Niger J Paed 2014; 41 (1):54 - 58  
Osinupebi OA  
Ogunlesi TA  
Fetuga MB  
Pattern of nosocomial infections in  
the special care baby unit of the  
Olabisi Onabanjo University  
Teaching Hospital, Sagamu, Nigeria  
Accepted: 27th September 2013  
Abstract Background: Sepsis  
contributes significantly to new-  
born deaths in Nigeria. A signifi-  
cant proportion of severe infec-  
tions in the newborn may be  
health care-related.  
Objective: To determine the  
prevalence, types and risk factors  
for nosocomial infections in the  
Special Care Baby Unit of a Nige-  
rian Tertiary Hospital.  
Method: A cross-sectional survey  
of consecutively admitted infants  
aged 0 to 28 days with signs of  
infections or who developed signs  
of infection following admission.  
Infants with or without nosoco-  
mial infections were compared for  
the clinical and laboratory details.  
Results: Out of 356 infants, 32  
A significantly higher proportion  
of babies with nosocomial infec-  
tions were inborn (p < 0.000) and  
stayed longer than 7 days on ad-  
mission (p = 0.034). Bacteraemia  
was significantly more frequent  
among babies with nosocomial  
infections (p = 0.014) while super-  
ficial skin and mucosal infections  
occurred to similar extents in both  
groups. Klebsiella and Proteus  
species were the leading isolates  
among babies with nosocomial  
infections. Nasogastric intubation  
was significantly more frequently  
performed among babies with  
nosocomial infections (p = 0.045).  
Conclusion: The present study  
revealed that hospital acquired  
infection is an important cause of  
morbidity in the newborn unit.  
Ogunlesi TA  
Fetuga MB  
Department of Paediatrics,  
Osinupebi OA  
Department of Medical Microbiology  
and Parasitology,  
College of Health Sciences,  
Olabisi Onabanjo University,  
P.M.B 2022, Sagamu-121001,  
Ogun State,  
Email: tinuade_ogunlesi@yahoo.co.uk  
8.9%) had between 1 and 3 noso-  
comial infections while 48  
13.5%) had community-acquired  
Keywords: Bacteraemia; Hospital  
-acquired infections; invasive pro-  
cedures; newborns  
infections. Half of babies with  
nosocomial infections were pre-  
term and weighed less than 2kg.  
newborn may also be associated with severe morbidities  
such as shock, necrotizing enterocolitis and meningitis  
which might leave severe sequelae in the survivors.  
Therefore, every measure required for the reduction of  
the burden of sepsis in the newborn needs to be put in  
Neonates are particularly susceptible to infection  
because of their immature immune system in addition to  
several invasive diagnostic and therapeutic procedur1e-s3  
and interventions required for their management.  
Studies from developed countries have reported that  
nosocomial infection is an important cause of morbidity  
and mortality among ne4o-6nates despite advances in  
neonatal intensive care.  
The overall contribution of nosocomial infections to  
cases of newborn sepsis varies in centres from different  
parts of the world, depending on the level of newborn  
care in practice. Although, referred infants are mostly  
A recent national epidemiological survey of newborn  
deaths in Nigeria suggested that severe infections were  
resp7onsible for 22% of all newborn deaths in the coun-  
try. This is very similar to 24.8% reported from a re-  
view of newborn deaths recorded at the Olabisi Ona-  
banjo University Teaching Hospit8al (OOUTH), Sagamu,  
Nigeria between 1996 and 2005. It is also important to  
add that sepsis-related case fatality rate at the same cen-  
tre in 9a study conducted between 2006 and 2008, was  
affected by severe infections, inborn babies may also be  
infected vertically or horizontally. The former com-  
monly presents as Early-Onset sepsis while the latter  
common10ly presents as Late-Onset sepsis or Nosocomial  
Surveillance reports between 1999 and 2001 at OOUTH,  
Sagamu, Nigeria revealed that the neonatal unit recorded  
all the cases of hospital-acq11uired infection in the pediat-  
ric service of the hospital. The Special Care Baby Unit  
2%. In addition to deaths, severe infections in the  
SCBU) of the hospital, with 20 cot and incubator  
Others included the type of infection, clinical findings,  
procedures, interventions and device use, as well as the  
results of bacteriologic examination of specimens. Cases  
of nosocomial infections were determined using modi-  
spaces, was established as a sub-unit of paediatric ser-  
vice in 1989. The unit offers basic care for acutely-ill  
newborn infants while facilities for intensive care such  
as mechanical ventilation, blood gas analysis and par-  
enteral nutrition are not available. The unit receives high  
fied CDC definitions. Inborn infants were the babies  
delivered in OOUTH while the out-born infants were  
referred babies. Data were analyzed using SPSS 17.0  
statistical package. Babies with nosocomial infections  
were statistically compared with babies with community  
acquired infections using clinical characteristics, pattern  
of infections and spectrum of isolates. Proportions and  
mean values were compared using the Chi-Squared test  
and Student’s t-test respectively. P-values less than 0.05  
defined statistical significance.  
-risk infants delivered in the maternity unit of the hospi-  
tal as well as infants referred from the lower levels of  
health care located within and outside Sagamu commu-  
nity. Thus, the unit is compartmentalized with inborn  
infants separated from referred infants to prevent cross-  
infections. A recent study conducted at the same centre  
between 2006 and 2008 showed that referred babies and  
inborn babies constituted 54.5% and 45.5%12respectively  
of the cases of Late-Onset sepsis in the unit.  
The role of surveillance as an important tool in the  
prevention of10hospital- acquired infection has been  
emphasized. This study is a step in that direction as it  
was designed to evaluate the prevalence, types and some  
of the factors related to health care associated infections  
in the SCBU of OOUTH, Sagamu, Nigeria.  
During the study period, 356 infants were admitted into  
the unit of whom 80 (22.5%) had infections. They con-  
sisted of 50 male and 30 female infants with male-to-  
female ratio of 1.7:1. The 80 infants with infections had  
a mean EGA of 35.9 ± 3.9 weeks and a mean weight of  
.3 ± 0.9kg.  
The general characteristics of these 80 infants are de-  
picted in Table 1. Most of the infants were males  
(62.5%), term (60.0%), single ton (87.5%), out-born  
(72.5%), weighed > 2.0kg (58.8%) and were hospital-  
ized for > 14 days (52.5%). The frequency of obvious  
infections in this population varied between one and  
three with a mean of 1.6 ± 0.6.  
Forty-eight babies (60.0%) had community-acquired  
infections while 32 (40.0%) had nosocomial infections.  
The infants with nosocomial infections constituted 8.9%  
(32/356) of the total admissions while infants with com-  
munity acquired infections constituted 13.5% (48/356)  
of the total admissions. The babies with community-  
acquired infections had higher EGA compared with  
those who acquired infection in the hospital. This differ-  
ence lacked statistical significance (36.6 ± 3.3 weeks Vs  
34.9 ± 4.5 weeks; t = 1.95, p = 0.055). Similarly, the  
mean weight of babies admitted with infections was  
similar to that of babies with nosocomial infections (3.4  
± 0.8kg Vs 3.2 ± 0.9kg; t = 1.04, p = 0.301).  
This cross-sectional survey was conducted in the Special  
Care Baby Unit (SCBU) of OOUTH, Sagamu, Nigeria  
between January and September, 2008. As mentioned  
above, this unit only provides basic care for high-risk  
infants with medical disorders such as preterm birth, low  
birth weight and the associated problems, intrapartum-  
related injuries, jaundice, infections and minor surgical  
conditions. The unit is managed by at least two Consult-  
ants, resident doctors, medical interns and nurses, most  
of whom are specialized in paediatric care. The scope of  
care available for infants with sep9sis had been described  
in detail in another publication. In addition, first-line  
antibiotics employed in the treatment of newborn sepsis  
in the unit include cefuroxime and gentami1c3in based on  
previous reports of antibiogram in the unit. The choice  
of antibiotics is usually tailored to the pattern of antibi-  
otic susceptibility reported by the laboratory. Microbio-  
logic services available in the hospital include routine  
bacteriologic analysis of specimens; facilities for an-  
aerobic bacterial, fungal or viral cultures are not rou-  
tinely available.  
Table 1: Clinical characteristics of 80 babies with infections  
Type of birth  
Frequency (%)  
The study population consisted of consecutive infants  
aged 0 to 28 days admitted with signs of infection or  
who developed features of infection whilst on admis-  
sion. The latter group was in turn, sub-divided into those  
who developed signs of infection within or after the first  
70 (87.5)  
10 (12.5)  
22 (27.5)  
58 (72.5)  
30 (37.5)  
50 (62.5)  
5 (6.2)  
13 (16.3)  
14 (17.5)  
48 (60.0)  
2 (2.5)  
Place of birth  
8 hours of admission. Informed consent was obtained  
from the primary care givers available at the point of  
recruitment into the study.  
Gestational Age (weeks)  
< 28  
9 – 32  
3 – 36  
Demographic and clinical data were obtained from all  
the infants admitted into the study. These data included  
age and body weight on admission, gender, estimated  
gestational age (EGA), place and type of delivery  
Birth weight (kg)  
0.5 – 1  
.1 – 1.5  
.6 – 2.0  
16 (20.0)  
15 (18.8)  
47 (58.9)  
Comparison of the babies admitted with and without  
obvious infections  
Table 3: Comparison of pattern of infections among babies  
admitted with infections and babies with nosocomial infections  
Babies admitted  
with infections  
Babies with p-  
nosocomial values  
Table 2 depicts the comparison of the clinical details of  
babies admitted with infections and babies with  
nosocomial infections. A significantly higher proportion  
of babies admitted with infections were out-born  
(n = 48)  
n = 32)  
13 (27.1)  
8 (25.0)  
p = 0.000) and stayed for 7 days or less on admission  
p = 0.034). Higher proportions of babies hospitalized  
Superficial skin  
11 (22.9)  
11 (22.9)  
6 (12.5)  
4 (12.5)  
5 (15.6)  
3 (9.4)  
with obvious infections also had records of prolonged  
rupture of membrane, prolonged labour and history  
suggestive of asphyxia but without statistical signifi-  
cance. Preterm birth, weight on admission < 2kg, multi-  
ple gestation, abdominal delivery and requirement for  
intubation at birth had no statistical association with  
development of nosocomial infections.  
2 (4.2)  
7 (21.9)  
Respiratory infec-  
Urinary tract infec- 1 (2.1)  
27 (56.3)  
3 (6.3)  
15 (46.9)  
2 (6.3)  
0 (0.0)  
3 (9.4)  
7 (14.6)  
Table 2: Comparison of babies with and without obvious  
Babies admit-  
Babies with P-  
ted with infec- nosocomial values  
Table 4: Comparison of bacterial isolate types among babies  
tions (n = 48)  
n =32)  
with and without nosocomial infections  
Babies admitted  
with infections  
(n = 48)  
Babies with p-  
nosocomial values  
Male sex  
Preterm birth  
29 (60.4)  
16 (33.3)  
17 (35.4)  
6 (12.5)  
42 (87.5)  
10 (20.8)  
15 (31.2)  
3 (6.3)  
21 (65.6)  
16 (50.0)  
16 (50.0)  
4 (12.5)  
16 (50.0)  
10 (31.3)  
6 (18.8)  
3 (9.4)  
Weight < 2kg  
Multiple births  
Caesarean Section  
Intubation at birth  
Prolonged labour  
Presence of asphyxia  
Duration of hospi-  
talization < 7 days  
(n = 32)  
Atypical coliforms  
3 (6.3)  
2 (6.3)  
2 (4.2)  
8 (16.7)  
2 (4.2)  
3 (6.3)  
2 (6.3)  
Klebsiella spp.  
Proteus spp.  
Pseudomonas spp.  
10 (31.3)  
5 (15.6)  
1 (3.1)  
12 (25.0)  
20 (41.7)  
14 (29.2)  
6 (18.8)  
11 (34.4)  
3 (9.4)  
Staph. Aureus  
8 (16.7)  
4 (12.5)  
Comparison of the possible factors predisposing babies  
to nosocomial infections  
*Pr olonged Rupture of Membranes  
Table 5 shows that nasogastric intubation was more  
frequently performed among babies with nosocomial  
infections with statistical significance (p = 0.045).  
Although intravenous infusion was more frequently  
carried out among babies admitted with obvious infec-  
tions, the difference did not reach level of significance.  
Intravenous cannulation, airway suctioning, assisted  
ventilation and clinical procedures like minor surgeries  
and blood transfusions were performed at similar rates  
in both groups of babies.  
Pattern of infections among babies admitted with or  
without obvious infections  
Bacteraemia was significantly more frequent among  
babies with obvious nosocomial infections (p = 0.014)  
as shown in Table 3. The prevalence of ophthalmia, om-  
phalitis, phlebitis, superficial skin infections and gastro-  
intestinal tract infections were similar in both groups.  
However, respiratory tract and urinary tract infection  
was remarkably rare in both groups.  
Table 5: Comparison of the procedures which may predispose  
to nosocomial infections  
Comparison of the pattern of clinical isolates among  
babies with obvious infections  
Babies admitted Babies with p-  
with infections  
nosocomial values  
(n = 48)  
In Table 4, Gram-negative bacilli, particularly Klebsiella  
species, were the leading isolates among babies in both  
groups. Staphylococcus aureus and Klebsiella species  
were the leading isolates among babies admitted with  
infections whereas Klebsiella and Proteus species were  
the leading isolates among babies with nosocomial  
infections. None of the comparisons of isolates reached  
the level of statistical significance.  
n = 32)  
Nasogastric Intuba-  
IV Cannulation  
Blood transfusion  
Intranasal oxygen  
Assisted ventilation  
Minor surgery  
19 (39.6)  
20 (62.5)  
29 (60.4)  
16 (33.4)  
18 (37.5)  
21 (43.8)  
20 (41.7)  
17 (35.4)  
1 (2.1)  
21 (65.6)  
14 (43.8)  
17 (53.1)  
11 (34.4)  
13 (40.6)  
8 (25.0)  
1 (3.1)  
Intravenous infusion 11 (22.9)  
3 (9.4)  
prevalence of Late-Onset sepsis was higher among  
infants delivered at E12GA less than 32 weeks and weigh-  
ing less than 1.5kg. Ordinarily, such small babies tend  
to have greater degree of immune immaturity and are  
more likely to require invasive procedures and stay  
longer on admission. Although, previous studies have  
suggested associations betwee1n5, 1m6 echanical ventilation  
that observation was  
not made in the present study. We speculate that the  
observation in the present study might be related to the  
small number of infants studied.  
The prevalence of nosocomial infections in the present  
study was 8.9%. This is comparable to 7.8% report1e5d in  
a neonatal intensive care unit (NICU) in Singapore but  
remarkably lower than 19.2% and 21.4% reporte1d6, 1i7n a  
Saudi NICU and an Egyptian NICU respectively.  
addition, the prevalence of hospital-acquired infections  
in different settings in Europe and North Am14,e1r8ica has  
and intravenous cannulation,  
been reported to vary between 7% and 24%.  
fore, the relatively lower prevalence of 8.9% in the pre-  
sent study might reflect the non-NICU setting of the  
study. It is important to add that the present study was  
conducted in a SCBU rather than NICU where nosoco-  
mial infections are expectedly more common. The dif-  
ference lies in the tendency to use more invasive proce-  
dures such as endotracheal intubation, mechanical venti-  
lation an1d0 total parenteral nutrition in NICU rather than  
The significant association between nasogastric intuba-  
tion and nosocomial infection in the present study may  
be borne out of the fact that this population of infants  
are often too small or too ill to feed by direct sucking.  
Therefore, tube feeding becomes a ready vehicle for  
nosocomial sepsis particularly when the tube is left in-  
situ for a long time just as total parenteral nutrition has  
been reported t1o6 be associated with nosocomial sepsis in  
other settings. For resource-poor settings where short-  
age of nursing personnel might militate against the in-  
sertion of a tube prior to every episode of feeding, fre-  
quent replacement of in-situ nasogastric tubes may be  
effective in the prevention of tube feeding-related sepsis.  
In addition, infection control practices have  
been institutionalized in our hospital since 1999 when  
the Infection Control Unit was established. This unit  
oversees the teaching and adoption of measures relevant  
to the prevention of nosocomial infections by all cadres  
of health workforce.  
A large number of the infections in the present study  
were blood stream infections. This is consistent with the  
findings in other studies in Nor1t5h-17America and Europe  
and few developing countries. Blood stream infec-  
tions are frequently reported in noso9comial neonatal  
infections in the intensive care unit. This is closely  
In conclusion, the present study revealed that hospital  
acquired infection is an important cause of morbidity in  
the newborn unit. The prevalence and types of nosoco-  
mial infections in a Nigerian Special Care Baby Unit are  
comparable to what had been reported in Neonatal  
Intensive Care Units in many parts of the world. This  
suggests that even in the absence of major invasive  
procedures, the morbidities associated with nosocomial  
infections in a Special Care Baby remains significant. It  
is important that every newborn unit has a written policy  
on guidelines for the control of hospital care-related  
infections. This may include training of all cadres of  
staff, environmental and equipment cleanliness, strin-  
gent control of antibiotic use, stringent hand hygiene  
practices and minimal use of invasive procedures. It is  
important that the existing infection control programmes  
should be strengthened.  
related to the use of intravenous catheters both periph-  
eral and central. The prominent role of superficial infec-  
tions like ophthalmia, phlebitis and omphalitis in the  
developin10g world setting has been extensively de-  
scribed. Cost-effective interventions such as stringent  
hand washing techniques, liberal use of disinfectants and  
prevention of ov1e0r,c2r0o, 2w1 ding are useful in the control of  
these infections.  
Most of the pathogens isolated from both groups of neo-  
nates were Gram negative bacilli. This finding is consis-  
tent with other reports that Gram negative bacilli are the  
m19,a2i2n,23pathogens in neonatal hospital acquired infections.  
Unlike other studies which reported the predomi-  
n15a, n16ce of Coagulase Negative Staphylococcus (CONS)  
there was no isolation of CONS in the present  
study. This is difficult to explain since another study  
from the same c2entre had earlier reported the role of  
CONS in LOS. The major challenge poised by these  
Author’s Contributions  
pathogens lies in the high incidence of multi-drug resis-  
tance resulting in limited9 therapeutic options and high  
morbidity and mortality. This calls for judicious use of  
antibiotics in newborn units to minimize the emergence  
of resistant strains of pathogens. Therefore, neonatal  
units need to have evidence-based protocols on the use  
of antibiotics.  
OOA and FMB conceived and designed the study.  
OTA analyzed and interpreted the data. All the authors  
drafted and edited the manuscript. All the authors made  
substantial contributions to the intellectual content of the  
Conflict of Interest: None  
Funding: None  
In the present study, the risk of nosocomial infection  
was not significantly higher among preterm babies an1d5  
babies weighing less than 2kg unlike previous reports.  
A previous study at the same centre suggested that the  
Ogunlesi TA, Ogunfowora OB.  
Predictors of mortality in neonatal  
septicaemia in an under-resourced  
setting. J Natl Med Ass 2010;  
17. Abdel-Wahab F, Ghoneim M,  
Khashaba M, El-Gilany A-H, Ab-  
del-Hady D. Nosocomial infections  
surveillance in an Egyptian neona-  
tal intensive care unit. J Hosp In-  
fect 2013; 83(3): 196-199.  
18. Schwab F, Geffers C, Barwolff S,  
Ruden H, Gastmeier P. Reducing  
neonatal blood stream infections  
through participation in a national  
surveillance system. J Hosp Infect  
2007; 65: 319-325.  
19. Couto RC, Carvalho EA, Pedrosa  
TM, Pedroso ER, Neto MC, Bis-  
cone FM. A 10 year prospective  
surveillance of nosocomial infec-  
tion in the neonatal intensive care  
units. Am J Infect Control. 2007;  
35: 183-189.  
20. Uwaezuoke SN, Obu HA. Nosoco-  
mial infections in neonatal inten-  
sive care units: cost-effective con-  
trol strategies in resource limited  
centres. Niger J Paediatr 2013; 40  
(2): 125-132.  
21. Gray J, Omar N. Nosocomial in-  
fections in neonatal intensive care  
units in developed and developing  
countries: how can we narrow the  
gap? J Hosp Infect 2013; 83: 193-  
22. Drews MB, Ludwig AC, Leititis  
JU, Daschner FD. Low birth  
weight and nosocomial infection in  
neonates in a neonatal intensive  
care unit. J Hosp Infect 1995;  
23. Landre –Peigne C, Ka AS, Peigne  
V, Bougere J, Seye MN, Imbert P.  
Efficacy of an infection control  
programme in reducing nosoco-  
mial bloodstream infections in a  
Senegalese neonatal unit. J Hosp  
Infect 2011; 79:161-165.  
Borguesies A, Stronati M. Strate-  
gies for the prevention of hospital  
acquired infections in the neonatal  
intensive care unit. J Hosp Infect  
008; 68:293-300.  
0. Ogunlesi TA, Dedeke O, Njo-  
kanma F. Nosocomial Infections.  
In: Common Infections in the  
Brito DV, de Brito CS, Resende  
DS, Moreira do OJ, , Abdallah  
VOS, Filho PPG. Nosocomial  
infections In a Brazilian neonatal  
intensive care unit: a 4 year sur-  
veillance study. Rev Soc Bras Med  
Trop 2010; 43 (6):633-647.  
GeffersC, Baewolf S, Schwab F,  
Gastmeier P. Incidence of health-  
care associated infections in high  
risk neonates: results from the  
German surveillance system for  
very low birth weight infants J  
Hosp Infect 2008:68:214-221.  
Boo NY, Chor CY. Six year trend  
of neonatal septicaemia in a large  
Malaysian maternity hospital. J  
Paediatr Child Health 1994; 30:23  
Newborn Period. 1 Edition. No-  
vascience Publishers, New York  
City. 2012: 53-66.  
1. Osinupebi AO, Tade AO, Oluwole  
AO, Eniola VO. Surveillance of  
nosocomial infection at Olabisi  
Onabanjo University Teaching  
Hospital. J Nig Infect Control Asso  
2. Ogunlesi TA, Ogunfowora OB,  
Osinupebi O, Olanrewaju DM.  
Changing trends in newborn sepsis  
in Sagamu, Nigeria: Bacterial aeti-  
ology, risk factors and antibiotic  
susceptibility. J Paediatr Child  
Health 2011; 47: 5-11.  
3. Njokanma OF, Olanrewaju DM,  
Akesode FA. Antibiotic resistance  
among bacterial isolates in neona-  
tal septicaemia. Niger J Paediatr  
Zaidi AK, Huskins WC, Thaver D,  
Bhutta ZA, AbbasZ, Goldmann  
DA. Hospital acquired neonatal  
infections in developing countries.  
Lancet 2005; 65:1175-1188.  
Macais AE, Munoz JM, Galvan A  
et al. Nosocomial bacteremia in  
neonates related to poor standards  
of care .Pediatric Infect DIs J  
994; 21: 47-51.  
4. Garner JJ, Jarvis WR, Emori TC,  
Hughes JM. CDC definitions for  
nosocomial infections. Am J Infect  
Control 1988; 16:128-140.  
5. Joseph CJ, Lian WB, Yeo LC.  
Nosocomial infections (Late Onset  
Sepsis) in the Neonatal Intensive  
Care Unit (NICU). Proceedings of  
Singapore Healthcare 2012; 21  
005; 24:713-716.  
Federal Republic of Nigeria Minis-  
try of Health. Saving Newborn  
Lives in Nigeria: Integrated Nater-  
nal, Newborn and Cnhdild Health  
Strategy. Revised 2 Edition,  
(4): 238-244.  
6. Mahfouz AA, Al-Azraqi TA, Ab-  
bag FI, Al-Gramal MN, Seef S,  
Bello CS. Nosocomial infections  
in a neonatal intensive care unit in  
South-western Saudi Arabia. East  
Mediterranean Health J 2010; 16  
Fetuga B, Ogunlesi T, Adekanmbi  
F, Olanrewaju D, Olowu A. Com-  
parative analyses of childhood  
deaths in Sagamu, Nigeria: Impli-  
cations for the Fourth MDG. South  
Afr J Child Health 2007; 1(3): 106  
(1): 40-44.