Postpartum Death in Malaysia 2013 - 2019 A Case Control Study

Background: Information regarding factors associated with postpartum death enhances the understanding of the root cause of the problem; hence, facilitate the formation of effective remedial measure. Literatures on postpartum death in Malaysia are limited despite the increasing trend of mortality in the state of Selangor. This study aimed to determine the associated factors with postpartum death in Malaysia. Method: A case-control was conducted using data from the Selangor State Health Department. All 114 postpartum death cases that were notied to the health department from 2013 to 2016 were selected. Controls (ratio case 1: control 4) were mothers who survived six weeks of postpartum within the same year range and selected via simple random sampling. A set of datasheets was used to gather information on; i) socio-demographic characteristics, ii) past medical/surgical history and iii) obstetrics history. Analyses were conducted using IBM Statistical Product and Service Solution Statistics Version 25. Results: The odds of postpartum deaths were almost three times higher among the non-Malaysian citizen (aOR 2.78, 95% CI: 1.35-5.72). The presence of medical/ surgical problems was associated with a 3-fold increase in the odds of postpartum death (aOR 3.22, 95% CI: 1.85-5.61). The odds of death were nine times higher for those who delivered in non-health facilities (home/en-route) (aOR 9.13,95% CI: 2.05-40.77) than those delivered in health facilities. Women who attended antenatal care at government health facilities and practised family planning had low odds for postpartum deaths [(aOR 0.35, 95% CI: 0.20-0.64) and (aOR 0.46, 95% CI: 0.25-0.85), respectively. Conclusions: Postpartum deaths could be prevented reduce by ensuring adequate health service delivery to the non-citizen, enhanced medical care on underlying medical/surgical illness and restrict delivery in non-health facilities. This reduction can be achieved by promoting attendance to antenatal care in government health facilities and practicing family planning. Future health programme aimed to curb this death should be designed with emphasis on education to these target group as early as pre-pregnancy throughout the antenatal


Introduction
The terms "postpartum period" and "postnatal period" are often used interchangeably but sometimes separately, when "postpartum" refers to issues pertaining to the mother and "postnatal" refers to those concerning the baby (1). The postnatal period begins immediately after the birth of the baby and extends up to six weeks (42 days) after birth (1). This crucial period is to optimize maternal health after delivery.
Home visit is one of the intervention to assess the wellbeing of mothers and babies, to provide education and support, to encourage breastfeeding or to provide practical support (2). Nevertheless, postpartum practices and beliefs are practiced throughout the world (3) , including among Malaysian. Asian postpartum practices aim to restore the balance of the new mother, as the delivery process is considered to make the mother weak and vulnerable (3). However, the community should be educated that taboos and traditions are not a substitute for medical care and consultation (3). Therefore, people should be educated to seek timely medical aid (3). Delay or failure in seeking medical treatment during postpartum period may result in maternal morbidity and mortality.
Retrospectively, a statistic in Malaysia  Additionally, ASEAN countries such as Myanmar and Indonesia (East Java Province) showed higher MMR compared to Malaysia, which is 282 per 100 000 live births in 2014 (6) and 89.6 per 100 000 livebirth in 2015, respectively (7) Even though the global MMR in 2017 representing a 38% reduction since 2000 (8), much effort is needed to achieve the Sustainable Development Goal (SDGs), aimed to reduce global maternal mortality ratio (MMR) to below 70 per 100 000 live birth by 2030.
Literature on postpartum death in Selangor was limited, hence, we were unable to ascertain the associated factors related to postpartum death. Therefore, this study aims to identify the sociodemographic characteristic, determine the trend and the major causes of death that contribute to postpartum death, and analyse the associated risk factors for postpartum death in Selangor.

Study Location
Selangor is one of the fourteen states of Malaysia. It is the west coast of peninsular Malaysia and is bordered by Perak to the North, Pahang to the East, Negeri Sembilan to the south and the Strait of Malacca to the West (9). Selangor is divided into nine districts. Petaling is the most central and urban with Gombak, Klang, and Sepang. Up north are the quaint district of Sabak Bernam, Kuala Selangor and Hulu Selangor whereas further south is Kuala Langat and Hulu Langat. In 2019, the estimation of the total population in Selangor was 6.53 million (10).

Study Design and Sample Size Measurement
It is a case-control study. The aim is to determine the association of risk factors and postpartum death.
The sample size calculated using Epi Info 7 (11). The following assumptions: 95% con dence, 80% power, ratio of control to case 4:1 and 20% contingency for non-responses used. It gave a total sample size of 114 cases and 456 controls. Therefore, the total sample size is 570 samples.

Selection of Case
Selangor State Maternal Death Technical Committee uses the World Health Organization International Classi cation of Disease de nition of maternal death "the death of a woman while pregnant or within 42 days of termination of pregnancy, irrespective of the duration and site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management but not from unintentional or incidental causes (8). All maternal death in Selangor were identi ed through the mandatory noti cation to the Selangor State Health Department. Death of any phase of pregnancy, include antenatal, intrapartum or puerperium on review of the death certi cation, must be reported by District Health O cer to Selangor State Technical Committee and subsequently to Development Division of Family Health, Ministry of Health within 24 hours from the time of death. All cases of maternal death must be investigated thoroughly at the district level and subsequently reviewed by state committee expert, which de ne the cause of death, death classi cation and remedial action that need to be implemented. There are a total of 184 of postpartum maternal death cases in these nine districts in Selangor from the year 2013 to 2019. The desired samples size that needed in this study is 114 case (obtained from a calculation using Epi Info). The cases were selected using strati ed random sampling (or Proportionate Strati cation), and we use districts as strata. We rst multiply the desired sample size (n) by the proportion of units in each stratum. As a result, we need to select 13 cases from Gombak, 21

Variable
The dependant variable was the postpartum maternal death in Selangor. The independent variables for the analysis were socio-demographic status included citizenship, ethnicity, maternal age, parity and associated risk factors includes maternal with pre-existing medical illness, place of delivery, mode of delivery, place of antenatal care and practice of family planning. This study also will describe the trend of postpartum deaths in Selangor according to the number of cases to years, the cause of death that contribute to postpartum death and the classi cation.

Statistical Analysis
The data obtained compared and analysed using SPSS version 25. The variables expressed as frequency and percentage while Simple and Multiple Logistic Regression used to compare variables. Level of signi cance "P" value evaluated, where P value < 0.05 was considered statistically signi cant.

The Trend of Maternal Mortality Ratio and Postpartum Death in Selangor 2013 to 2019
Socio-demographic characteristics A total of 570 women (114 cases and 456 controls) participated in this study. Most of the postpartum death (73.7%) were Malaysian citizenship, and most of them were Malays ethnicity (53.5%). The majority were 20 to 34 years old (65.8%), and 60.5% were multiparous (parity 2-4). Similarly, 69.3% had a preexisting medical or surgical illness and most had antenatal checked up at government health facilities (60.5%). The majority, 89.5% delivered in a health facility (government or private facilities) and most delivered via non-caesarean delivery (57.9%). 80.7% of women among postpartum death were not practising family planning.

Simple Logistic Regression
Page 6/23 The result from the Simple Logistic Regression analysis shows age was not statistically signi cantly associated with postpartum death and therefore, was not examined in the Multiple Logistic Regression.
From the Simple Logistic Regression test, non-Malaysian were 68% more likely to experience postpartum death compared to Malaysian (cOR 5.68, 95% CI: 3.21-10.04). Among ethnicity, other groups (consist of immigrants) associated with a 5-fold increase in the odds of postpartum death (cOR 5.08, 95% CI: 2.82-9.16). Primiparous women had a 61% reduction of postpartum death compared to grand-multiparous (cOR 0.39, 95% CI: 0.16-0.91). The presence of pre-existing medical/surgical problem was triple to experience postpartum death (cOR 3.44, 95% CI: 2.10-5.63). The odd death was 17 times higher for those who deliver in non-health facilities (home/en-route) (cOR 17.77, 95% CI: 4.92-64.10) than those who deliver in health facilities (government/private health facilities). Caesarean deliveries were 69% more likely to die from postpartum death (cOR 1.69, 95% CI: 1.11-2.58) compared to non-caesarean delivery. Women who went for antenatal checked up in government health facilities had a 64% reduction in postpartum death (cOR 0.36, 95% CI: 0.21-0.61) compared to those who had an antenatal check-up at private health facilities. Women who practised family planning had a 49% reduction of postpartum death compared to acceptor family planning (cOR 0.51, 95% CI: 0.30-0.89).  Table 1 shows that citizenship, ethnicity, parity, pre-existing medical/surgical illness, place of delivery, mode of delivery, place of antenatal care and family planning were statistically signi cant risk factors for postpartum death in Selangor from 2013 to 2019.

Multiple Logistic Regression
This analysis produced ve signi cant factors associated with postpartum death. The signi cant variables con rmed by the log-likelihood ratio (LR) test. From the analysis, we found that the odds of postpartum death were almost three times higher among non-Malaysian women (aOR 2.78,95% CI: 1.35-5.724) than Malaysian women. The presence of pre-existing medical or surgical problems was associated with a 3-fold increase in the odds of postpartum death (aOR 3.22,95% CI: 1.85-5.61). The odds of death were nine times higher for those who delivered in non-health facilities (home/en-route) (aOR 9.13,95% CI: 2.05-40.77) than those who delivered in health facilities. Women who went for antenatal checked up in government health facilities had a 65% reduction in postpartum death (cOR 0.35, 95% CI: 0.20-0.64) compared to those had an antenatal check-up at private health facilities. Women who practised family planning had a 54% reduction of postpartum death compared to non-acceptor of family planning (aOR 0.46, 95% CI: 0.25-0.85).    (15). Additionally, a study done in the United Kingdom revealed that 66.7% of postpartum death was preventable death (16) whereas preventable postpartum death in Selangor accounted for 54.9%.
As Selangor is a rapidly developing urban area, Selangor has become a focal point for the immigrant to ful l the jobs opportunities in various aspects. It is consistent with our result where the odds of postpartum death were almost three times higher among non-Malaysian citizenship (immigrant) (aOR 2.78,95% CI: 1.35-5.724) than Malaysian women. Immigrant women basically have a strong believe on traditional birth attendant. These women believe unregulated birth worker provides the best of both worlds which is support for physiological and birth care. It makes them feel safe as they believe that the traditional attendant has knowledge and skill to handle complication during delivery. This is a wrong perception because they are not trained to perform clinical and medical tasks therefore would not know how to handle complications (17) and eventually may lead to death. The ndings in our study were in accordance with the study conducted in Spain that show an excess risk of maternal death (cOR 2.19, 95% CI: 1.68-2.85) among immigrant women in their country (18). The scenario was even worst in the USA, where the majority was immigrant with racial and ethnic disparities exist for multiple adverse obstetric outcomes and types of obstetric care (19). Similarly to the ndings of the meta-analysis showed that migrant women in Western European countries have a doubled risk of dying during or after pregnancy when compared with indigenous born women, corresponding to additional risk of 9 maternal deaths per 100,000 deliveries among migrant women per year (20).
The major causes of postpartum death within the study period in Selangor was due to Associated Medical Condition, which accounted for 21.7% of all the total postpartum deaths. In Malaysia, between 2012-2014, Associated Medical Conditions continues to be the leading cause of maternal death (4.6 per 100 000 LB) followed by postpartum haemorrhage (3.9 per 100 000 LB) and pulmonary embolism (2.6 per 100 000 LB) (13). Among the deaths caused by an associated medical condition, half of the deaths caused by cardiac diseases (13).This nding is in accordance with our ndings that showed the presence of medical or surgical condition was associated with a 3-fold increase in the odds of postpartum death in Selangor. This nding is similar to the study conducted in the United Kingdom, where the odds of maternal mortality is seven times higher in cases with the pre-existing medical condition compared to the control group (21). Additionally, Godefay et al. reported that 88% of maternal death in Northern Ethiopia had a pre-existing medical condition (22) and study conduct by Nair et al. in the United Kingdom found that the odds of maternal death was six times higher in women with pre-existing medical illness (23).
Concerning the pre-existing medical/surgical illness, this nding re ects the gaps in service delivery of pre-pregnancy care and family planning. All women should be aware of the importance of being healthy optimally before embarking into the pregnancy (13). Thus, this should alert the need to strengthen the program of pre-pregnancy care and family planning in Selangor.
In the present study, delivery at non-health facilities identi ed to be strongly associated with postpartum death in Selangor. Non-health facilities de ned as delivery occur at the home, workplace, on the site of accident or injury and may be en-route to health facilities or brought in dead (BID) (13). In Malaysia, maternal deaths outside health facilities over ve years from 2009 to 2014, had shown an increasing trend from 19.7 to 35.7%. Generally, the number of pregnancy-related death outside health facilities was the highest in Selangor, followed by Johor and Sabah for three consecutive years (13). This nding corresponding to study conducted in Guinea (24) revealed that maternal death signi cantly associated in case of transfer from another hospital (OR 24.60, 95%CI 11.32-53.46). In contrast, Chinkumba et al. reported a study conducted in sub-Saharan Africa; the odds for maternal mortality who deliver in health facilities were two times higher compared to home delivery (25) The reason that might account for this result includes those seeking care at facilities may be complicated cases with a higher risk of mortality and referral-based secondary and tertiary facilities in treating women with more complex conditions. The World Health Organization (WHO) envisions a world where every pregnant woman and new born receives quality care throughout the pregnancy, childbirth and the postnatal period. Within the continuum of reproductive health care, antenatal care (ANC) provides a platform for critical health-care functions, including health promotion, screening, diagnosis and disease prevention (26). A study conducted in developing countries revealed the utilisation of antenatal care and skilled delivery service has signi cantly and negatively associated with MMR (27) Similarly, our study revealed that women who received antenatal care in government facilities were less likely experiencing postpartum death compared to those received in private facilities. Correspondingly to the study conducted in the United Kingdom, the odd of death was 23 times higher for women who have inadequate use of antenatal care (21). Besides, our study also in agreement with the study conducted in Nigeria, the likelihood of death was 68% lower for mothers who booked for antenatal care (28) while in Zambia, the likelihood of experiencing maternal death was 94% less among women who completed their scheduled antenatal care visits than for those who did not (29) The nding of our study observed women who practice family planning were less likely experiencing postpartum death compared to those who did not practice family planning. Despite contraceptive service more accessible in all health clinic or community clinic in Selangor, high-risk women in the reproductive age group tend to refuse to utilise the service for many reasons. A study conducted in Northern Ethiopia found that women who did not practice family planning before the last pregnancy was three times higher to experience maternal death compared to those who practised (22). Ahmed et al. report that without contraceptive use, the number of maternal deaths would have been 1·8 times higher than with contraceptive use, which means that contraceptive use averted 44·3% of maternal deaths in worldwide (30). Universal access to high-quality family planning is one of the vital action strategies under the Safe Motherhood Initiative.

Strength and Limitation
This study is the rst to analyse data of maternal mortality in the state of Selangor focussing on postpartum death in the study period of 2013 to 2019. The ndings of this study could make a baseline analysis for further planning and policies to reduce maternal death in Selangor and Malaysia. However, this study has several limitations. The number of postpartum deaths is small and it may affect the outcome of the analysis. Besides, the selection of control was con ned to one district only where it could avoid further bias by collecting data from all districts in Selangor state and preferable to match with the sample of the cases. Moreover, this study should exclude immigrant as it would have posed bias toward the analysis among Malaysian citizenship. In addition, the study should exclude postpartum death that was classi ed as fortuitous or undetermined.

Conclusions
Risk factors which were signi cantly associated with postpartum death in Malaysia includes non-Malaysian citizenship, the presence of pre-existing medical or surgical illness, place of delivery, place of antenatal care and family planning practices. Determining associated factors for postpartum death serves as a foundation for the development of an effective preventive strategy for prevention of maternal mortality. It is essential to enhance service delivery of antenatal care to these target group and strive for strict follow-up with emphasis on education. A comprehensive pre-pregnancy health promotion effort which extends into the antenatal period would reduce the incidence of postpartum death.        The leading cause of death in the seven years of the study period was medical (non-cardiac) and surgical condition (21.7%), infection and sepsis (18.5%), antepartum and postpartum bleeding (17.9%) and obstetric embolism (16.8%).