MAJOR ARTICLE

Impact of Enhanced Infection Control at 2 Neonatal

Intensive Care Units in The Philippines
Christopher J. Gill,1,2 Jose B. V. Mantaring,5 William B. Macleod,1 Myrna Mendoza,6 Sookee Mendoza,2
W. Charles Huskins,4 Donald A. Goldmann,3 and Davidson H. Hamer1,2
1

Center for International Health and Development, Department of International Health, Boston University School of Public Health, 2Infectious
Diseases Section, Department of Internal Medicine, Boston University School of Medicine, and 3Department of Pediatrics, Childrens Hospital
Boston, Boston, Massachusetts; 4College of Medicine, Mayo Clinic, Rochester, Minnesota; and 5Newborn Medicine Section and 6Department of
Medicine, Philippines General Hospital, and 7Department of Neonatology, Jose Fabella Memorial Hospital, Manila, The Philippines

Background. The growing burden of neonatal mortality associated with hospital-acquired neonatal sepsis in
the developing world creates an urgent need for cost-effective infection-control measures in resource-limited
settings.
Methods. Using a before-and-after comparison design, we measured how rates of staff hand-hygiene compliance, colonization with drug-resistant pathogens (defined as ceftazidime- and/or gentamicin-resistant gramnegative bacilli and drug-resistant gram-positive cocci), bacteremia, and overall mortality changed after the introduction of a simplified package of infection-control measures at 2 neonatal intensive care units (NICUs) in Manila,
The Philippines.
Results. Of all 1827 neonates admitted to the NICU, 561 (30.7%) arrived from delivery already colonized with
drug-resistant bacteria. Of the 1266 neonates who were not already colonized, 578 (45.6%) became newly colonized
with drug-resistant bacteria. Of all 1827 neonates, 358 (19.6%) became bacteremic (78.2% were infected with
gram-negative bacilli) and 615 (33.7%) died. Of 2903 identified drug-resistant colonizing bacteria, 85% were drugresistant gram-negative bacilli (predominantly Klebsiella species, Pseudomonas species, and Acinetobacter species)
and 14% were methicillin-resistant Staphylococcus aureus. Contrasting the control period with the intervention
period at each NICU revealed that staff hand-hygiene compliance improved (NICU 1: relative risk, 1.3; 95%
confidence interval 1.11.5; NICU 2: relative risk, 1.6; 95% confidence interval, 1.42.0) and that overall mortality
decreased (NICU 1: relative risk, 0.5; 95% confidence interval, 0.40.6; NICU 2: relative risk, 0.8; 95% confidence
interval, 0.70.9). However, rates of colonization with drug-resistant pathogens and of sepsis did not change
significantly at either NICU.
Discussion. Nosocomial transmission of drug-resistant pathogens was intense at these 2 NICUs in The Philippines; transmission involved mostly drug-resistant gram-negative bacilli. Infection-control interventions are feasible and are possibly effective in resource-limited hospital settings.
Nosocomial infections contribute a growing, yet underappreciated, share of neonatal mortality in the developing world [1]. Poor hand-hygiene practices [2, 3],
reuse of single-use medication vials and devices [4],
and inadequate sterilization of medical equipment [5]
are key proximate events that facilitate transmission of

Received 26 May 2008; accepted 8 September 2008; electronically published

24 November 2008.
Reprints or correspondence: Dr. Christopher J. Gill, Center for International
Health and Development, Boston University School of Public Health, 710 Albany
St., Boston, MA 02118 (cgill@bu.edu).
Clinical Infectious Diseases 2009; 48:1321
 2008 by the Infectious Diseases Society of America. All rights reserved.
1058-4838/2009/4801-0002$15.00
DOI: 10.1086/594120

nosocomial pathogens. Institutional factors, especially

inadequate resources to fund infection-control programs, also contribute to such transmission. Overuse
of empirical antibiotic therapy is simultaneously a consequence of and a contributing factor to transmission
[6]. A consequence of these factors has been the emergence of infection due to highly pathogenic multidrugresistant gram-negative bacilli and methicillin-resistant
Staphylococcus aureus (MRSA) in neonatal intensive
care units (NICUs) in developing countries, particularly
in South and Southeast Asia [710]. Simple, effective,
and affordable infection-control interventions that are
appropriate to settings in developing countries are urgently needed.
Enhanced Infection Control CID 2009:48 (1 January) 13

Downloaded from http://cid.oxfordjournals.org/ by guest on September 30, 2016

(See the editorial commentary by Apisarnthanarak and Fraser on pages 224)

In this context, we evaluated the effectiveness of a package

of infection-control interventions at 2 NICUs in Manila, The
Philippines; these interventions were selected because of their
proven efficacy in the developed world and their perceived
feasibility in the developing world [1]. Our primary outcome
was the rate of new colonization with drug-resistant bacteria
among neonates. Secondary outcomes included rates of bacteremia, mortality, and staff hand-hygiene compliance.
METHODS

Table 1. Baseline characteristics of 2 level-III neonatal intensive care units (NICUs) in Manila, The Philippines.
NICU
Characteristic

7000
825

35,000
900

25
Inborn patients only
Combined level II and III
80 (25 level III)

60
Inborn patients only
Combined level II and III
60

Multiple babies per basinet

Nurse-to-patient ratio

No
1:5

Yes (up to 3 per basinet)

1:13

Reuse of single-use vials

No. of functioning sinks in unit

Yes
7

Annual no. of deliveries at the hospital

Annual no. of level-III NICU admissions
Level-III NICU mortality during the previous year, %
Source of NICU admissions
Organization of unit
Maximum bed capacity

Hand-drying system

Yes
2 (with 1 additional sink located
outside the unit)

Ethanol handwash available in hospital

Cloth towels with irregular

replacement of rolls
No

Cloth towels with irregular

replacement of rolls
Yes (but seldom used)

Latex glove supply

Admixture of intravenous medications

Inconsistent
Performed in pharmacy

Inconsistent
Performed at bedside

First-line antibiotic regimen

Piperacillin-tazobactam plus
amikacin
Ciprofloxacin plus meropenem

Penicillin plus gentamicin

On-site soap and water, then

soaked in 2% glutaraldehyde
and rinsed in sterile water
Formula (mixed under laminar
flow hood)

On-site method unspecified

Common second-line antibiotic regimen

Reprocessing of ventilator equipment

Neonatal feedings

14 CID 2009:48 (1 January) Gill et al.

Second- and third-generation

cephalosporins, amikacin, and
carbapenems

Maternal breast milk only

Downloaded from http://cid.oxfordjournals.org/ by guest on September 30, 2016

Study setting. We enrolled all neonates who were admitted

to either NICU during the study period. The study was conducted at 2 inner-city, level-3 NICUs in Manila. The key baseline characteristics of the units are summarized in table 1. NICU
1 is located in the Philippines General Hospital, the largest
public teaching hospital in The Philippines. NICU 2 is located
in the Jose Fabella Memorial Hospital, a busy (100 deliveries
per day) obstetrical and gynecological charity hospital. The
protocol was approved by the Philippines General Hospital and
Boston University Medical Center institutional review boards.
Both NICUs accept only infants who are delivered at the respective hospitals.
Study design. The study used a quasi-experimental before-

and-after design. After a preparatory phase, we introduced the

interventions in staggered fashion, as shown in figure 1.
Preparatory phase. Over 2 months, we standardized indications and procedures for blood cultures and microbiology
laboratory procedures and trained our research assistants. A
baseline survey generated site-specific guidance for finalization
of the infection-control checklist.
Phase I: baseline surveillance. Our research assistants conducted 4 forms of surveillance. First, for each neonate, the
assistants tracked the duration of NICU stay, use of vascular
catheters and/or ventilators, dates and types of infection, and
antibiotic use. Second, they conducted regular hand-hygiene
compliance surveys. Third, they performed serial surveillance
cultures for all neonates. Fourth, they documented all blood
culture results.
Phase II: infection-control interventions. First, ethanol
handrub at each basinet was prepared using locally purchased
70% ethanol mixed with glycerin as an emollient. Second, preparatory workshops for key staff at each unit were implemented
at the onset of phase II. These workshops included lectures on
hand hygiene and infection control, interactive case discussions,
and a collaborative critique of a video of patient-staff interactions and infection-control activities at 1 of the NICUs. Third,

Table 2. Components of the daily infection-control checklist.

Checklist
Are the ethanol handwash dispensers stocked at each basinet?
For neonates with intravascular catheters, are catheter sites free
of visible infection and contamination? Are catheters free of mechanical defects? Is each catheter still required?
Can any urinary catheters be discontinued?
For neonates receiving mechanical ventilation, is ventilation required? Is there a weaning plan?
Are all multiuse medication vials and solutions being stored
appropriately?
For neonates receiving empirical antibiotic therapy, have 72 h
elapsed without a positive culture result? If so, can the drugs
be stopped?
For neonates being treated for a infection due to a defined pathogen, does the antibiotic-resistance profile accommodate the current regimen?
For neonates receiving parenteral nutrition, can this be changed to
enteral nutrition?
For neonates receiving intravenous lipids, can caloric requirements
be met without lipids? If so, is there a stop order?
Are appropriate contact precautions being used for all patients colonized or infected with a drug-resistant pathogen?

days thereafter, and on the day of hospital discharge. The same

schedule was used for screening for MRSA, omitting the day0 cultures. The reason for gram-negative bacilli surveillance on
day 0 was to determine the proportion of drug-resistant gramnegative bacilli colonization that occurred before NICU admission. Screening for drug-resistant gram-negative bacilli was
performed with using MacConkey agar plates impregnated with
8 mg/mL of gentamicin or 2 mg/mL of ceftazidime. Screening
to detect VRE was performed using bile-esculin agar plates with
6 mg/mL of vancomycin. Screening for MRSA was performed
using Mueller-Hinton agar plates containing 6 mg/mL of oxacillin. Plates were incubated for 24 h at 35C.
Blood culture. One milliliter of blood was inoculated in

Figure 1. Study schematic and timeline at 2 neonatal intensive care units (NICUs) in Manila, The Philippines
Enhanced Infection Control CID 2009:48 (1 January) 15

Downloaded from http://cid.oxfordjournals.org/ by guest on September 30, 2016

daily infection-control checklists were used. Implementation of

the daily checklist was the responsibility of the NICU teams.
Because of the large number of checklist items, only 1 item
was used per day. The item was selected by listing each item
on a note card and having the team pull a different card each
day (table 2). Fourth, monthly infection-control checklists were
used. The NICU attending physicians were responsible for implementation of these checklists (table 3).
Hand-hygiene assessments. No effort was made to blind
NICU staff to the purpose of the observations. One-hour observations were performed intermittently during the day or
night shift at a 2:1 ratio. A neonate was chosen, and then all
hygiene encounters for that patient and the adjacent 2 neonates
were monitored (3 neonates per observer period) [11].
Laboratory methods. Quality-control procedures were established using standard American Type Culture Collection
strains (George Mason University Research Laboratory). Gramnegative bacilli were categorized as enteric or nonenteric. Enteric bacteria were defined as bacteria for which the primary
reservoir is the human gut, such as Escherichia coli. We defined
nonenteric bacteria as bacteria that normally exist in the environment and only occasionally colonize the human gut, such
as Pseudomonas aeruginosa.
Surveillance cultures. All neonates underwent screening
for drug-resistant gram-negative bacilli and gram-positive cocci
according to a standard schedule. Drug-resistant gram-negative
bacilli were bacilli that were resistant to gentamicin and/or
ceftazidime, and drug-resistant gram-positive cocci were MRSA
or vancomycin-resistant enterococci (VRE). Culture of perianal
swab and/or stool specimens were used for surveillance of
gram-negative bacilli and VRE, and umbilical and anterior nare
swab specimens (Culturette tubes, Becton Dickinson) were used
for surveillance of MRSA. Surveillance for drug-resistant gramnegative bacilli commenced within 16 h after admission to an
NICU (day 0) and was performed on days 2 and 7, every 7

Table 3.

Components of the monthly infection-control checklist.

Checklist
Are ethanol handwash supplies sufficient?
Are clean hand towels stocked at each sink?
Are intravascular admixture solutions being prepared in a clean
environment or in the central pharmacy?
Are appropriate decontamination procedures and solutions being
applied to all reusable devices, such as ventilator tubing and
instruments?
Is reusable sterilized equipment completely dry before storage?
Are disinfectants being used according to the manufacturers
instructions?
Do neonatal intensive care unit personnel go from patient to
patient without changing their gowns?

mitted), incidence density of bacteremia (number of bloodstream infections per patient-days), cumulative mortality
(number of NICU deaths per number of NICU admissions),
and hand-hygiene compliance before and after implementation
of the infection-control program (number of observed contacts
between a health care provider and a neonate when appropriate
cleansing with soap and water or ethanol handrub was performed per the total number of contacts).
The sample size was determined for our primary outcome.
Assuming 10% baseline colonization, we estimated that 865
neonates per hospital would be needed to detect a 4% difference
in colonization rate.
RESULTS

16 CID 2009:48 (1 January) Gill et al.

Patient characteristics. From May 2003 through July 2004,

a total of 1827 neonates were admitted to 2 NICUs (925 to
NICU 1 and 903 to NICU 2); 768 (83.0%) of neonates who
were admitted to NICU 1 and 632 (70.0%) who were admitted
to NICU 2 were admitted directly after delivery. The mean
gestational age was 34.7 weeks (1139 [62.3%] were born before
36 weeks of gestation), and the mean birth weight was 2085.7
g (1249 [68.3%] weighed !2500 g, and 497 [27.2%] weighed
!1500 g); 1771 (96.9%) of neonates received antibiotics, 837
(45.8%) required central vascular catheters, and 1126 (61.6%)
required mechanical ventilation.
Colonizations. As shown in table 4, most of the isolates
confirmed by our colonization surveys were drug-resistant
gram-negative bacilli. Among 8986 colonized swab specimens,
2903 drug-resistant bacterial isolates were detected; 2476
(85.3%) of these isolates were drug-resistant gram-negative bacilli, and 427 (14.7%) were drug-resistant gram-positive cocci
(table 4). Moreover, all 2476 drug-resistant gram-negative bacillus isolates were, a priori, detected in only the 4055 rectal
swab and/or stool specimens; the crude positivity ratio was
61%. Many neonates (30.7%) had a stool and/or rectal specimen positive for drug-resistant gram-negative bacilli by culture
on the date of NICU admission, which implied exposure during
labor and delivery. Of the neonates who were not already colonized on NICU days 01, 45.6% became colonized with a
drug-resistant gram-negative bacillus during hospitalization in
the NICU. We identified 823 new colonizations with a gentamicin-resistant gram-negative bacillus and 771 new colonizations with ceftazidime-resistant gram-negative bacillus; 763
(92.7%) of the isolates were resistant to both drugs. Only 60
(7.3%) were resistant to gentamicin but not to ceftazidime, and
only 8 isolates (0.9%) were resistant to ceftazidime but not to
gentamicin.
The 5 most common drug-resistant colonizing pathogens
were Klebsiella pneumoniae (including K. pneumoniae pneumoniae and K. pneumoniae ozanae), Acinetobacter baumanii, P.
aeruginosa, Enterobacter species, and E. coli. We identified 11

Downloaded from http://cid.oxfordjournals.org/ by guest on September 30, 2016

each blood culture bottle for each neonate with suspected bacteremia. Most clinical specimens were initially submitted to the
respective microbiology laboratories at the 2 hospitals and were
referred to the Philippines General Hospital research microbiology laboratory for confirmation and/or characterization
with use of BACTEC (Becton Dickinson).
Antibiotic susceptibility testing. For all screening of clinical
isolates, we determined sensitivity by using Kirby-Bauer disk
diffusion. Antibiotic susceptibility classifications were determined on the basis of National Committee for Clinical Laboratory Standards breakpoints [12].
Statistical analyses. Analyses were conducted using SAS,
version 9.2 (SAS Institute), and SPSS software, version 11
(SPSS). In our conceptual model of nosocomially acquired neonatal sepsis, we assumed that colonization typically preceeds
invasive disease and that our interventions act primarily by
interrupting transmission and, hence, colonization. In accordance with these assumptions, our primary outcome was the
proportion of neonates who were newly colonized with the
target bacteria (drug-resistant gram-negative bacilli or grampositive cocci). Therefore, we excluded cases of colonization or
bacteremia during NICU days 01, under the assumption that
these reflected pre-NICU exposures. We calculated the incidence density for colonization or confirmed bacteremia (first
event per patient-days at risk) for each NICU separately. To
prevent double counting of identical isolates, we included only
the first new isolate from a given neonate, after which that
neonate was censored for only that isolate. Subsequent colonizations with a different isolate were still included. In addition,
we calculated incidence density for colonization and sepsis independent of each other. For example, if a neonate became
colonized with a given isolate on day 2 and became bacteremic
with that same isolate on day 4, the bacteremia calculation
would not have been adjusted by censoring from the colonization event on day 2.
Secondary outcomes included rate of bacteremia (number
of positive blood culture results per blood culture bottles sub-

Table 4. Results of surveillance for colonization with drug-resistant pathogens.

NICU 2

NICU 1
Variable
No. of neonates admitted
No. of swab specimens submitted
Rectal
Nasal
Umbilical
All
Positive isolates
All

Phase I

Phase II

328

Phase I

Phase II

597

597

Total

305

1827

638

1667

1096

654

4055

433
429
1500

1076
1076
3819

609
599
2304

355
354
1363

2473
2458
8986

261/1500 (17)

776/3819 (20)

1340/2304 (58)

526/1363 (39)

2903/8986 (32)

42/261 (16)
41

116/776 (15)
111

268/1340 (20)
263

11

219/261 (84)

660/776 (85)

1072/1340 (80)

525/526 (100)

2476/2903 (85)

88/219 (40)
33
8
6

595/660 (90)
216
265
45

555/1072 (52)
280
159
46

298/525 (57)
112
111
28

1536/2476 (62)
641
543
125

5
34
2

5
44
20

7
26
37

2
23
22

19
127
81

Gram-positive cocci
All
MRSA
VRE
Gram-negative bacilli

427/2903 (15)
416

Enteric
All
Klebsiella pneumoniae pneumoniae
K. pneumoniae ozanae
Escherichia coli
Citrobacter freundii
Enterobacter aerogenes
Other Enterobacter species
Nonenteric
All

131/219 (60)

65/660 (10)

517/1072 (48)

227/525 (43)

940/2476 (38)

Pseudomonas aeruginosa
Acinetobacter baumanii

77
16

10
25

189
261

56
117

332
419

Stenotrophomonas maltophilia
Alcaligenes faecalis

16
22

14
16

42
25

28
26

100
89

NOTE. Data are no. or proportion (%) of isolates, unless otherwise indicated. MRSA, methicillin-resistant Staphylococcus aureus; VRE, vancomycinresistant enterococci.

VRE isolates by culture of rectal specimens. MRSA was commonly isolated, particularly at NICU 2 (table 4).
Colonization rates did not change significantly at either
NICU between phase I and phase II (table 5), with regard to
ceftazidime resistance and/or gentamicin resistance. Of interest,
the types of detected gram-negative bacilli changed dramatically
at both NICUs; there was an increase in enteric gram-negative
bacilli and a decrease in nonenteric gram-negative bacilli (table
4). At NICU 1, the ratio of enteric to nonenteric gram-negative
bacilli was 40%:60% during phase I and 90%:10% during
phase II (P ! .001). At NICU 2, the ratio of enteric to nonenteric
gram-negative bacilli was 52%:48% during phase I and 57%:
43% during phase II (P p .06). MRSA essentially vanished
from NICU 2 during phase II (relative risk [RR], 0.01; 95%
CI, 0.000.05).
Bacteremia. Overall, 358 (19.6%) of 1828 neonates developed at least 1 episode of bacteremia during their NICU stay,
predominantly due to resistant gram-negative bacilli (78.2%

gram-negative bacilli vs. 11.8% gram-positive cocci; P ! .001)

(table 6). When we omitted duplicate isolates from the same
neonate, 510 (40.4%) of 1262 blood culture results were positive. The dominant isolates were Klebsiella species, Enterobacter
species, Pseudomonas species, and Alcaligenes faecalis (in approximately the same proportions as those found in the colonization surveys) (table 4). Similar to the colonization isolates,
there was a shift from nonenteric to enteric gram-negative bacilli at both NICUs during phase II (ratio of enteric to nonenteric bacilli at NICU 1, 11%:89% during phase 1 vs. 72%:28%
during phase 2; P ! .001; ratio of enteric to nonenteric bacilli
at NICU 2, 68%:32% during phase I vs. 78%:22% during phase
II; P p .04). Among the gram-positive cocci that caused bacteremia, coagulase-negative staphylococci predominated
(75%); there was only 1 MRSA isolate, and there were no VRE
isolates. Fungemia was uncommon and was primarily due to
nonCandida albicans Candida species.
Bacteremia rates at NICU 2 far exceeded those at NICU 1.
Enhanced Infection Control CID 2009:48 (1 January) 17

Downloaded from http://cid.oxfordjournals.org/ by guest on September 30, 2016

All

1/526 (0.2)
1

Table 5. Changes in incidence density for target pathogen colonization between phase I and phase II.
Incidence density, no. of new colonizations
per 1000 NICU-days at risk
NICU 1
Pathogen category

NICU 2

Phase I

Phase II

Phase I

Phase II

Ceftazidime-resistant gram-negative bacilli

Gentamicin-resistant gram-negative bacilli

148.2
166.0

158.9
147.1

.59
.35

343.9
472.9

310.8
434.6

.31
.38

Gram-negative bacilli resistant to both ceftazidime and gentamicin

MRSA

129.5
68.3

128.2
79.3

.94
.54

314.3
205.8

296.9
0.0

!.001

.56

NOTE. Vancomycin-resistant enterococci were too infrequently identified to calculate meaningful incidence rates. MRSA, methicillin-resistant
Staphylococcus aureus.

18 CID 2009:48 (1 January) Gill et al.

DISCUSSION
This project tested the effectiveness of a package of infectioncontrol interventions at 2 of the largest NICUs in The Philippines. The interventions were associated with an increased
rate of hand-hygiene compliance in general and of alcoholbased handrub use in particular. However, the overall incidences of colonization with resistant bacteria and of bacteremia
were unchanged. Therefore, although mortality rates decreased
substantially, we were unable to conclude that the interventions
were responsible for the decreases.
Our data provide some bleak insights into the epidemiology
of drug-resistant bacteria at these NICUs: colonization pressure
with drug-resistant pathogens was intense, with correspondingly high rates of bacteremia and mortality. Moreover, the
spectrum of pathogens causing sepsis was remarkable for its
dissimilarity to the spectrum at NICUs in the developed world.
Four of the most common pathogens were nonenteric pathogens: A. baumanii, P. aeruginosa, Stenotrophomonas maltophilia, and A. faecalis (an uncommon pathogen that is rarely
reported in the developing world) [1315]. The high frequency
of nonenteric gram-negative bacilli and the high rates of drug
resistance strongly imply that nosocomial transmission, rather
than mother-to-child transmission during delivery, was responsible for colonization. MRSA colonization was common
(although sepsis was rare), and VRE appeared to be an emerging
threat, which is worrisome, given the lack of prior reports of
VRE infection in The Philippines.
The agents that most commonly cause early neonatal sepsis
in the developed world are group B streptococci (GBS) and E.
coli. However, in our study, GBS sepsis was conspicuous for
its absence, echoing results from the International Infections
in Pregnancy study group, which found GBS colonization rates
to be one-half of those found in Dublin, Ireland, and one-third
of those found in Philadelphia [16, 17]. A recent review by
Zaidi et al. [18] of hospital-acquired neonatal sepsis in developing nations also found GBS in only 2.3% of sepsis episodes.

Downloaded from http://cid.oxfordjournals.org/ by guest on September 30, 2016

However, after adjustment for patient-time at risk, overall bacteremia rates did not change at either unit between phase I and
phase II (table 7). However, this adjustment concealed a decrease in the number of cases of nonenteric bacteremia, particularly at NICU 1, and there was an absolute and relative
increase in the number of cases of enteric bacteremia at both
NICUs. In addition, NICU 1 experienced a small but statistically significant increase in the incidence of bacteremia due to
coagulase-negative staphylococci and nonC. albicans Candida
species.
NICU mortality. Mortality rates were high at both NICUs
but decreased during phase II. Of 1828 neonates admitted to
the NICU, 615 (33.6%) died. At NICU 1, the risk of death
decreased during phase II (290 deaths per 1000 admissions
during phase I vs. 144 deaths per 1000 admissions during phase
II; RR, 0.50; 95% CI, 0.380.64); the absolute risk reduction
was 15% (95% CI, 9%20%). Mortality also decreased at NICU
2 during phase II (598 deaths per 1000 admission during phase
I vs. 481 deaths per 1000 admissions during phase II; RR, 0.81;
95% CI, 0.710.91); the absolute risk reduction was 12% (95%
CI, 6%18%).
Hand-hygiene compliance. Hand-hygiene compliance improved at both NICUs during phase II (table 8). Compliance
was evaluated for a mean of 18 h per NICU per month; therefore, there were 513 1-h observations and 5423 patient contacts.
In a comparison of phase II with phase I, the likelihood of
precontact hand-hygiene compliance improved at both units
(NICU 1: RR, 1.3; 95% CI 1.151.49; NICU 2: RR, 1.61; 95%
CI, 1.401.86). The phase II interventions had no impact on
the overall likelihood of postcontact hand-hygiene compliance
at NICU 1 (RR, 0.94; 95% CI, 0.791.11), although there was
a preferential shift toward ethanol-based handrub over soap
and water (P ! .001). At NICU 2, the likelihood of postcontact
hand-hygiene compliance increased during phase II (RR, 1.6;
95% CI, 1.311.98), again with preferential use of ethanol-based
handrub (P p .01).

Table 6. Bloodstream isolates, by neonatal intensive care unit (NICU) and study phase.
NICU 2

NICU 1
Variable
Total no. of blood cultures performed
Positive isolates

Phase I

Phase II

Phase I

Phase II

Total

197

504

289

272

1262

32 (16)

99 (20)

216 (75)

163 (60)

536 (42)

0/32 (0)
0
0
0

25/99 (25)
4
1
2

16/216 (7)
5
0
0

19/163 (12)
3
0
0

60/536 (11)
12
1
2

0
0

0
18

0
11

0
16

0
45

All
Enteric

27/32 (84)

32/99 (32)

198/216 (92)

142/163 (87)

417/536 (78)

All

3/27 (11)

24/32 (75)

138/198 (70)

122/142 (86)

287/417 (69)

0
0
0
3

9
0
0
14

76
16
2
44

78
8
2
34

163
24
4
95

25/27 (93)

9/32 (28)

64/198 (32)

32/142 (28)

130/417 (31)

0
0

5
1

52
3

15
4

72
8

0
22

3
0

4
2

4
6

11
30

0
2
1
5/32 (16)

0
0
0
42/99 (42)

3
0
0
2/216 (1)

3
0
0
2/163 (1)

6
2
1
59/536 (11)

All
Gram-positive cocci
All
MSSA
MRSA
VSE
VRE
Coagulase-negative staphylococci
Gram-negative bacilli

Proteus rettgeri
Nonenteric
All
Pseudomonas species
Acinetobacter baumanii
Stenotrophomonas maltophilia
Alcaligenes faecalis
Burkholderia cepacia
Flavobacterium species
Achromobacter species
Other pathogens

NOTE. Data are no. or proportion (%) of isolates, unless otherwise indicated. MRSA, methicillin-resistant Staphylococcus
aureus; MSSA, methicillin-susceptible S. aureus; VRE, vancomycin-resistant enterococci; VSE, vancomycin-susceptible enterococci.

This is problematic because current guidelines for neonatal

sepsis, such as the World Health Organizations pocket manual
for the management of hospital illnesses in resource-poor countries [19], assume that GBS and fully drug-susceptible enteric
gram-negative bacilli are the primary agents that cause early
neonatal sepsis.
Thirty percent of neonates arrived at the 2 NICUs already
colonized with drug-resistant gram-negative bacilli. This finding prompted us to go beyond the plan of the study and investigate what conditions and/or practices in the labor and
delivery unit might be contributing to transmission. Several
possibilities emerged; one was that latex gloves, frequently in
short supply, were often rinsed and reused after air-drying on
a rack. We also noted that it was standard practice for obstetricians to swab each mothers perineum with povidone-iodine
before delivery. These solutions were decanted from open large-

necked containers into multiuse bottles. P. aeruginosa and other

environmental gram-negative bacilli can be resistant to povidone-iodine. Therefore, contamination of the stock supplies
could systematically expose the neonates to iodine-resistant
pathogens during delivery [2022]. Additional environmental
studies would be helpful to test these observations and identify
other potential exposure sources.
We observed a shift from nonenteric to enteric gram-negative
bacilli isolates during phase II, particularly at NICU 1. Because
this shift occurred among both the invasive isolates and the
colonizing isolates, it probably represents a true shift in the
epidemiology of these nosocomial pathogens. Whether this occurred because of our interventions or because of factors external to our study is unclear.
The rate of hand-hygiene compliance at the start of the project was low but improved over time at both NICUs. Moreover,

Enhanced Infection Control CID 2009:48 (1 January) 19

Downloaded from http://cid.oxfordjournals.org/ by guest on September 30, 2016

Klebsiella species
Escherichia coli
Citrobacter species
Enterobacter species

Table 7. Changes in bacteremia rates between phase I and phase II.

Incidence density, no. of cases
of bacteremia per 1000 days at risk
NICU 1
Pathogen causing bacteremia
All
Gram-negative bacilli
Enteric
Nonenteric
Gram-positive cocci
Candida species

NICU 2

Phase I

Phase II

Phase I

Phase II

21.4
16.8

28.1
8.3

.23
.02

163.3
141.5

166.9
143.0

.85
.93

2.1

6.1

.07

90.5

112.6

.12

13.7
0.0

2.5
6.0

!.01

36.6

27.0

.12

!.01

9.0

14.7

.19

3.5

14.6

!.01

0.6

1.7

.40

NOTE. NICU, neonatal intensive care unit.

downstream effects. The fact that colonization and bacteremia

rates remained stable during phase II conflicts with this model
and calls into question whether our interventions or external
factors caused mortality rates to decrease; the result in our study
was similar to that in a recent study by Rupp et al. [27], in
which hygiene compliance improved without changing transmission rates. A notable exception was the disappearance of
MRSA colonization from NICU 2 during phase II. This was
surprising and hard to reconcile because of the very high rates
of MRSA colonization during phase I at NICU 2 and stable
MRSA colonization rates at NICU 1.
Our study had 2 main limitations. First, before-and-after
comparisons are inherently vulnerable to confounding by intercurrent seasonal or secular events. However, randomization
would have been realistic only in the context of a large multicenter trial. Second, resource constraints did not allow us to
collect process indicators to monitor whether other aspects of
our intervention, such as the daily and monthly checklists, were
being implemented as intended.
In conclusion, this study at 2 of the largest NICUs in The
Philippines demonstrates that enhanced infection-control interventions are feasible and possibly effective in developing
settings. The epidemiology of neonatal sepsis at these NICUs
is characterized by intense colonization pressure with multi-

Table 8. Hand-hygiene compliance rates, by neonatal intensive care unit (NICU) and study period.
NICU 2

NICU 1
Compliance rate, %
Variable

Compliance rate, %

No. of
observations

No
compliance

Ethanol-based
handrub

Soap
and water

No. of
observations

No
compliance

Ethanol-based
handrub

Soap
and water

440
952

67.1
56.8

2.1
24.6

30.9
18.6

852
476

76.5
62.2

16.3
28.4

7.2
9.5

440
946

75.9
74.1

1.4
9.7

22.5
16.0

851
475

86.9
79

8.3
12.6

4.6
7.8

Before patient contact

Phase I
Phase II
After patient contact
Phase I
Phase II

20 CID 2009:48 (1 January) Gill et al.

Downloaded from http://cid.oxfordjournals.org/ by guest on September 30, 2016

the modest absolute increase in hand-hygiene compliance rates

concealed a much larger proportional increase in ethanol-based
handrub use, a modality vastly superior to soap and water [23,
24]. Although it was discouraging that hand-hygiene compliance rates did not increase more, such results are typical [11].
In a trial at Childrens Hospital Boston (Boston, MA), Harbarth
et al. [25] saw the rate of hand-hygiene compliance increase
after their intervention, but the effect was transientthe rate
decreased from 42.5% to 35.1% within months after their study
began. Other studies documented postintervention compliance
rates as low as 14% [26], although typical postintervention
compliance rates are 40%60% [11]. Such findings emphasize
the variability in responses to behavioral interventions and the
enormous challenges inherent in effecting sustained behavioral
change.
Ultimately, we were unable to conclude definitively that our
interventions were effective. On one hand, improvements in
hand hygiene compliance demonstrate the feasibility of infection control through behavioral change in a resource-limited
setting. Furthermore, mortality decreased significantly at both
NICUs during phase II, which was a positive, although surprising, outcome. On the other hand, our conceptual model
assumes that improved hand hygiene reduces colonization pressure, with decreased bacteremia and mortality rates being

drug-resistant gram-negative bacilli, with correspondingly high

rates of bacteremia and mortality. Additional analyses have been
conducted to better characterize these pathogens and to identify
factors related to the transmission of drug-resistant gram-negative bacilli at these NICUs. MRSA and VRE appear to be
emerging threats. By contrast, the absence of GBS suggests that
this pathogen is unimportant in this context. Reducing the
burden of drug-resistant nosocomial infections in developing
hospital settings must be a priority for regional and global
public health agencies. Additional research is needed to understand what kinds of barriers to improving hand-hygiene
compliance exist for the staff at these NICUs, to identify reservoirs for environmental pathogens as a key step in interrupting their transmission, and to explore the cost implications
of implementing infection-control programs in resource-poor
hospitals.

We thank our research assistants, J. Estrada, R. Canseco, E. Aduan, and

A. Geraldez.
Financial support. Boston University and the Office of Health and
Nutrition of the United States Agency for International Development (cooperative agreement GHS-A-00-03-00020-00) and National Institute of Allergy and Infectious Diseases, National Institutes of Health (K23 AI 62208
to C.J.G.).
Potential conflicts of interest. All authors: no conflicts.

References
1. Goldmann DA, Huskins WC. Control of nosocomial antimicrobialresistant bacteria: a strategic priority for hospitals worldwide. Clin Infect Dis 1997; 24(Suppl 1):S13945.
2. Parry MF, Hutchinson JH, Brown NA, Wu CH, Estreller L. Gramnegative sepsis in neonates: a nursery outbreak due to hand carriage
of Citrobacter diversus. Pediatrics 1980; 65:11059.
3. Pawa AK, Ramji S, Prakash K, Thirupuram S. Neonatal nosocomial
infection: profile and risk factors. Indian Pediatr 1997; 34:297302.
4. Archibald LK, Ramos M, Arduino MJ, et al. Enterobacter cloacae and
Pseudomonas aeruginosa polymicrobial bloodstream infections traced
to extrinsic contamination of a dextrose multidose vial. J Pediatr
1998; 133:6404.
5. Zaidi M, Angulo M, Sifuentes-Osornio J. Disinfection and sterilization
practices in Mexico. J Hosp Infect 1995; 31:2532.
6. Isturiz RE, Carbon C. Antibiotic use in developing countries. Infect
Control Hosp Epidemiol 2000; 21:3947.
7. Babay HA, Twum-Danso K, Kambal AM, Al-Otaibi FE. Bloodstream
infections in pediatric patients. Saudi Med J 2005; 26:155561.
8. Tseng YC, Chiu YC, Wang JH, Lin HC, Su BH, Chiu HH. Nosocomial
bloodstream infection in a neonatal intensive care unit of a medical
center: a three-year review. J Microbiol Immunol Infect 2002; 35:
16872.
9. Chandrashekar MR, Rathish KC, Nagesha CN. Reservoirs of nosocomial pathogens in neonatal intensive care unit. J Indian Med Assoc
1997; 95:727.

Enhanced Infection Control CID 2009:48 (1 January) 21

Downloaded from http://cid.oxfordjournals.org/ by guest on September 30, 2016

Acknowledgments

10. Espino Hernandez M, Couto Ramos MJ, Fiol Ferrer N, Rojas Hernandez N. Resistance to antimicrobials and combination therapy assessment in neonatal sepsis [in Spanish]. Rev Panam Salud Publica
2003; 13:21421.
11. Centers for Disease Control and Prevention. Guideline for hand hygiene in health-care settings: recommendations of the healthcare infection control practices advisory committee and the HICPAC/SHEA/
APIC/IDSA hand hygiene task force. MMWR Morb Mortal Wkly Rep
2002; 51:156.
12. NCCLS. Performance standards for antimicrobial susceptibility testing:
9th informational supplement. Wayne, PA: NCCLS, 1991.
13. Berry AM. B. alcaligenes faecalis septicemia and meningitis in the newborn: report of an unusual case. Indian J Pediatr 1967; 34:2424.
14. Khetarpal SK. B. alcaligenes faecalis septicemia in the newborn. Clin
Pediatr (Phila) 1964; 3:10810.
15. Kumhar GD, Ramachandran VG, Gupta P. Bacteriological analysis of
blood culture isolates from neonates in a tertiary care hospital in India.
J Health Popul Nutr 2002; 20:3437.
16. Whitney CG, Daly S, Limpongsanurak S, et al. The international infections in pregnancy study: group B streptococcal colonization in
pregnant women. J Matern Fetal Neonatal Med 2004; 15:26774.
17. Thinkhamrop J, Limpongsanurak S, Festin MR, et al. Infections in
international pregnancy study: performance of the optical immunoassay test for detection of group B streptococcus. J Clin Microbiol
2003; 41:528890.
18. Zaidi AK, Huskins WC, Thaver D, Bhutta ZA, Abbas Z, Goldmann
DA. Hospital-acquired neonatal infections in developing countries.
Lancet 2005; 365:117588.
19. World Health Organization. Pocket book of hospital care for children:
guidelines for the management of common illnesses with limited resources. Geneva, Switzerland: World Health Organization, 2005.
20. Panlilio AL, Beck-Sague CM, Siegel JD, et al. Infections and pseudoinfections due to povidone-iodine solution contaminated with Pseudomonas cepacia. Clin Infect Dis 1992; 14:107883.
21. Wisplinghoff H, Schmitt R, Wohrmann A, Stefanik D, Seifert H. Resistance to disinfectants in epidemiologically defined clinical isolates
of Acinetobacter baumannii. J Hosp Infect 2007; 66:17481.
22. Berkelman RL, Anderson RL, Davis BJ, et al. Intrinsic bacterial contamination of a commercial iodophor solution: investigation of the
implicated manufacturing plant. Appl Environ Microbiol 1984; 47:
7526.
23. Guilhermetti M, Hernandes SE, Fukushigue Y, Garcia LB, Cardoso CL.
Effectiveness of hand-cleansing agents for removing methicillin-resistant Staphylococcus aureus from contaminated hands. Infect Control
Hosp Epidemiol 2001; 22:1058.
24. Kjolen H, Andersen BM. Handwashing and disinfection of heavily
contaminated handseffective or ineffective? J Hosp Infect 1992; 21:
6171.
25. Harbarth S, Pittet D, Grady L, et al. Interventional study to evaluate
the impact of an alcohol-based hand gel in improving hand hygiene
compliance. Pediatr Infect Dis J 2002; 21:48995.
26. Bischoff WE, Reynolds TM, Sessler CN, Edmond MB, Wenzel RP.
Handwashing compliance by health care workers: the impact of introducing an accessible, alcohol-based hand antiseptic. Arch Intern
Med 2000; 160:101721.
27. Rupp ME, Fitzgerald T, Puumala S, et al. Prospective, controlled, crossover trial of alcohol-based hand gel in critical care units. Infect Control
Hosp Epidemiol 2008; 29:815.