Globally, studies indicate that institutional delivery has led to improved maternal and child health outcomes [4,5,6,7]. Evidence from India shows that introduction of government programs focused on incentivizing delivery at health facility has been associated with increases in in-facility births and significant reductions in perinatal and neonatal mortality . Within this context, our study explores factors that influence institutional delivery and the role of the ASHA worker. Previous research documents importance of several individual and system level factors as determinants of institutional delivery [11,12,13,14,15,16, 30]. Contributing to this body of evidence, in the discussion that follows, we highlight some of the main findings.
First, we found that frequency of ANC contacts is a crucial determinant of institutional delivery. More specifically, increasing ANC contacts to eight or more was significantly associated with increased odds of institutional delivery. Antenatal care has long been regarded as a core component of routine maternal and child health services and as a strategy to reduce maternal and neonatal mortalities . Evidence from India and other countries in South Asia show that antenatal care is associated with improved maternal and child health outcomes [18,19,20, 23, 31]. In particular, studies demonstrate that increased frequency of ANC contacts are associated with reduced risk of neonatal mortality, and increased odds of institutional delivery [18,19,20]. The reason for this may be that apart from routine check-ups, women who have had higher number of ANC contacts would also be more likely to receive information regarding benefits of delivery at a health facility and knowledge about danger signs and obstetric complications. Most pregnant women in developed nations including the United States are recommended to make far more than eight visits during pregnancy; women in developed countries may make 13 to 14 visits to a health care professional before they deliver . Developing countries considered prenatal visits more seriously only since the United Nations conference on Safe Motherhood Initiative held in Nairobi, Kenya in 1987. This conference drew attention to the fact that over half a million women were dying from preventable maternity complications and recommended member countries to reduce maternal mortality by 50% by 2000. It called for specific actions: a pregnant woman without any complications should make at least 4 visits to a skilled health attendant and more visits as needed if the pregnancy is complicated; and that delivery be done at a health facility attended by skilled health personnel . More recently, studies comparing the reduced visit ANC model (at least 4 contacts) with the recent standard (at least 8 contacts) model found that standard model with eight or more contacts was associated with better maternal, fetal and neonatal outcomes [33, 34]. Additionally, evidence shows that women from high, medium and low resource settings prefer more contact with antenatal services and do not like reduced visit schedules; particularly, valuing the opportunity to build supportive relationships during their pregnancy . Based on this evidence, the WHO recommends eight or more ANC contacts as crucial to reducing perinatal death . A 2018 study in Bangladesh found a significant positive association between eight or more ANC contacts and receipt of increased number of ANC services . While studies in India have examined the association between frequency of ANC contacts and maternal and child health outcomes [18,19,20], none of these studies have focused on the effect of eight or more contacts. Although the current guidelines in India recommend a minimum of 4 or more contacts , our results highlight the importance of the recent WHO recommendation of eight or more ANC contacts, and of the link between frequency of ANC contacts and institutional delivery. Also, to reduce global health disparities in maternal deaths, going forward, it makes sense to aim even higher and encourage women in developing countries like India to increase the number of ANC contacts to eight or more.
Second, ASHA workers are crucial for increasing institutional delivery. While we found that exposure to ASHA worker is associated with increased odds of institutional delivery, addition of ASHA worker variable into the model also increased effect of ANC contacts. Community health workers are key to bridging the gap between health services and utilization, and there is a global need to increase numbers of community health workers and to integrate them directly into health systems . In India, ASHA workers are the cornerstone of the National Rural Health Mission, and are crucial to counselling and mobilizing women to access reproductive health services. Evidence shows that presence of ASHA workers is associated with increased frequency of ANC contacts, increase in skilled birth attendance and facility births, decreased unmet need for family planning and increase in immunization coverage [26, 27, 30]. Qualitative studies indicate that the ASHA workers played an important role in health promotion, awareness generation and counseling on safe delivery, as well as facilitating access to health care services through arrangement of transportation to the hospital . Overall, while studies suggest a positive influence of ASHA workers on maternal and child health outcomes, there is some conflicting evidence on effectiveness of the ASHA program in relation to institutional delivery. A recent study using national level data from India showed that increased placement of ASHA workers had no direct impact on institutional delivery, but was associated with increase in other aspects of healthcare utilization . In contrast, another study found that receipt of ASHA services was associated with increase in maternal health service utilization including antenatal care and institutional delivery . Our study results lend support to this body of evidence and demonstrates that exposure to ASHA worker is associated with increased odds of institutional delivery, controlling for frequency of ANC contacts and socio-demographic factors. Further, we found that the strength of association between frequency of ANC contacts and institutional delivery increased in Model II where we included the variable on exposure to ASHA worker. As previous studies have found a negative correlation between ASHA workers and ANC contacts , we suspected a suppression effect of ASHA worker variable. Several authors have called a variable a suppressor when a suppressor and another predictor are positively correlated with the outcome variable but are negatively correlated with each other . Using the Chi-Square Test of Independence, we found that the variables measuring frequency of ANC contacts and exposure to ASHA worker are negatively, significantly correlated, indicating that ASHA workers served as a suppressor, suppressing outcome-irrelevant variance in ANC contacts resulting in its increase in regression weight [40, 41]. These findings contribute to the body of evidence demonstrating the positive effect of ASHA worker on service utilization, and additionally highlight their potential role in awareness generation and as a motivating factor to access antenatal care. Despite the positive effect of ASHA workers on maternal and child health outcomes, there is considerable regional variation in reported receipt of ASHA services  and our study results show that approximately three-fourths of the sample reported no exposure to an ASHA worker. Future research should focus on understanding the reasons why the ASHA workers have been unable to reach certain sections of the population and what strategies may be used to increase accessibility.
Third, consistent with previous studies, our study results found that socio-demographic factors such as women’s age, spouse’s age, women’s age at marriage, number of children, education, caste, religion and annual household wage were significantly associated with institutional delivery. Education emerges as a strong predictor of institutional delivery in previous research [11, 14] and within our study we found that women who had primary or secondary level education had almost twice the odds of institutional delivery as compared with those who reported having no education. It is likely that educated women would have greater awareness about maternal health care services and the utilization of these services , greater ability to afford health care, and freedom to make health-related decisions . Furthermore, our study results also lend support to prior research which shows that husband’s education level and knowledge about pregnancy-related care emerge as a crucial determinant of utilization of maternal health services [44, 45]. Within our study, we recoded age at marriage into a dichotomous variable and our results indicate that women married after 18 years of age are 26% more likely to have an institutional delivery. Evidence shows that age at marriage influences reproductive health outcomes  and maternal health service utilization ; with risks of stillbirth, miscarriage, and complications during pregnancy being more pronounced among women married below 14 years of age . While the government has implemented programs to increase age at marriage, there is a need for sustained efforts towards increasing knowledge, awareness and access to healthcare among married adolescents. Additionally, compared to Hindus, Muslim women were less likely to have institutional delivery; in contrast, compared with those belonging to forward caste (Brahmin), women who belong to Other Backward Classes (OBC) were more likely to have an institutional delivery. These social conditions vary between states and within different communities in each state. Future research should focus on addressing the underlying reasons of religion and caste based inequities in utilization of maternal health services, in addition to state-level policy formulation that is tailored to specific contexts .
This study has some limitations. The chosen study population only includes women who had at least one birth post 2005, and thus excludes those who gave birth prior to 2005 and their health behavior. This study does not allow us to capture the influence of cultural norms and beliefs as determinants of institutional delivery. Further, we were unable to account for travel time and distance to health facility within our model, which may influence decisions regarding choice of place of delivery. Lastly, we were unable to control for Janini Suraksha Yojana (JSY) beneficiaries within the model and thus cannot separate the effect of visit by ASHA worker and impact of conditional cash transfer through JSY scheme. Despite these limitations, our study makes an important contribution by using national level data to examine the determinants of institutional delivery. This study calls attention to the importance of promoting eight or more ANC contacts among pregnant women in India as a key strategy to improve maternal health outcomes. Also, we find that the role ASHA worker in improving maternal health is further validated. We have utilized the 2012 IHDS data as it contains the required variables on ANC contacts and exposure to ASHA worker.