Bacteriological Profile of Sepsis Outbreak in the NICU of a Tertiary Care Hospital in Western Nepal

Address for correspondence Dr. Kalpana Malla A-7, Staff quarters, Manipal Teaching Hospital Pokhara, nepal E-mail: kalpana01malla@gmail.com How to cite this article ? Malla KK, Malla T, Rao KS. Bacteriological Profi le of Sepsis Outbreak in the NICU of a Tertiary Care Hospital in Western Nepal. J Nepal Paediatr Soc 2013;33(1):8-14. Bacteriological Profile of Sepsis Outbreak in the NICU of a Tertiary Care Hospital in Western Nepal


Introduction
N eonatal septicemia remains one of the most important causes of mortality despite considerable progress in hygiene, introduction of new antimicrobial agents, and advanced measures for early diagnosis and treatment 1,2 .Several studies 3,4 including Nepal 5 have observed outbreak of sepsis in their Neonatal Intensive Care Unit (NICU) but the microbials and sensitivity pattern in each varies.Therefore to overcome this problem active surveillance of sepsis and antimicrobial sensitivity of responsible micro-organisms is mandatory in defi ning the empiric antibiotic regimens.This is the largest tertiary care hospital in Western region of Nepal, and serves as a referral center for the population of this region.Hence a heavy nursing workload is there during overloaded admissions which are a major risk factor for sepsis, especially Nosocomial infection (NI).We describe a scenario of such an outbreak.

Materials and Methods
This was a prospective observational study conducted in the NICU unit of MTH, Pokhara.The unit has 22 beds, and its occupancy varies from 12 to 22 throughout the year.The study period was from 1 st April 2011 -15 th August 2011.Ethical approval and informed consent from parents were obtained before starting the study.The inclusion criteria for sepsis was positive Creactive protein plus presence of one or more clinical signs consistent with sepsis, lethargy, refusal of feeds, abdominal distension, vomiting, grunting, respiratory distress, hypothermia, hyperthermia or sclerema, seizures, apnea, color changes in skin, petechia with or without supporting evidence of risk factors such as prematurity, low birth weight, birth asphyxia, maternal chorioamnionitis (maternal fever and/or foul smelling vaginal discharge) and prolonged rupture of membranes (>18hrs).Culture positive cases were the gold standard for diagnosis but in conditions where culture was negative other indirect screening parameters 6 like leukopenia -<5000/mm3, leukocytosis ->20,000/ mm3, absolute neutropenia <1,000, immature/total neutrophil ratio >0.2 were also considered.Patients with respiratory distress syndrome (RDS), gross congenital anomalies, suspected intrauterine infections like Toxoplasmosis Rubella, Cytomegalovirus, Herpes simplex virus infections were excluded.Study patients were categorized as having early onset (72 hours of life) or late onset (>72hrs -28 days of life) septicemia.NI was considered when features of sepsis with positive CRP were noted 48 hours after admission.The environmental sample which is routinely sent for cultures was also taken under consideration.The isolated organisms in newborn (NB) and environmental sample (ES) with antimicrobial susceptibility were analyzed.Epi Info version 3.5.2was used and test applied was F-Test and Chi square test for data analysis.A p value <.05 was considered statistically signifi cant.

Discussion
The outbreak of sepsis is noticed in almost all NICUs.The presumed source of infection is an infected neonate who subsequently causes cross-infection and colonization of the nursery environment.Despite routine hand washing, minimal handling of newborns and environmental decontamination procedures which is practiced in our NICU, we observed a sepsis outbreak.The possible contributing factors for sepsis outbreak can be organisms continuing to spread between neonates via the contaminated hands of health workers, Cots lined up with no space in between, nursing staffs not routinely washing hands while handling newborns, lack of adequate facilities, understaffi ng and overcrowding.It is estimated that up to 20% of neonates develop sepsis and approximately 1% die of sepsis related causes 7 .During three and half month study period we observed 57.22% episodes of sepsis.In same institute in year 2007 5 , a study over 6 years had shown 49.43% episodes of neonatal sepsis which is less than this outbreak.Only 9.31% of sepsis was noted in NICU of Taipei, Taiwan over a period of two years, 3 this indicates that scenario may vary in different settings.Sepsis was more in PT (84.8%) babies.The reason for this is premature babies have low immunity; they need additional supports like ventilation, intravenous fl uids, or blood products, and stay in hospital for longer time.Sepsis has got a special challenge for neonatologists and can be very serious leading to very high mortality.In this study also there was 70% (14/20) sepsis deaths.Forty percent (11) and 7.22% (3) mortality was observed by other authors which is lower than ours.The reason for higher mortality in our study is those patients are referred very late in our hospital when they fail to manage in other centers.
Infections occurring more than 48 hours after admission are usually considered nosocomial 8 .NI occur worldwide, both in the developed and developing world.In this study there was 39.8% of NI.Patients usually acquire the infection during the procedure itself, either endogenously from fl ora on the skin or exogenously from air, medical equipment, doctors, or other staff.In other studies 24.6% 9 and 45% 4 of NI was detected.Another major issue for NI is nurse to patient ratio which was 1:10 to 1:11 in this study.Similar patient to nurse ratio was reported by other authors 10 .In the United States, guidelines recommend a nurse-to-patient ratio of 1:1 for patients who are unstable and severely ill, 1:2 for patients who are stable but severely ill, and 1:4 for patients who are stable 10 .In this study there were 21.35% culture positive sepsis.In another study 35% and 34.88% had a positive culture 11,12 .which is similar to our fi ndings 15 .Though the percentage of culture positive cases are similar the isolated organisms differ in different settings.Group B streptococcal (GBS) sepsis is the most important cause of neonatal sepsis in Europe and North America 13 , but there is a preponderance of gram-negative organisms in tropical and developing countries 14 .The epidemiology of neonatal septicemia within a geographical location, however, also may change with time 15,16 .Group B Streptococcus (GBS) was not isolated in this study.The insignifi cance of GBS as a pathogen in many developing countries is supported by a number of other studies 17,18 .This may be attributable to low prevalence of GBS colonization of pregnant mothers in this area or possibly, to the presence of strains with low virulence.In this study, E. coli was the predominant organisms in newborn sample followed by S.aureus and klebsiella while S. aureus and Klebsiella was predominant organism isolated in ES.The prevalence of E. coli may be due to the fact that it is commonly found as part of the intestinal and vaginal fl ora, and most deliveries are conducted at home, presumably under conditions of poor hygiene.Similar scenario was also observed in an Indian study 19 , but unlike our study they also had high incidence of fungal infection causing sepsis.Yet in another study the 19.2% of fungal infections was reported 20 .Klebsiella and S.aureus were reported also by other authors 21 .The organisms grown in the NB during the outbreak were associated with similar environmental growth 6/9 (66.66%).This signifi es high incidence of nosocomial infection.
Antibiotic susceptibilities: It is diffi cult to compare antibiotic susceptibility patterns between countries because the epidemiology of neonatal sepsis is extremely variable.Most of the isolates in this study showed high rates of resistance to almost all cephalosporins both in NB and ES.Only few isolates like CoNS, S. aureus and E.coli were highly sensitive to Cefuroxime, Ceftazidime, Cefoperazone (R) to third generation cephalosporins (> 80%) was also observed in another study 19 .They also observed (R-50-75%) to aminoglycosides which differed from our study.In our study aminoglycosides were highly sensitive.Netilmycin was S-100% for most isolates [CoNS, Enterococcus, S.aureus, E.coli, Pseudomonas and citrobacter], Gentamycin was S-71.42%-100% to CoNS and S.aureus, Amikacin and Tobramycin was S-100% for pseudomonas and S-50 -80% for Klebsiella, Enterococcus, E.coli.We found Acintobacter highly resistant to all antibiotics except imipenem [S-100% = NB&ES] and Tetracyclin [S-100% = ES].But in other study they were also sensitive to ciprofl oxacin (96.2%), amikacin (92.4%) and gentamycin (87.3%) 22 .Imipenem was also highly sensitive to S. aureus, Klebsella in our study.A 20% resistance to Imipenem was observed in another study, 17 but our study showed no resistance to imipenem.A fi nding similar to ours was noted in another study 12  A study done in the same unit 5 years ago also showed resistance to most cephalosporins, penicillins, aminoglycosides and effective antibiotics were imipenem and cefi pime 5 .
To control infections, prolonged use of broadspectrum antibiotics is often encountered, which leads to the resurgence of multidrug-resistant organisms.Therefore, preventive antibiotics should be used as little as possible, while therapeutic antibiotics should be specifi c and used for short period of time.Possible strategies to be considered might include simple infection control methods such as hand washing and barrier nursing, promotion of clean deliveries, exclusive breast feeding, judicious use of antibiotic, and rationalization of admissions to and discharges from neonatal units.

Conclusion
In conclusion, different NICUs have different epidemiologies of nosocomial infections.Collection of up-to-date data is mandatory for appropriate use of antibiotics, and strategies to avoid the resurgence of multidrug resistant strains should be established.Every unit must follow the bacterial spectrum and antibacterial resistance patterns to choose their specifi c empirical treatment strategy for sepsis.Due to the small sample size and hospital-based design of this study, we recommend additional community-based studies of local patterns and antibiotic sensitivity of pathogens of neonatal septicemia in order to formulate rational antibiotic use policies.

Table 2 &
3 shows the antibiogram of microorganims in NB and ES.The NICU was fully loaded during the outbreak giving nurse-to-patient ratio of 1:10 to 1:11 in each shift (20 to 22 NB and 4 nurses in morning shift, 3 in afternoon and 3 in night shift.There were two registered nurse and rest volunteers (with less experience and no training), rostered on duties.

Term (n=64) Preterm(n=39) F statistics p value
Weight distribution and Sepsis in different gestation age in study population

Table 1 :
Organisms isolated in babies and environmental