Revisiting “3 Delays Model” for Critical Analysis of Hypertensive Maternal Mortality in a Tertiary Care Centre

Abstract

Background

Hypertensive disorders in pregnancy are responsible for 14% of maternal deaths. Maternal mortality is not only due to medical disorder but also due to underlying social–cultural factors lead to delays in getting timely and appropriate intervention. This study aimed to analyse three delays, to identify socio-cultural, demographic and accessibility factors that contributed to maternal deaths due to hypertension in pregnancy.

Method

It was a prospective observational study. Verbal autopsy was done for each maternal death that occurred due to hypertensive disorder. The “3 delay model” concept was used in collection and analysis of data for each maternal death. The Chi-square test (χ²) and multivariable logistic regression model were used to accomplish the study objectives.

Results

Out of 326 maternal deaths, 52 deaths were due to hypertensive disorders. Pulmonary oedema, eclamptic encephalopathy, sepsis and cerebrovascular accident were the major causes of maternal mortality. In thirty-nine (75%) type I delay was the most significant contributor of maternal death. Twenty-four (46.15%) and 14 (26.9%) had Type II/III delay. The reasons for level I delay were failure to recognize the danger signs, delay in decision-making, no ANC registration, Type II delay being lack of timely transport facility, type III delay being lack of adequate monitoring, delay in initiating treatment and substandard care in hospital were observed.

Conclusions

Socio-cultural and demographic factors significantly contribute to the direct or indirect medical cause of maternal death. Analysing these factors provide deep insight into the maternal death and will provide useful information for guiding the policy makers to take corrective actions and direct their programmes and policies such as strengthening referral networking, widespread coverage of healthcare facility and raising awareness of danger signs of pregnancy by reaching at grass root levels.

Introduction

Every day approximately 800 women die from preventable causes related to pregnancy and childbirth.

Recent WHO report says that pregnancy and birth-related complications contribute to 300,000 maternal deaths worldwide. Ninety per cent of these deaths occurred in LMICs of which 99 per cent is contributed by Saharan Africa and South Asia. A total of 45,000 maternal deaths occurred during 2015 in India, and this figure drives India into the hub of the second largest number of maternal deaths after Nigeria [1].

Almost 75% of maternal deaths (6057) occurred in the Empowered Action Group (EAGA) States, with Uttar Pradesh and Uttarakhand contributing approximately 22% (2059) of these deaths, followed by other EAGA states [2]. In 2017, the global maternal mortality ratio (MMR) was 211, reflecting a 38% decline from 2000 [3].

The “three delay model” by Thaddeus and Maine (1990) [4] is a valuable tool and is broadly acknowledged as a system for critical analysis of hindrances in obstetric care and their contribution in maternal mortality [4]. The findings of this model may help in the formulation of strategies for preventive measures.

The types of delay described by this model are as follows.

Type 1 delay—delayed recognition of a pregnancy complication and decision.

Type 2 delay—delays in reaching an emergency obstetric care facility.

Type 3 delay—lack of receipt of timely, adequate and appropriate obstetric care at a healthcare facility.

There is a paucity of literature on direct relationship between delay and severity of maternal outcomes. Therefore, the aim of this study is to find facility-based and community-based factors, which contributed to maternal death by applying 3 delay model.

Methods

It was a descriptive observational study conducted at the largest tertiary care hospital and medical college in northern India. The hospital offers 24X7 comprehensive emergency obstetric care with 12 bedded obstetric ICU with well-equipped neonatal care unit, blood bank and dialysis facility.

Women with diagnosis of preeclampsia and eclampsia admitted were included in the study. Preeclampsia and eclampsia were diagnosed using ACOG 2020 criteria [5].

Maternal death defined as death of a woman while pregnant or within 42 days of termination of pregnancy, irrespective of the duration and site of pregnancy from any cause related or aggravated by pregnancy or its management but not from accidental or incidental causes (ICD-10).

Maternal deaths that met the WHO criteria were included in the study, while deaths due to accidental or incidental causes were excluded.

Structured questionnaire was used for documentation. Data were collected by interviewing the primary caregiver who accompanied the women by using facility-based Maternal Death Review pro forma (as per the NHM MDSR guidelines) [3] and verbal autopsy questionnaire and Maternal Death Review (MDR) case summary. “3 delay model” concept as given in MDR case summary was used in collection and analysis of each data. Data on basic demography, referral details, treatment at various levels (including emergency antihypertensive therapy and eclampsia prophylaxis with magnesium sulphate), mode of delivery, circumstances leading to death, causes of death and delays contributing to mortality were recorded.

The cross-tabulations with Chi-square test (χ2) were used to show the association between various attributes. The multivariable logistic regression model was applied to estimate the odds ratio with 95% confidence intervals, and p-value less than 0.05 was considered statistically significant. The data analyses were performed with the statistical software STATA version 13.0 software. The English transcripts were imported into the QSR N-Vivo 10 (categorized and analysed by adopting thematic approach).

Results

During the study period, there were 10,920 obstetric admissions and 5930 deliveries. Out of this, 1198 were preeclampsia and eclampsia cases giving prevalence of 10.97%. A total number of maternal mortalities during this period were 326 out of which 52 maternal deaths were due to hypertensive disorders of pregnancy. Out of this, 57.69% (30/52) deaths occurred in eclampsia patients and 42.3% (22/52) in preeclampsia patients.

Sociodemographic and clinical characteristic of the women with preeclampsia and eclampsia are depicted in Table 1. The mean age of women with preeclampsia and eclampsia being 27.8 ± 4.7 years and 25.09 ± 4.9 years, respectively. The mean BMI was 23.8 ± 3.0 kg/m² for those with preeclampsia and 22.08 ± 2.0 kg/m² for those with eclampsia.

Table 1.

Distribution of subjects according to sociodemographic and clinical characteristics

			Preeclampsia (n = 22)	Eclampsia (n = 30)
1	Mean age (Yrs)		27.8 ± 4.7	25.09 ± 4.9
2	Mean BMI		23.8 ± 3	22.08 ± 2.0
3	Socioeconomic status	Lower	52.01%	90.61%
4	Number of ANC visit	< 4 visit	47.8%	81.2%
		4 and more visit	52.8%	18.8%
5	Parity	Primi	53.5%	67.3%
6	Residence	Rural	55.19	92.2%
7	Hypertension in previous pregnancy	Yes	6.6%	5.3%
8	Additional medical risk factors and comorbidities	Severe anaemia	31.4%	12.7%
		Maternal hypothyroidism	26.3%	3.2%
		Jaundice	5.7%	4.9%
		COVID pneumonia	15.2%	2.04%
		IVF conceived	6%	0.40%
		Autoimmune disease	0.18%	0
		APLA positive	4.2%	0

A large proportion of these women were from rural areas (55.19% for preeclampsia and 92.2% for eclampsia).

Notably, more than half of the women who died had access to some form of antenatal care, although the quality of care provided was suboptimal. The substandard quality of care was reflected in several areas, including insufficient screening for high-risk conditions, such as hypertension and gestational diabetes, and inadequate management of complications such as preeclampsia and eclampsia. Furthermore, essential treatments, such as antihypertensive, were either not administered on time or were administered incorrectly. Additionally, many healthcare facilities were inadequately equipped, lacked essential medications, and faced a shortage of trained staff, all of which further compromised the quality of care delivered. There was also a lack of proper follow-up visits, poor coordination between primary and secondary healthcare levels, and in some cases weak referral system leading to delayed intervention.

Delays were also observed in accessing timely care, with 29.1% of women facing delays in obtaining transportation and 32% reporting the unavailability of trained healthcare providers at the first-level health facility.

Gestational age at the time of admission in our facility ranged between 32 and 36 weeks. About 48% (25/52) women delivered by caesarean section, 17.30 (19/52) delivered vaginally and 15.38% (8/52) women expired undelivered. The present study reported high referral rate (92.30%), and most of the cases were referred from district hospital (26.92%), community health centre (19.23%) and private hospital (34%).

We noticed that the main reasons for referral were non-availability of anaesthetist (3.8%), lack of skilled personnel (42.30%), blood bank facility (13.46%), non-availability of ICU (32.69%) and NICU facility (7.69%) at first referral site.

In the present study, most frequent cause of maternal mortality due to hypertensive disorder was pulmonary oedema (42.9%) followed by sepsis (17.31%), COVID pneumonia (15.38%), eclamptic encephalopathy (11.54%), cerebrovascular accident (7.6%) and haemorrhage (5.7%) (Fig. 1).

Fig. 1.

Causes of maternal mortality due to hypertensive disorder

Most of the cases presented late, and the mean admission to death interval was 50.12 hours. About 61% women had more than 3 referrals before reaching the tertiary care centre.

On analysing the hypertensive maternal deaths on 3 Delay model, we observed that Type 1 delay, i.e. delay in decision-making to seek care was found in 75% maternal death followed by Type 2 delay in 46.15% women and delay in receiving appropriate care, Type 3 was found in 26.9%. All three of the delays present in 7.9% of maternal deaths (Table 2). Combined Type 1 and Type 2 delay was present in 46.5% women. Combined Type 1 and Type 3 delay in 30% and combined type 2 and 3 delays were present in 11.5% maternal deaths.

Table 2.

Distribution of subject according to type of delay. More than one type of delay possible in maternal mortality presented in order of frequency

Type of delay	Reason for delay	No. of women	Percentage (%)
Type I Delay—delay in seeking care (n = 39 75%)	Unawareness of danger signs	28	71.7%
	Illiteracy and ignorance	32	42.6%
	Delay in decision-making	12	16%
	No birth preparedness	8	10.6%
	Belief and customs	5	6.6%
	Non-availability of healthcare p[professional	24	32%
	Any other	9	12%
Type 2 Delay—delay in reaching first-level health facility (n = 24, 46.15%)	Delay in getting transport	7	29.1%
	Delay in mobilizing funds	8	33.3%
	Not reaching appropriate facility in time Multiple referrals	20	83.3%
Type 3 Delay—Delay in receiving adequate care in facility (n = 14, 26.9%)	Delay in initiating treatment	12
	Substandard care in hospital
	-BP not measured -Mgso4 not given -Mgso4 not given	4 6	28.5% 42.8%
	Antihypertensive not given	7	50%
	Lack of blood, equipment and drugs	2	14.28%

In this study, a univariate analysis was done to identify the potential predictor for maternal mortality using the logistic regression analysis. In univariate analysis, the potential predictors for maternal mortality identified were education, socioeconomic status, booking status, ANC visits, mode of delivery, presenting symptoms like breathlessness, GCS, pulmonary oedema, AKI, jaundice, PRESS, investigations like haemoglobin, liver function test, serum LDH and kidney function test were found to be associated with risk of maternal mortality in current study population (Table 3).

Table 3.

Univariable analysis to predict maternal mortality

Variables	Odds ratio	Lower 95% CI	Upper 95% CI	P value
Education	0.67	0.52	0.86	0.002
Breathlessness	17.2	9.5	31	< 0.001
GCS	2.5	2.05	3.21	< 0.001
Pulmonary oedema	11.2	5.6	22.07	< 0.001
AKI	5.5	2.3	13.3	< 0.001
Jaundice	5.08	2.1	12.09	< 0.001
PRESS	45.8	4.08	513	0.002
Socioeconomic status	4.7	2.03	11.1	< 0.001
Booked/unbooked	0.27	0.13	0.55	< 0.001
ANC visits	0.75	0.64	0.88	0.001
Hb	0.87	0.77	0.98	0.028
SBT	2.12	1.65	2.7	< 0.001
SBD	1.56	1.29	1.88	< 0.001
SGPT	1.003	1.001	1.005	0.003
Alkaline phosphatase	1.00009	1.00003	1.00016	0.003
sLDH	1.0004	1.000002	1.0007	0.035
Serum uric acid	1.03	1.01	1.06	< 0.001
S urea	1.03	1.02	1.04	< 0.001
S creatinine	1.78	1.4	2.2	< 0.001

Multivariable analysis indicated that breathlessness was associated with 11.6 times higher risk, unstable general condition at the time of admission 2.07 times, AKI 3.79 times and low socioeconomic status 4.26 times higher risk for maternal mortality in North India population (Table 4).

Table 4.

Multivariable analysis to predict the maternal mortality

Maternal mortality	Odds ratio	Std. Err	z	P value	[95% Conf.	Interval]
Breathlessness	11.63	4.69	6.08	< 0.001	5.28	25.63
GCS	2.07	0.31	4.79	< 0.001	1.54	2.79
AKI	3.79	2.37	2.13	0.03	1.11	12.91
SES	4.26	2.58	2.39	0.02	1.30	13.99
SBD	1.48	0.22	2.70	0.01	1.11	1.97

Discussion

Hypertensive disorders of pregnancy are multifactorial, life-threatening disorders specific to pregnancy. Most maternal deaths occur due to the preventable causes [5]. Of the various factors affecting maternal death, those which have an impact on the time period between onset of obstetric complications and its outcome are the most important (Table 5).

Table 5.

Distribution of delays in various studies

Type of delay	Current study 2021–22	Srivastava9et al. Varanasi (2019)	Gumare J.P10 et al. Maharashtra (2018)	Ankit Anand11et al. (2019)	JOGI8 et al. (2020)
Type 1 delay	75%	81%	27%	48.6%	95.54%
Type 2 delay	46.12%	54.5%	52%	33.8%	70.54%
Type 3 delay	26.9%	45.5%	21%	18.9%	47.32%

In our study, prevalence of hypertensive disorder of pregnancy was 10.97%, and maternal mortality due to hypertension was observed in 15.95% of total mortality. Maity S et al. [6] and other authors also reported similar prevalence. Similar to previous reported literature [6–8], majority of the deaths in the present study also occurred in young, primi gravida, belonging to low socioeconomic who lacked antenatal care. Panda et al. [8] reported that out of 7.4% cases of hypertension in pregnancy, 2/9% ended in maternal death. Of these 2.6% were primi gravida and 3.8% belonged to poor socioeconomic status.

First delay is delay in “seeking care” plays an important role when maternal mortality is considered in developing country like India [8]. The major causes ascribed to first delay (75% in our study) are illiteracy and ignorance (42.6%), unawareness of danger signs (71.7%), customs and belief (6.6%) and absence of birth preparedness (10.6%). Similar observation was made by Jogi et al. 2020 [9] and Kumari et al. 2019 [10], in their study (Chhattisgarh & Varanasi). They reported type I delay in 95.54% and 81.8% cases. In contrast with this, Gumare et al. 2018 [11] in their study from Maharashtra recorded type 1 delay in 27% cases. State-wise variation could be due to the fact that both Uttar Pradesh and Chhattisgarh belong to EAGA states which continue to be deficient in sociodemographic indices pointing to the need for focal intervention in these areas. Even when women recognized the danger signs they often neglected it or they failed to perceive severity of complications to seek care on time [12, 13]. Sk et al. [12] reported that 32.1% women failed to recognize the danger signs. This highlights serious gaps in the knowledge and service provided by the ASHA workers to the pregnant women and their families as the ASHA workers are an important link between pregnant women families, and healthcare facilities.

There is a need for more comprehensive training of ASHA workers focused on hypertension and its warning signs.

To overcome barriers due to Type 1 delay, Government of India launched Mother and Child Tracking System (MCTS) in 2009 to track all pregnant women and neonate and infant for quality antenatal, intranatal and postnatal care. However, recent reports have suggested that there is a strong need to strengthen the performance of MCTS scheme across the states [14].

Although not much found in isolation, level 2 delay was found in 46.15% of cases with other delay or in isolation. Sk et al. [13] observed type 2 delay in 31.3% cases out of 17.4% was because of unavailability of transport, and 16.1% were due to financial constraints to avail transport and treatment. Singh et al. [15] recorded type 2 delay in 65% cases¹⁶ due to delay in getting transport, bad roads, poor vehicle conditions, delay in mobilising funds and not reaching facility in time due to multiple and delayed referrals indicating unassured referral linkage.

Jogi et al. [9] reported 68.5% women travelled a distance of more than 25 km to reach the health facility leading to delay in life saving interventions. Sk et al. [12] observed that the delay in reaching health facility was more common in second gravidas and primiparous women, which was found among 36.4 per cent each in these groups. The current study revealed that 29.1% women faced difficulty in getting the transport to reach the health facility which is less than the previously reported studies [12, 13]. This could be because of the steps taken by Government of India to provide free transport facility in the form of 102 and 108 ambulance services.102 and 108 National Ambulance services under NHM, to facilitate 24 × 7 free transport services to pregnant women and newborns to reach healthcare facility, is showing results. Pawalia K et al. [16] reported that the number of ambulances in India increased by 57.5% from 2012 to 2022. Kumutha et al. [17] observed that since the introduction of the 108 and 102 ambulance services, institutional deliveries have increased in all states and union territories where the ambulances have been deployed.

Singh et al. reported 22% high-risk pregnant women in Andhra Pradesh and 27% from Himachal Pradesh used ambulance service to reach the health facility in time [18]. Another recent study from Maharashtra reported that 38.22% women used emergency ambulance services indicating significant increase compared to previous reports [19].

A recent study from state of Tamil Nadu reported that 100% respondents were aware of 108 ambulance service while knowledge about 102 ambulance services was seen in 26.8% respondents. Despite awareness, dismal 15% utilized this service [20]. It is suggested that awareness creation among women during antenatal care through advertisements, mass media and IEC will motivate all the women to use 108/102 ambulance facilities at the government health facilities.

To promote institutional deliveries, Government of India has launched conditional cash transfer scheme called Janani Suraksha Yojana (JSY) and JSSK (Janani Shishu Suraksha Karyakram) to eliminate out of pocket expenditure for pregnant women.

In the current study, third delay was observed in 26.9% of the cases, i.e. delay in receiving adequate care in the facility due to delay in initiating treatment or substandard care given. Similar observations were made by and Jogi et al. study [9] (47.32%) and Kumari et al. [10] (45.5%). Gumare et al. [11] and Sk et al. [12] reported comparatively lower percentage of cases having third delay, 21% and 18.9% cases, respectively.

Some of the startling observations in the current study were that basic step of measuring B.P was omitted in 26% women. Antihypertensive was not started in 20.3% women when raised BP was found. A few referrals also documented using sublingual Nifedipine, to control very high blood pressure which is now known to be a dangerous practice. Magnesium sulphate was either not given or incomplete loading dose given (only intramuscular dose), and women were referred without proper documentation of the same.

In Uttar Pradesh, various training programmes have been initiated by NHM & UP State Technical Unit (UPTSU), and emphasis is made on strengthening of FRU by training the medical officers and healthcare workers in various emergencies. Subsequent mentoring visits by the experienced faculty from parent medical colleges are proving to be a successful strategy to improve the quality of care [21].

The maternal mortality rate in state of Uttar Pradesh has been decreasing, with the rate in 2019–2020 dropping to 97 per 100,000 live births from 130 per 100,000 live births in 2014–15.

However, attention is needed in all three levels to reach our target of less than 70 per 1 lac live birth by 2030.

Conclusion

Type 1 and type 2 delays are significant contributors of maternal mortality due to hypertensive disorders of pregnancy. Early identification of danger sings, knowledge and behaviour change towards seeking early medical care, better transport could further decrease our MMR.

Educational training for healthcare providers who deliver antenatal care should be strengthened. Furthermore, education on the danger signs of pregnancy and during childbirth should be provided not only to pregnant women and their husbands but also to the community.

The use of telemedicine and mobile health technologies to enhance monitoring and management of hypertension during pregnancy will address gaps in care and will offer scalable solutions in resource-limited settings.

Main outcome of the study will help to alarm the society and will enable the policy makers to understand the magnitude of problem and the focus areas, to develop guidelines for FRUs, to take corrective actions and to take steps improve awareness regarding free transportation, to strengthen the referral network.

Universal use of Gestosis score for prediction of preeclampsia, endorsement of low-dose aspirin in high-risk cases as per FOGSI and FIGO guidelines and calcium supplementation in low-resource settings for prevention of preeclampsia can go a long way in reducing maternal mortality due to hypertension in pregnancy.

Author contribution

Dr Neha helped in data collection. Prof Seema Mehrotra worked in concept and design of study. Dr Vandana Solanki contributed to lay out of study and writing. Prof Urmila Singh helped in formal analysis of study.

Funding

The study was not supported by any funding.

Declarations

Conflicts of interest

There are no conflicts of interest in the study.

Ethical approval

Institutional ethical committee clearance was obtained for study (ethical number102nd ECM II B-Thesis/P22). All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

Informed consent

Informed consent was obtained from every individual participant included in the study.