHE2 score were adjusted, the overall model was p = 0.001, showing a significant overall Cox regression model. It was found from the data that the risk of mortality in cases not receiving VRE antibiotic treatment was 5.2 times that in cases receiving treatment, which reached a statistically significant level (p < 0.0001). Moreover, each one-point increase in the APACHE2 score caused the risk of mortality to increase by 1.1 times, which reached a statistically significant level (p = 0.047) (as shown in Table 5).
Table 5Analysis of the factors of invasive VRE infection cases affecting survival using Cox regression
Survival
Mortality
Test Analysis
(n = 17)
(n = 31)
HR(95%CI)
P-value
Gender
1.7(0.76-3.94)
0.189
Male
11(64.7)
20(64.5)
Female
6(35.3)
11(35.5)
Age
56 ± 18.2 (52)
63 ± 18 (61)
1.0(0.99-1.03)
0.493
CCI
1.2 ± 1.5 (0)
1.6 ± 1.6 (1)
1.0 (0.78-1.31)
0.954
APACHE2
15.8 ± 6.6 (15)
18.9 ± 6.7 (18)
1.1 (1.00-1.13)
0.047*
VRE treatment
Treated
14(82.4)
22(71)
Reference
Not treated
3(17.6)
9(29)
5.2(2.08-12.90)
<0.0001*
Note:* p < 0.05; the () in the category variable is the percentage; data presented as continuous variables are the average ± standard deviation (median). VRE: vancomycin-resistant Enterococcus; VSE: vancomycin sensitive Enterococcus; CCI: charlson comorbidity index; APACHE 2: acute physiological and chronic health evaluation 2.
In addition, exploration of the influence of the early use of related antibiotic treatment on the survival of VRE infection cases found that, after Kaplan-Meier testing, there was no significant difference (p = 0.752) in the adoption of related early antibiotic treatment on survival days (Table 6).
Table 6Influence of the treatment time of VRE infection cases on survival
Invasive VRE infection does cause increased mortality, and hospitalization costs should not be underestimated. The risk factor of VRE infection is a history of receiving chemotherapy (p = 0.042) (Table 4) before an infection after controlling for age, gender, and disease severity. It is therefore suggested that active screening in the future should be started in oncology wards, as the risk of patient mortality could be effectively reduced if the VRE colonization could be confirmed early and related isolation and protection measures were immediately taken to reduce the invasive infection. Regarding the bacteremia data, the crude mortality rate was 73.1%, indicating that the mortality rate of bacteremia would be higher than other invasive infections, which is in accord with studies by the National Taiwan University Hospital in 2001
After age, gender, and disease severity were controlled, we found that the use of 3rd-generation cephalosporins, glycopeptides, and metronidazole-category medicines before the infection had a higher correlation with VRE infection (p < 0.05)(Table 4), which is similar to other studies (Dai et al., 2010).
Many people think the early use of the correct antibiotics for treatment can increase survival rate. The lack of an effect of an early antibiotic treatment could be related to the relatively low number of patients evaluated in the study. According to this study, most antibiotic treatments start after 72 hours, accounting for 64.3%, while those starting within 24 hours account for 14.3%. However, Kaplan-Meier analysis demonstrated that the adoption of early antibiotic treatment did not significantly affect patients prognosis (p = 0.752) (Table 6), which is consistent with the study results of Han et al. in 2009.
In addition to this, we also found that for survival analyzed on the invasive VRE infection cases and VSE infection cases, which was not significant (as shown in Fig. 1), the main reason may be the small number of study samples. Therefore, although the number of survival days of the VRE cases was significantly lower than that for the VSE infection cases, it did not reach the statistically significant level.
This study analyzed the risk factors of infection, prognosis status, and antibiotic usage costs by taking invasive VRE infection cases as the main object of study, and is the first in Taiwan to conduct related studies on the antibiotic costs of invasive VRE infections. In this study, the invasive VRE infections were mainly of E.faecium, accounting for 95.8% (Table 1), which is much higher than the invasive E.faecium infection rate (42.9%) found in the five year study by Chiayi Chang Gung Memorial Hospital,
Our study was limited by the fact that it was a retrospective study of medical records, thus some deviations in data might occur in the event that some situations were not noted on the medical records. Moreover, since the data under study belong to the same hospital, and the number of samples is small, the representativeness of the sample would be affected.
Currently, there is no standard treatment for VRE infection. A prospective randomized placebo-controlled trial should be designed to determine the optimal antibiotic choice to improve patient survival and outcome.
Acknowledgements
None reported.
Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Conflicts of interest: None.
Ethical Approval: The protocol has been approved by the Tzu Chi General Hospital Research Ethics Committee. The committee is organized under, and operates in accordance with, the Good Clinical Practice guidelines and governmental laws and regulations.
References
Tacconelli E.
Cataldo M.A.
Vancomycin-resistant enterococci (VRE): transmission and control.
International Journal of Antimicrobial Agents (ELSEVIER).2008; 31: 99-106
National Health Insurance Administration Ministry of Health and Welfare: National health insurance drug price adjustment paid job.http://www.nhi.gov.tw.Accessed2013/06/01.
31222-X&id=gr1.jpg){kind=link}