A retrospective study of transition-specific risk factors of maternal morbidity in northeast India: a multistage modelling approach

Abstract

We examine the transition pattern and transition-specific risk factors of maternal morbidities by employing the multistage modelling approach using data from the fifth round of the National Family Health Survey. The study includes 26,887 women aged 15–49 years who gave birth to their last child alive in the five years preceding the survey. We found that the prevalence of intrapartum and postpartum complications is higher among women who had complications in the preceding stages. Women who have attained higher education are less likely to have maternal complications. Nulliparous women and pregnancies resulting in multiple births are more likely to suffer from maternal complications. The risk of postpartum complications is higher among women whose index pregnancy was not desired. Women with caesarean delivery before index pregnancy also have a higher risk of maternal complications. Women are less likely to have complications if they seek antenatal care and delivery assistance from health facilities and skilled personnel. Our study concludes that women having complications in any stage of pregnancy are more likely to have complications in the subsequent stages. Nulliparity, unwanted pregnancy, caesarean delivery in the previous delivery, multiple births, and not attending antenatal care substantially increase maternal morbidity risk.

KEYWORDS:

• Transition-specific
• maternal morbidity
• antepartum
• intrapartum
• postpartum

Introduction

Maternal morbidity is the health condition that arises during the three stages of a reproductive period, i.e., antepartum, intrapartum and postpartum. It circumscribes a wide range of complications that could be fatal if not treated promptly with appropriate medications (Callaghan et al. 2008; Kuklina et al. 2009). The ratio between women dying due to pregnancy complications and women experiencing acute or chronic complications leading to permanent after-effects to their physical, mental and sexual health is reported to be in the range of 1 in 20 to 1 in 30 (Ashford 2002; Reichenheim et al. 2009; Pacagnella et al. 2010). Similar to the estimates of maternal mortality, the maternal morbidity estimate is also high in low- and middle-income countries (LMICs), particularly among those women belonging to the most impoverished socioeconomic strata (Storeng et al. 2010). In low-income countries, the estimated lifetime risk for maternal mortality is 1 in 52 compared to high-income countries, where the risk is 1 in 3400 (Filippi et al. 2016). The major problem in identifying and curbing the burden of maternal morbidity is the absence of a common definition, standard classification criteria and its multitudinous and complex causes (Firoz et al. 2013).

The burden of maternal morbidity remains challenging for India due to its vast geographical variation across the states and union territories. For a vast country like India, achieving Sustainable Development Goal-3 significantly depends on reducing inter and intra-state variations in maternal morbidities and maternal mortality. In past decades, India has introduced many region-specific or economic-specific health sector reforms aiming to improve the availability, accessibility, and affordability of quality healthcare, particularly maternal care. A notable example is the introduction of the Janani Suraksha Yojana (JSY), now renamed Janani Sishu Suraksha Karyakram (JSSK), a conditional cash transfer scheme to promote institutional delivery under the National Rural Health Mission (NRHM), now integrated with National Health Mission (NHM). After the reforms, India saw a rapid rise in institutional delivery: 39% in 2005–2006 to 89% in 2019–2021 (International Institute for Population Sciences (IIPS) and ICF 2021). India also witnessed a substantial maternal mortality ratio (MMR) decline: 212 per 100,000 live births in 2007-09 to 103 per 100,000 live births in 2017–2019 (SRS. 2011; 2022). But Gupta et al. (2012) made an alarming call reporting an increasing burden of maternal morbidity, which is a precursor to maternal mortality. Despite the setback in reducing maternal morbidity, the efforts made to reduce maternal morbidity in India that resulted in a widespread increase in access to maternal and reproductive health care are worth mentioning. As a result, many Indian states have made some progress (Sanneving et al. 2013). However, the northeastern states and some other regions of India remain one of the most underdeveloped areas in terms of maternal healthcare in India. Comparatively, India’s southern and western states are demographically and economically more forward than the northern and northeastern states (Bose 1991; Bhat and Zavier 1999; Pathak et al. 2010). Therefore, regional context plays a vital role in this respect, as in India, a wide range of cultural factors influence maternal and child healthcare practices (Shariff and Singh 2002). Northeast (NE) India comprises of eight small states: Arunachal Pradesh, Assam, Manipur, Meghalaya, Mizoram, Nagaland, Sikkim and Tripura. The healthcare system and socioeconomic conditions in NE India are relatively poor compared to other developed Indian states (Albert et al. 2015; Saikia and Das 2016). Due to these reasons, India’s NE states were included among the eighteen highly focused states under the NRHM (Kumar 2005; Singh et al. 2012). According to International Institute for Population Sciences (IIPS) and ICF (2021), six out of eight NE states are among the bottom ten states in India, with the lowest percentage of the population receiving information about pregnancy complications. Similarly, International Institute for Population Sciences (IIPS) and ICF (2021) also reported six NE states, namely, Nagaland (46%), Meghalaya (58%), Arunachal Pradesh (79%), Manipur (80%), Assam (84%) and Mizoram (86%) have the lowest prevalence of health facility delivery, of which two are far below the national average of 89%. Nagaland and Arunachal Pradesh also have the lowest prevalence of postnatal checks for mothers within two days of delivery. Thus, considerable attention is necessary to find factors responsible for poor maternal health status in NE India to reduce inter-regional variation and lighten the burden of maternal morbidities in India.

In maternal morbidity studies, there is a possibility that a woman experiencing complications in the first stage may experience complications in the succeeding stages resulting in repeated measures (Islam et al. 2004). Thus, keeping in view the importance of NE states of India in reducing the burden of maternal morbidities at the national level and the nature of dependency between complications at different stages, the paper aims to examine the pattern of transition of maternal morbidities between the three different stages of a reproductive period by employing a multistage model which is necessary to identify the transition-specific risk factors associated with maternal morbidities effectively. Identifying the transition-specific risk factors of maternal morbidities will be beneficial in the strategic planning of intervention programs at the regional and national levels.

Methods and materials

Data

The study extensively used information on maternal complications, which was collected retrospectively in the fifth round of the National Family Health Survey (NFHS-5). NFHS-5 is a nationally representative survey conducted across all states and union territories of India in 2019–2021. However, the present study includes only the eight NE states of India. NFHS-5 adopted a two-stage stratified sampling design to collect information from the study respondents. A total of 724,115 Indian women, 15 to 49 years of age, in 636,669 surveyed households were interviewed, with a response rate of 97%. See International Institute for Population Sciences (IIPS) and ICF (2021) for detailed information on the sampling procedure of the survey.

Analytical sample

The study population includes 26,887 married women aged 15 to 49 years who gave birth to their last child alive five years preceding the survey residing in the NE states of India.

Outcome variable

A women’s reproductive cycle consists of three main stages: antepartum, intrapartum and postpartum. We categorized maternal morbidities based on these three stages, i.e., complications that arise during the antepartum, intrapartum and postpartum periods. Antepartum complications refer to complications or morbidities that occur between conception to delivery. Intrapartum complications are the types of complications or morbidities that arise during delivery. Postpartum complications refer to complications or morbidities that occur within 42 days after delivery. NFHS-5 retrospectively collected information on eight specific maternal morbidities and is broadly categorized as (International Institute for Population Sciences (IIPS) and ICF 2021),

Table

Antepartum Complications:	difficulty in vision during daylight, convulsion not from fever, swelling of legs or body or face
Intrapartum Complications:	breech presentation, prolonged labour, haemorrhage
Postpartum Complications:	haemorrhage, very high fever

Explanatory variables

The following explanatory variables are considered to study the transition-specific risk factors of maternal morbidities at different stages of the reproductive cycle: age of the woman when her first child was born, maternal level of education, whether the woman had experienced caesarean delivery before the indexed pregnancy, whether the indexed pregnancy resulting in multiple births, whether the indexed pregnancy was desired, whether the women had terminated pregnancy before indexed pregnancy, number of antenatal visits for the indexed pregnancy, place where antenatal care and delivery was performed, personnel assisted during antenatal care and delivery, wealth quintile as the socioeconomic status of the household and place of residence.

Statistical analysis

Generally, such analysis inculpates the problem of repeated observations and censoring. For instance, the proportional hazard model helps in modelling partially censored data having single-time observation for each individual (Cox 1972). The theory of the proportional hazard model was extended for competing causes of failure by Holt (1978), Prentice et al. (1978) and Farewell (1979). Its theory is then further extended for several transitions, reverse transitions and repeated transitions and shows the method of testing the equality of parameters of transitions and repeated transitions (Beck 1979; Kay 1982; Andersen 1986; Andersen and Rasmussen 1986; Ataharul Islam and Singh 1992; Islam 1994).

Here in our study, there is a possibility that women experiencing complications in the first stage may also experience complications in the second and third stages, resulting in repeated measures (refer to Figure 1). And repeated measures will lead to correlated observations resulting in understated standard errors (Andersen 1986; Islam et al. 2004). In such conditions, examining separate models for different stages will not be sufficient to bring out the dynamics of the relationship between the risk factors and the outcome variables (Kay 1982; Andersen 1986; Islam et al. 2004). Thus, to assess the transition-specific effects of risk factors leading to maternal morbidities, we consider the multistage model (extended proportional hazard) proposed by Islam et al. (2004). This cause-specific hazard model is employed under the competing risk framework (Islam et al. 2004).

Fig. 1 Flow diagram for transition of maternal morbidity.

In stage 1, the women were assumed to be free from complications at conception. After that, two possible outcomes that the women may experience are (i) they advance towards stage 2 without any complication during stage 1, which is denoted by the transition 0-0, and (ii) they suffer from any complications during the stage 1 and proceed towards stage 2, which is denoted by the transition 0-1. Similarly, there will be four sets of outcomes in stage 2, two each for every outcome in stage 1. Therefore, the possible transition in stage 2 are 0-0-0, 0-0-1, 0-1-0 and 0-1-1. And the possible outcomes in stage 3 are 0-0-0-0, 0-0-0-1, 0-0-1-0, 0-0-1-1, 0-1-0-0, 0-1-0-1, 0-1-1-0 and 0-1-1-1. The transition ending with an event at each stage is considered uncensored, while others are considered censored.

The model is based on the above transitions for different stages. For each transition in different stages, the hazard function is defined for covariate vector X as, (1) $$\lambda (t,X)=\underset{\Delta t\to 0}{\lim }\frac{P[t\le T\le t+\Delta t|T\ge t;X(t)]}{\Delta t}$$(1)

Where $X(t)=\left[X_1(t),X_2(t),\dots ,X_p(t)\right]$ is a regression vector of p covariates at time t. The above hazard function is used for single absorbing and single transient states which were extended for competing causes (Prentice et al., 1978; Farewell 1979).

Let us consider that a group of women make the transition from a state of origin r (r = 0, 1) to a state of destination s (s = 0, 1) in stages k (k = 1, 2, 3). Where r = 0 and s = 0 denote that the women we are considering do not have any complications, while r = 1 and s = 1 denote the presence of complications. Denote $X_{rs|j}^k(t)$ as the regression vector at time t for those women who make the transition from r at (k-1)th stage to s at kth stage for given state j at (k-2)th stage ($j=$ 0, 1). Then, the hazard function can be defined as, (2) $${\lambda }_{rs|j}^k(tX)=\underset{\Delta t\to 0}{\lim }\frac{P\left[t\le T\le t+\Delta t,Z^k\left(t+\Delta t\right)=S/T\ge t,Z^{k-1}\left(t\right)=r;Z^{k-2}\left(t\right)=j;X_{rs|j}^k\left(t\right)\right]}{\Delta t}$$(2)

Where, $Z^k(t)$ is the stochastic process on states 0, 1 at stage k; ${\lambda }_{rs|j}^k(t, X)$ is the hazard functions for the transitions r at stage (k-1) to s at stage k for the given value of $Z^{k-2}(t)=j$ at stage (k-2).

Equation (2)(2) $${\lambda }_{rs|j}^k(tX)=\underset{\Delta t\to 0}{\lim }\frac{P\left[t\le T\le t+\Delta t,Z^k\left(t+\Delta t\right)=S/T\ge t,Z^{k-1}\left(t\right)=r;Z^{k-2}\left(t\right)=j;X_{rs|j}^k\left(t\right)\right]}{\Delta t}$$(2) can also be expressed in its generalized form as, (3) $${\lambda }_{rs\text{|}j}^k\left(t,X\right)={\lambda }_{0rs\text{|}j}^k\left(t\right)ex{p}^{X_{rs\text{|}j}^k\left(t\right){\beta }_{rs\text{|}j}^k}$$(3)

Where ${\beta }_{rs|j}^k$ is the regression coefficient corresponding to the covariate vector $X_{rs|j}^k$ for the transitions r at stage (k-1) to s at stage k for the given value of $Z^{k-2}(t)=j$ at stage (k-2).

We used Quantin et al. (1996) test to test the proportionality assumption of the hazard function in equation (3)(3) $${\lambda }_{rs\text{|}j}^k\left(t,X\right)={\lambda }_{0rs\text{|}j}^k\left(t\right)ex{p}^{X_{rs\text{|}j}^k\left(t\right){\beta }_{rs\text{|}j}^k}$$(3) . Quantin et al. (1996) test is best suited for proportionality test assumption for the extended proportional hazard model employed under the competing risk framework. In other words, it can assess the proportionality assumption of different transitions simultaneously (Islam et al. 2004).

First, we checked for the significance of the association between the explanatory variables and maternal complications and provided the prevalence of maternal complications at different stages of transition. We also estimated the overall prevalence of complications at different stages of transition. Then we fit the model to assess the transition-specific risk factors of maternal complications. The model is further tested for its proportionality assumptions. All analyses in this study are performed in STATA 16.0 (StataCorp L 2019) using appropriate sampling weights and adjusting for the complex survey design of NFHS-5.

Results

We considers seven states of transition provided in Figure 1, they are: occurrence of complications during antepartum period (0-1), occurrence of complications during intrapartum period among those women who had no complications during antepartum period (0-0-1), occurrence of complication in intrapartum period among those women who had complications during antepartum period (0-1-1), occurrence of complications during postpartum period among those women who had no complications during both antepartum and intrapartum period (0-0-0-1), occurrence of complications during postpartum period among those women who had complications during intrapartum period but not in antepartum period (0-0-1-1), occurrence of complications during postpartum period among those women who had complications during antepartum period but not in intrapartum period (0-1-0-1), and occurrence of complications during postpartum period among those women who had complications during antepartum and intrapartum period (0-1-1-1).

Table 1 presents the transition-specific prevalence of complications during the reproductive cycle. Around 37% of pregnant women suffer from antepartum complications. Among women with no antepartum complications, 30% suffer from intrapartum complications. In comparison, 47% of women suffered from intrapartum complications among those women who had antepartum complications. Among women who have not experienced antepartum and intrapartum complications, only 9% have postpartum complications. The prevalence of postpartum complications is 27% and 17% among women with intrapartum and antepartum complications, respectively. And 39% of women with both antepartum and intrapartum complications suffer from postpartum complications.

Table 1 Transition-specific prevalence of complications during the reproductive cycle.

	Complications
	No		Yes
Type of transition	%	N	%	N	Total
0-1	63.21	16,995	36.79	9,892	26,887
0-0-1	69.74	11,852	30.26	5,143	16,995
0-1-1	52.76	5,219	47.24	4,673	9,892
0-0-0-1	91.11	10,798	8.89	1,054	11,852
0-0-1-1	72.56	3,732	27.44	1,411	5,143
0-1-0-1	82.97	4,330	17.03	889	5,219
0-1-1-1	61.35	2,867	38.65	1,806	4,673

Note: complications during antepartum period (0-1); complications during intrapartum period among those women who had no complications during antepartum period (0-0-1); complication in intrapartum period among those women who had complications during antepartum period (0-1-1); complications during postpartum period among those women who had no complications during both antepartum and intrapartum period (0-0-0-1); complications during postpartum period among those women who had complications during intrapartum period but not in antepartum period (0-0-1-1); complications during postpartum period among those women who had complications during antepartum period but not in intrapartum period (0-1-0-1); and complications during postpartum period among those women who had complications during antepartum and intrapartum period (0-1-1-1).

Table 2 presents the transition-specific prevalence of complications during the reproductive cycle by selected demographic and socioeconomic characteristics. Most antepartum complications happen among women who give birth to their first child at 30 years and above. In comparison, intrapartum and postpartum complications are comparatively higher among women who gave birth to their first child at less than 19 years of age. The prevalence of intrapartum and postpartum complications is relatively lower among women who gave birth to their first child in the age group 19 to 29 years and 30 to 34 years. Antepartum complications are also somewhat higher among women with a caesarean delivery history and a history of pregnancy termination. Similarly, the prevalence of antepartum and intrapartum complications is higher among the women whose index pregnancy resulted in multiple births. The prevalence of postpartum complications is higher among nulliparous women who had complications during pregnancy and delivery. Women who receive their antenatal care assistance from elsewhere and unskilled personnel have a higher prevalence of maternal complications at all stages of the reproduction cycle. However, the prevalence of maternal complications is higher among women who had delivered their index child in an institution, irrespective of transition states. Women from the higher and lower wealth quintile suffered complications more than those from the middle wealth quintile. And women living in rural areas have a higher prevalence of complications in all transition states.

Table 2 Transition-specific prevalence of complications during the reproductive cycle by selected demographic and socioeconomic characteristics.

	0-1	0-0-1	0-1-1	0-0-0-1	0-0-1-1	0-1-0-1	0-1-1-1
	Prevalence	Prevalence	Prevalence	Prevalence	Prevalence	Prevalence	Prevalence
Age at first birth
Less and equal 18	33.83**	31.53*	50.29**	10.99**	27.49	22.01	46.83
19 to 29 years	35.72	28.23	43.24	10.31	28.26	18.67	39.45
30 to 34 years	39.55	30.56	41.88	10.02	26.60	13.88	37.58
35 and above	38.19	31.97	43.81	2.99	30.87	24.15	41.64
Maternal level of schooling
No schooling	35.25**	25.36**	43.07	11.00	26.94*	22.63	39.19
Primary	35.92	28.56	47.59	10.79	26.57	18.72	41.13
Secondary	34.99	30.44	45.74	9.97	29.12	18.78	42.93
Higher	37.70	28.45	38.21	12.30	22.46	18.27	34.82
Prior caesarean delivery
No	35.36**	29.31	44.94	10.41	27.97	19.28	41.69
Yes	36.86	27.05	51.06	11.53	29.27	20.08	33.15
Number of prior births
Nulliparous	35.25**	31.42**	46.09*	10.76**	28.42*	17.80	43.10**
1 to 2 births	35.04	28.07	43.44	10.66	26.84	20.58	40.69
3+ birth	37.04	26.93	47.38	8.55	30.85	19.00	39.79
Multiple birth
Single	35.33**	29.31	44.95*	10.45	27.96	19.35	41.51
Multiple	42.94	23.54	55.98	6.18	33.93	8.70	45.87
Desire to be pregnant
No	36.91**	32.06*	51.74**	12.78	27.93	25.56**	40.22
Yes	35.26	29.07	44.49	10.26	27.99	18.85	41.67
Pregnancy terminated
No	34.77**	29.60	45.22	10.06	28.43	18.96	41.77
Yes	39.52	26.89	43.87	13.00	24.40	21.16	40.22
No. of ANC visits
No visit	36.11**	30.53**	48.09**	10.77*	26.58**	19.73**	40.51**
1 to 3 visits	35.61	33.76	47.87	11.32	33.63	22.84	47.11
4+ visits	35.08	25.73	42.31	9.78	22.77	16.77	36.99
Place of antenatal care
Elsewhere	34.44	31.72**	43.13	10.16**	29.89	22.15**	45.65
Institutional	35.58	28.74	45.42	10.48	27.52	18.66	40.73
Assistance during antenatal care
Un-skilled	35.52**	30.32**	48.51*	12.18	29.22	22.01**	46.21**
Skilled	35.36	29.09	44.40	10.12	27.75	18.83	40.63
Place of delivery
Elsewhere	32.71**	25.38**	42.93**	7.81**	23.72**	19.86	39.21**
Institutional	35.91	30.09	45.41	11.01	28.73	19.18	41.95
Assistance during delivery
Un-skilled	33.75**	27.64**	44.92*	8.55**	26.59	19.79	45.01*
Skilled	35.67	29.57	45.05	10.78	28.22	19.20	40.97
Wealth quintile
Poorest	35.06**	28.29**	42.97**	10.43	29.71	23.31**	42.51
Poorer	33.99	31.51	51.40	9.96	28.03	15.96	39.47
Middle	35.80	30.96	43.44	11.27	27.85	17.28	44.75
Richer	37.87	24.54	38.69	9.01	19.67	17.54	37.51
Richest	43.34	22.33	38.02	15.06	27.77	16.52	47.95
Place of residence
Urban	40.95**	28.63*	39.76**	10.79	20.76**	15.43**	38.28
Rural	34.36	29.38	46.18	10.37	29.14	20.23	42.16
Total (N)	26887	16995	9892	11582	5143	5219	4673

Note: ** p-value < 0.01 & * $0.01<$ p-value < 0.05; HR - Hazard Ratio; CI–Confidence Interval; complications during antepartum period (0-1); complications during intrapartum period among those women who had no complications during antepartum period (0-0-1); complication in intrapartum period among those women who had complications during antepartum period (0-1-1); complications during postpartum period among those women who had no complications during both antepartum and intrapartum period (0-0-0-1); complications during postpartum period among those women who had complications during intrapartum period but not in antepartum period (0-0-1-1); complications during postpartum period among those women who had complications during antepartum period but not in intrapartum period (0-1-0-1); and complications during postpartum period among those women who had complications during antepartum and intrapartum period (0-1-1-1).

Table 3 presents the estimated effects of risk factors on complications in different transition states. In all transition states, the likelihood of suffering from maternal complications decreases as the age of the women at their first birth increases. Compared to women who have not attended any formal schooling, women who have attended primary (HR = 1.27; 95% CI = [1.18, 1.36]) and secondary (HR = 1.42; 95% CI = [1.34, 1.52]) level of schooling have a higher risk of suffering from antepartum complications. Similarly, the risk of intrapartum complications is higher among women who have attended primary and secondary education, irrespective of whether women have suffered or not from complications before the delivery. Women who have attended secondary-level schooling also have a higher risk of postpartum complications, irrespective of transition states. Women who had prior caesarean delivery are more likely to suffer from antepartum (HR = 1.46; 95% CI = [1.24, 1.71]) and intrapartum (HR = 1.45; 95% CI = [1.14, 1.85]) complications. Similarly, a higher risk of intrapartum (HR = 1.35; 95% CI = [1.08, 1.69]) and postpartum (HR = 1.66; 95% CI = [1.01, 2.76]) complications were observed among women who had prior caesarean deliveries and antepartum complications. The risk of postpartum complications was also higher among women who had prior caesarean deliveries and had complications during pregnancy and delivery (HR = 1.57; 95% CI = [1.13, 2.18]). Compared to nulliparous women before the index pregnancy, women with at least one child are less likely to suffer from maternal complications. Women whose index pregnancy results in multiple births are more likely to suffer from antepartum (HR = 2.02; 95% CI = [1.61, 2.52]) and intrapartum (HR = 2.05; 95% CI = [1.46, 2.89]) complications. Similarly, women whose index pregnancy resulted in multiple births and had complications before deliveries have a higher risk of intrapartum complications (HR = 2.24; 95% CI = [1.67, 3.02]). The risk of postpartum complications is lesser among women whose index pregnancy is desired even if she had antepartum complications (HR = 0.65; 95% CI = [0.52, 0.81]). Women who have attended four or more antenatal care visits have a lower risk of intrapartum complications (HR = 0.88; 95% CI = [0.78, 0.99]) than their counterparts. And women who have attended antenatal care in an institution (public or private hospital) have a lower risk of suffering from maternal complications, irrespective of transition states. However, women delivering the index child in an institution have a higher risk of maternal complications in all transition states. The risk of suffering from intrapartum complications among those women who had complications during pregnancy was lower among women who were assisted by skilled personnel during delivery (HR = 0.80; 95% CI = [0.70, 0.93]). Similarly, the risk of postpartum complications was lower among women who were assisted by skilled personnel during delivery, even if they had had complications during pregnancy and delivery (HR = 0.64; 95% CI = [0.51, 0.81]). Compared to the household with the poorest socioeconomic status, the likelihood of suffering from maternal complications decreases as the level of socioeconomic status increases. And compared to those women living in urban areas, women living in rural areas are more likely to suffer from intrapartum complications (HR = 1.15; 95% CI = [1.05, 1.25]). Similarly, women living in rural areas have a higher risk of intrapartum complications if they have complications during pregnancy (HR = 1.16; 95% CI = [1.06, 1.27]). The risk of postpartum complications is also higher among women living in rural areas, whether they have had complications during pregnancy or delivery.

Table 3 Transition-specific effects of selected demographic and socioeconomic characteristics on complications at different stages of transition during the reproductive cycle.

	0-1		0-0-1		0-1-1		0-0-0-1		0-0-1-1		0-1-0-1		0-1-1-1
	HR	95% CI	HR	95% CI	HR	95% CI	HR	95% CI	HR	95% CI	HR	95% CI	HR	95% CI
Age at first birth
Less and equal 18 ®
19 to 29 years	0.29**	(0.27, 0.30)	0.25**	(0.23, 0.27)	0.26**	(0.24, 0.28)	0.25**	(0.21, 0.29)	0.26**	(0.23, 0.30)	0.28**	(0.23, 0.33)	0.25**	(0.23, 0.29)
30 to 34 years	0.06**	(0.06, 0.07)	0.05**	(0.04, 0.06)	0.05**	(0.04, 0.05)	0.04**	(0.03, 0.06)	0.04**	(0.03, 0.06)	0.06**	(0.04, 0.08)	0.04**	(0.03, 0.05)
35 and above	0.02**	(0.02, 0.02)	0.02**	(0.01, 0.02)	0.02**	(0.01, 0.02)	0.01**	(0.00, 0.02)	0.02**	(0.01, 0.03)	0.02**	(0.01, 0.04)	0.02**	(0.01, 0.02)
Maternal level of schooling
No schooling ®
Primary	1.27**	(1.18, 1.36)	1.35**	(1.22, 1.49)	1.17**	(1.05, 1.30)	1.15	(0.92, 1.43)	1.15	(0.94, 1.41)	1.13	(0.89, 1.44)	1.13	(0.95, 1.34)
Secondary	1.42**	(1.34, 1.52)	1.53**	(1.40, 1.67)	1.32**	(1.21, 1.45)	1.30**	(1.07, 1.57)	1.44**	(1.22, 1.71)	1.29*	(1.05, 1.59)	1.34**	(1.16, 1.55)
Higher	1.02	(0.92, 1.12)	1.11	(0.97, 1.28)	0.94	(0.82, 1.09)	0.88	(0.64, 1.20)	0.89	(0.67, 1.17)	1.08	(0.78, 1.49)	0.79	(0.63, 1.01)
Prior caesarean delivery
No ®
Yes	1.46**	(1.24, 1.71)	1.45**	(1.14, 1.85)	1.35**	(1.08, 1.69)	1.51	(0.84, 2.70)	1.52	(0.99, 2.33)	1.66*	(1.01, 2.76)	1.57**	(1.13, 2.18)
Number of prior births
Nulliparous ®
1 to 2 births	0.23**	(0.22, 0.24)	0.23**	(0.21, 0.25)	0.22**	(0.20, 0.24)	0.21**	(0.18, 0.25)	0.22**	(0.19, 0.25)	0.27**	(0.23, 0.32)	0.21**	(0.19, 0.24)
3+ birth	0.06**	(0.06, 0.07)	0.06**	(0.05, 0.06)	0.05**	(0.04, 0.06)	0.05**	(0.04, 0.06)	0.05**	(0.04, 0.06)	0.05**	(0.04, 0.06)	0.05**	(0.04, 0.06)
Multiple birth
Single ®
Multiple	2.02**	(1.61, 2.52)	2.05**	(1.46, 2.89)	2.24**	(1.67, 3.02)	0.62	(0.15, 2.56)	1.63	(0.84, 3.17)	1.15	(0.41, 3.21)	1.32	(0.78, 2.22)
Desire to be pregnant
No ®
Yes	0.95	(0.88, 1.02)	0.92	(0.82, 1.02)	0.91	(0.82, 1.01)	0.94	(0.74, 1.19)	1.11	(0.89, 1.38)	0.65**	(0.52, 0.81)	0.97	(0.82, 1.14)
Pregnancy terminated
No ®
Yes	1.03	(0.98, 1.10)	0.87	(0.75, 1.01)	0.88	(0.79, 1.00)	0.99	(0.82, 1.20)	0.74	(0.51, 1.10)	0.98	(0.81, 1.18)	0.94	(0.81, 1.08)
No. of ANC visits
No visit ®
1 to 3 visits	1.06	(0.98, 1.15)	1.03	(0.92, 1.16)	1.19**	(1.07, 1.32)	1.29	(0.99, 1.68)	1.14	(0.92, 1.40)	1.12	(0.87, 1.44)	1.25**	(1.06, 1.46)
4+ visits	1.04	(0.96, 1.13)	0.88*	(0.78, 0.99)	0.98	(0.88, 1.09)	1.11	(0.85, 1.45)	0.94	(0.76, 1.17)	0.79	(0.61, 1.02)	0.99	(0.84, 1.17)
Place of antenatal care
Elsewhere ®
Institutional	0.92**	(0.87, 0.98)	0.87**	(0.81, 0.95)	0.89**	(0.82, 0.96)	0.79**	(0.67, 0.93)	0.83*	(0.72, 0.96)	0.74**	(0.62, 0.89)	0.83**	(0.73, 0.95)
Assistance during antenatal care
Un-skilled ®
Skilled	1.03	(0.95, 1.11)	1.15*	(1.03, 1.29)	0.89*	(0.80, 0.98)	0.88	(0.70, 1.11)	0.90	(0.74, 1.09)	0.84	(0.66, 1.05)	0.72**	(0.63, 0.83)
Place of delivery
Elsewhere ®
Institutional	1.16**	(1.06, 1.28)	1.33**	(1.16, 1.52)	1.34**	(1.17, 1.55)	1.55**	(1.15, 2.09)	1.38*	(1.05, 1.83)	1.12	(0.83, 1.51)	1.86**	(1.47, 2.36)
Assistance during delivery
Un-skilled ®
Skilled	0.91	(0.83, 1.01)	0.93	(0.81, 1.07)	0.80**	(0.70, 0.93)	0.96	(0.70, 1.31)	0.81	(0.61, 1.07)	1.15	(0.84, 1.58)	0.64**	(0.51, 0.81)
Wealth quintile
Poorest ®
Poorer	0.86**	(0.82, 0.91)	0.99	(0.92, 1.06)	0.96	(0.89, 1.03)	0.87	(0.75, 1.01)	0.81**	(0.71, 0.92)	0.69**	(0.58, 0.82)	0.77**	(0.68, 0.87)
Middle	0.78**	(0.74, 0.83)	0.84**	(0.77, 0.91)	0.81**	(0.74, 0.89)	0.70**	(0.58, 0.85)	0.72**	(0.62, 0.85)	0.70**	(0.57, 0.86)	0.8**	(0.69, 0.92)
Richer	0.74**	(0.68, 0.80)	0.69**	(0.61, 0.77)	0.73**	(0.65, 0.82)	0.61**	(0.47, 0.78)	0.68**	(0.55, 0.84)	0.52**	(0.40, 0.69)	0.81*	(0.67, 0.97)
Richest	0.75**	(0.67, 0.85)	0.58**	(0.48, 0.71)	0.64**	(0.54, 0.77)	0.58*	(0.38, 0.88)	0.48**	(0.32, 0.73)	0.42**	(0.27, 0.64)	0.79	(0.60, 1.06)
Place of residence
Urban ®
Rural	0.99	(0.93, 1.05)	1.15**	(1.05, 1.25)	1.16**	(1.06, 1.27)	1.01	(0.84, 1.22)	1.26*	(1.06, 1.50)	1.34**	(1.08, 1.65)	1.22**	(1.06, 1.40)

Note: **p-value < 0.01 & * $0.01<$ p-value < 0.05; HR - Hazard Ratio; CI - Confidence Interval; complications during antepartum period (0-1); complications during intrapartum period among those women who had no complications during antepartum period (0-0-1); complication in intrapartum period among those women who had complications during antepartum period (0-1-1); complications during postpartum period among those women who had no complications during both antepartum and intrapartum period (0-0-0-1); complications during postpartum period among those women who had complications during intrapartum period but not in antepartum period (0-0-1-1); complications during postpartum period among those women who had complications during antepartum period but not in intrapartum period (0-1-0-1); and complications during postpartum period among those women who had complications during antepartum and intrapartum period (0-1-1-1).

The model estimates presented in Table 3 are further tested for proportionality assumption violations. The mother’s age at first birth and the number of prior births affected the proportional assumptions in four transition states, 0-1, 0-0-1, 0-0-1-1 and 0-1-1-1 (see Chi-Square statistic in Table 4).

Table 4 Test for proportionality assumptions.

Type of Transition	Chi-Square	d.f.	p-value
0-1	10.33	23	0.011
0-0-1	11.38	23	0.021
0-1-1	19.11	23	0.305
0-0-0-1	13.18	23	0.052
0-0-1-1	11.83	23	0.027
0-1-0-1	12.09	23	0.031
0-1-1-1	15.60	23	0.128

Note: d.f.–degrees of freedom; complications during antepartum period (0-1); complications during intrapartum period among those women who had no complications during antepartum period (0-0-1); complication in intrapartum period among those women who had complications during antepartum period (0-1-1); complications during postpartum period among those women who had no complications during both antepartum and intrapartum period (0-0-0-1); complications during postpartum period among those women who had complications during intrapartum period but not in antepartum period (0-0-1-1); complications during postpartum period among those women who had complications during antepartum period but not in intrapartum period (0-1-0-1); and complications during postpartum period among those women who had complications during antepartum and intrapartum period (0-1-1-1).

After that, we fitted the same model stratified by the mother’s age at first birth and number of prior births to adjust for the proportionality assumption violation. Interestingly, we found no significant change in the estimates of the stratified model compared to the model estimates presented in Table 3 (see Appendix Table A1).

Discussion

Our study discusses the transition-specific risk factors of maternal morbidities during women’s reproductive periods of their recent births in northeast India from a quantitative viewpoint. Thus, reinforcing the knowledge about maternal morbidity in northeast India. Our study adopted a more robust multistage modelling approach introduced by Islam et al. (2004) to identify or study the transition-specific risk factors of maternal morbidity in northeast India. A multistage model is necessary for this study, as for studies that deal with multiple stages of complications or morbidities; if fitted separately, there is a high chance that the information attributable to the transition during different stages of a reproductive period may affect the estimates of the study (Islam 1994; Islam et al. 2004). Moreover, the estimated standard error of the estimates will be biased if the model is not adjusted for repeated measures, i.e., the dependency between different stages of the reproductive period (Cox 1972; Kay 1982; Andersen 1986; Islam et al. 2004). The estimates from different transitions show a significant association between different stages of pregnancy and childbirth.

Although the current study considers maternal morbidity as a continuous process consisting of three stages, the findings from the study are consistent with those from past studies.

We found that women are more likely to have complications during the intrapartum period if they had at least one of the complications during the antepartum period. Similarly, women who had at least one complication during the antepartum period are more likely to have postpartum complications, irrespective of having suffered from intrapartum complications or not. Similar findings have been noted by Islam et al. (2004) in their study of Bangladesh population, an immediate neighbor to the NE region of India. Additionally, the study reveals that among women who have had complications during the antepartum period, complications are more likely to occur in the intrapartum period than in the postpartum period.

The likelihood of suffering from maternal morbidity is less among women whose age at first birth is high and women who have attained higher education irrespective of transition between difference stages from pregnancy to childbirth. Sikder et al. (2014) also reported a significant role of the age of women on pregnancy complications with increased risk among women with age less than 18 years. Interestingly, the study found that the higher the household socioeconomic status, the fewer women suffer from complications irrespectively of the different stages of transition. It shows that increasing the household socioeconomic status may reduce maternal morbidity, a statement that is supported by Gazmararian et al. (1996), Kahn et al. (2000), Johnson et al. (2002), and Salam and Siddiqui (2006). Also, this may be attributed to the inability to provide better healthcare services during these periods among women with poorer socioeconomic status. And most importantly, women belonging to rural areas are found to be more vulnerable to maternal complications than those in urban areas.

Studies have reported that the number of prior births significantly contributes to various maternal complications such as vaginal bleeding, premature rupture of membranes, puerperal endometritis, anaemia, and eclampsia (Conde-Agudelo and Belizán 2000; Sikder et al. 2014). Finding from our study is in complete alignment with the conclusions of the previous studies. Our study found that nulliparous women before the index pregnancy were more likely to suffer from maternal complications at any stage of a reproductive period. Sikder et al. (2014) also reported that nulliparous women prior to index pregnancy pose a higher risk of complications during the index pregnancy.

Evidence from the past studies have shown that both mother and newborn are at risk if the pregnancy is not desired, and the pregnancy results in multiple birth (Zwart et al. 2008; Sikder et al. 2014; Mahran et al. 2017; Monden and Smits 2017). Our study also shows the risk of postpartum complications is higher among those women whose index pregnancy was not desired. Similarly, the risk of antepartum and intrapartum complications is observed to be higher among women whose index pregnancy resulting in multiple birth. We also found that women who had caesarean delivery before index pregnancy have a significantly higher risk of maternal complications at any stage of pregnancy to childbirth, irrespective of having suffered from complications in the past. Similar findings were reported by Kayiga et al. (2018) and Wen et al. (2005).

The place and person from whom women seek antenatal care, and delivery assistance play a vital role in reducing maternal morbidity (Basinga et al. 2011; Mannava et al. 2015). The study also found that women are less likely to suffer from complications at any stage if the women seek antenatal care and delivery assistance from institutional and skilled personnel. Women who had at least four antenatal care visits are also less likely to suffer from these complications. However, women who delivered their index pregnancy at an institution are more likely to suffer complications, which contradicts the previous statement that women seeking delivery at an institution are less likely to suffer complications. But this contradiction may be due to the influence of an individual’s treatment-seeking behaviour (delayed treatment practices) and their autonomy in the socioeconomic condition of the household. A similar argument on Individual’s treatment seeking behaviour has been noted by Bloom et al. (2001), Rani and Bonu (2003), and Mistry et al. (2009).

The primary strength of the study is the methodological ability to identify the transition-specific risk factors of maternal morbidity by considering maternal morbidity as a continuous process consisting of three stages (Islam et al. 2004). By treating maternal morbidity as a continuous process involving three stages (antepartum, intrapartum and postpartum period), the paper presents more robust estimates of the factors associated with maternal morbidity. And identifying the transition-specific risk factors of maternal morbidity will help in the strategic planning of interventions at the regional and national levels. However, the study of maternal morbidity is not easy as it circumscribes several complications at different stages. Therefore, the study’s shortcoming is the restriction of maternal morbidities to a few broad categories and curtailing risk factors to only that is available in the fifth round of the National Family Health Survey. And the complications that we are addressing in the study are self-reported. And finally, the information on maternal complications was collected retrospectively in the survey, which may be affected by the recall bias. To reduce the effect of recall bias, we restricted our study to the recent births. In fact, NFHS-5 also collects information on maternal morbidities or complications for recent births only (International Institute for Population Sciences (IIPS) and ICF 2021).

Conclusion

The study highlight’s the interrelationship between maternal complications during different stages of pregnancy and childbirth. Women who suffer from one or more of the major antepartum complications are more likely to have postpartum complications, irrespective of the women had or not had intrapartum complications. Similarly, women who had antepartum complications are more likely to have intrapartum complications. The major risk factors contributing to the maternal complications at different stages of pregnancy are age at first birth less tha 18 years, caesarean delivery, nulliparity prior to index pregnancy, unwanted pregnancy, index pregnancy resulting in multiple births, and neglecting or not receiving the antenatal and delivery care from skilled personal.

Authors’ Contributions

WBM and LL conceptualized the study. WBM conducted the analysis and prepared the first draft. Both WBM and LL read and revised the draft for final submission.

Ethical Statement

Ethical Statement is not required as the data used in the study is available for public use.

Data Availability Statement

NFHS-5 data is available for public use in Demographic and Health Survey data repository (https://dhsprogram.com/data/dataset/India_Standard-DHS_2020.cfm?flag=0).

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

No funding was received to conduct this study.

Appendix

Table A1 Transition-specific effects of selected demographic and socioeconomic characteristics on complications at different stages of transition during a reproductive cycle adjusted for proportionality assumption by stratification.

	0-1		0-0-1		0-1-1		0-0-0-1		0-0-1-1		0-1-0-1		0-1-1-1
	HR	95% CI	HR	95% CI	HR	95% CI	HR	95% CI	HR	95% CI	HR	95% CI	HR	95% CI
Age at first birth
Less and equal 18 ®
19 to 29 years					0.26**	(0.24, 0.28)	0.25**	(0.21, 0.29)					0.25**	(0.23, 0.29)
30 to 34 years	Adjusted		Adjusted		0.05**	(0.04, 0.05)	0.04**	(0.03, 0.06)	Adjusted		Adjusted		0.04**	(0.03, 0.05)
35 and above					0.02**	(0.01, 0.02)	0.02**	(0.01, 0.03)					0.02**	(0.01, 0.02)
Maternal level of schooling
No schooling ®
Primary	1.32**	(1.23, 1.42)	1.42**	(1.29, 1.57)	1.17**	(1.05, 1.30)	1.15	(0.92, 1.43)	1.16	(0.95, 1.42)	1.12	(0.88, 1.42)	1.13	(0.95, 1.34)
Secondary	1.67**	(1.57, 1.77)	1.85**	(1.69, 2.01)	1.32**	(1.21, 1.45)	1.30**	(1.07, 1.57)	1.38**	(1.16, 1.63)	1.24*	(1.01, 1.53)	1.34**	(1.16, 1.55)
Higher	1.25**	(1.14, 1.36)	1.48**	(1.29, 1.70)	0.94	(0.82, 1.09)	0.88	(0.64, 1.2)	0.90	(0.68, 1.19)	1.11	(0.80, 1.54)	0.79	(0.63, 1.01)
Prior caesarean delivery
No ®
Yes	0.87	(0.75, 1.02)	0.83	(0.66, 1.06)	1.35**	(1.08, 1.69)	1.51	(0.84, 2.7)	1.50	(0.97, 2.31)	1.63	(0.98, 2.70)	1.57**	(1.13, 2.18)
Number of prior births
Nulliparous ®
1 to 2 births	Adjusted		Adjusted		0.22**	(0.20, 0.24)	0.21**	(0.18, 0.25)	Adjusted		Adjusted		0.21**	(0.19, 0.24)
3+ birth					0.05**	(0.04, 0.06)	0.05**	(0.04, 0.06)					0.05**	(0.04, 0.06)
Multiple births
Single ®
Multiple	1.18	(0.94, 1.47)	1.17	(0.83, 1.64)	2.24**	(1.67, 3.02)	0.62	(0.15, 2.56)	1.63	(0.84, 3.16)	1.20	(0.43, 3.35)	1.32	(0.78, 2.22)
Desire to be pregnant
No ®			Adjusted
Yes	0.98	(0.91, 1.06)			0.91	(0.82, 1.01)	0.94	(0.74, 1.19)	1.08	(0.87, 1.34)	0.65**	(0.52, 0.82)	0.97	(0.82, 1.14)
Pregnancy terminated
No ®
Yes	1.03	(0.98, 1.10)	0.80**	(0.73, 0.87)	0.88**	(0.81, 0.96)	0.99	(0.82, 1.2)	0.77**	(0.64, 0.92)	0.98	(0.81, 1.18)	0.94	(0.81, 1.08)
No. of ANC visits
No visit ®
1 to 3 visits	1.07	(0.98, 1.16)	1.03	(0.91, 1.15)	1.19**	(1.07, 1.32)	1.29	(0.99, 1.68)	1.11	(0.91, 1.37)	1.08	(0.84, 1.38)	1.25**	(1.06, 1.46)
4+ visits	1.03	(0.95, 1.12)	0.89	(0.79, 0.99)	0.98	(0.88, 1.09)	1.11	(0.85, 1.45)	0.91	(0.73, 1.12)	0.76*	(0.59, 0.98)	0.99	(0.84, 1.17)
Place of antenatal care
Elsewhere ®
Institutional	0.82**	(0.78, 0.87)	0.77**	(0.71, 0.83)	0.89**	(0.82, 0.96)	0.79**	(0.67, 0.93)	0.87	(0.76, 1.01)	0.76**	(0.63, 0.91)	0.83**	(0.73, 0.95)
Assistance during antenatal care
Un-skilled ®
Skilled	0.99	(0.92, 1.07)	1.11	(0.99, 1.24)	0.89*	(0.80, 0.98)	0.88	(0.7, 1.11)	0.93	(0.77, 1.12)	0.87	(0.69, 1.10)	0.72**	(0.63, 0.83)
Place of delivery
Elsewhere ®
Institutional	1.4**	(1.27, 1.54)	1.76**	(1.53, 2.03)	1.34**	(1.17, 1.55)	1.55**	(1.15, 2.09)	1.43*	(1.09, 1.89)	1.07	(0.79, 1.44)	1.86**	(1.47, 2.36)
Assistance during delivery
Un-skilled ®
Skilled	0.96	(0.87, 1.06)	0.94	(0.81, 1.09)	0.80**	(0.70, 0.93)	0.96	(0.70, 1.31)	0.78	(0.60, 1.03)	1.17	(0.85, 1.6)	0.64**	(0.51, 0.81)
Wealth quintile
Poorest ®
Poorer			1.04	(0.97, 1.12)	0.96	(0.89, 1.03)	0.87	(0.75, 1.01)	0.85*	(0.74, 0.97)	0.71**	(0.60, 0.85)	0.77**	(0.68, 0.87)
Middle			0.93	(0.85, 1.01)	0.81**	(0.74, 0.89)	0.70**	(0.58, 0.85)	0.78**	(0.66, 0.91)	0.73**	(0.60, 0.90)	0.80**	(0.69, 0.92)
Richer	Adjusted		0.75**	(0.67, 0.84)	0.73**	(0.65, 0.82)	0.61**	(0.47, 0.78)	0.74**	(0.60, 0.91)	0.58**	(0.44, 0.76)	0.81*	(0.67, 0.97)
Richest			0.58**	(0.48, 0.70)	0.64**	(0.54, 0.77)	0.58*	(0.38, 0.88)	0.54**	(0.36, 0.81)	0.44**	(0.29, 0.68)	0.79	(0.60, 1.06)
Place of residence
Urban ®
Rural	1.08**	(1.03, 1.14)	1.18**	(1.08, 1.29)	1.16**	(1.06, 1.27)	1.01	(0.84, 1.22)	1.22*	(1.03, 1.46)	1.32*	(1.07, 1.63)	1.22**	(1.06, 1.40)

Note: Adjusted - Adjusted for proportionality assumption by stratification; ** p-value <0.01 & * 0.01 <p-value <0.05; HR - Hazard Ratio; CI - Confidence Interval; complications during antepartum period (0-1); complications during intrapartum period among those women who had no complications during antepartum period (0-0-1); complication in intrapartum period among those women who had complications during antepartum period (0-1-1); complications during postpartum period among those women who had no complications during both antepartum and intrapartum period (0-0-0-1); complications during postpartum period among those women who had complications during intrapartum period but not in antepartum period (0-0-1-1); complications during postpartum period among those women who had complications during antepartum period but not in intrapartum period (0-1-0-1); and complications during postpartum period among those women who had complications during antepartum and intrapartum period (0-1-1-1).