Preventable contributors to the neonatal healthcare-associated infections: a uni-center analytical study from South India.

Introduction

Reports from UNICEF and the World Bank 2018 showed a reduction in Neonatal Mortality Rate (NMR) among all the participating countries. India’s NMR is 23 as per 2017 report.^1,2 Each year nearly 0.748 million newborn deaths occur in India, contributing to 26% or 1/3rd of the world’s neonatal death. As per WHO, in developing countries like South East Asia, HAIs are responsible for nearly 50% of mortality in Neonatal Intensive Care Unit (NICU).³ The rate of healthcare-associated infections (HAIs) reported by the World Health Organization (WHO) in South Asian countries ranges from 11.3-23.6%.⁴

According to Liu et al., dominating causes of neonatal deaths are preterm (35%), birth asphyxia (20%), pneumonia (16%), sepsis (15%), and other causes (<10% each group), these are also the primary reasons for ICU admission and development of HAIs.⁵ Indian studies have reported bloodstream HAIs, Ocular infection (Healthcare-associated conjunctivitis), urinary tract infection, and skin infection among neonates.^6–11

Newborns needing critical care and support get hospitalized in Neonatal Intensive care units (NICUs) are at increased risk of acquiring Healthcare-associated infections (HAIs).¹² HAIs have a detrimental effect on the recovery, length of stay in NICUs, and immune system acting as a vicious cycle.¹³

Neonatal HAIs is one of the preventable leading cause for neonatal mortality. Developed countries are putting efforts to identify the risk factors to neonatal HAIs and its preventable measures.^14,15 The contributing factors to neonatal HAIs in India are unexplored among neonates.¹⁶ The rate of HAIs and their related mortality and morbidity is also not explored in Indian literature. To contain the spread of microorganism, domestic preventable contributors to neonatal HAIs needs to be explored. The objective of the study was to identify the contributors to neonatal HAIs in one study centre.

Methods

The study collated and analysed the contributing factors from December 2016 - June 2018. The study was conducted in one of the largest tertiary care teaching referral hospital in south India. Manipal Academy of Higher Education (MAHE) ethics committee provided Institutional Ethics Committee (IEC) approval was taken, approval ID: MUEC/014/2016-17. The study setting is equipped with a 30-bedded level III Neonatal Intensive Care Unit (NICU) with mechanical ventilation, indwelling catheters, intravenous fluid, phototherapy, and an advanced monitoring facility.

ICU hospitalized neonates were followed to development of neonatal HAIs. A case-control study design was adopted with controls matching the gestational age to identify maternal and neonatal direct contributors. The neonates were defined from the day of birth ‘0’ days to ‘28’ days; however, each day of life was counted from 24 hours of birth, not at midnight.

In order to compute the sample size, the probability of HAIs in current settings was obtained from a pilot study data of the same study setting and found to be 8 per 100 admissions >48 hrs. We assumed the probability of acquiring neonatal HAIs (p) as 0.08 and the probability of not acquiring neonatal HAIs (q) as 1-p = 0.92. With a 95% confidence interval using estimation of single proportion sample size calculation method d (accepted error) = 0.05. Sample size computation was made as below: $\mathit{s}\mathit{=}{\mathit{Z}}^{\mathbf{2}}\frac{\mathit{pq}}{{\mathit{d}}^{\mathbf{2}}}$ = 113 neonates with HAIs for infinite population and 106 for the finite population. We assumed 10% missed events; hence another 12 cases were added in the sample pool leading to 118 samples of neonates acquiring HAIs with similar numbers in the control group.

The case detection and confirmation were carried out for any newborn admission to NICU >48 hrs using a combination of CDC and WHO recommended clinical findings and diagnostic findings, where the presence of at least any two variables of each category confirms the case label as HAIs.^17,18 The following clinical and diagnostic criteria were used to determine the neonate acquiring HAIs. The presence of any two clinical and diagnostic tests was considered as the case of positive neonatal HAIs (Table 1). The Case Record Form (CRF) was prepared keeping objectivity as priority to clearly discriminate between neonate acquiring HAIs versus other sick babies. The CRF was pilot tested for construct, content and criterion validity before the commencement of the study.

Neonates admitted from other hospitals with sepsis and maternal infections like UTI/Chorioamnionitis/Pneumonia were excluded from the study. Each morning of a working day the duty clinician as well as the nurse in charge was enquired for any probable case of HAIs by the researcher. If it was a holiday or weekend the next working day the enquiry was raised on suspected case of neonatal HAIs who was in NICU since last 48 hours. Only infections originating at the study site after 48 hours of admission were identified and included in the study as per the clinician ascertainment and the diagnostic criteria determining HAIs. The data on these neonates were obtained from medical records.

Bed occupancy days, admission and discharge in total per month, and further inborn and outborn categorization is reported in numbers per month. The rate of HAIs was calculated per 100 neonatal admissions for >48 hours.

Rate of HAIs per 100 neonates admitted for >48 hours = (# of new HAIs/# of all admissions in respective month >48 hours) × 100

Bacteraemia rate was calculated by dividing the number of new cases by the total number of bed occupancy days and multiplying it by 1000.

Bacteraemia rate per 1000 days = (#of HAIs per month/Total number of Bed Occupancy days in the respective month) × 1000

A Chi-square test was performed on the birthing place and HAIs to find any statistical significance among two variables where the level of significance was fixed at <0.05.

The variables were entered in Microsoft Excel sheet within 48 hours of data collection and the neonate was followed till its outcome as improved/succumbed/discharge against medical advice. If the any variable input was missed the clinician and nurse care giver was asked to provide the relevant information on missed data. Missed data was manually filled post completion of the enquiry. The descriptive variables were reported using mean and standard deviation when normally distributed. Median with interquartile range was reported for skewed data. The frequency and percentage table were used for the nominal and categorical variables. Boxplot was used for skewed data to identify and report the outliers, however, further analysis with or without outliers was not performed.

Results

Identifying the neonatal contributors causing neonatal HAIs

Eighty-three (70%) neonates acquiring HAIs were delivered through cesarean section (C-Section), similar to studies reported from India.¹⁹ Most of the neonates were improved and discharged (71%) from NICU; however, mortality (25%) and discharge against medical advice (4%) outcome was recorded, in control group there was 19 (16%) neonates died and 99 (84%) neonates improved (Table 2). The overall mortality during the study period in the entire NICU was 5% among all neonatal admissions, whereas there was 25% mortality among the neonates acquiring HAIs, contributing to nearly 1/4^th mortality (24%) among all the admissions to NICU. The Odds of mortality were found to be insignificant (Odds ratio=1.777; 95% CI=0.93-3.39; p=0.0814) among cases with HAIs compared to non-HAIs cases.

The male gender (59%) acquired HAIs more than the female gender (41%). Moderate to late preterm neonates (38%) acquired HAIs more compared to the rest of the categories. Mean gestational age was 32±4.3; the gestational age showed near-normal distribution among neonates with HAIs.

On univariate analysis, we did not find any birth weight category as a contributor to neonatal HAIs both the Odds ratio and chi-square test were insignificant to HAIs; ELBW (p=0.339), VLBW (p=0.7833), LBW (p=0.6353), NBW (p=0.09).

Most of the neonates acquiring HAIs were born through C-section (71%), were VLBW with a median birth weight of 1370 (IQR=530, 3860) grams. The majority of the neonate acquiring HAIs were male gender (59.3%) and were moderate to late preterm (39%) with VLBW (35%). Further analysis between gender and birthweight showed that the male gender with VLBW (19%) was more prone to HAIs than females in a similar birth weight category (16%) (Table 2).

Overall extreme preterm neonates (38%) had detrimental outcome (mortality) as compared to the rest of the gestational age group. Neonate with moderate to late preterm with LBW acquiring HAIs had higher mortality than any other gestational age or birth weight. Neonates had birth weight ranging from 520 g to 3850 g showed marked improvement in health conditions over time and were discharged (Figure 1).

Figure 1. Outcome, gestational age and birth weight distribution of the neonates with HAIs (1K = 1000 grams).

Discussion

As per a systematic review burden of HAIs ranged from 3.6 to 11.6 per 100 neonatal admissions, whereas in low-middle income countries, it ranges from five to 19 per 100 neonatal admissions.⁴ The incidence density of HAIs in the USA and Europe ranges from 13.0 to 20.3 incidence per 1000 hospital bed days. A multicentre study from Canada and another one from Germany reported the rate of HAIs as 23.5% and 12.3% among 100 neonatal admissions.^20,21

In Brazil rate of HAIs ranged from 12.3 % in NBW neonates and up to 51.9% in very low to extreme low birth weight neonates (ELBW). The overall incidence density bacteremia rate was 24.9 per 1,000 hospital bed days.^22,23 Another large study from Germany has reported BSI incidence as 6.5 per 1000 hospital bed days.²⁴ In South Asia, the incidence density of neonatal HAIs reportedly is 9.8 per 1000 live births.²⁵

The rate of HAIs and the bacteremia rate in the current study was similar to a study conducted in Italy where the rate of HAIs was 9%, but incidence density was lower 3.5 per 1000 hospital bed days compared to the current study.²⁶

Neonates acquiring HAIs and delivered to mother with PROM had no different outcome, however, is considered as a risk for HAIs^27,28 and were born to very preterm and moderate to late gestational age. Maternal peripartum infection,²⁹ UTI and leaking per vaginal had no different outcomes among neonates acquiring HAIs. However, the number of cases with HAIs were very few to analyse.

Mothers with preterm labor observed high neonatal mortality (35%), India witnesses the highest preterm deliveries in the globe and 50% of neonatal death occurs in preterm (<37 weeks gestation).³⁰ We observed high mortality among moderate to late preterm neonates (34%) born to mothers with preterm labor, probably due to the high number of neonates in this category acquiring HAIs.

Neonates acquiring HAIs and delivered to mother with PROM had no different outcome, however, is considered as a risk for HAIs^27,28 and were born to very preterm and moderate to late gestational age. Maternal peripartum infection,²⁹ UTI and leaking per vaginal had no different outcomes among neonates acquiring HAIs. However, the number of cases with HAIs were very few to analyse.

Mothers with preterm labor observed high neonatal mortality (35%), India witnesses the highest preterm deliveries in the globe and 50% of neonatal death occurs in preterm (<37 weeks gestation).³⁰ We observed high mortality among moderate to late preterm neonates (34%) born to mothers with preterm labor, probably due to the high number of neonates in this category acquiring HAIs.

Another known contributor to neonatal HAIs is mothers unpasteurized breast milk that was contaminated with MRSA leading to neonatal HAIs in one of the studies, however, we did not find any such contributors during our study.³¹

Preterm labour, that was significantly associated with neonatal HAIs alone may not be able to contribute to HAIs, it can lead to lower gestational age and LBW delivery of the neonate that might contribute to a more significant extent on acquiring HAIs.

Basu et al.³² found no significant association between the occurrence of HAIs and mode of delivery but contradictory to another study published from North-eastern India where they found a significant association between vaginal delivery (p=0.002) and occurrence of HAIs,³³ we did not find any association with mode of delivery.

The mortality of near 25% was similar to the study by Bammigatt et al.³⁴ where they also reported 24% mortality with no statistically significant (p>0.05) among cases with HAIs.

HAIs are always known to cause higher related mortality and cause of concern among all the risk factors for mortality.^12,35 Neonatal mortality among developing nations due to HAIs ranges from 4% to 56% of all causes of mortality during the neonatal period,³⁶ and our study found 24% of mortality was due to HAIs.

We found a predominance of male (59.3%) gender in neonatal HAIs in similar studies from the same region where they found a male predominance of 62.3%, and in another study, it was found to be 1.3:1 for the male to female ratio in acquiring HAIs in India.^8,28 A systematic review on neonatal sepsis found that the male gender (OR: 1.3, 95% CI: 1.02, 1.68) is a risk factor.³⁷

Those born preterms had a 92% increase in the risk of getting HAIs compared to other gestational ages, as reported in studies from developed and developing nations where preterm birth is a risk of acquiring HAIs.^25,38,39 The number of moderate to late preterm neonates who acquired HAIs was more than other gestational age groups, and the reason could be the higher number of admission for 48 hours and beyond to NICU of moderate to late preterm neonates compared to other gestational age.^40,41 The majority of neonates in this category was admitted for respiratory distress syndrome.

The neonatal HAIs among neonates with birth weight >1500 grams (62%) was reported as significant from north India.⁴² Another large study from Germany has reported a high prevalence of HAIs among VLBW neonates.²⁴

LBW neonates are more susceptible to neonatal HAIs than NBW categories,²⁶ but our findings were in contrast. In our study, most of the neonates were in the VLBW category (35%); there was an equal proportion of birth weight distribution among ELBW (23%) and LBW (23%); however, the NBW neonates (19%) were very few in numbers.

VLBW and prematurity were identified as factors contributing to neonatal HAIs in a study from western India.⁴³ VLBW with very preterm neonates and moderate to low birth weight with moderate to late preterm were found susceptible to HAIs in our study, similar to a study from Northern India that reported LBW as a risk factor to neonatal HAIs.⁴⁴

However, in this study, we found only three VAP cases, and the majority were bloodstream infections still, the presence of mechanical ventilation was identified as a contributor to HAIs. Mechanical ventilation with intubation was found to be a risk to neonates causing HAIs.^45,46 The neonates exposed to non-invasive ventilation had an insignificant risk of acquiring HAIs, similar to other studies showing that the increased duration of NIV utilization poses a higher risk of acquiring HAIs.^24,47

In this study, 97 (82%) of the neonates were kept nil per oral (NPO) before developing HAIs that could have been a reason for lower immunity and infection as reported in the literature.^48,49

The placement of the umbilical catheter line did not show significant risk to neonates for HAIs in contrast to other studies where the author found an association in umbilical catheter insertion to HAIs.^40,50 We did not find any study evaluating association with the cannulation procedure in less than four days to greater than one day to the occurrence of HAIs.

We found a frequent change of IV cannula (for > 3 times) leading to multiple pricks in the skin and use of mechanical ventilation (< 3 days) was found frequent among neonates acquiring HAIs. However, further analysis could not be carried out. A study by M. Takrouri et al. suggests changing the IV cannula every 48 hours to prevent colonization and further infection to the patient,⁵¹ but our study finds a frequent change in IV cannula poses a higher risk to neonatal HAIs. Although umbilical catheters had no significant association with HAIs, changing the central venous catheter/umbilical catheter every 10 days or less to prevent HAIs may have a better outcome.⁵¹ Different therapeutic procedures, both invasive and non-invasive, significantly contribute to neonatal HAIs as per the large multicentre study of the German Neonatal Network.⁵² To curb the development of HAIs, it is necessary to reduce the number of procedures, duration of invasive and non-invasive procedures, especially to neonates born due to premature labor and very low birth, as they pose a higher risk of acquiring HAIs.

This study had a few limitations as data collection was carried out in one study centre. In order to identify potential case of HAIs the researcher was dependent on the discretion of the duty physician and nurse. Physician initiated studies and in closed ICU would have provided further insight to the outcome. It would have been better to have cohort study rather case-control study design keeping track of all the patients irrespective of the underlying disease or HAIs. The record on number of IV site pricks and reason to remove existing IV line for all the neonate was not available and hence further analysis on this potential factor could not be carried out. Maximum data was captured from medical records of the neonate rather prospectively collected. There could be miss cases of HAIs due to selection bias created by the study data collection method that could affect the results of the study. There was no practice on collection of swab samples from various parts of ICU except on two occasions, where there was suspected outbreak of two microorganism during the study. During these two occasions identified source was from the cradle of the neonate and washbasin of NICU. Probability of missing source of infections cannot be ignored while interpreting the findings. There are many other cofactors that could be contributor to neonatal HAIs like hand hygiene of healthcare workers,⁵³ this was analysed and reported elsewhere but its establishment with cases of HAIs was not ascertained. There are other factors that were identified as contributor to neonatal HAIs elsewhere like ELBW, MV, umbilical catheter, were not found as contributor in this study setting.

As this was a uni-centre study, the results cannot be generalized to the whole population. The explored factors are similar to the published literature but some new factors that were identified and highlighted need further exploration to determine as the contributor to neonatal HAIs. These contributors could be related to only this study setting considering the policy and practices of healthcare workers. Furhter multicente cohort studies considering capturing the data on this list of factors could help to bring out associated preventable contributors. There are variabilities on neonatal care practices that could be a potential contributor requiring further analytical studies like decision on removal of a central line/umbilical catheter. There is a dearth of published literature from India requiring further research and reporting on factors contributing to neonatal HAIs.

Conclusion

Bloodstream infection (83%) was prevalent, causing neonatal HAIs. Mechanical ventilation, NIV, and PICC line on univariate analysis contributed to neonatal HAIs. Although on univariate analysis, mechanical ventilation for > three days duration, NIV for > five days, and PICC line insertion procedure were contributors to neonatal HAIs but on multivariate analysis, NIV, PICC line, preterm labor, and LBW were found as contributors to neonatal HAIs. IV cannulation more than three times in four consecutive days was associated with neonatal HAIs; this needs further studies to find any correlation as a contributor to the neonatal HAIs.

However, mechanical ventilation, umbilical catheter, PROM, ELBW, LBW, and NBW did not pose a statistically significant risk to acquiring HAIs. A larger multicentric study from India will be required to establish further evidence.

Data availability

Authors’ contributors

All the authors conceptualized the research idea and contributed to the manuscript. Usha Rani conceptualized, designed, analyzed, and reported the data. Leslie E. Lewis carried out technical guidance, data analysis verification, and data reporting. Usha Rani and Kiran Chawla developed a data extraction form, and later interpreted the analyzed data. Usha Rani, and Anup Naha, analyzed and reported the data. Usha Rani prepared the discussion based on the results. All the authors have proofread and has provided their intellectual contributions in the manuscript. All authors have approved the final results of the manuscript for publication.