The study is based on a publically available secondary data set with no identifiable information on the survey participants. This dataset is available in the public domain for research use and hence no formal approval from the institutional review board is required. So, no ethics statement required for this work.

We used data from the third round of the District Level Household Survey (DLHS-3) conducted in 2007–08 in 601 districts from 34 states and union territories of India. DLHS-3 adopted a multi-stage stratified systematic sampling design which resulted in national and state-representative samples after applying sampling weights to control for complex survey design [7]. A total of 643,944 ever married women aged 15–49 years and 166,260 unmarried women aged 15–24 years were interviewed in the survey. Data on postnatal care were collected from all women who had a birth in the five years preceding the survey date, irrespective of whether they sought delivery care at home or in a health facility.

The outcome variable considered in the analysis was the type and timing of postnatal care. The availability of detailed information in DLHS-3 provided us with opportunities to analyze inequalities in type and timing of postnatal care by place of birth, complications during birth, and cesarean birth. In addition, the analysis considered antenatal care and care at birth (home or facility).

The main explanatory variable of interest was a proxy measure of household economic status. Direct data on income or expenditure are not available in the DLHS-3 and in circumstances where such data available in retrospective surveys are subject to reporting bias, a wealth index is computed based on the ownership of household assets and consumer durables [20]–[26]. Other demographic and social variables that influence maternal health care behaviour were also added including woman’s level of education, her place and region of residence, birth order, and gender status of the child.

Two measures were used to assess the extent of inequalities in the use of PNC: the rich-poor ratio (computed by dividing the top quintile by the bottom quintile of the wealth index) and the concentration index (CI - which measures the relationship between accumulated proportions of mothers ranked by their socio-economic status against the cumulative proportion of postnatal care use) [25], [27]–[29]. The values of the CI range from −1 to 1. A value equal to 1 or −1 indicates that only the richest or the poorest mothers use postnatal care respectively. These measures were also used to compare the differences in PNC with those receiving antenatal care and care at birth between home and facility and between births with and without maternal complications. Binary logistic regression models were fitted to assess the adjusted effect of wealth quintile on the likelihood of using PNC after adjusting for antenatal and childbirth care, and other relevant socio-economic and demographic characteristics. The explanatory variables were initially screened for potential multicollinearity before considering those in the regression models.

Only 44% of the mothers interviewed in the survey received any PNC check-up within 48 hours of giving birth (Table 1). Moreover, only 45% of the newborns were checked within 24 hours. Around 62% of the babies did, however eventually receive two or more check-ups within the first 10 days after birth. As expected, a majority of these babies were examined in a private facility (55%).

In contrast antenatal care was utilised by more women – about 73% of the mothers reported availing some form of care for their recent birth. However, only about 42% received the recommended four or more visits. Only 46% of women received advice on institutional care for childbirth. About 57% of the births took place at home or at places other than a health facility and about 58% of the mothers reported experiencing complications during childbirth. The majority of women had vaginal births while about 8% were cesarean sections.

Rich-poor ratios are presented in Table 2. Findings suggest enormous socio-economic inequalities in use of PNC services. Moreover, the socio-economic inequalities also varied across the continuum of care consisting of antenatal, birth and postnatal care services respectively. For instance, the rich-poor ratio varied from as low as 1.7 in case of receipt of any form of ANC to as high as 3.4 as regard to the compliance with 4 or more ANC visits. The rich were also four times more likely to have had an institutional birth when compared to the poor. The use of PNC was particularly unequal with the rich utilizing these services three times more compared with the poor. The rich-poor ratio in case of babies receiving two or more check-ups within the first 10 days after birth, however, was only 1.8. Further, babies belonging to the richer households were more likely to be examined in a private facility compared to the poor who were more likely to be examined in the government facilities.

Estimated concentration indices further confirm the wide socio-economic inequalities in the use of selected services (Table 2). The highest inequality was found in the use of PNC for newborns within 24 hours of birth. The rich were much more likely than the poor to get their babies seen by a health worker within 24 hours of birth. In addition, they were also more likely than the poor to get their newborns examined in a private facility. On the other hand, poor were more likely to receive a newborn check-up in a government facility. The rich were also relatively better in terms of receiving ANC advice than the poor. Findings clearly suggest that PNC was particularly unequal compared to other services.

The differentials in use of PNC by place of birth were substantial; the use being much higher in the case of facility births compared to home births. Not surprisingly, the use of postnatal care was limited for mothers who gave birth at home (Figure 1). The findings based on concentration indices suggest higher socio-economic inequalities in selected components of postnatal care in case of home births when compared to facility births (Table 3). For example, the concentration index for a newborn check-up within 24 hours of births was much higher in case of mothers who had birth at home compared with mothers who had birth at health facility suggesting higher socio-economic inequalities in this aspect of PNC relating to home births in contrast with births at a facility. Comparing home births with facility births, the inequalities in receipt of two or more check-ups for the baby before 10 days of life were found to be more pronounced in case of home births. An interesting finding emerges when we examine the concentration indices for choice of facility for the newborn’s check-up. The socio-economic inequalities were much higher in choice of facility for PNC check-up of newborn in the case of facility births as compared with births at home.

The adjusted probabilities of PNC use, as derived from logistic regression model are presented in Table 4. Apart from the fact that use of PNC was distinctly lower among home births as against facility births, the poor-rich difference was maintained at almost similar levels. Across all components of PNC and the entire spectrum of wealth score there remained an advantage for use of PNC in case of facility births. The two components i.e. check-up at government facility and private facility depict a pattern conforming to the preference of the rich for private facility. Indeed the socio-economic inequalities in the use of PNC remained high even after adjusting for important socio-economic, demographic and residence related variables. Overall, the predicted probabilities suggest significantly higher use by the rich compared to the poor irrespective of the place of birth of newborn. However, the gap between the richest and poorest was found to be higher in case of facility births compared to home births.

Literature suggests that cesarean births and births with maternal complications are considered to be at high risk during the first few weeks of life. Though low birth weight babies and babies of small size are also at high risk, but they were not included in the analysis due to the unavailability of such information in the DLHS – 3 dataset. Therefore, the analysis focusses only on use of PNC among those who had a cesarean birth and those with maternal complications.

Figure 2 presents the use of PNC for cesarean deliveries. For comparison we have contrasted the same against the use of PNC for normal deliveries. Findings clearly suggest higher use of PNC for those who had cesarean deliveries compared to those with normal deliveries. Check-ups for mother and baby within 24 hours of birth were almost universal in the case of cesarean births. On the contrary, such examinations were limited in the case of normal births. However, even in the case of cesarean deliveries not all babies received two-or more check-ups within first 10 days of life. This percentage was even lower in the case of normal deliveries. Another observation relates to babies born from cesarean sections being more likely to get a check-up at a private facility as opposed to normal births examined at government facilites.

Estimated concentration indices suggest considerable socio-economic inequalities in the use of PNC even for mothers who had cesarean births. This is an important finding given the fact that cesarean sections occur in health facilities and women remain in the facility for a prolonged period of time after the procedure. This should in theory facilitate closer interaction between the mother and the health provider and results in better opportunities to receive appropriate PNC and related counselling services. Though the levels of inequalities were low in three out of five components of PNC, the inequalities were particularly pronounced in choice of place for PNC. The poor were more likely than the rich to access care in a government facility, whereas, the rich tend to utilize private facilities for PNC (Table 5).

The adjusted probabilities of accessing PNC by wealth quintile are presented in Table 6. The results adjusted for socio-economic, demographic and residence related variables suggest that rich were much more likely than the poor to utilize PNC and that were also more likely to utilize private facility for availing such care.

Unlike the case of cesarean births, no consistent association was found between maternal complications at the time of birth and subsequent use of PNC (Figure 3). Some of the components had greater use among mothers who did not have complications during birth, while the others were more likely to be used by those mothers with reported complications. Births with maternal complications were no different from normal births in receiving the recommended set of PNC. This finding contradicts our belief that mothers with complications during birth should be using PNC more often than those without complications.

However, we observe socio-economic inequalities in use of PNC even among mothers who experienced complications during birth. The concentration indices for the first three components were 0.074, 0.236 and 0.109, suggesting significant socio-economic inequality in the use of these three components (Table 5). Even in case of births with maternal complications, the poor were more likely than the rich to utilize government facilities whereas the rich were more likely to utilize private facilities. The logistic regression results confirm the descriptive findings that the rich availed PNC more than the poor. In some components the differences were of the order of 0.57–0.58 (Table 6).