The prevalence of NNM was determined to be 79 per 1,000 live births in Koshi Hospital, Nepal, using a combination of pragmatic and management criteria. Multiparity and caesarean section decreased the likelihood of NNM. SMM and mothers with no education were found to increase the risk of NNM.
Consensus is lacking regarding a standardized period in which NNM is agreed to occur across countries, which makes it difficult to compare NNM-rates between studies. Some studies have used a near-miss period of 0–6 days (7, 10, 13, 22, 23), while others have utilized 0–27 days (15, 17, 19, 20, 24, 25). Kale et al. recommend extending extrauterine life from seven to 28 days to increase sensitivity of near miss criteria. A decrease in sensitivity was, however, found when it was applied to 0–364 days (26). In the current study, a period of seven days was used because four-fifths of neonatal deaths still occur within the first week of life, with one quarter taking place in the first 24 hours (27). Besides, the chance of information bias increases if NNM information is obtained from parents in the community after hospital discharge.
The prevalence of NNM in this study, when compared, was within the range of previous studies of 45.1 to 72.5 per 1,000 live births, that used the same definition proposed by Pileggi-Castro, et al. (7, 13) . A population-based study in Nepal reported NNM prevalence of 22 per 1,000 live births, which is lower than in the current hospital-based study (18). Possible reasons may be due to differences in used NNM definition and criteria as well as the study settings (15, 20).
Prevalence of NNM was 87.6 per 1,000 live births in two studies from India, using only pragmatic criteria (19, 20). This is higher than the 65 per 1,000 live births (pragmatic criteria only) in our study. A possible explanation is that a survival period of 28 days was used in India; hence, sensitivity increased owing to the more extended survival period.
Multiparity decreased the likelihood of NNM in the current study, similar to other studies (17, 25). Studies from southern and northern Ethiopia, however, reported that multiparity was a risk factor for NNM (28, 29). A recent prospective cohort study in Ethiopia reported that grand multiparity was a risk factor for perinatal mortality among women with MNM (30).
Both nulliparous and grand-multiparous mothers were at high risk of developing complications during birth (31, 32), which places neonates at risk of adverse outcomes (22, 33, 34). Nulliparity among mothers ≥35 years was a risk factor for adverse perinatal outcomes (35, 36). Neonates born to advanced aged nulliparous women had a higher likelihood of admission to NICU (36-38). In this study, however, the proportion of women aged ≥35 years and with >4 children were small, preventing to draw further conclusions.
Literature reviews suggest that a nulliparous women are at increased risk of hypertension and lack experience in childbirth (39, 40). This elucidates the likelihood of NNM affecting nulliparous women (22, 34). Prior studies have shown that first-born infants are at higher risk of neonatal mortality than second- or third-born infants (39, 41). In some studies, however, parity was not shown to be associated with neonatal mortality (42).
Elsewhere, a high chance of NNM affected women undergoing caesarean section (25, 28, 43-45). In recent studies in India and Ethiopia, although NNM occurred more frequently in women who underwent caesarean birth, a direct association could not be established (7, 20). Contrary to these studies, in our study caesarean section was protective against NNM. In support of this finding, caesarean section reduced neonatal mortality in preterm births in the United States (46). A study in the Gambia found that in women with SAMM risks of stillbirth among vaginal birth increased four-fold compared to caesarean birth (47).
WHO recommends caesarean section only when medically necessary and recommends an upper population limit of 15% (48). In Koshi Hospital, the percentage of caesarean birth was 17%, which is higher than 12% in public hospitals in Nepal (4). The proportion of caesarean sections performed in mothers with SMM was two times higher than in mothers without SMM (31% versus 16%). Previous literature shows SMM to be significantly associated with higher percentages of caesarean section and higher numbers of preterm birth occurred in mothers with SMM than without (49, 50). An increase in fetal deaths and higher numbers of babies admitted to NICU for ≥7 days was found together with increased in numbers of caesarean birth (51). A systematic review and meta-analysis showed that maternal and perinatal outcomes were often linked (52). Mothers at high risk of maternal complications often gave birth by second stage caesarean sections to babies with low Apgar scores at 5 min. They were more likely to be admitted to NICU than after caesarean section during the first stage of labor (52).
Risks of intraoperative complications and haemorrhage following caesarean birth are high in low and middle income countries (52). Timely intervention can prevent adverse neonatal outcome among women with SAMM (52, 53). Caesarean section could be a confounder if an operational procedure is performed only among fetus with a greater likelihood of being born alive (54). However, overall, there is a lack of consensus in the literature that neonatal mortality and morbidity are higher in infants delivered by caesarean section (52, 55-58).
In the current study, maternal secondary education decreased the likelihood of NNM. Prior studies have not established a significant association between NNM and maternal education (7, 17, 20, 28, 43). A universal association, however, between maternal education and neonatal mortality, especially in low income countries, has been demonstrated and supports current study’s findings [39, 59, 61, 62]. In addition, educated mothers more likely have a higher socio-economic status, have better knowledge of healthy behaviours, have a more informed approach to self-care, make better health-related choices and utilize healthcare appropriately (33, 59).
The current study found an association between SMM and NNM, consistent with a study in Ethiopia (7), but contradictory to one in Brazil (60). Very few studies, however, have explored the relationship between MNM and NNM. One study showed a strong association (OR 17.15; 95% CI 1.85–159.12), whereas others have not demonstrated a significant association between MNM and NNM (17, 45). Mixed associations existed between haemorrhage and hypertensive disorders during pregnancy and NNM in southern Ethiopia (28) and Brazil (25). In support of the current study, an association between MNM and higher rates of adverse perinatal outcome was found (47, 54, 61, 62). Tura et al. claim that adverse perinatal outcomes among SAMM women is self-evident given the fact that SAMM are identified using severe clinical criteria along with organ disfunction (30).
Among women with SAMM, also NNM is higher (22, 63-65). A considerable number of newborn infants with severe hypoxia, low birth weight and neonatal asphyxia were born to women with MNM [57, 67]. A two-fold increase of stillbirths was found among women with ≥2 complications in the Gambia (47). Similarly, maternal obstetric complications have been shown to play a role in the underlying causes of neonatal deaths (41, 66). Therefore, early screening for poor obstetric conditions during ANC and appropriate management of intrapartum complications are crucial to ensure a reduction of NNM.
The current study did not establish any association between ANC and NNM, unlike a study in southern Ethiopia, where adequate ANC visits were associated with less NNM (28). Attending ≥4 ANC sessions was protective, whereas an inadequate number of antenatal visits was associated with risk of neonatal mortality and adverse birth outcome (65). Possible explanations for non-association in our study were, firstly, that only a quarter (24%) of women in Nepal received all seven components of ANC (67). The majority of Nepal public institutions lack basic ultrasonography and laboratory facilities (blood and urine testing) and most pregnant women only receive health education, iron supplementation, blood pressure measurements and anti-tetanus toxoid (67). Secondly, there is poor compliance by pregnant mothers with ANC advice (68). Hence, women with or without attending ≥4 ANC sessions did not show any association with NNM.
With advancing maternal age, the prevalence of pre-existing conditions appear to increase and also the risk of caesarean birth, contributing to increased fetal risks (69). Advanced maternal age and age <18 or >20 was significantly associated with NNM [15, 27, 38-39, 57-58, 62, 69]. Secondary analysis of the WHO multi-country survey on maternal and newborn health showed that advanced maternal age significantly increased the risk of stillbirths and perinatal deaths (69). No association, however, was established between maternal age and NNM in our study.
Strengths and limitations
Results of the current study are generalizable to births in government institutions in Nepal. To the best of our knowledge, this is the only study that explored factors that impact NNM in South Asia, using both pragmatic and management criteria. The study also had several limitations. The cross-sectional nature meant that causal associations could not be proved. Seventeen of the 44 neonates with conditions requiring admission to NICU were self-referred to private hospitals with unavailable data. Date of the last menstrual period was used to calculate gestational age, possibly introducing incorrect estimations due to recall bias.
Recommendations
Nulliparity and SMM should be considered high-risk obstetric conditions. Therefore, screening should be performed during ANC and, if indicated, referral should be made to hospitals with adequate facilities. Future studies should explore contributory factors to NNM in illiterate women and those with communication barriers, as well as quality of ANC. Evaluation of the risk of NNM is recommended in specific maternal populations, such as advanced aged women, and those with multiple pregnancies. Near misses are lives saved due to timely intervention, so future studies should standardize its definition.