- Email: [email protected]
Case-control analysis of endemic Acinetobacter baumannii bacteremia in the neonatal intensive care unit
American Journal of Infection Control 42 (2014) 23-7
Contents lists available at ScienceDirect
American Journal of Infection Control
American Journal of Infection Control
journal homepage: www.ajicjournal.org
Major article
Case-control analysis of endemic Acinetobacter baumannii bacteremia in the neonatal intensive care unit Jen-Fu Hsu MD a, b, Shih-Ming Chu MD a, b, Reyin Lien MD a, b, Cheng-Hsun Chiu MD, PhD b, c, Ming-Chou Chiang MD a, b, Ren-Huei Fu MD, PhD a, b, Chiang-Wen Lee MD, PhD d, Hsuan-Rong Huang MD a, b, Ming-Horng Tsai MD b, d, e, * a
Division of Pediatric Neonatology, Department of Pediatrics, Chang Gung Memorial Hospital, Taoyuan, Taiwan College of Medicine, Chang Gung University, Taoyuan, Taiwan Division of Pediatric Infectious Disease, Department of Pediatrics, Chang Gung Memorial Hospital, Taoyuan, Taiwan d Chang Gung University of Science and Technology, Chiayi, Taiwan e Division of Neonatology and Pediatric Hematology/Oncology, Department of Pediatrics, Chang Gung Memorial Hospital, Yunlin, Taiwan b c
Key Words: Late-onset sepsis Gram-negative bacteremia Multidrug resistance Neonate Recurrence
Background: We aimed to characterize the clinical manifestations and outcomes of patients with Acinetobacter baumannii bacteremia in the neonatal intensive care unit (NICU). Methods: All patients with A baumannii bacteremia in our NICU from 2004 to 2010 were reviewed. A matched case-control study was performed by comparing each case of A baumannii to 2 uninfected controls and all cases of Escherichia coli and Klebsiella bacteremia, respectively. Results: Thirty-seven cases with A baumannii bacteremia were identified. Multidrug-resistant isolate was noted in only 2 cases (5.4%), and the overall mortality rate was 8.1%. Compared with matched, uninfected controls, infants with A baumannii were more likely to have had a central vascular catheter (CVC) (P ¼ .009), use of total parenteral nutrition (TPN) (P ¼ .021), longer duration of ventilator use (P ¼ .002), and hospitalization (P ¼ .010). Compared with E coli or Klebsiella bacteremia, infants with A baumannii bacteremia had lower birth weight (median of 1,090 g vs 1,300 g, P ¼ .044) and a higher rate of CVC and TPN use (both P < .001) at the time of infection. Conclusion: A baumannii bacteremia occurs endemically or sporadically in the NICU, primarily in lowbirth-weight infants on TPN use and with CVC in situ. Although A baumannii does not often cause mortality, and multidrug-resistant A baumannii is uncommon, it contributes significantly to longer hospitalization. Crown Copyright Ó 2014 Published by Elsevier Inc. on behalf of the Association for Professionals in Infection Control and Epidemiology, Inc.
Acinetobacter baumannii is a nonfermentative, gram-negative coccobacillus, which is increasingly found in the intensive care unit.1,2 This microorganism has been one of the most important pathogens causing hospital-acquired infections for decades.3,4 It contributes to 8% to 25% of all gram-negative bacillary (GNB) lateonset sepsis (LOS) in the neonatal intensive care unit (NICU).5-7 In addition, A baumannii accounts for 4.6% of ventilator-associated pneumonia, 5.7% of catheter-related bloodstream infections (BSIs), and 4.3% of all BSIs.8-10 A baumannii can possibly become highly resistant to all antibiotics,11 which makes it deserve special attention and emphasis of infection control. Several recent reports
* Address correspondence to Ming-Horng Tsai, MD, Division of Neonatology and Pediatric Hematology/Oncology, Department of Pediatrics, Chang Gung Memorial Hospital, No. 707 Gongye Road, Sansheng, Mailiao Township, Yunlin, Taiwan, R.O.C. E-mail address: [email protected] (M.-H. Tsai). Conflicts of interest: None to report.
have described the emergence or outbreak of multidrug-resistant (MDR) A baumannii in the NICU that is associated with higher morbidity or mortality.3,12-14 Recent studies focused on clinical features and successful control of MDR A baumannii outbreak11-15; however, non-MDR A baumannii is rarely investigated. In adult or surgical intensive care unit, an A baumannii outbreak has traditionally been linked to environmental sources, such as contaminated medical devices or prolonged mechanical ventilation,16,17 but little relevant information was available in the NICU.3 The purpose of this case-control analysis was to identify and compare cases of endemic A baumannii bacteremia in a NICU over a 7-year period with matched uninfected and other GNB-infected controls to identify risk factors, presenting signs and symptoms, and outcomes specific to A baumannii that may facilitate the diagnosis and development of interventions for the control of A baumannii.
0196-6553/$36.00 - Crown Copyright Ó 2014 Published by Elsevier Inc. on behalf of the Association for Professionals in Infection Control and Epidemiology, Inc. http://dx.doi.org/10.1016/j.ajic.2013.06.017
24
J.-F. Hsu et al. / American Journal of Infection Control 42 (2014) 23-7
PATIENTS AND METHODS Hospital characteristics and patients Between January 2004 and December 2010, all neonates with at least 1 episode of A baumannii bacteremia during their hospitalization in the NICU of Chang Gung Memorial Hospital (CGMH) were identified from the electronic database. The NICU of CGMH is a tertiary-level medical center in northern Taiwan and contains a total of 49 beds equipped with mechanical ventilator and 28 beds with special care nurseries. This study was approved by the Institutional Review Board of CGMH.
not normally distributed. Student t test and the Mann-Whitney test were used to compare continuous variables and the c2 test or Fisher exact test was used to compare categorical variables. All P values were 2 tailed, and P values < .05 were considered to be statistically significant. Dichotomous variables were analyzed using odds ratio and bivariate logistic regression analysis. Despite the relatively small sample size and risk of model breakdown, independent covariates with a P value of < .10 as identified by bivariate analysis were subjected to multivariate logistic regression analysis. Statistical analyses were performed using SPSS version 15.0 (SPSS Inc, Chicago, IL).
Control case selection and matching criteria
RESULTS
To identify potential risk factors associated with acquisition of A baumannii bacteremia, 2 uninfected controls were selected and matched to each case of A baumannii bacteremia using birth weight (100 g), gender, and gestational age (1 week). Next, all cases of E coli and Klebsiella spp bacteremia during the same perioddafter excluding extended-spectrum b-lactamase-producing straind were enrolled as the infected controls.
Patient characteristics
Date collection and analysis Clinical information and microbiologic data were collected and included demographic characteristics, presence of central venous catheters (CVCs), invasive diagnostic and therapeutic procedures, total parenteral nutrition (TPN), mechanical ventilation, previous antibiotic exposure, and current usage. Definitions LOS was defined as positive microbial growth in 1 culture of blood samples obtained after 72 hours of life with accompanying clinical signs of sepsis. Recurrent bacteremia was defined as any new episode of documented BSI by the same or different pathogens occurring at least 2 weeks after the initial episode, a period during which blood cultures were negative. Prolonged intubation was defined as the need for mechanical ventilation for more than 14 days prior to infection. Prolonged hospitalization was defined as an infant who stayed in the NICU after the postconceptional age of 42 weeks. Cholestasis was defined according to the definition of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition.18,19 Death was considered attributable to sepsis if it occurred within 7 days of the positive blood culture or if clinical signs and symptoms of sepsis were documented in the medical record as the direct cause of death. All comorbidities of prematurity, including respiratory distress syndrome, intraventricular hemorrhage, bronchopulmonary dysplasia, and necrotizing enterocolitis were based on the latest updated diagnostic criteria in the standard textbook of neonatology.20 In our NICU, all blood cultures were assessed using the Bactec 9240 system (Becton Dickinson Diagnostic Instrument Systems, Sparks, MD), and antibiotic susceptibility patterns were tested using disk diffusion methods on Mueller-Hinton agar plates, using calibrated inoculum of the isolates based on the McFarland standard.21,22 MDR A baumannii was defined when the isolate was resistant to 3 of the following 4 antibiotics: ceftazidime, ciprofloxacin, gentamicin, and imipenem. Statistical analysis Continuous variables were summarized using means and standard deviations if normally distributed and median and ranges if
Thirty-seven cases of A baumannii bacteremia were identified in neonates hospitalized in the NICU of CGMH between January 2004 and December 2010. Infants with A baumannii bacteremia had median gestational age and birth weight of 28 weeks and 1,090 g, respectively. During the same period, a total of 137 neonates who experienced Escherichia coli or Klebsiella bacteremia, after excluding extended-spectrum b-lactamase-producing bacteria, was identified as the control group. The uninfected controls consisted of 74 neonates. The demographic characteristics of cases and selected controls are summarized in Table 1. Organism characteristics and antibiotic susceptibility patterns of A baumannii Of all episodes, 19 cases occurred endemically into 4 major outbreaks (defined as significantly more episodes within a short period when compared with a 6-month interval by c2 test), and others occurred sporadically over this 7-year period (Fig 1). Only 2 isolates were MDR strain, and none of the others were resistant to gentamicin or third-generation cephalosporins. Most cases (35/37, 94.6%) were on TPN and/or intralipids nutritional support and with a CVC in situ, and 81.1% (30/37) were under ventilator support at the time of infection. Comparison of cases of A baumannii and uninfected controls The overall mortality was significantly higher among infected patients (8.1% vs 0%, respectively, P ¼ .014, Table 1). Patients with A baumannii were more likely to have longer duration of ventilator use and hospitalization (P ¼ .002 and P ¼ .010, respectively). None of perinatal risk factors, bronchopulmonary dysplasia, intraventricular hemorrhage, or necrotizing enterocolitis were found to be associated with acquisition of A baumannii bacteremia. Although neonates with A baumannii were comparable with the uninfected controls with regard to the day of first attempt at feeding, they used significantly longer duration of TPN and/or intralipids and thus much later reached full feeding amount (both P < .001). The bivariate logistic regression analysis (Table 2) revealed that patients with A baumannii were more likely to have an indwelling CVC (OR, 11.22; 95% CI: 1.43-88.21, P ¼ .021), use of TPN and/or intralipids (OR, 9.27; 95% CI: 1.17-73.40, P ¼ .035), use of high frequency oscillatory ventilator after birth (OR, 3.11; 95% CI: 1.148.44, P ¼ .026), prolonged intubation (OR, 2.59; 95% CI: 1.12-5.97, P ¼ .025), and cholestasis (OR, 6.35; 95% CI: 2.47-16.30, P < .001). Multivariate logistic regression confirmed cholestasis as the only significant independent risk factor for A baumannii bacteremia (OR, 4.50; 95% CI: 1.61-12.54, P ¼ .004).
J.-F. Hsu et al. / American Journal of Infection Control 42 (2014) 23-7
25
Table 1 Characteristics of cases of A baumannii sepsis versus uninfected, sepsis with other gram-negative bacilli controls
Characteristics
A baumannii (n ¼ 37)
Uninfected controls (n ¼ 74)
E coli or Klebsiella spp (n ¼ 137)*
P valuey
1,090 (420-2,618) 29.4 3.8 17/20 8 (21.6) 23.0 (13.3-40.3) 0 (0) 4 (10.8) 19 (51.4) 0 (0) 0 (0) 18 (48.6) 14 (37.8) 3.0 (2.0-7.0) 37.0 (20.0-72.0) 35 (94.6) 32 (12.0-57.0) 50 (3.0-75.3) 35.5 (14.5-66.5) 10 (27.0) 15 (40.5) 85.5 (53.8-111.0) 3 (8.1)
1,100 (472-2,600) 29.6 3.4 38/36 14 (18.9) 0 (0) 1 (1.4) 29 (39.1) 1 (1.4) 1 (1.4) 29 (39.2) 11 (14.9) 3.0 (2.0-6.0) 20.0 (13.8-30.0) 55 (74.3) 11.5 (5.0-20.5) 19.0 (2.0-50.3) 17.7 (9.0-25.7) 9 (12.1) 60.5 (40.3-87.5) 0 (0)
1,300 (540-3,870) 30.6 4.3 76/61 57 (41.6) 25.0 (16.0-46.0) 2 (1.5) 19 (13.9) 60 (43.8) 3 (2.2) 7 (5.1) 61 (44.5) 42 (30.7) 6.0 (3.0-12.3) 39.0 (24.0-71.5) 85 (62.1) 21.0 (10.5-45.0) 30.0 (2.8-68.3) 22.0 (14.5-48.5) 21 (15.3) 32 (23.4) 69.0 (46.0-97.0) 11 (8.0)
.98, .044 .71, .13 .54, .26 .80, .034 n/a, .50 n/a, .45 .026, .59 .15, .44 .48, .36 .48, .15 .22, .85 <.001, .68 .45, .087 .001, .84 .009, <.001 <.001, .026 .002, .088 <.001, .096 n/a, .077 <.001, .032 .010, .16 .014, .97
Birth body weight (g), median (range) Gestational age (wks) Sex (m/f) Vaginal delivery, n (%) Day of life at positive culture, median (IQR) Perinatal asphyxia and/or hypoxic ischemic encephalopathy, n (%) Surgical history, n (%)z 5-Minute Apgar score 7, n (%) Necrotizing enterocolitis (stage II), n (%) Intraventricular hemorrhage (grade III), n (%) Bronchopulmonary dysplasia, n (%) Cholestasis, n (%) Days of trying oral feeding, median (IQR) Days of reaching full feeding, median (IQR) Use of TPN and/or intralipids, n (%) Duration of TPN and/or intrafat, days, median (IQR) Duration of ventilator use, days, median (IQR) Duration of central venous catheter, days, median (IQR) Recurrent sepsis after bacteremia, n (%)z Prolonged hospitalization, n (%) Days of hospitalization, median (IQR) Final mortality, n (%)
IQR, interquartile range; n/a, not available; TPN, total parenteral nutrition. *Including all cases of late-onset sepsis caused by E coli and Klebsiella spp, excluding extended-spectrum b-lactamase-producing bacteria during the study period. y P values are expressed as comparisons between A baumannii and uninfected controls, A baumannii and E coli/Klebsiella spp. z Defined as blood culture-proven bloodstream infection after the cases of A baumannii or E coli and Klebsiella spp bacteremia.
mortality, except cases with A baumannii were more likely to have thrombocytopenia (defined as platelet count < 50,000/mL) (45.9% vs 27.7%, respectively, P ¼ .040). Despite relatively a small number of cases and controls for analysis, a multivariable logistic regression was performed and the TPN use determined to be a significant independent covariate for A baumannii bacteremia as compared with E coli and Klebsiella (adjusted OR, 6.59; 95% CI: 1.44-30.13, P ¼ .015). For final outcome analysis, although the overall mortality rates were comparable between A baumannii bacteremia and E coli and Klebsiella bacteremia, patients with A baumannii bacteremia had relatively high rate of recurrent LOS and significantly higher rate of prolonged hospitalization (OR, 2.25; 95% CI: 1.04-4.88, P ¼ .039). DISCUSSION
Fig 1. Distribution of 37 episodes of Acinetobacter baumannii bacteremia and 4 major outbreaks (marked by star at the top) in the neonatal intensive care unit during 2004 to 2010.
Comparison of cases of A baumannii and E coli and Klebsiella controls Infants with A baumannii bacteremia had lower birth weight (median of 1,090 g vs 1,300 g, respectively, P ¼ .044) than those with E coli and Klebsiella LOS. Although their gestational ages were comparable, A baumannii bacteremia occurred more frequently in infants with gestational age of less than 29 weeks (OR, 2.27; 95% CI: 1.08-4.78, P ¼ .032). Infants with A baumannii bacteremia were more likely to have a CVC in place (OR, 4.35; 95% CI: 1.23-10.43, P ¼ .021) and be receiving TPN and/or intralipids (OR, 7.29; 95% CI: 1.66-31.95, P ¼ .008) at the time of infection. There were no differences in personal risk factors, presenting symptoms and signs, or
The main objective of this study was to investigate the risk factors and impact of A baumannii bacteremia in the NICU. We found that A baumannii accounted for 4.7% of all episodes of LOS and 14.0% of all GNB-related LOS in our NICU. From our data, although A baumannii had a tendency toward clustered episodes of bacteremia, it could also occur sporadically. Very low-birth-weight neonates were at higher risk of A baumannii bacteremia, especially when they were on TPN use and with CVC in situ. Because most patients had mechanical ventilator at the time of infection, the presumptive sources of infection, as previous studies identified, may include the medical devices and supportive apparatuses.16,17 Although A baumannii bacteremia in our NICU was mostly antibiotic susceptible and was not associated with higher mortality, it deserves investigation because A baumannii bacteremia was significantly associated with longer hospitalization, which potentially increase patient risk to further adverse events, costs, and family bonding. The prolonged hospital stay may be associated with the relatively higher rate of recurrent LOS after A baumannii bacteremia and the existences of more TPN and CVC use in neonates with A baumannii bacteremia. Although cholestasis was confirmed by multivariate logistic regression analysis to be the
26
J.-F. Hsu et al. / American Journal of Infection Control 42 (2014) 23-7
Table 2 Bivariable logistic regression analysis of risk factors and outcomes associated with cases of A baumannii sepsis versus uninfected controls Characteristic
A baumannii (n ¼ 37)
Host and perinatal factors Male sex Vaginal delivery Prolonged rupture of membranes Maternal fever Chorioamnionitis Five-minute Apgar score 7 Personal factors Bronchopulmonary dysplasia Surgical history Cholestasis Procedures and conditions Prolonged intubation Use of HFOV after birth Central venous catheter Use of TPN and/or intralipids Clinical outcome Prolonged hospitalization Death
17 8 5 2 1 19
Uninfected controls (n ¼ 74)
(45.9) (21.6) (13.5) (5.4) (2.7) (51.4)
38 14 11 2 0 29
OR (95% CI)
(51.3) (18.9) (14.9) (2.7) (0) (39.1)
(0.37-1.69) (0.43-3.05) (0.29-2.81) (0.28-15.28) 1.84 (0.81-4.20)
.50 .79 .85 .48 .14
18 (48.6) 4 (10.8) 14 (37.8)
29 (39.2) 1 (1.4) 11 (14.9)
1.68 (0.74-3.80) 8.63 (0.93-80.30) 6.35 (2.47-16.30)
.21 .058 <.001
19 11 36 35
22 9 53 55
(51.4) (29.7) (97.3) (94.6)
15 (40.5) 3 (8.1)
0.76 1.14 0.89 2.06
P value
(29.7) (12.2) (71.6) (74.3)
2.59 3.11 11.22 9.27
9 (12.1) 0 (0)
(1.12-5.97) (1.14-8.44) (1.43-88.21) (1.17-73.40)
.025 .026 .021 .035
6.43 (2.31-17.90)
<.001
CI, confidence interval; HFOV, high-frequency oscillatory ventilator; OR, odds ratio; TPN, total parenteral nutrition.
Table 3 Bivariable logistic regression analysis of risk factors and outcomes associated with cases of A baumannii sepsis versus other gram-negative, late-onset sepsis controls Characteristic Host and perinatal factors Extremely low birth body weight (BBW < 1,000 g) Very preterm (GA < 29 weeks) Male sex Delivery by cesarean section Prolonged rupture of membranes Maternal fever Chorioamnionitis Five-minute Apgar score 7 Personal factors Bronchopulmonary dysplasia Use of HFOV after birth Surgical history Cholestasis Procedures and conditions Intubated Prolonged intubation Central venous catheter TPN and/or intralipids Presenting signs and symptoms Fever Apnea and/or bradycardia and cyanosis Abdominal distension and/or feeding intolerance Hypoglycemia or hyperglycemia Tachycardia Hypotension Coagulopathy Disseminated intravascular coagulopathy Thrombocytopenia (platelet count < 50,000/mL) Clinical outcome Recurrent late-onset sepsis Prolonged hospitalization Death
A baumannii (n ¼ 37), n (%)
E coli or Klebsiella spp (n ¼ 137), n (%)
OR (95% CI)
P value*
15 20 17 29 5 2 1 19
(40.5) (54.1) (45.9) (78.4) (13.5) (5.4) (2.7) (51.4)
39 48 76 80 24 4 0 60
(28.5) (35.0) (55.5) (58.4) (17.5) (2.9) (0) (43.8)
1.76 2.27 0.62 2.56 0.71 1.85
(0.82-3.76) (1.08-4.78) (0.29-1.30) (1.09-6.03) (0.25-2.01) (0.32-10.54) 1.42 (0.67-3.03)
.15 .032 .21 .032 .52 .49 .36
18 11 4 14
(48.6) (29.7) (10.8) (37.8)
61 37 19 42
(44.5) (27.0) (13.9) (30.7)
1.13 1.20 0.73 1.54
(0.53-2.39) (0.54-2.66) (0.23-2.30) (0.71-3.35)
.75 .66 .73 .27
19 19 36 35
(51.4) (51.4) (97.3) (94.6)
55 57 99 85
(40.1) (41.6) (70.8) (62.1)
1.49 1.35 4.35 7.29
(0.70-3.17) (0.64-2.88) (1.23-10.43) (1.66-31.95)
.30 .43 .02 .008
13 28 15 18 6 12 11 5 17
(35.1) (75.7) (40.5) (48.6) (16.2) (32.4) (29.7) (13.5) (45.9)
41 101 58 46 34 37 35 11 38
(29.9) (73.7) (42.3) (33.5) (24.8) (27.0) (25.5) (8.0) (27.7)
10 (27.0) 15 (40.5) 3 (8.1)
21 (15.3) 32 (23.4) 11 (8.0)
.41 .98 .74 .12 .28 .29 .68 .33 .04 2.09 (0.87-4.96) 2.25 (1.04-4.88) 1.03 (0.27-3.89)
.095 .039 .97
BBW, birth body weight; CI, confidence interval; GA, gestational age; HFOV, high-frequency oscillatory ventilator; OR, odds ratio; TPN, total parenteral nutrition. *For presenting signs and symptoms, P value was calculated by c2 test.
independent risk factor of A baumannii bacteremia, it may be an indicator of more and prolonged TPN use and slower enteral feeding, which were both found significantly in the infants with A baumannii bacteremia, rather than the direct cause of susceptibility of A baumannii infection. Episodes of A baumannii bacteremia in the intensive care unit have been reported as clustered epidemics with contamination of environmental source or hand carriage by health care workers leading to colonization and subsequent infection.3,4,12,23 Reinforcement of
hand hygiene appears to be responsible for less frequent outbreaks of A baumannii in our NICU3; however, near half of our cases have occurred sporadically without a definite outbreak over the past 7 years. The route of transmission and main reservoirs of these organisms cannot be identified. Therefore, further studies by molecular genotyping and active surveillance cultures of environmental specimens or handwashings are needed to identify the source of this pathogen in order to better create prevention strategies and manipulation of these potential sources.
J.-F. Hsu et al. / American Journal of Infection Control 42 (2014) 23-7
Several studies have concluded that antibiotic-resistant A baumannii in the NICU is associated with higher risk of morbidity and mortality.2,24,25 In our study, the mortality rate of A baumannii bacteremia was comparable with that of overall neonates with bacteremia but significantly lower than the reported mortality rate of 37% to 47.5% for MDR A baumannii infection in previous studies.24,25 Based on these results, avoiding emergence of MDR A baumannii may reduce the mortality rate. Most A baumannii isolates in our NICU (94.6%) were antibiotic susceptible, which may be due to our policy of unusual administration of third-generation cephalosporin as the prophylactic or therapeutic regimen for hospitalized patients. Our study is consistent with previous reports that very low-birthweight infants were found to be associated with an increasing risk of Acinetobacter infection.3,12,14,15 We further demonstrated that infants with A baumannii bacteremia had significantly lower birth weight and relatively lower gestational age than those with E coli and Klebsiella bacteremia. This is probably due to their extended requirement of medical devices and supportive apparatuses, including mechanical ventilator and CVC, which were considered the presumptive sources of A baumannii. Therefore, removal of unnecessary CVC, as well as decreased use of TPN, may be the outbreak control strategy of A baumannii in the NICU. There were some limitations in our study, given its retrospective nature and inadequate sample size. Therefore, only 4 variables (both Tables 2 and 3) with a P value < .1 in the bivariate analysis were entered in to multivariate logistic regressions. Although the matched case-control study was chosen to control the potential confounding effects of prematurity and birth weight on the susceptibility to infection, there might be some selection bias in the uninfected controls. Infants with E coli and Klebsiella bacteremia were chosen as the control group because they were the most common gram-negative pathogens causing LOS in the NICU. However, the impact of A baumannii bacteremia may be more obvious if the infected controls were chosen based on the distributed spectrum in the NICU, in which gram-positive pathogens were the majority. In addition, none of the isolates were available for comparison analysis of organism DNA by pulse-field gel electrophoresis, thus whether or not the episodes from the suspected outbreak were caused by an identical strain cannot be documented. In conclusion, A baumannii bacteremia can occur sporadically in the NICU in addition to some outbreaks. Compared with E coli or Klebsiella, A baumannii tends to infect more premature infants and those with extremely low birth weight, especially those on TPN use and with a CVC in situ. Although A baumannii is not associated with higher mortality rate and MDR A baumannii is uncommon, it contributes significantly to longer hospitalization.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11. 12.
13.
14.
15.
16.
17.
18.
19. 20. 21.
Acknowledgments The authors thank Chiao-Ching Chiang for keeping the database of our NICU and all nursing staff working in our NICUs for keeping extremely detailed patient records, which contributed greatly to the completion of this research. References 1. Markogiannakis A, Fildisis G, Tsiplakou S, Ikonomidis A, Koutsoukou A, Pournaras S, et al. Cross-transmission of multidrug-resistant Acinetobacter
22.
23.
24.
25.
27
baumannii clonal strains causing episodes of sepsis in a trauma intensive care unit. Infect Control Hosp Epidemiol 2008;29:410-7. Simmonds A, Munoz J, Aguero-Rosenfeld M, Carbonaro C, Montecalvo M, Clones B, et al. Outbreak of Acinetobacter infection in extremely low birth weight neonates. Pediatr Infect Dis J 2009;28:210-4. Huang YC, Su LH, Wu TL, Leu HS, Hsieh WS, Chang TM, et al. Outbreak of Acinetobacter baumannii bacteremia in a neonatal intensive care unit: clinical implications and genotyping analysis. Pediatr Infect Dis J 2002;21: 1105-9. Thom KA, Johnson JK, Lee MS, Harris AD. Environmental contamination because of multidrug-resistant Acinetobacter baumannii surrounding colonized or infected patients. Am J Infect Control 2011;39:711-5. Cordero L, Rau R, Taylor D, Ayers LW. Enteric gram-negative bacilli bloodstream infections: 17 years’ experience in a neonatal intensive care unit. Am J Infect Control 2004;32:189-95. Cohen-Wolkowiez M, Moran C, Benjamin DK, Cotten CM, Clark RH, Benjamin DK Jr, et al. Early and late onset sepsis in late preterm infants. Pediatr Infect Dis J 2009;28:1052-6. Brodie S, Sands K, Gray J, Parker R, Goldmann D, Davis R, et al. Occurrence of nosocomial bloodstream infections in six neonatal intensive care units. Pediatr Infect Dis J 2000;19:56-62. Auriti C, Ronchetti MP, Pezzotti P, Marrocco G, Quondamcarlo A, Seqanti G, et al. Determinants of nosocomial infection in 6 neonatal intensive care units: an Italian multicenter prospective cohort study. Infect Control Hosp Epidemiol 2010;31:926-33. Tsai MH, Lien R, Wang JW, Huang HR, Chiang CC, Chu SM, et al. Complication rates with central venous catheters inserted at femoral and non-femoral sites in very low birth weight infants. Pediatr Infect Dis J 2009;28:966-70. Srinivasan R, Asselin J, Gildengorin G, Wiener-Kronish J, Flori HR. A prospective study of ventilator-associated pneumonia in children. Pediatrics 2009;123: 1108-15. Poirel L, Nordmann P. Carbapenem resistance in Acinetobacter baumannii: mechanisms and epidemiology. Clin Microbiol Infect 2006;12:826-36. Chan PC, Huang LM, Lin HC, Chang LY, Chen ML, Lu CY, et al. Control of an outbreak of pandrug-resistant Acinetobacter baumannii colonization and infection in a neonatal intensive care unit. Infect Control Hosp Epidemiol 2007; 28:423-9. Munoz-Price LS, Zembower T, Penugonda S, Schreckenberger P, Lavin MA, Welbel S, et al. Clinical outcomes of carbapenem-resistant Acinetobacter baumannii bloodstream infections: study of a 2-state monoclonal outbreak. Infect Control Hosp Epidemiol 2010;31:1057-62. McGrath EJ, Chopra T, Abdel-Haq N, Preney K, Koo W, Asmar BI, et al. An outbreak of carbapenem-resistant Acinetobacter baumannii infection in a neonatal intensive care unit: investigation and control. Infect Control Hosp Epidemiol 2011;32:34-41. Melamed R, Greenberg D, Porat N, Karplus M, Zmora E, Golan A, et al. Successful control of an Acinetobacter baumannii outbreak in a neonatal intensive care unit. J Hosp Infect 2003;53:31-8. Caricato A, Montini L, Bello G, Michetti V, Maviqlia R, Bocci MG, et al. Risk factors and outcome of Acinetobacter baumannii infection in severe trauma patients. Intensive Care Med 2009;35:1964-9. Playford EG, Craig JC, Iredell JR. Carbapenem-resistant Acinetobacter baumannii in intensive care unit patients: risk factors for acquisition, infection and their consequences. J Hosp Infect 2007;65:204-11. Moyer V, Freese DK, Whitington PF, Olson AD, Brewer F, Colletti RB, et al. Guideline for the evaluation of cholestatic jaundice in infants: recommendations of the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition. J Pediatr Gastroenterol Nutr 2004;39:115-28. Emerick KM, Whitington PF. Neonatal liver disease. Pediatr Ann 2006;35: 280-6. Taeusch HW, Ballard RA, Gleason CA. Avery’s Disease of the Newborn. 8th ed. Philadelphia [PA]: Elsevier Saunders; 2006. Bizzarro MJ, Raskind C, Baltimore RS, Gallagher PG. Seventy-five years of neonatal sepsis at Yale: 1928-2003. Pediatrics 2005;116:595-602. National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial disk susceptibility tests. Approved standard. 7th ed. NCCLS document M2eA7. Wayne [PA]: National Committee for Clinical Laboratory Standards; 2000. Hong KB, Oh HS, Song JS, Lim JH, Kang DK, Son IS, et al. Investigation and control of an outbreak of imipenem-resistant Acinetobacter baumannii infection in a pediatric intensive care unit. Pediatr Infect Dis J 2012;31: 685-90. Al Jarousha AM, El Jadba AH, Al Afifi AS, El Qouqa IA. Nosocomial multidrugresistant Acinetobacter baumannii in the neonatal intensive care unit in Gaza City, Palestine. Int J Infect Dis 2009;13:623-8. Saleem AF, Ahmed I, Mir F, Ali SR, Zaidi AK. Pan-resistant Acinetobacter infection in neonates in Karachi, Pakistan. J Infect Dev Ctries 2009;4:30-7.
Contents lists available at ScienceDirect
American Journal of Infection Control
American Journal of Infection Control
journal homepage: www.ajicjournal.org
Major article
Case-control analysis of endemic Acinetobacter baumannii bacteremia in the neonatal intensive care unit Jen-Fu Hsu MD a, b, Shih-Ming Chu MD a, b, Reyin Lien MD a, b, Cheng-Hsun Chiu MD, PhD b, c, Ming-Chou Chiang MD a, b, Ren-Huei Fu MD, PhD a, b, Chiang-Wen Lee MD, PhD d, Hsuan-Rong Huang MD a, b, Ming-Horng Tsai MD b, d, e, * a
Division of Pediatric Neonatology, Department of Pediatrics, Chang Gung Memorial Hospital, Taoyuan, Taiwan College of Medicine, Chang Gung University, Taoyuan, Taiwan Division of Pediatric Infectious Disease, Department of Pediatrics, Chang Gung Memorial Hospital, Taoyuan, Taiwan d Chang Gung University of Science and Technology, Chiayi, Taiwan e Division of Neonatology and Pediatric Hematology/Oncology, Department of Pediatrics, Chang Gung Memorial Hospital, Yunlin, Taiwan b c
Key Words: Late-onset sepsis Gram-negative bacteremia Multidrug resistance Neonate Recurrence
Background: We aimed to characterize the clinical manifestations and outcomes of patients with Acinetobacter baumannii bacteremia in the neonatal intensive care unit (NICU). Methods: All patients with A baumannii bacteremia in our NICU from 2004 to 2010 were reviewed. A matched case-control study was performed by comparing each case of A baumannii to 2 uninfected controls and all cases of Escherichia coli and Klebsiella bacteremia, respectively. Results: Thirty-seven cases with A baumannii bacteremia were identified. Multidrug-resistant isolate was noted in only 2 cases (5.4%), and the overall mortality rate was 8.1%. Compared with matched, uninfected controls, infants with A baumannii were more likely to have had a central vascular catheter (CVC) (P ¼ .009), use of total parenteral nutrition (TPN) (P ¼ .021), longer duration of ventilator use (P ¼ .002), and hospitalization (P ¼ .010). Compared with E coli or Klebsiella bacteremia, infants with A baumannii bacteremia had lower birth weight (median of 1,090 g vs 1,300 g, P ¼ .044) and a higher rate of CVC and TPN use (both P < .001) at the time of infection. Conclusion: A baumannii bacteremia occurs endemically or sporadically in the NICU, primarily in lowbirth-weight infants on TPN use and with CVC in situ. Although A baumannii does not often cause mortality, and multidrug-resistant A baumannii is uncommon, it contributes significantly to longer hospitalization. Crown Copyright Ó 2014 Published by Elsevier Inc. on behalf of the Association for Professionals in Infection Control and Epidemiology, Inc.
Acinetobacter baumannii is a nonfermentative, gram-negative coccobacillus, which is increasingly found in the intensive care unit.1,2 This microorganism has been one of the most important pathogens causing hospital-acquired infections for decades.3,4 It contributes to 8% to 25% of all gram-negative bacillary (GNB) lateonset sepsis (LOS) in the neonatal intensive care unit (NICU).5-7 In addition, A baumannii accounts for 4.6% of ventilator-associated pneumonia, 5.7% of catheter-related bloodstream infections (BSIs), and 4.3% of all BSIs.8-10 A baumannii can possibly become highly resistant to all antibiotics,11 which makes it deserve special attention and emphasis of infection control. Several recent reports
* Address correspondence to Ming-Horng Tsai, MD, Division of Neonatology and Pediatric Hematology/Oncology, Department of Pediatrics, Chang Gung Memorial Hospital, No. 707 Gongye Road, Sansheng, Mailiao Township, Yunlin, Taiwan, R.O.C. E-mail address: [email protected] (M.-H. Tsai). Conflicts of interest: None to report.
have described the emergence or outbreak of multidrug-resistant (MDR) A baumannii in the NICU that is associated with higher morbidity or mortality.3,12-14 Recent studies focused on clinical features and successful control of MDR A baumannii outbreak11-15; however, non-MDR A baumannii is rarely investigated. In adult or surgical intensive care unit, an A baumannii outbreak has traditionally been linked to environmental sources, such as contaminated medical devices or prolonged mechanical ventilation,16,17 but little relevant information was available in the NICU.3 The purpose of this case-control analysis was to identify and compare cases of endemic A baumannii bacteremia in a NICU over a 7-year period with matched uninfected and other GNB-infected controls to identify risk factors, presenting signs and symptoms, and outcomes specific to A baumannii that may facilitate the diagnosis and development of interventions for the control of A baumannii.
0196-6553/$36.00 - Crown Copyright Ó 2014 Published by Elsevier Inc. on behalf of the Association for Professionals in Infection Control and Epidemiology, Inc. http://dx.doi.org/10.1016/j.ajic.2013.06.017
24
J.-F. Hsu et al. / American Journal of Infection Control 42 (2014) 23-7
PATIENTS AND METHODS Hospital characteristics and patients Between January 2004 and December 2010, all neonates with at least 1 episode of A baumannii bacteremia during their hospitalization in the NICU of Chang Gung Memorial Hospital (CGMH) were identified from the electronic database. The NICU of CGMH is a tertiary-level medical center in northern Taiwan and contains a total of 49 beds equipped with mechanical ventilator and 28 beds with special care nurseries. This study was approved by the Institutional Review Board of CGMH.
not normally distributed. Student t test and the Mann-Whitney test were used to compare continuous variables and the c2 test or Fisher exact test was used to compare categorical variables. All P values were 2 tailed, and P values < .05 were considered to be statistically significant. Dichotomous variables were analyzed using odds ratio and bivariate logistic regression analysis. Despite the relatively small sample size and risk of model breakdown, independent covariates with a P value of < .10 as identified by bivariate analysis were subjected to multivariate logistic regression analysis. Statistical analyses were performed using SPSS version 15.0 (SPSS Inc, Chicago, IL).
Control case selection and matching criteria
RESULTS
To identify potential risk factors associated with acquisition of A baumannii bacteremia, 2 uninfected controls were selected and matched to each case of A baumannii bacteremia using birth weight (100 g), gender, and gestational age (1 week). Next, all cases of E coli and Klebsiella spp bacteremia during the same perioddafter excluding extended-spectrum b-lactamase-producing straind were enrolled as the infected controls.
Patient characteristics
Date collection and analysis Clinical information and microbiologic data were collected and included demographic characteristics, presence of central venous catheters (CVCs), invasive diagnostic and therapeutic procedures, total parenteral nutrition (TPN), mechanical ventilation, previous antibiotic exposure, and current usage. Definitions LOS was defined as positive microbial growth in 1 culture of blood samples obtained after 72 hours of life with accompanying clinical signs of sepsis. Recurrent bacteremia was defined as any new episode of documented BSI by the same or different pathogens occurring at least 2 weeks after the initial episode, a period during which blood cultures were negative. Prolonged intubation was defined as the need for mechanical ventilation for more than 14 days prior to infection. Prolonged hospitalization was defined as an infant who stayed in the NICU after the postconceptional age of 42 weeks. Cholestasis was defined according to the definition of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition.18,19 Death was considered attributable to sepsis if it occurred within 7 days of the positive blood culture or if clinical signs and symptoms of sepsis were documented in the medical record as the direct cause of death. All comorbidities of prematurity, including respiratory distress syndrome, intraventricular hemorrhage, bronchopulmonary dysplasia, and necrotizing enterocolitis were based on the latest updated diagnostic criteria in the standard textbook of neonatology.20 In our NICU, all blood cultures were assessed using the Bactec 9240 system (Becton Dickinson Diagnostic Instrument Systems, Sparks, MD), and antibiotic susceptibility patterns were tested using disk diffusion methods on Mueller-Hinton agar plates, using calibrated inoculum of the isolates based on the McFarland standard.21,22 MDR A baumannii was defined when the isolate was resistant to 3 of the following 4 antibiotics: ceftazidime, ciprofloxacin, gentamicin, and imipenem. Statistical analysis Continuous variables were summarized using means and standard deviations if normally distributed and median and ranges if
Thirty-seven cases of A baumannii bacteremia were identified in neonates hospitalized in the NICU of CGMH between January 2004 and December 2010. Infants with A baumannii bacteremia had median gestational age and birth weight of 28 weeks and 1,090 g, respectively. During the same period, a total of 137 neonates who experienced Escherichia coli or Klebsiella bacteremia, after excluding extended-spectrum b-lactamase-producing bacteria, was identified as the control group. The uninfected controls consisted of 74 neonates. The demographic characteristics of cases and selected controls are summarized in Table 1. Organism characteristics and antibiotic susceptibility patterns of A baumannii Of all episodes, 19 cases occurred endemically into 4 major outbreaks (defined as significantly more episodes within a short period when compared with a 6-month interval by c2 test), and others occurred sporadically over this 7-year period (Fig 1). Only 2 isolates were MDR strain, and none of the others were resistant to gentamicin or third-generation cephalosporins. Most cases (35/37, 94.6%) were on TPN and/or intralipids nutritional support and with a CVC in situ, and 81.1% (30/37) were under ventilator support at the time of infection. Comparison of cases of A baumannii and uninfected controls The overall mortality was significantly higher among infected patients (8.1% vs 0%, respectively, P ¼ .014, Table 1). Patients with A baumannii were more likely to have longer duration of ventilator use and hospitalization (P ¼ .002 and P ¼ .010, respectively). None of perinatal risk factors, bronchopulmonary dysplasia, intraventricular hemorrhage, or necrotizing enterocolitis were found to be associated with acquisition of A baumannii bacteremia. Although neonates with A baumannii were comparable with the uninfected controls with regard to the day of first attempt at feeding, they used significantly longer duration of TPN and/or intralipids and thus much later reached full feeding amount (both P < .001). The bivariate logistic regression analysis (Table 2) revealed that patients with A baumannii were more likely to have an indwelling CVC (OR, 11.22; 95% CI: 1.43-88.21, P ¼ .021), use of TPN and/or intralipids (OR, 9.27; 95% CI: 1.17-73.40, P ¼ .035), use of high frequency oscillatory ventilator after birth (OR, 3.11; 95% CI: 1.148.44, P ¼ .026), prolonged intubation (OR, 2.59; 95% CI: 1.12-5.97, P ¼ .025), and cholestasis (OR, 6.35; 95% CI: 2.47-16.30, P < .001). Multivariate logistic regression confirmed cholestasis as the only significant independent risk factor for A baumannii bacteremia (OR, 4.50; 95% CI: 1.61-12.54, P ¼ .004).
J.-F. Hsu et al. / American Journal of Infection Control 42 (2014) 23-7
25
Table 1 Characteristics of cases of A baumannii sepsis versus uninfected, sepsis with other gram-negative bacilli controls
Characteristics
A baumannii (n ¼ 37)
Uninfected controls (n ¼ 74)
E coli or Klebsiella spp (n ¼ 137)*
P valuey
1,090 (420-2,618) 29.4 3.8 17/20 8 (21.6) 23.0 (13.3-40.3) 0 (0) 4 (10.8) 19 (51.4) 0 (0) 0 (0) 18 (48.6) 14 (37.8) 3.0 (2.0-7.0) 37.0 (20.0-72.0) 35 (94.6) 32 (12.0-57.0) 50 (3.0-75.3) 35.5 (14.5-66.5) 10 (27.0) 15 (40.5) 85.5 (53.8-111.0) 3 (8.1)
1,100 (472-2,600) 29.6 3.4 38/36 14 (18.9) 0 (0) 1 (1.4) 29 (39.1) 1 (1.4) 1 (1.4) 29 (39.2) 11 (14.9) 3.0 (2.0-6.0) 20.0 (13.8-30.0) 55 (74.3) 11.5 (5.0-20.5) 19.0 (2.0-50.3) 17.7 (9.0-25.7) 9 (12.1) 60.5 (40.3-87.5) 0 (0)
1,300 (540-3,870) 30.6 4.3 76/61 57 (41.6) 25.0 (16.0-46.0) 2 (1.5) 19 (13.9) 60 (43.8) 3 (2.2) 7 (5.1) 61 (44.5) 42 (30.7) 6.0 (3.0-12.3) 39.0 (24.0-71.5) 85 (62.1) 21.0 (10.5-45.0) 30.0 (2.8-68.3) 22.0 (14.5-48.5) 21 (15.3) 32 (23.4) 69.0 (46.0-97.0) 11 (8.0)
.98, .044 .71, .13 .54, .26 .80, .034 n/a, .50 n/a, .45 .026, .59 .15, .44 .48, .36 .48, .15 .22, .85 <.001, .68 .45, .087 .001, .84 .009, <.001 <.001, .026 .002, .088 <.001, .096 n/a, .077 <.001, .032 .010, .16 .014, .97
Birth body weight (g), median (range) Gestational age (wks) Sex (m/f) Vaginal delivery, n (%) Day of life at positive culture, median (IQR) Perinatal asphyxia and/or hypoxic ischemic encephalopathy, n (%) Surgical history, n (%)z 5-Minute Apgar score 7, n (%) Necrotizing enterocolitis (stage II), n (%) Intraventricular hemorrhage (grade III), n (%) Bronchopulmonary dysplasia, n (%) Cholestasis, n (%) Days of trying oral feeding, median (IQR) Days of reaching full feeding, median (IQR) Use of TPN and/or intralipids, n (%) Duration of TPN and/or intrafat, days, median (IQR) Duration of ventilator use, days, median (IQR) Duration of central venous catheter, days, median (IQR) Recurrent sepsis after bacteremia, n (%)z Prolonged hospitalization, n (%) Days of hospitalization, median (IQR) Final mortality, n (%)
IQR, interquartile range; n/a, not available; TPN, total parenteral nutrition. *Including all cases of late-onset sepsis caused by E coli and Klebsiella spp, excluding extended-spectrum b-lactamase-producing bacteria during the study period. y P values are expressed as comparisons between A baumannii and uninfected controls, A baumannii and E coli/Klebsiella spp. z Defined as blood culture-proven bloodstream infection after the cases of A baumannii or E coli and Klebsiella spp bacteremia.
mortality, except cases with A baumannii were more likely to have thrombocytopenia (defined as platelet count < 50,000/mL) (45.9% vs 27.7%, respectively, P ¼ .040). Despite relatively a small number of cases and controls for analysis, a multivariable logistic regression was performed and the TPN use determined to be a significant independent covariate for A baumannii bacteremia as compared with E coli and Klebsiella (adjusted OR, 6.59; 95% CI: 1.44-30.13, P ¼ .015). For final outcome analysis, although the overall mortality rates were comparable between A baumannii bacteremia and E coli and Klebsiella bacteremia, patients with A baumannii bacteremia had relatively high rate of recurrent LOS and significantly higher rate of prolonged hospitalization (OR, 2.25; 95% CI: 1.04-4.88, P ¼ .039). DISCUSSION
Fig 1. Distribution of 37 episodes of Acinetobacter baumannii bacteremia and 4 major outbreaks (marked by star at the top) in the neonatal intensive care unit during 2004 to 2010.
Comparison of cases of A baumannii and E coli and Klebsiella controls Infants with A baumannii bacteremia had lower birth weight (median of 1,090 g vs 1,300 g, respectively, P ¼ .044) than those with E coli and Klebsiella LOS. Although their gestational ages were comparable, A baumannii bacteremia occurred more frequently in infants with gestational age of less than 29 weeks (OR, 2.27; 95% CI: 1.08-4.78, P ¼ .032). Infants with A baumannii bacteremia were more likely to have a CVC in place (OR, 4.35; 95% CI: 1.23-10.43, P ¼ .021) and be receiving TPN and/or intralipids (OR, 7.29; 95% CI: 1.66-31.95, P ¼ .008) at the time of infection. There were no differences in personal risk factors, presenting symptoms and signs, or
The main objective of this study was to investigate the risk factors and impact of A baumannii bacteremia in the NICU. We found that A baumannii accounted for 4.7% of all episodes of LOS and 14.0% of all GNB-related LOS in our NICU. From our data, although A baumannii had a tendency toward clustered episodes of bacteremia, it could also occur sporadically. Very low-birth-weight neonates were at higher risk of A baumannii bacteremia, especially when they were on TPN use and with CVC in situ. Because most patients had mechanical ventilator at the time of infection, the presumptive sources of infection, as previous studies identified, may include the medical devices and supportive apparatuses.16,17 Although A baumannii bacteremia in our NICU was mostly antibiotic susceptible and was not associated with higher mortality, it deserves investigation because A baumannii bacteremia was significantly associated with longer hospitalization, which potentially increase patient risk to further adverse events, costs, and family bonding. The prolonged hospital stay may be associated with the relatively higher rate of recurrent LOS after A baumannii bacteremia and the existences of more TPN and CVC use in neonates with A baumannii bacteremia. Although cholestasis was confirmed by multivariate logistic regression analysis to be the
26
J.-F. Hsu et al. / American Journal of Infection Control 42 (2014) 23-7
Table 2 Bivariable logistic regression analysis of risk factors and outcomes associated with cases of A baumannii sepsis versus uninfected controls Characteristic
A baumannii (n ¼ 37)
Host and perinatal factors Male sex Vaginal delivery Prolonged rupture of membranes Maternal fever Chorioamnionitis Five-minute Apgar score 7 Personal factors Bronchopulmonary dysplasia Surgical history Cholestasis Procedures and conditions Prolonged intubation Use of HFOV after birth Central venous catheter Use of TPN and/or intralipids Clinical outcome Prolonged hospitalization Death
17 8 5 2 1 19
Uninfected controls (n ¼ 74)
(45.9) (21.6) (13.5) (5.4) (2.7) (51.4)
38 14 11 2 0 29
OR (95% CI)
(51.3) (18.9) (14.9) (2.7) (0) (39.1)
(0.37-1.69) (0.43-3.05) (0.29-2.81) (0.28-15.28) 1.84 (0.81-4.20)
.50 .79 .85 .48 .14
18 (48.6) 4 (10.8) 14 (37.8)
29 (39.2) 1 (1.4) 11 (14.9)
1.68 (0.74-3.80) 8.63 (0.93-80.30) 6.35 (2.47-16.30)
.21 .058 <.001
19 11 36 35
22 9 53 55
(51.4) (29.7) (97.3) (94.6)
15 (40.5) 3 (8.1)
0.76 1.14 0.89 2.06
P value
(29.7) (12.2) (71.6) (74.3)
2.59 3.11 11.22 9.27
9 (12.1) 0 (0)
(1.12-5.97) (1.14-8.44) (1.43-88.21) (1.17-73.40)
.025 .026 .021 .035
6.43 (2.31-17.90)
<.001
CI, confidence interval; HFOV, high-frequency oscillatory ventilator; OR, odds ratio; TPN, total parenteral nutrition.
Table 3 Bivariable logistic regression analysis of risk factors and outcomes associated with cases of A baumannii sepsis versus other gram-negative, late-onset sepsis controls Characteristic Host and perinatal factors Extremely low birth body weight (BBW < 1,000 g) Very preterm (GA < 29 weeks) Male sex Delivery by cesarean section Prolonged rupture of membranes Maternal fever Chorioamnionitis Five-minute Apgar score 7 Personal factors Bronchopulmonary dysplasia Use of HFOV after birth Surgical history Cholestasis Procedures and conditions Intubated Prolonged intubation Central venous catheter TPN and/or intralipids Presenting signs and symptoms Fever Apnea and/or bradycardia and cyanosis Abdominal distension and/or feeding intolerance Hypoglycemia or hyperglycemia Tachycardia Hypotension Coagulopathy Disseminated intravascular coagulopathy Thrombocytopenia (platelet count < 50,000/mL) Clinical outcome Recurrent late-onset sepsis Prolonged hospitalization Death
A baumannii (n ¼ 37), n (%)
E coli or Klebsiella spp (n ¼ 137), n (%)
OR (95% CI)
P value*
15 20 17 29 5 2 1 19
(40.5) (54.1) (45.9) (78.4) (13.5) (5.4) (2.7) (51.4)
39 48 76 80 24 4 0 60
(28.5) (35.0) (55.5) (58.4) (17.5) (2.9) (0) (43.8)
1.76 2.27 0.62 2.56 0.71 1.85
(0.82-3.76) (1.08-4.78) (0.29-1.30) (1.09-6.03) (0.25-2.01) (0.32-10.54) 1.42 (0.67-3.03)
.15 .032 .21 .032 .52 .49 .36
18 11 4 14
(48.6) (29.7) (10.8) (37.8)
61 37 19 42
(44.5) (27.0) (13.9) (30.7)
1.13 1.20 0.73 1.54
(0.53-2.39) (0.54-2.66) (0.23-2.30) (0.71-3.35)
.75 .66 .73 .27
19 19 36 35
(51.4) (51.4) (97.3) (94.6)
55 57 99 85
(40.1) (41.6) (70.8) (62.1)
1.49 1.35 4.35 7.29
(0.70-3.17) (0.64-2.88) (1.23-10.43) (1.66-31.95)
.30 .43 .02 .008
13 28 15 18 6 12 11 5 17
(35.1) (75.7) (40.5) (48.6) (16.2) (32.4) (29.7) (13.5) (45.9)
41 101 58 46 34 37 35 11 38
(29.9) (73.7) (42.3) (33.5) (24.8) (27.0) (25.5) (8.0) (27.7)
10 (27.0) 15 (40.5) 3 (8.1)
21 (15.3) 32 (23.4) 11 (8.0)
.41 .98 .74 .12 .28 .29 .68 .33 .04 2.09 (0.87-4.96) 2.25 (1.04-4.88) 1.03 (0.27-3.89)
.095 .039 .97
BBW, birth body weight; CI, confidence interval; GA, gestational age; HFOV, high-frequency oscillatory ventilator; OR, odds ratio; TPN, total parenteral nutrition. *For presenting signs and symptoms, P value was calculated by c2 test.
independent risk factor of A baumannii bacteremia, it may be an indicator of more and prolonged TPN use and slower enteral feeding, which were both found significantly in the infants with A baumannii bacteremia, rather than the direct cause of susceptibility of A baumannii infection. Episodes of A baumannii bacteremia in the intensive care unit have been reported as clustered epidemics with contamination of environmental source or hand carriage by health care workers leading to colonization and subsequent infection.3,4,12,23 Reinforcement of
hand hygiene appears to be responsible for less frequent outbreaks of A baumannii in our NICU3; however, near half of our cases have occurred sporadically without a definite outbreak over the past 7 years. The route of transmission and main reservoirs of these organisms cannot be identified. Therefore, further studies by molecular genotyping and active surveillance cultures of environmental specimens or handwashings are needed to identify the source of this pathogen in order to better create prevention strategies and manipulation of these potential sources.
J.-F. Hsu et al. / American Journal of Infection Control 42 (2014) 23-7
Several studies have concluded that antibiotic-resistant A baumannii in the NICU is associated with higher risk of morbidity and mortality.2,24,25 In our study, the mortality rate of A baumannii bacteremia was comparable with that of overall neonates with bacteremia but significantly lower than the reported mortality rate of 37% to 47.5% for MDR A baumannii infection in previous studies.24,25 Based on these results, avoiding emergence of MDR A baumannii may reduce the mortality rate. Most A baumannii isolates in our NICU (94.6%) were antibiotic susceptible, which may be due to our policy of unusual administration of third-generation cephalosporin as the prophylactic or therapeutic regimen for hospitalized patients. Our study is consistent with previous reports that very low-birthweight infants were found to be associated with an increasing risk of Acinetobacter infection.3,12,14,15 We further demonstrated that infants with A baumannii bacteremia had significantly lower birth weight and relatively lower gestational age than those with E coli and Klebsiella bacteremia. This is probably due to their extended requirement of medical devices and supportive apparatuses, including mechanical ventilator and CVC, which were considered the presumptive sources of A baumannii. Therefore, removal of unnecessary CVC, as well as decreased use of TPN, may be the outbreak control strategy of A baumannii in the NICU. There were some limitations in our study, given its retrospective nature and inadequate sample size. Therefore, only 4 variables (both Tables 2 and 3) with a P value < .1 in the bivariate analysis were entered in to multivariate logistic regressions. Although the matched case-control study was chosen to control the potential confounding effects of prematurity and birth weight on the susceptibility to infection, there might be some selection bias in the uninfected controls. Infants with E coli and Klebsiella bacteremia were chosen as the control group because they were the most common gram-negative pathogens causing LOS in the NICU. However, the impact of A baumannii bacteremia may be more obvious if the infected controls were chosen based on the distributed spectrum in the NICU, in which gram-positive pathogens were the majority. In addition, none of the isolates were available for comparison analysis of organism DNA by pulse-field gel electrophoresis, thus whether or not the episodes from the suspected outbreak were caused by an identical strain cannot be documented. In conclusion, A baumannii bacteremia can occur sporadically in the NICU in addition to some outbreaks. Compared with E coli or Klebsiella, A baumannii tends to infect more premature infants and those with extremely low birth weight, especially those on TPN use and with a CVC in situ. Although A baumannii is not associated with higher mortality rate and MDR A baumannii is uncommon, it contributes significantly to longer hospitalization.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11. 12.
13.
14.
15.
16.
17.
18.
19. 20. 21.
Acknowledgments The authors thank Chiao-Ching Chiang for keeping the database of our NICU and all nursing staff working in our NICUs for keeping extremely detailed patient records, which contributed greatly to the completion of this research. References 1. Markogiannakis A, Fildisis G, Tsiplakou S, Ikonomidis A, Koutsoukou A, Pournaras S, et al. Cross-transmission of multidrug-resistant Acinetobacter
22.
23.
24.
25.
27
baumannii clonal strains causing episodes of sepsis in a trauma intensive care unit. Infect Control Hosp Epidemiol 2008;29:410-7. Simmonds A, Munoz J, Aguero-Rosenfeld M, Carbonaro C, Montecalvo M, Clones B, et al. Outbreak of Acinetobacter infection in extremely low birth weight neonates. Pediatr Infect Dis J 2009;28:210-4. Huang YC, Su LH, Wu TL, Leu HS, Hsieh WS, Chang TM, et al. Outbreak of Acinetobacter baumannii bacteremia in a neonatal intensive care unit: clinical implications and genotyping analysis. Pediatr Infect Dis J 2002;21: 1105-9. Thom KA, Johnson JK, Lee MS, Harris AD. Environmental contamination because of multidrug-resistant Acinetobacter baumannii surrounding colonized or infected patients. Am J Infect Control 2011;39:711-5. Cordero L, Rau R, Taylor D, Ayers LW. Enteric gram-negative bacilli bloodstream infections: 17 years’ experience in a neonatal intensive care unit. Am J Infect Control 2004;32:189-95. Cohen-Wolkowiez M, Moran C, Benjamin DK, Cotten CM, Clark RH, Benjamin DK Jr, et al. Early and late onset sepsis in late preterm infants. Pediatr Infect Dis J 2009;28:1052-6. Brodie S, Sands K, Gray J, Parker R, Goldmann D, Davis R, et al. Occurrence of nosocomial bloodstream infections in six neonatal intensive care units. Pediatr Infect Dis J 2000;19:56-62. Auriti C, Ronchetti MP, Pezzotti P, Marrocco G, Quondamcarlo A, Seqanti G, et al. Determinants of nosocomial infection in 6 neonatal intensive care units: an Italian multicenter prospective cohort study. Infect Control Hosp Epidemiol 2010;31:926-33. Tsai MH, Lien R, Wang JW, Huang HR, Chiang CC, Chu SM, et al. Complication rates with central venous catheters inserted at femoral and non-femoral sites in very low birth weight infants. Pediatr Infect Dis J 2009;28:966-70. Srinivasan R, Asselin J, Gildengorin G, Wiener-Kronish J, Flori HR. A prospective study of ventilator-associated pneumonia in children. Pediatrics 2009;123: 1108-15. Poirel L, Nordmann P. Carbapenem resistance in Acinetobacter baumannii: mechanisms and epidemiology. Clin Microbiol Infect 2006;12:826-36. Chan PC, Huang LM, Lin HC, Chang LY, Chen ML, Lu CY, et al. Control of an outbreak of pandrug-resistant Acinetobacter baumannii colonization and infection in a neonatal intensive care unit. Infect Control Hosp Epidemiol 2007; 28:423-9. Munoz-Price LS, Zembower T, Penugonda S, Schreckenberger P, Lavin MA, Welbel S, et al. Clinical outcomes of carbapenem-resistant Acinetobacter baumannii bloodstream infections: study of a 2-state monoclonal outbreak. Infect Control Hosp Epidemiol 2010;31:1057-62. McGrath EJ, Chopra T, Abdel-Haq N, Preney K, Koo W, Asmar BI, et al. An outbreak of carbapenem-resistant Acinetobacter baumannii infection in a neonatal intensive care unit: investigation and control. Infect Control Hosp Epidemiol 2011;32:34-41. Melamed R, Greenberg D, Porat N, Karplus M, Zmora E, Golan A, et al. Successful control of an Acinetobacter baumannii outbreak in a neonatal intensive care unit. J Hosp Infect 2003;53:31-8. Caricato A, Montini L, Bello G, Michetti V, Maviqlia R, Bocci MG, et al. Risk factors and outcome of Acinetobacter baumannii infection in severe trauma patients. Intensive Care Med 2009;35:1964-9. Playford EG, Craig JC, Iredell JR. Carbapenem-resistant Acinetobacter baumannii in intensive care unit patients: risk factors for acquisition, infection and their consequences. J Hosp Infect 2007;65:204-11. Moyer V, Freese DK, Whitington PF, Olson AD, Brewer F, Colletti RB, et al. Guideline for the evaluation of cholestatic jaundice in infants: recommendations of the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition. J Pediatr Gastroenterol Nutr 2004;39:115-28. Emerick KM, Whitington PF. Neonatal liver disease. Pediatr Ann 2006;35: 280-6. Taeusch HW, Ballard RA, Gleason CA. Avery’s Disease of the Newborn. 8th ed. Philadelphia [PA]: Elsevier Saunders; 2006. Bizzarro MJ, Raskind C, Baltimore RS, Gallagher PG. Seventy-five years of neonatal sepsis at Yale: 1928-2003. Pediatrics 2005;116:595-602. National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial disk susceptibility tests. Approved standard. 7th ed. NCCLS document M2eA7. Wayne [PA]: National Committee for Clinical Laboratory Standards; 2000. Hong KB, Oh HS, Song JS, Lim JH, Kang DK, Son IS, et al. Investigation and control of an outbreak of imipenem-resistant Acinetobacter baumannii infection in a pediatric intensive care unit. Pediatr Infect Dis J 2012;31: 685-90. Al Jarousha AM, El Jadba AH, Al Afifi AS, El Qouqa IA. Nosocomial multidrugresistant Acinetobacter baumannii in the neonatal intensive care unit in Gaza City, Palestine. Int J Infect Dis 2009;13:623-8. Saleem AF, Ahmed I, Mir F, Ali SR, Zaidi AK. Pan-resistant Acinetobacter infection in neonates in Karachi, Pakistan. J Infect Dev Ctries 2009;4:30-7.