Prompt Control of an Outbreak Caused by Extended-Spectrum β-Lactamase–Producing Klebsiella pneumoniae in a Neonatal Intensive Care Unit

Prompt Control of an Outbreak Caused by Extended-Spectrum b-Lactamase–Producing Klebsiella pneumoniae in a Neonatal Intensive Care Unit Joseph B. Cantey, MD1, Pranavi Sreeramoju, MD2,3, Mambarambath Jaleel, MD1, Sylvia Trevi~no, MT3, Rita Gander, PhD4, Linda S. Hynan, PhD5, Jennifer Hill, BSN6, Cari Brown, BSN6, Wendy Chung, MD, MPH7, Jane D. Siegel, MD1, and Pablo J. Sanchez, MD1 Objectives To assess the effectiveness of a set of multidisciplinary interventions aimed at limiting patient-topatient transmission of extended-spectrum b-lactamase-producing Klebsiella pneumoniae (ESBL-KP) during a neonatal intensive care unit (NICU) outbreak, and to identify risk factors associated with ESBL-KP colonization and disease in this setting. Study design A 61-infant cohort present in the NICU during an outbreak of ESBL-KP from April 26, 2011, to May 16, 2011, was studied. Clinical characteristics were compared in infected/colonized infants and unaffected infants. A multidisciplinary team formulated an outbreak control plan that included (1) staff reeducation on recommended infection prevention measures; (2) auditing of hand hygiene and environmental services practices; (3) contact precautions; (4) cohorting of infants and staff; (5) alleviation of overcrowding; and (6) frequent NICU-wide screening cultures. Neither closure of the NICU nor culturing of health care personnel was instituted. Results Eleven infants in this level III NICU were infected/colonized with ESBL-KP. The index case was an 18-day-old infant born at 25 weeks’ gestation who developed septicemia from ESBL-KP. Two other infants in the same room developed sepsis from ESBL-KP within 48 hours; both expired. Implementation of various infection prevention strategies resulted in prompt control of the outbreak within 3 weeks. The ESBL-KP isolates presented a single clone that was distinct from ESBL-KP identified previously in other units. Being housed in the same room as the index infant was the only risk factor identified by logistic regression analysis (P = .002). Conclusion This outbreak of ESBL-KP affected 11 infants and was associated with 2 deaths. Prompt control with eradication of the infecting strain from the NICU was achieved with multidisciplinary interventions based on standard infection prevention practices. (J Pediatr 2013;163:672-79).

O

utbreaks caused by multidrug-resistant organisms (MDROs) are a significant cause of morbidity and mortality in infants in neonatal intensive care units (NICUs) worldwide. Specifically, multidrug-resistant gram-negative bacilli, including extended-spectrum b-lactamase (ESBL)-producing organisms, have been responsible for an increasing number of NICU outbreaks,1-3 resulting in worse outcomes, including death, in affected infants as well as higher health care costs.4-7 When an outbreak is recognized in a NICU, prompt multidisciplinary investigation led by the infection prevention and control team and implementation of risk mitigation strategies are recommended.8 Coordination and implementation of such control measures are often time-consuming and difficult, leading to ongoing transmission of the causative organism and occasionally necessitating closure of the unit to admissions.9-11 In 2011, an outbreak due to ESBL-producing Klebsiella pneumoniae (ESBL-KP) occurred in the NICU at Parkland Memorial Hospital (PMH) in Dallas. Here we report our multidisciplinary management of this outbreak, which resulted in the prompt control and eradication of this organism from the NICU, which remained open to all admissions. We also report our investigation of risk factors associated with ESBL-KP colonization and disease.

Methods PMH is a 672-bed public tertiary care center that provides a wide range of services, including high-risk obstetrics. The PMH NICU is a 90-bed level 3C,

ESBL HCP KP MDROs NICU PMH

Extended-spectrum b-lactamase Health care personnel Klebsiella pneumoniae Multidrug-resistant organisms Neonatal intensive care unit Parkland Memorial Hospital

From the Departments of 1Pediatrics and 2Internal Medicine, University of Texas Southwestern Medical Center; 3Infection Prevention, Parkland Health and Hospital System; Departments of 4Pathology and 5 Clinical Sciences, University of Texas Southwestern Medical Center; 6Women and Infants’ Specialty Health, Parkland Health and Hospital System; and 7Dallas County Health and Human Services, Dallas, TX The authors declare no conflicts of interest. Presented in part at the Pediatric Academic Societies’ Annual Meeting, Boston, MA, April 27 to May 1, 2012. 0022-3476/$ - see front matter. Copyright ª 2013 Mosby Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2013.03.001

672

Vol. 163, No. 3  September 2013 predominantly inborn unit with approximately 1200 admissions annually (Figure 1; available at www.jpeds.com). The NICU consists of 7 bays (“rooms” hereinafter) each with a maximum capacity of 12 infants depending on acuity (Figure 1, rooms A to G; Table I). There is 1 separate single bed, in a negatively ventilated isolation room (room H). A proximity-operated sink is located at the entrance to the NICU, and additional sinks are positioned at the entrance and rear of each room. Wall-mounted hand gel dispensers are available at the entrance to the NICU, at the entrance to each room, and at each bedside. Neither gowning nor a 3-minute surgical scrub is required for entrance into the NICU. PMH’s continuing-care nursery is an intermediate care unit on the same hospital floor but geographically separate from the main NICU area. It consists of 2 rooms with a maximum capacity of 10 infants per room (Table I, rooms I and J), and a separate, negatively ventilated isolation room capable of housing 3 infants. PMH’s newborn nursery is a level-one nursery that spans 2 hospital floors and is geographically separate from both the continuing-care nursery and the NICU. The study cohort comprised infants who were present in the NICU on the day of onset of illness in the index patient (April 26, 2011) and underwent at least 1 surveillance culture. The infants were followed to the time of hospital discharge or death. On May 2, 3 distinct infant cohorts, housed in separate rooms, were established: ESBL-KP–colonized infants, non– ESBL-KP–colonized infants who had been in the NICU when colonized infants were identified, and new admissions to the NICU after the outbreak was recognized. These latter infants who were admitted to the NICU after cohorting began on May 2 were not included in our study cohort. We defined a case as an infant in the NICU in whom ESBL-KP was isolated from bacterial cultures obtained from normally sterile sites (eg, blood, cerebrospinal fluid, urine), clinical samples taken from nonsterile sites (eg, tracheal aspirates, conjunctiva), or rectal/throat swab cultures obtained for surveillance. On May 3, a multidisciplinary team was formed consisting of neonatologists (1 of whom was the medical director of the NICU), pediatric infectious disease physicians, the nurse director, nurse managers, respiratory therapy managers, medical director of infection prevention and control, infection preventionists, microbiologists, and the director of environmental services. On May 13, epidemiologists from the Dallas County Health and Human Services and the Texas Department of Health and Human Services were invited to assist the NICU multidisciplinary team with management and analysis. Clinical cultures were obtained from infants at the discretion of the health care personnel (HCP). A schedule for obtaining surveillance cultures was initiated on May 2, 2011, after recognition of the outbreak. Rectal and throat swab cultures were obtained from each infant present in the NICU between May 2 and May 20; subsequently, only rectal swab cultures were obtained twice-weekly between May 20 and

June 13, weekly between June 13 and August 1, and every other week thereafter, continuing to August 2012. The last infant in the cohort was discharged on September 9, 2011. Specimens were processed in PMH’s microbiology laboratory by plating samples onto MacConkey agar containing ceftazidime to select for b-lactamase production. ESBL-KP isolates were analyzed for relatedness with the DiversiLab Klebsiella DNA Fingerprinting Kit (bioMerieux, Durham, North Carolina).12 Isolates were also sent to the Texas Department of Health and Human Services and the Centers for Disease Control and Prevention for confirmatory testing by pulsed-field gel electrophoresis.13 The specific ESBL enzymes were identified by high-fidelity amplification polymerase chain reaction (50 Prime, Gaithersburg, Maryland), as described previously.14,15 We conducted a cohort study to identify risk factors associated with ESBL-KP colonization or infection among all infants present in the NICU at any time between April 26 and May 2, 2011. Demographic, clinical, laboratory, and outcome data were collected for each infant in the cohort. Other data collected included daily room assignments, medical staff assignments, duration of invasive devices (eg, endotracheal tube for mechanical ventilation, central venous catheter), duration of parenteral nutrition, duration and type of enteral nutrition, surgical procedures performed in the operating room or at the bedside, radiographic studies, ophthalmology examinations, medication use (including surfactant, acid-blocking therapy, and antibiotics), transfusions of blood products, length of NICU stay, number and type of infections, and mortality. The number of ESBL-KP patient-exposure days was calculated by multiplying the number of days in any given room by the number of infected or colonized infants in that room; for example, being in the same room with 3 colonized infants for 4 days would equal 12 patient-exposure days. Data are reported as median and IQR were provided for continuous measures, and infected or colonized infants were compared with uninfected infants using MannWhitney U statistics. Data are reported as number and percentage for categorical measures, and the c2 test was used to compare the infant groups except where noted. Multivariate logistic regression using the forward stepwise method was used to predict infection group using demographic characteristics. The Hosmer-Lemeshow goodness-of-fit P value was used to describe the fit of the model to the data, with P > .40 considered to indicate a good fit. The variables included in the stepwise model were gestational age in days, birth weight, days of humidity, days in a bassinet, days with an umbilical arterial catheter in place, days with an umbilical venous catheter in place, days with a peripherally inserted central catheter in place, days of room air, number of abdominal ultrasound examinations, doses of surfactant received, days admitted to room E, and number of patientexposure days. SPSS version 19 (IBM, Armonk, New York) and StatXact-8 (Cytel, Cambridge, Massachusetts) were used to analyze the data. All tests were 2-sided, with P < .05 considered to indicate significance. 673

THE JOURNAL OF PEDIATRICS



www.jpeds.com

Vol. 163, No. 3

Table I. Comparison of infants infected or colonized with ESBL-KP and those not colonized during outbreak Infants Variable

Infected/colonized (n = 11)

Noncolonized (n = 50)

P value

Male sex, n (%) Gestational age, weeks, median (IQR) Birth weight, g, median (IQR) Apgar score, median (IQR) 1 minute 5 minute Physician team, n (%) Red Blue Purple Mother admitted <3 months before delivery, n (%) Mother received antibiotics <3 months before delivery, n (%) Mother received intrapartum antibiotics, n (%) Days of intrapartum antibiotics, median (IQR) Ampicillin Gentamicin Clindamycin Age at time of culture, days, median (IQR) Duration of bed use, days, median (IQR) Isolette Isolette with humidity Bassinet Duration of total parenteral nutrition, days, median (IQR) Hyperalimentation Intralipid Duration of central line use, days, median (IQR) Umbilical venous catheter Umbilical arterial catheter Peripherally inserted central venous catheter Surgical central line Peripheral arterial line Duration of respiratory support, days, median (IQR) Mechanical ventilation Inhaled nitric oxide Continuous positive airway pressure Nasal cannula Room air Duration of enteral nutrition, days, median (IQR) Breast milk Formula Fortifier Surgical procedures, n, median (IQR) Operating room Bedside Hypothermia therapy, n (%) Off-unit radiographic studies, n, median (IQR) Ophthalmologic examination, n, median (IQR) Sonography, n, median (IQR) Cranial Abdominal Echocardiography Other Surfactant use, n (%) Duration of antiacid therapy, days, median (IQR) H2 blocker Proton pump inhibitor Duration of antibiotic use, days, median (IQR) Ampicillin Gentamicin Oxacillin Cefotaxime Piperacillin/tazobactam Vancomycin Meropenem Transfusions, n. median (IQR) Packed red blood cells Platelets

5 (45) 28 (25-32) 790 (730-1224)

23 (46) 32 (29-36) 1614 (1153-2250)

>.999 .027 .002

674

4 (3-5) 6 (6-8) 3 (27) 6 (55) 2 (18) 3 (27) 1 (9) 8 (73)

6 (2-8) 8 (6-9) 13 (26) 23 (46) 14 (28) 22 (44) 14 (28) 33 (66)

.275 .249 .911* .500* .265* >.999*

1 (0-3) 0 (0-0) 0 (0-0) 18 (11-40)

1 (0-1) 0 (0-0) 0 (0-0) 23 (10-40)

.342 .839 .710 .829

13 (11-37) 6 (0-10) 0 (0-0)

12 (0-37) 0 (0-5) 4 (0-19)

.284 .005 .019

12 (7-17) 9 (7-17)

7.5 (0-16) 7 (0-14)

.461 .416

7 (4-8) 4 (0-5) 5 (0-16) 0 (0-0) 0 (0-0)

0 (0-6) 0 (0-3) 0 (0-10) 0 (0-0) 0 (0-0)

.014 .064 .099 .278 .735

6 (0-11) 0 (0-0) 3 (0-23) 0 (0-0) 0 (0-0)

1 (0-9) 0 (0-0) 2 (0-8) 0 (0-6) 7 (0-21)

.179 .504 .506 .160 .005

10 (4-23) 7 (0-18) 3 (0-13)

8 (0-26) 7 (0-19) 5 (0-21)

.639 .872 .521

0 (0-0) 0 (0-0) 0 (0) 0 (0-1) 0 (0-2)

0 (0-0) 0 (0-1) 3 (6) 0 (0-1) 0 (0-1)

.159 .039 >.999* .675 .855

2 (2-3) 0 (0-0) 0 (0-1) 0 (0-0) 8 (73)

1 (0-3) 0 (0-1) 0 (0-1) 0 (0-0) 15 (30)

.379 .040 .621 .639 .014*

0 (0-0) 0 (0-0)

0 (0-0) 0 (0-0)

.504 .913

2 (2-4) 4 (2-7) 2 (0-2) 0 (0-0) 0 (0-0) 0 (0-0) 0 (0-0)

2 (2-3) 3 (2-7) 0 (0-2) 0 (0-0) 0 (0-0) 0 (0-0) 0 (0-0)

.883 .615 .436 >.999 .639 .336 .435

0 (0-2) 0 (0-0)

2 (0-3) 0 (0-0)

.113 .336 (continued )

Cantey et al

ORIGINAL ARTICLES

September 2013

Table I. Continued Infants Variable

Infected/colonized (n = 11)

Noncolonized (n = 50)

0 (0-0), 9 0 (0-0), 3 0 (0-0), 7 0 (0-0), 7 11 (0-18), 21 0 (0-0), 21 0 (0-0), 0 0 (0-0), 0 0 (0-0), 0 0 (0-0), 4 0 (0-0), 14 11 (100) 39 (6-56) 100 (20-127) 2 (18) 2 (18)

0 (0-0), 14 0 (0-3), 21 0 (0-2), 15 0 (0-0), 14 0 (0-0), 16 0 (0-0), 21 0 (0-0), 21 0 (0-0), 10 0 (0-0), 13 0 (0-3), 13 0 (0-0), 14 45 (90) 8 (3-19) 41 (20-88) 0 (0) 1 (2)z

Days in room, median (IQR), maximum A B C D E† F G H I J Newborn nursery Housed in room with colonized infant at any time, n (%) Patient-days of exposure, median (IQR) NICU stay, days, median (IQR) Death due to infection, n (%) Death from any cause, n (%)

P value .442 .136 .513 .713 .002 .903 .159 .504 .159 .198 .986 .574* .009 .367 .030* .072*

*Fisher exact test. †Index room. zDeath secondary to Trisomy 18.

Results The index patient (infant 1; Table II; available at www.jpeds. com) was an 18-day-old infant born at 25 weeks’ gestation (birth weight, 790 g) who developed respiratory distress, hypotension, and metabolic acidosis on April 26. Aerobic cultures were obtained from a peripherally inserted central venous catheter and peripheral blood, urine, and endotracheal tube aspirate, and antibiotic therapy with oxacillin and gentamicin was initiated. On April 28, Gramnegative rods were identified from both blood cultures. Oxacillin was discontinued, and piperacillin-tazobactam was added. On April 28, 2 other infants in the same room exhibited signs of sepsis, one (infant 2) with apnea and the other (infant 3) with hypotension and thrombocytopenia. Peripheral and central venous catheter blood cultures and urine cultures were obtained in both infants, along with cerebrospinal fluid culture in infant 2. Both infants were started on oxacillin and gentamicin therapy. On April 29, ESBL-KP was identified from the index infant’s blood cultures obtained on April 26, and antibiotic therapy was changed to meropenem (minimum inhibitory concentration #1). ESBL-KP also was isolated from a second set of blood cultures obtained on April 28 during gentamicin therapy; however, a third set of blood cultures obtained after the addition of piperacillin/tazobactam but before meropenem were sterile. On April 30, ESBL-KP was identified from the urine culture obtained from infant 2; blood and cerebrospinal fluid cultures were sterile. Based on in vitro susceptibilities, amikacin therapy (minimum inhibitory concentration #16) was begun. By May 1, infant 1 was improving, but infants 2 and 3 had deteriorated clinically. Infant 2 developed fulminant necrotizing enterocolitis and expired; autopsy confirmed necrotizing

enterocolitis totalis. Infant 3 developed septic shock and expired despite meropenem therapy; initial cultures on April 28 were sterile, but repeat blood cultures obtained on May 1 yielded ESBL-KP. Surveillance cultures from all infants in the NICU performed on May 2 revealed that 3 of 71 infants (4%)were colonized with ESBL-KP. A second set of surveillance cultures obtained 1 week later (May 9) detected colonization in another 3 of 59 infants (5%). The final 2 colonized infants were identified from surveillance cultures of 60 infants (3%) performed on May 16 (Figure 2). On May 1, empiric antibacterial therapy for suspected sepsis was changed from oxacillin and gentamicin to meropenem for infants colonized with ESBL-KP or exposed to such infants. Infants 4 and 7, known to be colonized with ESBL-KP, later developed an ESBL-KP urinary tract infection on May 13 and May 16, respectively; both were treated successfully with 7 days of meropenem therapy. There were no subsequent deaths associated with ESBL-KP after May 1. All infected or colonized infants had a clonally identical or highly related strain of ESBL-KP. The outbreak strain harbored 2 b-lactamase genes, CTX-M 15 and TEM-1. In addition, a gene encoding KP carbapenemase-3 was detected in an isolate recovered from infant 3, but this isolate was susceptible to meropenem by in vitro testing (minimum inhibitory concentration #1). On the 16th day of the outbreak, a rectal culture from the father of the index infant revealed ESBL-KP that was highly related, but not identical, to the outbreak strain. Rectal cultures obtained from 3 other parents of colonized infants did not yield ESBL-KP. In addition, 10 ESBLKP isolates from adult patients hospitalized at PMH that were typed were not genetically related to the outbreak strain. Multiple interventions were implemented within 1 week of recognition of the outbreak (Table III; available at www. jpeds.com).

Prompt Control of an Outbreak Caused by Extended-Spectrum b-Lactamase–Producing Klebsiella pneumoniae in a Neonatal Intensive Care Unit

675

THE JOURNAL OF PEDIATRICS



www.jpeds.com

Vol. 163, No. 3

Figure 2. Timeline showing the number of infants screened for ESBL-KP and number of infected or colonized infants.

Hand Hygiene, Education, and Environmental Cleaning All NICU staff members were reeducated on proper hand hygiene. The charge nurse inspected the hands of all staff members at the beginning of each shift, and audits of hand hygiene practices were performed weekly by both infection prevention and NICU staff members. Existing policy of keeping nails trimmed to within 1/4 inch of the fingertip and prohibiting artificial nails was reinforced.16 Bracelets, watches, complex rings, and nail polish were prohibited.17,18 Educational information was developed and disseminated to all NICU HCP. Compliance with hand hygiene improved from 78%88% in February-April 2011 to 91%-100% in May-July 2011, and has been maintained at a median of 98% (IQR, 90%-100%) ever since the outbreak ended. Environmental services employees received further training on bed turnover and cleaning; weekly audits of cleaned rooms and patient areas were performed using a fluorescent tagging method (Glo Germ powder; Glo Germ, Moab, Utah).19 Any bed spaces deemed deficient prompted recleaning and additional training of the environmental services team. At the beginning of the outbreak, 1 isolette space was tagged with the Glo Germ and of 5 items tested, only 1 was adequately cleaned. Later during the outbreak, adequate cleaning improved to a median of 90% of tested items, and it has been maintained at a median of 92% since the outbreak ended. Contact Precautions, Staffing, Improved Space, and Cohorting Contact precautions were observed, with audits, for all ESBL-KP–infected/colonized infants and continued for the duration of the NICU stay. Staff assignments were reviewed daily, and the recommended number of infants assigned to nurses was maintained according to American Academy of Pediatrics perinatal guidelines.20 The NICU was not closed to new admissions, but rather newly admitted infants were housed in a room that had been vacated, had been thoroughly cleaned, and had multiple surface cultures negative for ESBL-KP. 676

Three rooms (5 beds per room) of 1 newborn nursery were converted to additional NICU space once cohorting began on May 2, in an effort to decrease overcrowding. In addition, the maximum number of infants allowed in each NICU room was decreased from 12 to 8, to provide more space between isolettes and bassinets (Table III). In the continuing-care nursery, the number of beds was halved, from 16 to 8 per room. These interventions increased the space around each infant in the high-acuity rooms from 77 ft2 per infant at the beginning of the outbreak to 111 ft2 per infant by the end, and specifically, from 69 ft2 per infant to 125 ft2 per infant in the outbreak room (Figure 1, rooms E and F). Infants were assigned to cohorts according to colonization and exposure status. Colonized infants were cohorted into 1 room, with dedicated nurses and respiratory therapists who had no contact with other infants during a work shift. Exposed infants were managed in the same way. For personnel who examined infants in all groups (eg, neonatologists, nurse practitioners, residents, consultants), unexposed infants were seen first, followed by exposed infants, and finally colonized infants. Medical rounds were conducted outside the bays housing exposed or colonized infants. No room-to-room transfers were permitted for either colonized or exposed infants for the duration of their stay. Cohort Study Results Sixty-two infants were present in the NICU on April 26. One infant was subsequently discharged before the first surveillance culture was performed and thus was excluded from our analysis, leaving a total of 61 infants in the cohort. On univariate analysis (Table I), infected or colonized infants were more likely than noncolonized infants to have been admitted to the index room (P = .002) and had significantly more patient-exposure days (P = .009). Infected or colonized infants also were more likely to be of lower gestational age and birth weight, to have been in a humidified isolette, to have had an umbilical line placed, and to have received surfactant. There was no cephalosporin use in the cohort. On multivariate analysis, only length of stay in the index room remained statistically significant (P = .009), although there Cantey et al

ORIGINAL ARTICLES

September 2013

Table IV. Relative risk of infection or colonization by total number of patient-days of exposure Patient-days of exposure

Uninfected

Infected

Proportion infected

Relative risk

0-5 6-28 29-42 >42

21 23 3 3

1 3 2 5

0.0455 0.1154 0.4000 0.6250

1.0 2.5 8.8 13.7

was a trend toward lower birth weight in the infected or colonized infants (P = .06). A dose-effect was observed with exposure days (Table IV).

Discussion It has been suggested that there is a small but consistent background rate of gram-negative MDROs that colonize infants in NICUs, but with the maintenance of optimal infection prevention and control practices, horizontal transmission is low and colonization is short-lived.21,22 In our NICU, ESBL-KP had not been identified previously from any clinical culture, although surveillance cultures had not been performed routinely before the outbreak. Although the source of the outbreak remains unknown, ESBL-producing organisms have been identified in colonizing flora of the gastrointestinal tract of healthy individuals with no hospital exposure.23,24 It is possible that as the prevalence of MDROs increases in the community, the likelihood of its introduction into a NICU will increase as well. The father of the index infant had rectal colonization with a strain related to the outbreak strain, but the culture was not obtained until 16 days into the outbreak, and so it is possible that he acquired it from the index infant. The primary risk factor identified for ESBL-KP colonization or infection in our study cohort was days spent in the room in which the index case was housed when the ESBLKP infection was identified. This room is used for the highest-acuity infants. Seven of the 11 infected or colonized infants (64%) had been admitted to this particular room, and the other 4 infants had been housed in a room with 1 or more of the 7 colonized infants before the outbreak was identified. A trend toward lower birth weight as a risk factor for acquisition of ESBL-KP also was seen, as has been reported previously.25,26 Seven of the 11 (64%) ESBL-KP–colonized infants in our cohort had a birth weight <1000 g, and many of these infants were cared for in the high-acuity room in which the index infant had been housed. This clustering might have contributed to horizontal transmission of ESBL-KP, given that a high density of low birth weight infants in a room is a well-established risk factor for NICU outbreaks.27,28 Overcrowding can facilitate NICU outbreaks with various pathogens, including methicillin-resistant Staphylococcus aureus, Klebsiella species, and other Enterobacteriaceae.27,29-31 Efforts to alleviate crowding by decreasing the number of infants in each room and opening new NICU rooms in the former newborn nursery area likely

contributed to the successful eradication of the outbreak strain from the NICU. Despite these efforts, the available square footage per infant did not reach the American Academy of Pediatrics–recommended 150 ft2 for infants in intensive care owing to the space limitations of our NICU.32 A new PMH building with a NICU that meet this requirement is under construction, with completion anticipated in early 2015. Importantly, the outbreak occurred despite ongoing antibiotic stewardship efforts that minimized the use of thirdgeneration cephalosporin and piperacillin/tazobactam antimicrobial agents; no infants in our cohort had received the former and only 1 had received the latter during hospitalization. Other previously reported predictors of ESBL-KP colonization, including days of mechanical ventilation,33 total parenteral nutrition,30 length of stay,34 and use of central venous catheters,35 were not identified as independent risk factors in our cohort. The absence of other significant associations on logistic regression analysis reinforces the importance of horizontal transmission of ESBL-KP within the NICU, and attests to the effectiveness of cohorting infants and staff. The outbreak was controlled successfully within 18 days of implementing interventions (Table III) developed by a multidisciplinary team. The multidisciplinary team approach included frequent meetings to review interval changes and allowed for open communication and implementation of recommended interventions suggested by the infection prevention specialists. Though not specifically tested in this outbreak, the importance of the multidisciplinary team has been demonstrated previously both during outbreaks and in ongoing prevention.8,36 Participation of senior administrators facilitated authorization of environmental changes needed, avoiding unnecessary bureaucratic delays. In addition, the long-term clinical and economic benefits of maintaining “outbreak-level” practices have been well described.37 Notably, the outbreak was controlled, and the ESBL-KP was eradicated from the NICU without the need to resort to extreme measures, such as closing the nursery to new admissions or performing cultures of HCP. The separation of infants into 3 distinct and geographically separate cohorts was done within days of recognition of the outbreak, with newly admitted infants housed in separate rooms from the exposed and infected infants. Horizontal transmission of MDROs on the hands of HCPs has been described in several outbreaks,4,38,39 and specific indications for performing hand cultures have been suggested,40 but the role of HCP hand cultures remains controversial. Considering that the risk factors for transmission of outbreak strains by hands of HCP include failure to perform appropriate hand hygiene, use of artificial nails, chipped nail polish, and chronic dermatitis, we instead focused on education on proper hand hygiene, with audits and visual inspection of hands of each staff member by the charge nurse at the beginning of each shift and prohibition of nail polish. Given that introduction of MDROs into NICUs will likely continue, future infection prevention efforts

Prompt Control of an Outbreak Caused by Extended-Spectrum b-Lactamase–Producing Klebsiella pneumoniae in a Neonatal Intensive Care Unit

677

THE JOURNAL OF PEDIATRICS



www.jpeds.com

should focus on strict hygiene procedures, with audits and feedback and active surveillance with the goal of eliminating horizontal transmission. n We thank Brian Wickes, PhD (Department of Microbiology and Immunology, University of Texas Health Science Center at San Antonio) for help with the molecular analysis of the b-lactamase enzymes of the outbreak strain. We also thank all staff members of the NICU, as well as members of the multidisciplinary team, whose efforts were invaluable in the successful management of the outbreak: Thomas Button, RNC, BSN, NE-BC, CIC (Director, Infection Prevention); Glenn Metoyer, BS, RRT, RCP (Manager, NICU Respiratory Care); Edward Best, MBA, RRT, RCP (Director, Respiratory Care); Linda Byrd, MT(ASCP)SM and the microbiology laboratory; Linda Castleman, RN, Monique Barksdale, RN, BSN, and Molly Curtis, RN (Associate Unit Managers, NICU); Jared Furner (Inpatient Director, Environmental Services); Isaac Egharevba (Associate Director, Environmental Services); Pamela Ford, RN (Unit Manager, Newborn Nursery); Gregory Jackson, MD (Clinical Director, Newborn Nursery); Dale Kemp, LVN (Infection Prevention); Paula Turicchi (Vice-President, Women and Infants’ Specialty Health); Joyce Thompson, RN (Patient Safety and Risk); Anne E. Tudhope, RNC, BSN, MS (Director, Maternal and Infants’ Specialty Health); and Neil Pascoe, RN, BSN, CIC (Epidemiologist, Texas Department of State Health Services). Submitted for publication Aug 1, 2012; last revision received Jan 25, 2013; accepted Mar 1, 2013. nchez, MD, Department of Pediatrics, University Reprint requests: Pablo J. Sa of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390-9063. E-mail: [email protected]

Vol. 163, No. 3

10.

11.

12.

13.

14.

15.

16. 17. 18.

References 19. 1. Mahieu LM, Buitenweg N, Beutels P, De Dooy JJ. Additional hospital stay and charges due to hospital-acquired infections in a neonatal intensive care unit. J Hosp Infect 2001;47:223-9. 2. Maragakis LL, Winkler A, Tucker MG, Cosgrove SE, Ross T, Lawson E, et al. Outbreak of multidrug-resistant Serratia marcescens infection in a neonatal intensive care unit. Infect Control Hosp Epidemiol 2008;29:418-23. 3. Tamma PD, Savard P, Pal T, Sonnevend A, Perl TM, Milstone AM. An outbreak of extended-spectrum b-lacatamase producing Klebseilla pneumoniae in a neonatal intensive care unit. Infect Control Hosp Epidemiol 2012;33:631-4. 4. Cassettari VC, Silveira IR, Balsamo AC, Franco F. Outbreak of extendedspectrum b-lactamase–producing Klebsiella pneumoniae in an intermediate-risk neonatal unit linked to onychomycosis in a healthcare worker. J Pediatr (Rio J) 2006;82:313-6. 5. de Oliveira Garcia D, Doi Y, Szabo D, Adams-Haduch JM, Vaz TM, Leite D, et al. Multiclonal outbreak of Klebsiella pneumoniae–producing extended-spectrum b-lactamase CTX-M-2 and novel variant CTX-M-59 in a neonatal intensive care unit in Brazil. Antimicrob Agents Chemother 2008;52:1790-3. 6. Stone PW, Gupta A, Loughrey M, Della-Latta P, Cimiotti J, Larson E, et al. Attributable costs and length of stay of an extended-spectrum blactamase–producing Klebsiella pneumoniae outbreak in a neonatal intensive care unit. Infect Control Hosp Epidemiol 2003;24:601-6. 7. Hosoglu S, Hascuhadar M, Yasar E, Uslu S, Aldudak B. Control of an Acinetobacter baumannii outbreak in a neonatal ICU without suspension of service: a devastating outbreak in Diyarbakir, Turkey. Infection 2012; 40:11-8. 8. Ransjo U, Lytsy B, Melhus A, Aspevall O, Artinger C, Eriksson BM, et al. Hospital outbreak control requires joint efforts from hospital management, microbiology and infection control. J Hosp Infect 2010;76:26-31. 9. van Ogtrop ML, van Zoeren-Grobben D, Verbakel-Salomons EM, van Boven CP. Serratia marcescens infections in neonatal departments: 678

20. 21.

22.

23.

24.

25.

26.

27.

28.

description of an outbreak and review of the literature. J Hosp Infect 1997;36:95-103. Zawacki A, O’Rourke E, Potter-Bynoe G, Macone A, Harbarth S, Goldmann D. An outbreak of Pseudomonas aeruginosa pneumonia and bloodstream infection associated with intermittent otitis externa in a healthcare worker. Infect Control Hosp Epidemiol 2004;25: 1083-9. Siegel JD, Rhinehart E, Jackson M, Chiarello L, Healthcare Infection Control Practices Advisory Committee. Management of multidrugresistant organisms in health care settings, 2006. Am J Infect Control 2007;35:S165-93. Brolund A, Haeggman S, Edquist PJ, Gezelius L, Olsson-Liljequist B, Wisell KT, et al. The DiversiLab system versus pulsed-field gel electrophoresis: characterisation of extended-spectrum b-lactamase–producing Escherichia coli and Klebsiella pneumoniae. J Microbiol Methods 2010;83: 224-30. Hansen DS, Skov R, Benedi JV, Sperling V, Kolmos HJ. Klebsiella typing: pulsed-field gel electrophoresis (PFGE) in comparison with O: K-serotyping. Clin Microbiol Infect 2002;8:397-404. Rasheed JK, Biddle JW, Anderson KF, Washer L, Chenoweth C, Perrin J, et al. Detection of the Klebsiella pneumoniae carbapenemase type 2 carbapenem-hydrolyzing enzyme in clinical isolates of Citrobacter freundii and K. oxytoca carrying a common plasmid. J Clin Microbiol 2008;46: 2066-9. Lewis JS II, Herrera M, Wickes B, Patterson JE, Jorgensen JH. First report of the emergence of CTX-M type extended-spectrum b-lactamases (ESBLs) as the predominant ESBL isolated in a US health care system. Antimicrob Agents Chemother 2007;51:4015-21. Gupta A. Hospital-acquired infections in the neonatal intensive care unit: Klebsiella pneumoniae. Semin Perinatol 2002;26:340-5. Ward DJ. Hand adornment and infection control. Br J Nurs 2007;16: 654-6. Jepson AP, McDougall C, Clark A, Bateman A, Williamson G, Kaufmann ME. Finger rings should be removed prior to scrubbing. J Hosp Infect 2006;64:197-8. Carling PC, Briggs JL, Perkins J, Highlander D. Improved cleaning of patient rooms using a new targeting method. Clin Infect Dis 2006;42: 385-8. Polin RA, Denson S, Brady MT. Strategies for prevention of health careassociated infections in the NICU. Pediatrics 2012;129:e1085-93. Toltzis P, Dul MJ, Hoyen C, Salvator A, Walsh M, Zetts L, et al. Molecular epidemiology of antibiotic-resistant gram-negative bacilli in a neonatal intensive care unit during a nonoutbreak period. Pediatrics 2001; 108:1143-8. Mammina C, Di Carlo P, Cipolla D, Giuffre M, Casuccio A, Di Gaetano V, et al. Surveillance of multidrug-resistant gram-negative bacilli in a neonatal intensive care unit: prominent role of cross-transmission. Am J Infect Control 2007;35:222-30. Abdul Rahman EM, El-Sherif RH. High rates of intestinal colonization with extended-spectrum lactamase-producing Enterobacteriaceae among healthy individuals. J Investig Med 2011;59:1284-6. Qin X, Zerr DM, Weissman SJ, Englund JA, Denno DM, Klein EJ, et al. Prevalence and mechanisms of broad-spectrum b-lactam resistance in Enterobacteriaceae: a children’s hospital experience. Antimicrob Agents Chemother 2008;52:3909-14. Pessoa-Silva CL, Meurer Moreira B, Camara Almeida V, Flannery B, Almeida Lins MC, Mello Sampaio JL, et al. Extended-spectrum b-lactamase–producing Klebsiella pneumoniae in a neonatal intensive care unit: risk factors for infection and colonization. J Hosp Infect 2003;53: 198-206. Abdel-Hady H, Hawas S, El-Daker M, El-Kady R. Extended-spectrum blactamase–producing Klebsiella pneumoniae in neonatal intensive care unit. J Perinatol 2008;28:685-90. Eckmanns T, Rath A, Ruden H, Gastmeier P, Daschner F. Outbreak of Enterobacter cloacae related to understaffing, overcrowding, and poor hygiene practices. Infect Control Hosp Epidemiol 2000;21:305-7. Sarvikivi E, Lyytikainen O, Salmenlinna S, Vuopio-Varkila J, Luukkainen P, Tarkka E, et al. Clustering of Serratia marcescens

Cantey et al

ORIGINAL ARTICLES

September 2013

29.

30.

31.

32. 33.

34.

35.

infections in a neonatal intensive care unit. Infect Control Hosp Epidemiol 2004;25:723-9. Dalben M, Varkulja G, Basso M, Krebs VL, Gibelli MA, van der Heijden I, et al. Investigation of an outbreak of Enterobacter cloacae in a neonatal unit and review of the literature. J Hosp Infect 2008;70:7-14. Huang Y, Zhuang S, Du M. Risk factors of nosocomial infection with extended-spectrum b-lactamase–producing bacteria in a neonatal intensive care unit in China. Infection 2007;35:339-45. Haley RW, Cushion NB, Tenover FC, Bannerman TL, Dryer D, Ross J, et al. Eradication of endemic methicillin-resistant Staphylococcus aureus infections from a neonatal intensive care unit. J Infect Dis 1995;171:614-24. Lockwood CJ, Lemons JA, eds. Guidelines for perinatal care. 6th ed. Elk Grove Village (IL): American Academy of Pediatrics; 2007. Boo NY, Ng SF, Lim VK. A case-control study of risk factors associated with rectal colonization of extended-spectrum b-lactamase–producing Klebsiella sp in newborn infants. J Hosp Infect 2005;61:68-74. Crivaro V, Bagattini M, Salza MF, Raimondi F, Rossano F, Triassi M, et al. Risk factors for extended-spectrum b-lactamase–producing Serratia marcescens and Klebsiella pneumoniae acquisition in a neonatal intensive care unit. J Hosp Infect 2007;67:135-41. Kuo KC, Shen YH, Hwang KP. Clinical implications and risk factors of extended-spectrum b-lactamase–producing Klebsiella pneumoniae infec-

36.

37.

38.

39.

40.

tion in children: a case-control retrospective study in a medical center in southern Taiwan. J Microbiol Immunol Infect 2007;40:248-54. Enoch DA, Summers C, Brown NM, Moore L, Gillham MI, Burnstein RM, et al. Investigation and management of an outbreak of multidrug carbapenem-resistant Acinetobacter baumannii in Cambridge, UK. J Hosp Infect 2008;70:109-18. Piednoir E, Thibon P, Borderan GC, Godde F, Borgey F, Le Coutour X, et al. Long-term clinical and economic benefits associated with the management of a nosocomial outbreak resulting from extended-spectrum blactamase–producing Klebsiella pneumoniae. Crit Care Med 2011;39: 2672-7. de Vries JJ, Baas WH, van der Ploeg K, Heesink A, Degener JE, Arends JP. Outbreak of Serratia marcescens colonization and infection traced to a healthcare worker with long-term carriage on the hands. Infect Control Hosp Epidemiol 2006;27:1153-8. Prospero E, Barbadoro P, Esposto E, Manso E, Martini E, Savini S, et al. Extended-spectrum b-lactamase Klebsiella pneumoniae: multimodal infection control program in intensive care units. J Prev Med Hyg 2010;51: 110-5. Grant PS, Charns LG, Rawot BW, Benedetti SG. Consideration to culture health care workers related to increased methicillin-resistant Staphylococcus aureus activity in a neonatal intensive care unit. Am J Infect Control 2008;36:638-43.

50 Years Ago in THE JOURNAL OF PEDIATRICS Sudden and Unexpected Death in Infants: II. Viral Infections as Causative Factors Valdes-Dapena MA, Hummeler K. J Pediatr 1963;63:398-401

T

his article reports a series of investigations led by pathologist Marie A. (Molly) Valdes-Dapena into the possible causes of sudden death in 109 Philadelphia infants on whom she performed autopsies, focusing on evaluation for viral infections. The investigative methods were extraordinarily rigorous. Tissue samples from heart, both lungs, spleen, kidney, liver, colon, larynx, and brain were inoculated intracerebrally into adult mice, intraperitoneally into suckling mice, into the amniotic fluid of chick embryos, and into tissue cultures of HeLa cells and monkey kidney cells. Only a single patient had a virus isolated—a Coxsackie B virus from the lung. The findings were pivotal in turning investigations away from infectious diseases as an important cause of sudden, unexpected death. This was correct. See Valdes-Dapena et al1 for the first installment from these studies, as well as personal insights into Dr Valdes-Dapena’s career. Sarah S. Long, MD Department of Pediatrics St Christopher’s Hospital for Children Philadelphia, Pennsylvania http://dx.doi.org/10.1016/j.jpeds.2013.03.050

Reference 1. Valdes-Dapena MA, Eichman MF, Ziskin L. Sudden and unexpected death in infants: I. gamma globulin levels in the serum. J Pediatr 1963;63:290-4.

Prompt Control of an Outbreak Caused by Extended-Spectrum b-Lactamase–Producing Klebsiella pneumoniae in a Neonatal Intensive Care Unit

679

THE JOURNAL OF PEDIATRICS



www.jpeds.com

Vol. 163, No. 3

Figure 1. Floor plan of the NICU, with letters denoting separate infant bays (“rooms”). *E was the room of the index patient. The maximum number of beds per room was reduced to prevent crowding: rooms A and B, 12 to 9 beds; room C, 12 to 8 step-down beds or 5 intensive care beds; room D, 6 to 5 beds; room E, 10 beds before and after; room F, 9 to 6 beds; room G, 8 to 6 beds; and room H, 1 separate single bed.

679.e1

Cantey et al

September 2013

Infant

Age at time of Birth Weight at time of culture, days weight, g culture, g

1

18

790

940

2 3

24 11

1090 740

1150 850

4

13 25 8 11 12 19 74 21 37 58

690

750 1130 730 1730 1310 1540 2450 2610 1350 1740

5 6 7 8 9 10 11

780 1600 1224 950 2072 650 730

Signs when first identified Acidosis, hypotension, respiratory distress Apnea Acidosis, hypotension, thrombocytopenia None Respiratory distress None None None Apnea None None None None

Infection Date of positive (type)/colonization culture

Initial site(s) of ESBL-KP culture

Antibiotic therapy

Infection (CLABSI)

4/26/11

Infection (UTI, NEC) Infection (CLABSI)

4/28/11 5/1/11

Blood, endotracheal Gentamicin/piperacillin/tazobactam/meropenem aspirate Urine Gentamicin/amikacin/meropenem Blood Gentamicin/meropenem

Colonization Infection (UTI)* Colonization Colonization Colonization Infection (UTI)* Colonization Colonization Colonization Colonization

5/1/11 5/13/11 5/1/11 5/2/11 5/9/11 5/16/11 5/9/11 5/9/11 5/16/11 5/16/11

Rectum Urine Rectum Rectum Rectum Urine Rectum Rectum Rectum Rectum

None Meropenem None None None Meropenem None None None None

Status at hospital discharge Alive Deceased Deceased Alive Alive Alive Alive Alive Alive Alive Alive

CLABSI, central line–associated bloodstream infection; NEC, necrotizing enterocolitis; UTI, urinary tract infection. *Known to be colonized with ESBL-KP before infection.

679.e2

ORIGINAL ARTICLES

Prompt Control of an Outbreak Caused by Extended-Spectrum b-Lactamase–Producing Klebsiella pneumoniae in a Neonatal Intensive Care Unit

Table II. Characteristics of the 11 infants with ESBL-KP infection or colonization during the outbreak, April-May 2011

THE JOURNAL OF PEDIATRICS



www.jpeds.com

Vol. 163, No. 3

Table III. Interventions implemented after identification of outbreak caused by ESBL-KP in the NICU Intervention Hand hygeine All staff members reeducated on proper hand hygiene Hand audits performed weekly by nurse managers and infection preventionists Existing policy of keeping fingernails trimmed to 1/4 inch reinforced Artificial nails prohibited Bracelets, watches, complex rings, and nail polish prohibited Visual inspection of all hands at beginning of shift by charge nurse Environmental services Environmental services received additional training on bed space cleaning Weekly audits performed on bed spaces using fluorescein powder Failure of audit prompted repeat cleaning and further training Three cohorts created: colonized, noncolonized, and new admissions Contact isolation for all colonized or exposed infants Staff cohorted; personnel caring for all groups (neonatologists, consultants) saw unexposed first, then exposed, then colonized No room-to-room transfers permitted for colonized or exposed infants Crowding alleviated No computers on wheels allowed in rooms of colonized infants Only the minimum necessary number of examiners allowed inside rooms of colonized infants New bed spaces opened in unused newborn nursery unit Clean equipment no longer stored in rooms that contained patients Focus on communication Timely and frequent communication between members of multidisciplinary team Disclosure of outbreak and discussion of control measures with all parents by medical team Frequent discussion and updates with parents of infants and hospital administrators

Level of evidence* I-B I-B II-B I-B II-B III-B II-B II-B II-A I-A II-A III-B II-B II-A II-A III-B II-A Regulatory III-A

*US Public Health Service grading system (strength of recommendation: A, good evidence to support recommendation; B, moderate evidence to support a recommendation; C, poor evidence to support a recommendation; Quality of evidence: I, evidence from 1 or more properly randomized, controlled trial(s); II, evidence from 1 or more well-designed clinical trial(s) without randomization, from cohort or case-controlled analytic studies [preferably from more than 1 center], from multiple time series, or from dramatic results from uncontrolled experiments; III, evidence from opinions of respected authorities, based on clinical experience, descriptive studies, or reports of expert committees).

679.e3

Cantey et al