Incidence, bacteriological profile and antibiotic sensitivity pattern of neonatal sepsis in a tertiary health facility in Abuja, North-central Nigeria. CURRENT STATUS:

Background : Neonatal sepsis is commonly caused by bacteria in the first 28 days of life. If prompt management is not instituted, it could lead to death within hours of onset. Due to diagnostic limitation in developing settings, prompt laboratory identification of causative organism is usually a challenge. To prevent mortality, clear knowledge of bacteria and their antibiotics sensitivity pattern is important for prompt empirical treatment. Methods : This prospective and cross-sectional study enrolled 339 newborns and were admitted for probable sepsis to the special care unit of the university of Abuja Teaching Hospital. Socio-demographic profile and Blood culture was done from every enrolled newborn using BACTEC 9050. The pattern of the clinical features, incidence, bacteriological profile and antibiotic sensitivity pattern of newborns with confirmed neonatal sepsis were documented. Results : A total of 645 newborn were admitted for probable sepsis during the study period based on clinical features and initial laboratory work-up. Forty-six of the 645 newborns (46/645) had laboratory confirmed sepsis based on blood culture resulting in a neonatal sepsis incidence rate of 71.3 (95%CI 50.7-91.9) per 1000 admitted newborn . Seventeen of the 46 confirmed sepsis cases were among the 1322 newborns delivered within the study facility during the study period giving an in-hospital neonatal sepsis incidence rate of 12.9 (95% CI 6.7-19.0) per 1000 live birth. Amongst the 46 babies with positive blood culture, 27/46 (58.7%) had normal white cell count while the remaining 19/46 (41.3%) had abnormal results. In all, 52 counts of bacteria categorized into 11 bacteria species were isolated from the 46 positive blood cultures. Enterococcus spp and streptococcus species were the commonest gram-positive while Escherichia coli and Chryseomonas luteola were the commonest gram-negative bacteria isolates. Imipenem (all bacterial isolates except Vibrio fluvialis ), Augmentin ( Streptococcus spp , Staphylococcus spp, Escherichia spp,


Vancomycin (Streptococcus spp, Staphylococcus spp, Escherichia spp, Enterococcus spp),
and Ofloxacin (all except Vibrio fluvialis and Citrobacter freundii) had the widest coverage of bacteria isolated from newborn with sepsis.
Conclusion: Sepsis in newborn is still prevalent in our environment and compared to previous documented isolates and sensitivity pattern, the bacteria causes, and their antibiotic sensitivity patterns appears to be changing.

Background
Neonatal sepsis is a life threatening emergency and any delay in treatment may result to septic shock and death. 1 According to a 2006 report of the World Health Organization, it is estimated that 1.6 million deaths occur globally every year due to neonatal sepsis and it is responsible for 30 to 40% of all neonatal deaths occurring in developing countries. 2− 4 Nigeria ranks number one in annual newborns death in Africa and second highest in the world. 5 The neonatal mortality in Nigeria stands at 33/1000 live births. 6 In developing countries, bacterial pathogens are the most common cause of neonatal sepsis, 7 causing a wide variety of infections including meningitis, pneumonia, urinary tract infection and sepsis. 3,5 These infections can run a rapid course with death occurring in less than 24 hours, if prompt effective empirical treatment is not instituted. 3,8 The development of effective empirical antibiotic protocols depend on the knowledge of the prevailing bacterial pathogens in that locality. In the neonatal period, it has been documented that the commonest blood-culture isolates in many low income countries were Staphylococcus aureus, Streptococcus pyogenes, and Escherichia coli. 7,9 Tinuade et al in Sagamu and Olateju et al in Gwagwalada, described Klebsiella specie and coagulase-negative Staphylococci featuring prominently in Nigeria. 10,11 The spectrum of organisms that cause neonatal sepsis changes over time and varies from region to region, even in the same center from time to time. 3,10 These organisms have also developed increasing multi-drug resistance over the last two decades. 3 Therefore, current knowledge of center specific pattern of bacterial isolates and their antimicrobial susceptibility pattern is useful for prompt management of patients. The present study was undertaken to highlight the local pattern of bacterial isolates in neonates and their antimicrobial sensitivity in a tertiary care hospital in Abuja.

Study design and area
This prospective and cross-sectional study was conducted over an 8-month period in 2014 at the Special Care Baby Unit (SCBU) of the University of Abuja Teaching Hospital (UATH). The UATH is a 350-bed tertiary hospital that runs primary and secondary in addition to tertiary health care services. The hospital is in Gwagwalada Area Council of the Federal Capital Territory (FCT), Abuja; North Central region of Nigeria. 12 It is the main referral centre for neonatal care in north-central region. UATH has annual deliveries of 2,500 babies and an average annual SCBU admission of 968 babies using the year 2012 records. The SCBU is 30-bedded unit with separate inborn and outborns sections manned by a neonatologist, four senior and junior residents each, six house officers and twenty nurses with further training in neonatology. One thousand, three hundred and twenty-two (1322) newborns were delivered in the center during the study period.

Management overview of neonatal sepsis in UATH
On presentation to the SCBU, historical assessment and detailed physical examination are done. After collection of the relevant samples, the subjects are commenced on empiric antibiotics of amoxicillin/clavulanic acid at 15 mg/kg body weight and gentamicin at 2.5 mg /kg body weight per dose every 12 hours. Necessary changes are made based on clinical response and/or result of blood culture and sensitive test. Sometimes, third generation cephalosporin mostly cefotaxime and ceftazidime are used for severe infections and in babies who presented with relevant clinical features and/or clinical examination suggestive of meningitis. Other supportive measures where necessary are also initiated.

Sample size determination and sampling technique
The sample size was calculated using the Cochrane formula, 13 based on a prevalence of neonatal sepsis from a previous, 11 a confidence level of 95% and a level of precision estimated to be ± 5% of the assumed prevalence in addition to a non-response rate of 10%. This resulted to a minimum sample size of 331.Consecutive neonates who met the inclusion criteria were enrolled into the study. The inclusion criteria included neonates admitted to the SCBU with features suggestive of sepsis (such as fever, jaundice, lethargy, poor suck etc) and those whose parents/guardians gave consent for enrolment of their newborn in the study. Excluded from this study were newborns with major congenital malformations (this was to avoid mortality not directly linked to sepsis) and neonates who have had administration of antibiotics for more than 24 hours, prior to presentation to the hospital.

Blood and data collection method
All the neonates that were enrolled into the study had their blood specimens drawn for culture before treatment was initiated or latest within 12 hours of admission. The skin over the site of blood collection was cleaned thoroughly with 70% alcohol, allowed to dry and cleaned with Povidone iodine for two minutes before blood sample was collected. Two millilitres of blood were drawn by trained assistants from a peripheral vein. One millilitre was for blood culture, and the other for complete blood count (CBC ii.
Length of each baby enrolled in to the study, was measured using inelastic tape measure, by placing the baby on a hard surface, lying supine and the lower limbs fully stretched.
The frontal and occipital bony prominences were identified before measurements were taken using an inelastic tape measure.
iv. Gestational age of each baby was determined using modified Ballard score. 14 v.
The newborns family's Socio-economic status was determined using Olusanya's social economic classification. 15 vi. Neonatal sepsis for the sake of this study was defined as blood stream infection with positive blood culture result. For gram staining procedure, a single colony was picked and emulsified in a drop of normal saline on a clean glass slide and allowed to dry. The slide was then fixed with heat, stained with crystal violet for one minute, followed by procedure mordant with Lugo's iodine solution for 30 seconds. The slide was then washed with water. The process of decolourisation was carried out with acetone for one to two seconds before washing with water. Counter-staining was with safranin, added for two minute and later washed with water and allowed to air dry. The slide was then examined microscopically. Gram positive bacteria (GPB) were presumed if the identified pathogen appeared dark purple, while a red colour suggested Gram negative bacteria (GNB).
Antimicrobial susceptibility testing was done using modified Kirby-Bauer disc diffusion method, 17  seizures. For the sake of this study, normal white cell count on CBC was defined as cell count between 9-30,000/mm 3 (or 9-30 × 10 9 /L). Values above or below this range were considered abnormal.

Data analysis
The raw data were inputted into the Excel spread sheet and analysed using the Statistical Package for Social Sciences (SPSS) version 20 Microsoft USA. Frequency tables were generated for simple proportions and descriptive analysis.

Results
Characteristics of newborns enrolled for study Figure 1 shows a summary of the recruitment process. A total of 645 newborns were admitted to the SCBU during the study period. About two-thirds (474) of the 645 admitted were born within the study center while the remainder (177) were referred to it (Fig. 1).
After initial review, 339 of the 645 admitted newborns were enrolled with a presumptive diagnosis of neonatal sepsis. Other isolated organisms were almost evenly distributed among the categories of newborns. There were no significant differences in newborns socio-demographic characteristics and bacteria isolates pattern seen (Table 3).     Table 3.

Discussion
Bacterial isolation rate of 13.6% was observed in this study, this is lower than what was reported in studies within and outside the country. In Ilorin, a bacteria isolation rate of 30%, 18 was reported and 34% in India. 19 However, both the Ilorin and Indian studies excluded subjects with any prior exposure to antibiotics unlike this current study which included newborns with 24 hours or less antibiotic exposure. This could be responsible for the high isolation in the two referenced studies. There may be need for BACTEC use in more studies at various centres to validate its applicability in the tropics where antibiotics is readily available over the counter. 9 Our study found that more than half (53.9%,) of the bacteria isolates were Gram positive organisms, while Gram negative organisms made up 46.1%. This pattern was found in Qatar 20 where a predominance of Gram-positive bacteria was noted (66%) and 16.2% for Gram negative bacteria, in Karachi (54.1%) 21 for Gram positive bacteria and 45.9% for GNB. In Gwagwalada, a report of 58% for GPB and 42% for GNB. 11 Some other studies also revealed preponderance of GPB over GNB. 22,23  organisms. These variations are possibly due to organisms acquired from the birth canal for the early onset sepsis and from the community for the late onset sepsis.
Enterococcus faecalis was the most common gram-positive organism isolated in this current study, while Escherichia coli were the commonest gram-negative organisms. This is a departure from findings from a similar study that was carried out in the unit 10 years earlier, where Staphylococcus aureus was found to be the commonest organism and Klebsiella was the leading Gram negative organism. 11 This further corroborates the fact that, the organisms responsible for neonatal sepsis vary from time to time, even in the same unit, and possible changing pattern of neonatal sepsis. The Enterococcus faecalis isolated did not demonstrate multi-resistant pattern, whuch favours the possibility of a non-nosocomial source of the isolates, even though it is usually associated with nosocomial infection. Additionally, verbal consent was obtained for older children who were able to give approval to be enrolled in the study. Participation in the study was entirely voluntary, and no financial inducement whatsoever was involved. Participants were informed that voluntary withdrawal at any stage of interaction was guaranteed for them without any adverse effect to them. All information was handled with strict confidentiality.

Consent for publication
Not applicable

Availability of data and material
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.