India reports the largest number of neonatal deaths worldwide with a mortality rate of 22 per 1,000 [1]. Although there has been a steady decline in child mortality rate over the last 25 years, neonatal mortality remains high owing to its slower reduction rate [2]. Provision of high-quality newborn care is vital to achieve the United Nation’s Sustainable Development Goal for reducing neonatal mortality to 12 per 1,000 live births by 2030 [3]. Newborns who survive life-threatening complications due to improper perinatal care portray several characteristics similar to the newborns who die of such complications. This similarity has led to the concept of neonatal near-miss (NNM), a criterion to evaluate and improve perinatal care, which allows the detection of risk factors associated with serious newborn complications and death [4]. It is predicted that the number of babies who survive a near-miss is three to six times more than that of neonatal deaths [5]. NNM is a scenario where a neonate with a lifethreatening condition survives the first 27 days of life [6]. PileggiCastro et al. [4] from Brazil developed the criteria for determining NNM based on two global multi-country surveys established by the World Health Organization (WHO). The commonly used and widely accepted NNM measures include three pragmatic criteria based on birth weight, gestational age, and Apgar score at 5 minutes after birth and a management criterion based on the procedures performed and drugs prescribed to the newborn [4].

It is important to identify the predictors or risk factors of NNM in the antenatal period itself, as it enables healthcare professionals and policy makers formulate targeted interventions to prevent or address these risk factors, mitigate health complications in the newborn, promote best practices, and improve the quality of perinatal care. However, there is a paucity of published medical literature on NNM and its predictors which states that more than 80% of the newborn mortality occurs in low to middle income countries. There is a need to address the lacunae in our understanding of what determines NNM in the Indian sociocultural context if we are to achieve the target of single-digit neonatal mortality as outlined by the India Newborn Action plan, by the year 2030 [7]. Therefore, this study was carried out to identify predictors of NNM at a tertiary hospital in Bangalore city.

This was an unmatched case-control study set in with a longitudinal study that aimed to estimate the incidence of NNM over a period of 1 year, conducted in a medical college hospital in Bangalore city in 2018. Institutional Ethics Committee approval was obtained prior to the start of the study. The study involved infants born in the hospital and their sample size was calculated using the unmatched case-control option in Epi-info 7 Stat Calc, with reference to a previous study in Brazil, with advanced maternal age (≥35 years) as the exposure factor [8]. The percentage of exposure among cases of NNM was 15.8% and 5.4% within the controls. With a 95% level of confidence, power of 80%, 10% non-response, and 1:1 ratio of cases to control, we estimated the minimum sample size to test a two-sided hypothesis, which turned out to be 120 newborns with NNM and 120 controls. The NNM cases involved newborns at 7 days after birth, fulfilling Pileggi-Castro’s criteria for NNM [4], either the pragmatic criteria (birth weight <1,750 g, gestational age <33 weeks, Apgar score <7 at 5 minutes of life) or the management criteria (parenteral antibiotic therapy, intubation or mechanical ventilation, phototherapy within 24 hours of life, cardiopulmonary resuscitation, use of vasoactive drugs, anticonvulsants, surfactants, blood products, steroids, or any surgical procedure). For every case that was enrolled, a newborn who did not meet any of the NNM criteria and was delivered on the same day or previous day, was selected as a control. Cases and controls were consecutively enrolled in the study until the sample size was achieved. Only live births were included in this study. Neonates born elsewhere and admitted to the neonatal intensive care unit (NICU) later were excluded from our study. Written informed consent was obtained from the postpartum mothers, and a pre-tested, structured questionnaire was designated. This tool was facevalidated for content by two experts in the field of neonatology and public health. The documentation consisted of three parts: (1) socio-demographic details of the mother; (2) maternal profile, including present and past obstetric history, antenatal care (ANC), investigations, supplements taken, and complications during the current pregnancy; and (3) neonatal profile, including delivery details, birth weight, sex, gestational age at delivery, Apgar scores, complications during and after delivery, and any interventions performed for the newborn. Data were obtained through a combination of interviews with the mothers and through the in-patient hospital records of the mother and baby. Socioeconomic status was determined using the modified BG Prasad classification [9].

A total of 240 newborns (120 NNM cases and 120 controls) were included in the study. The mean maternal age of cases were 26.16±3.48 years, while that of the controls were 22.42±3.48 years. Majority of the mothers were homemakers, among the cases (75.8%) as well as the controls (84.2%). Among the cases, 73 (60.8%) mothers had normal delivery, 45 (37.5%) underwent cesarean section, and two (1.7%) had forceps delivery. Among the controls, 91 (75.8%) newborns were delivered normally and 29 (24.2%) mothers underwent cesarean section. All newborns in our study had singleton births.

Within the NNM cases, 65% had birth weight <1,750 g, 43.3% were of gestational age <33 weeks, 36.7% had Apgar score <7 at 5 minutes of birth, 65% received parenteral antibiotics, 35% had intubation or mechanical ventilation, and 18.3% received phototherapy within 24 hours of birth (Table 1).

The odds of NNM were twice as high among newborns with mothers aged ≥30 years (OR, 2.47; 95% CI, 1.12 to 5.47; P=0.023). It was nearly three times higher for those with maternal education only up to high school (OR, 2.88; 95% CI, 1.71 to 4.86; P<0.001), nearly double for newborns from lower socioeconomic class (OR, 1.76; 95% CI, 1.02 to 3.02; P=0.041) and nuclear families (OR, 1.97; 95% CI, 1.16 to 3.36; P=0.016) (Table 2).

Various components of ANC were found to have a statistically significant relationship with NNM. IFA supplementation for <100 days (OR, 8.97; 95% CI, 4.38 to 18.35; P<0.001) and <3 ultrasound scans (OR, 8.78; 95% CI, 4.89 to 15.78; P<0.001) were both associated with a nearly nine-fold increase in risk of NNM. High ORs were observed for not receiving full ANC (OR, 26.12; 95% CI, 12.49 to 54.66; P<0.001). As compared to ≥8 ANC visits, babies born to mothers with four to seven ANC visits during pregnancy had 27 times greater risk of NNM (OR, 27.32; 95% CI, 3.58 to 52.51; P<0.001), and when mothers had <4 ANC visits, the babies had nearly 70 times greater risk of NNM (OR, 69.23; 95% CI, 9.50 to 110.21; P<0.001) (Table 3).

Multiparity (OR, 2.62; 95% CI, 1.53 to 4.46; P<0.001), hypertensive disorders in pregnancy (OR, 2.20; 95% CI, 1.90 to 2.54; P<0.001), PPROM (OR, 2.46; 95% CI, 2.09 to 2.91; P<0.001), prolonged labor (OR, 2.11; 95% CI, 1.84 to 2.42; P<0.001), obstructed labor (OR, 2.92; 95% CI, 1.10 to 7.75; P<0.001), and malpresentation (OR, 2.08; 95% CI, 1.82 to 2.38; P<0.001), each was found to have twice the risk of NNM, and this was statistically significant. Diabetes in pregnancy was not found to be a statistically significant factor. There was no correlation between NNM and maternal weight gain or previous abortion/intrauterine death/stillbirth (Table 4).

Some factors that were significant in the bivariate analysis lost their significance after multiple logistic regression analysis. These factors include education, socioeconomic status, type of family, IFA supplementation, and parity. The predictors that retained significance included the presence of one or more obstetric complication (hypertensive disorder in pregnancy, diabetes in pregnancy, PPROM, prolonged labor, obstructed labor, and malpresentation), which was associated with a fourtimes greater risk of NNM (AOR, 4.34; 95% CI, 1.26 to 14.95; P=0.02) and maternal anemia in pregnancy (AOR, 18.96; 95% CI, 3.10 to 116.02; P=0.001). Receiving inadequate ANC carried an eight times higher risk of NNM (AOR, 8.35; 95% CI, 1.98 to 51.12; P=0.032). Compared to mothers with ≥8 ANC visits, mothers who had four to seven ANC visits had a five times greater risk of NNM (AOR, 5.45; 95% CI, 1.86 to 14.56; P=0.035) and those with <4 ANC visits had a 17 times greater risk of NNM (AOR, 17.3; 95% CI, 2.89 to 103.68; P=0.035). Having <3 ultrasound scans during pregnancy (AOR, 12.5; 95% CI, 1.60 to 97.27; P=0.016) and a maternal age ≥30 years (AOR, 5.32; 95% CI, 1.12 to 9.29; P=0.041) also had significantly high adjusted odds (Table 5).

An extensive review of literature revealed an insufficiency of data on NNM in the Indian scenario. Studies on NNM by PileggiCastro et al. [4] have used both the pragmatic and management criteria, whereas a study in Vadodara, India, used only the pragmatic criteria to estimate the NNM rate [18]. In our study, we used both pragmatic and management criteria to determine NNM, as this combination allows evaluation of a larger number of surviving newborns considered to be under risk [4]. This is supported by a systematic review by Santos et al. [19], which recommends the application of a combination of the three pragmatic and management criteria (indicating dysfunction or failure of organs and systems) for identification of NNM. Our study highlights not only medical and obstetric predictors of NNM, but also social and economic factors specific to the Indian sociocultural context.

We found that newborns with mothers aged ≥30 years were five times more likely to have NNM compared to mothers aged <30 years. This data was similar to a prospective cohort study conducted in Uganda [20]. Advanced maternal age is linked to adverse neonatal outcomes such as preterm delivery and NICU admissions, as well as to complications such as hypertension and gestational diabetes, which in turn acts as risk factor for neonatal mortality and NNM [21,22].

Our study showed that factors like low education and low income were significantly associated with NNM, which was also comparable with the Ugandan study [20]. Social factors such as education and income are both linked to child mortality and morbidity and shares a reciprocal relationship with mothers having low education and low income or vice versa [23]. This is probably indicative of the fact that low education can result in poorer maternal awareness of health and consequently poor health-seeking and perpetuation of unhealthy practices during pregnancy and peripartum period. Whereas, low income acts as a barrier for maternal nutrition and access to quality ANC and perinatal care. Our study found that newborns from lower socioeconomic classes had nearly twice the likelihood of NNM than those from the upper or middle class. This was also reported in a prospective study in southern Ethiopia [24].

In our study, neonates born in nuclear families had nearly double the risk of NNM compared to those from joint families. In the Indian socio-cultural context, joint or extended family systems are common and can be a potential source of social support to the mother. Family members may encourage or actively conciliate in maternal health care of the mother or may facilitate care by paying for health care costs, accompanying them to health care facilities, and communicating about healthrelated informations [25].

In the present study, complications in pregnancy, such as hypertensive disorders, PPROM, prolonged labor, obstructed labor, and malpresentation, each carried twice the risk of NNM. The presence of any one obstetric complication carried a fourfold increase in the risk of NNM after adjusting other covariates. A WHO multi-country survey on maternal and newborn health, with data from 359 health facilities across 29 countries also confirmed that obstetric complications were predictors of neonatal mortality [26]. This emphasizes the need for screening for complications in pregnancy, with timely referral and appropriate management to reduce NNM.

Our study revealed that ANC is an important predictor of NNM. Although the WHO 2016 guidelines for ANC increased the recommended minimum number of ANC visits from four to eight [27], this has not been implemented across India, as four ANC visits are still considered the minimum. Our study has generated evidence to support a minimum of eight ANC visits to prevent NNM. We found five times higher risk of NNM among mothers with four to seven ANC visits and 17 times higher risk of NNM among mothers with <4 ANC visits, even after adjustment of covariates. Our study found a stronger association between the number of ANC visits and NNM as compared to other studies, such as Tekelab et al. [24] in southern Ethiopia, which found that mothers with <4 ANC visits had 2.4 times higher risk of NNM [22]. As part of essential ANC, women are screened for high-risk pregnancies at each visit and either receive treatment or get referred to a higher level of care accordingly. Fewer ANC visits could result in inadequate screening and management of highrisk pregnancies, and consequently, higher rates of NNM. This is also reflective of the fact that mothers who had <3 ultrasound scans during pregnancy had 12 times greater adjusted risk of NNM. Typically, as a part of routine ANC in urban areas of India, most women undergo at least three obstetric scans: first trimester scan for dating and nuchal translucency, second trimester scan to detect fetal anomalies, and third trimester scan for interval growth and placental anomalies. These scans provide an opportunity to further diagnose high-risk pregnancies and obstetric complications that could benefit from timely interventions, thereby reducing NNM.

Our analysis showed that mothers who did not receive full ANC had eight times higher adjusted odds of newborns with NNM. Antenatal visits, iron-folic acid supplementation, and TT vaccination are well established evidence-based measures to protect against neonatal mortality [28,29]. A minimum of 100 days of IFA is deemed necessary to prevent anemia during pregnancy. Our study revealed that IFA supplementation for <100 days was associated with a nine-fold increase in the risk of NNM. This is a major public health implication as the 4th National Family Health Survey in India shows that only 30% of pregnant women in the country consume IFA for 100 or more, indicating a lacuna in routine ANC [30]. This situation needs to be urgently addressed, as we also found that mothers who were anemic during pregnancy had nearly 20 times higher adjusted risk of NNM than nonanemic mothers.

Our study has revealed important modifiable predictors of NNM, which have implications in public health policy and practice in India. Our findings point to an urgent need for economic development in the healthcare sector to ensure conventional screening, referral, and management of obstetric complications. Ultrasound scans should be included as part of the essential ANC package and should be available, accessible, and affordable for women from lower socioeconomic backgrounds. Policy makers must ensure coverage and adherence for simple, time-tested antenatal interventions, such as IFA supplementation, while including the new WHO recommendation of eight ANC visits as part of the essential ANC package. The quality of routine ANC needs to be upgraded at all levels of the health care system by providing adequate monitoring and mentoring by specialists, a model that has been successfully shown to improve newborn care in North Karnataka, India [31].

The notable limitation of our study is that it was conducted in an urban metropolitan tertiary hospital setting and thereby the socioeconomic predictors may not be extrapolated to smaller towns and rural areas.

In conclusion, our study identified modifiable predictors of NNM, such as maternal age ≥30 years, lack of full ANC, <8 ANC visits, <3 obstetric ultrasounds, anemia in pregnancy, and the presence of any obstetric complications such as hypertensive disorder in pregnancy, gestational diabetes, PPROM, prolonged labor, obstructed labor, and malpresentation. All of these can be diagnosed, treated, and followed-up as a part of routine ANC or obstetric care. Practice of WHO recommendations of eight ANC visits, along with systems strengthening and capacity building, is needed at all levels of health care in order to effectively address these predictors. Pregnant women must be made aware of the need for regular and frequent ANC, and health care providers must utilize the opportunity offered by ANC visits to manage modifiable predictors to prevent NNM.