The neonatal period carries the highest risk of mortality per day than any other period during childhood. The daily risk of mortality in the first 4 weeks of life is ~30-fold higher than the post-neonatal period, that is, from 1 month to 59 months of age, [1]. The Neonatal sepsis accounts for 25% of neonatal deaths. Neonatal mortality rate in 2017 was 25.4 deaths per 1000 newborns [2]. Among these neonatal sepsis is one of the commonest preventable causes of mortality. In a recent study by Muthukumaran N, sepsis remained the commonest cause of death followed by respiratory distress syndrome and asphyxia [3]. Immediate complications of neonatal sepsis are systemic inflammatory response syndrome, disseminated intravascular coagulation, septic shock and multiple organ dysfunction syndromes which is why it contributes to neonatal mortality. Long-term complications are many and include respiratory, gastrointestinal complications, growth retardation and neurological sequelae [4].
In the NICU, the appropriate management of newborns who require intensive care indicates the need for insertion of an indwelling central venous catheter for fluid, drug and blood component administration. Maintaining a sterile central venous catheter during prolonged stay is a challenge in itself and the line itself getting colonised and eventually becoming the source of nosocomial sepsis is becoming increasingly common. Neonates in the NICU have delayed initiation of breastfeeding and have poor supplemental nutrition during the intensive care. This too contributes to poor defence against infection. Micronutrients play an important role in fighting against infection, and among them, zinc is one micronutrient that is known for its role in immunity and gastrointestinal health.
Materials and Methods
This was a prospective hospital-based observational study, conducted in the NICUs of the Government Hospital of Dakshina Kannada, Karnataka, India from February 2016 to September 2017. Institutional Ethics Committee approval was obtained (IEC number: IEC KMC MLR 01-16/03) to include the newborn including both inborn and outborn admitted at the Government NICU, in the study. Written, informed consent was taken from the parents prior to the recruitment of babies in the study. The sample size was calculated using the below formula, to be 132.
N=4pq/D2where, p=proportion of interest (i.e., 18%), q=1-p and D=relative precision
A minimum sample size of 128 was calculated assuming a confidence level of 95%, power of 80%, relative precision of 7% and Central Venous Catheter (CVC) tip colonisation to be 18% and a non-response rate of 10% [5].
Inclusion criteria: All neonates with a central venous catheter whose initial septic parameters are negative on admission to NICU (clinical sepsis and lab parameters).
Exclusion criteria: Neonates who were positive for sepsis at admission and those who did not require a central venous catheter were excluded from the study.
Methodology
In all the study neonates, their birth details, including maternal details were filled in a semi-structured proforma. Maternal risk factors like chorioamnionitis, maternal fever, urinary tract infection, foul-smelling liquor/vaginal discharge, prematurity and mode of delivery were recorded. Neonatal details like birth weight, perinatal asphyxia and indication requiring NICU admission was recorded. Once admitted in NICU, the following parameters were sent on Day One and again repeated on Day Five:
Total count, Differential count and Immature to total neutrophil (I:T) ratio was calculated
C-Reactive Protein (CRP)
Blood culture and sensitivity
Catheter site culture and sensitivity (drawn from the central venous catheter)
Positivity of total count, differential count (with I:T ratio) and/or CRP were considered as positive for sepsis parameters. All the neonates were observed for the development of sepsis using the Clinical Sepsis Criteria as this was a study planned in a resource limited setting of the government hospitals [6]. The definition of clinical sepsis (CSEP) includes:
Patient ≤1 year of age has at least one of the following clinical signs or symptoms with no other recognised cause: fever (>38°C), hypothermia (<36°C), apnoea or bradycardia/clinical symptoms or signs of sepsis.
AND
Blood culture not done or no organism or antigen detected in blood.
AND
No apparent infection at any other site.
AND
Physician institutes treatment for sepsis.
With the development of sepsis, the treatment for the same for instituted and the blood from the catheter site was sent for culture and sensitivity. The serum zinc levels were measured on Day 5 for all newborns in the study using the colourimetric method using Zinc Kit (Tulip, Coral Clinical Systems, Goa, India). The newborns were considered septic if they qualified for clinical sepsis according to the CSEP criteria. Blood parameters (total count, differential count, I:T ratio and CRP) were considered as additional supportive evidence for sepsis, a positive blood culture as confirmed sepsis and all these factors analysed against their serum zinc levels.
Statistical Analysis
The results were then analysed for the maternal and/or perinatal risk factors causing development of neonatal sepsis and/or catheter site sepsis and correlated with their serum zinc levels. Data were analysed using SPSS version 17.0 by the Mann Whitney test, Wilcoxon W test. The Z score and p-value were calculated. (p-value of <0.005 was considered statistically significant).
Results
A total of 132 newborns were included in the study and they were classified as per their gestational age, birth weight, maternal risk factors and perinatal asphyxia to tabulate their demography as in [Table/Fig-1]. Serum zinc levels in the newborns were calibrated as >55 μg/dL being normal and <55 μg/dL as deficiency [5] following which 19 newborns (14.3%) were found to be zinc deficient. The newborns who were proven aseptic on Day 1 were followed-up on Day 5 for the signs of clinical sepsis along with the sepsis markers. Out of 132 newborns, 51 developed clinical sepsis as per the above mentioned CSEP criteria.
Demographic Characteristics of the study population.
| Demographic Characteristics | Frequency (Percentage) |
|---|
| Gender | Male | 72 (54.5%) |
| Female | 60 (45.5%) |
| Gestational age | Term | 46 (34.8%) |
| Pre-term | 86 (65.1%) |
| Weight | Small for GestationalAge (SGA) | 12 (9%) |
| Appropriate for GestationalAge (AGA) | 103 (78%) |
| Large for Gestational Age (LGA) | 17 (12%) |
| Zinc levels(Normal 60-110 mcg/dL) | <55 (μg/dL) | 19 (14.4%) |
| >55 (μg/dL) | 113 (85.6%) |
| Perinatal Asphyxia | Present | 86 (65.1%) |
| Absent | 46 (43.9%) |
| Maternal risk factors for early onset neonatal sepsis | No risk factors | 85 (64.4%) |
| Risk factors for neonatal sepsis present | 47 (35.6%) |
When the sepsis markers were tabulated, 68 showed a positive CRP and 28 a positive I:T ratio, as detailed in [Table/Fig-2]. The blood culture reports sent showed positivity in 32 cases. The blood culture positivity in the study was 24.24%. The organisms that were cultured were variable, hence could not be correlated with the other parameters (Klebsiella, Staphylococcus aureus, Candida sps, Acinetobacter, Pseudomonas, Escherichia coli, Streptococcus and Enterococcus). The catheter site culture also grew similar organisms (31 cases), with a catheter site culture positivity of 23.4% but the same organism was cultured from both blood and catheter in only 15 cases (6 being Candida, 5 Acinetobacter, 4 showing staphylococcus growth). 12 of the cases had either one (blood or catheter site) showing positive growth, and in five cases, the cultures yielded different organisms wherein the catheter site culture was considered as the colonised organism as it grew nosocomial organisms like Acinetobacter and Candida. The newborns were treated for sepsis and the haematological parameters were correlated with their serum zinc levels. The zinc level in the study population was low in 19 (14.4%) newborns, the cut-off was taken as 55 μg/dL as per a study done by Terrin G et al., [5].
Details of clinical sepsis and sepsis markers on Day 5.
| Sepsis/sepsis marker on Day five | Number | Percentage |
|---|
| Clinical Sepsis (CSEP criteria) | Present | 51 | 38.3% |
| Absent | 81 | 61.3% |
| CRP (normal <6 mg/dL) | Positive | 68 | 51.5% |
| Negative | 64 | 48.5% |
| I/T Ratio | <0.2 | 108 | 81.8% |
| >0.2 | 24 | 18.2% |
| Blood Culture | Positive | 32 | 24.2% |
| Negative | 100 | 75.8% |
| Catheter site C/S | Positive | 31 | 23.5% |
| Negative | 101 | 76.5% |
Upon comparing the parameters of sepsis with the serum zinc levels [Table/Fig-3], newborns born to mothers with risk factors for sepsis showed a lower serum zinc level (134 μg/dL) as compared to those born to mothers without any such risk factors; however, this was not statistically significant. Similarly, the gestational age of the baby did not show any difference in the zinc levels of term and preterm newborns. Clinical sepsis positive newborns had also lower serum zinc values (138 μg/dL) as against clinically aseptic newborns (145 μg/dL) but without statistical significance. The presence of perinatal asphyxia in the newborn along with the development of sepsis had lower serum zinc values as compared to non-asphyxiated newborns and this showed statistical significance. Among the haematological markers, I:T ratio in the septic newborns, blood culture positivity and central catheter site culture positivity also showed similar lower zinc levels but did not exhibit statistical significance.
Correlation between the risk factors for development of neonatal sepsis along with the serum zinc levels.
| Parameter | Number/Percentage | Zn levels | Mann Whitney Test |
|---|
| Median IQR values | Z-value | p-value | Statistical significance |
|---|
| Maternal risk factors | Normal mothers | 85 (64.4%) | 152.87 | 118.5 (77.16-163.95) | -0.684 | 0.494 | Not significant |
| High risk mothers | 47 (35.6%) | 132.1 | 99.83 (81.4-168.3) |
| Gestational age | Term | 46 (34.8%) | 147.86 | NA | NA | 0.329 | Not significant |
| Pre-term | 86 (65.2%) | 145.29 | NA |
| Clinical sepsis | Present | 51 (38.6%) | 138.7 | NA | NA | 0.454 | Not significant |
| Absent | 81 (61.3%) | 145.6 | NA |
| Perinatal Asphyxia | Present | 86 (65.1%) | 126.3 | 94.2 (69.13-118.82) | -3.009 | 0.003 | Significant |
| Absent | 46 (34.8%) | 156.8 | 131.75 (87.3-172.02) |
| CRP | Positive | 68 (51.5%) | 130.8 | NA | NA | 0.679 | Not significant |
| Negative | 64 (48.5%) | 138.7 | NA |
| I:T ratio | >0.2 | 24 (18.2%) | 123.1 | 107.9 (67.81-149) | -0.861 | 0.389 | Not significant |
| <0.2 | 108 (81.8%) | 130.23 | 111.25 (80.83-168.3) |
| Blood culture | Culture positive | 32 (24.2%) | 142.4 | NA | NA | 0.776 | Not significant |
| Culture negative | 100 (75.8%) | 153.9 | NA |
| Catheter site C/S | Positive | 31 (23.5%) | 146.1 | NA | NA | 0.893 | Not significant |
| Negative | 101 (76.5%) | 140.9 | NA |
Discussion
In this study, 132 newborns were included who required NICU care and had an indwelling central venous catheter. The demography of the present study population was similar to a study done by Tsuzuki S and Morimoto N, where 44 newborns were included; serum zinc was measured at birth and on Day 5 [6]. There was no correlation between the zinc levels and the birth weight in that study. The prevalence of zinc deficiency in newborns in this study was 14.4%. The risk factors for the development of sepsis in newborns that were analysed were the presence of maternal risk factors, perinatal asphyxia and gestational age. The presence of maternal risk factors for early neonatal sepsis were not found to be statistically significant with serum zinc levels in the study and no specific references could be found for this, though study by Terrin G et al., have shown that supplementation of zinc has been beneficial in reducing the sepsis rates in the presence of maternal risk factors [7]. In this study, mean birth weight was 2.17 kg and mean Zinc level of preterm neonates (145.29 μg/dL) was nearly same as that of term neonates (147.86 μg/dL). In a study conducted by Iqbal AS et al., showed no significant correlation between gestational age and serum zinc levels [8]. Other study done by Terrin G et al., state that serum zinc levels in cord blood of preterm neonates is higher than term neonates and falls rapidly thereafter, which could explain the similar values that were noted in the present study on day 5 [5].
Similarly, when the occurrence of perinatal asphyxia was analysed with the serum zinc levels, it was found that neonates with perinatal asphyxia had lower zinc levels as compared to neonates without perinatal asphyxia. This association is statistically significant with a p-value of 0.003. However, authors could not find any similar reference articles with this co-relation. Development of clinical sepsis and serum zinc levels did not show statistical significance though in the study it was observed that aseptic newborns had higher zinc levels than clinically septic newborns. However, the role of zinc in immunity has been elaborated by Terrin G et al., and this can give scope to further studies on supplementation of zinc in the NICU and analysing the development of sepsis [5]. Upon analysing the haematological parameters, Low I:T ratio and positive CRP were found associated with lower serum zinc levels. This can be explained by the concordance of clinical sepsis with a positive sepsis screen.
When culture positivity was analysed, numerous organisms had grown in the blood cultures of the septic newborns and hence could not be analysed with respect to the organism. The culture positivity in the study was 24.24%. In other studies, blood culture positivity ranges from 25% to 42% but this value is dependent on multiple factors like volume of blood inoculated, the bacteraemia in the blood, empirical antibiotic usage and laboratory issues [9]. In low levels of bacteraemia the culture shows as less as <1 CFU/mL and hence this explained the low culture yield in this study [10]. Catheter site cultures showed similar positivity of 23% however there was concordance in only 15 cases (50%). The serum zinc levels in newborns who showed positive blood culture and catheter cultures were lower than those with sterile cultures, though statistically not significant. These findings imply the possibility of another study with zinc supplementation and neonatal sepsis.
Limitation
The analysis of sepsis could not be done with specific organisms in view of the varied culture growths. This study had a small sample size. Larger sample size with age-matched controls will be useful in deciding the need for zinc supplementation in asphyxiated newborns to reduce sepsis.
Conclusion
The risk factors for the development of sepsis in newborns with indwelling catheters were the presence of perinatal asphyxia in the presence of zinc deficiency. Other risk factors like the presence of maternal risk factors, gestational age, birth weight had lower zinc levels, as did newborns who developed signs of clinical sepsis, however, this was not statistically significant. Culture positivity (blood/catheter site) was found to be 24% and did not show any statistical significance with zinc deficiency. Further studies are needed to establish the role of zinc in neonatal sepsis.