Volume 105, Issue 1 p. 82-85
Averting maternal death and disability
Free Access

Reducing maternal mortality among Afghan refugees in Pakistan

Susan Purdin

Corresponding Author

Susan Purdin

The International Rescue Committee, New York, USA

Corresponding author.Search for more papers by this author
Tila Khan

Tila Khan

International Rescue Committee, Pakistan

Search for more papers by this author
Roxanne Saucier

Roxanne Saucier

The International Rescue Committee, New York, USA

Search for more papers by this author
First published: 20 February 2009
Citations: 25



The International Rescue Committee (IRC) strove to reduce maternal mortality among Afghan refugees in Hangu district of Pakistan by improving access to emergency obstetric care (EmOC), community knowledge of danger signs of pregnancy, and the use of health information.


IRC established EmOC centers, trained community members on safe motherhood, linked primary health care with education on danger signs of pregnancy and the importance of skilled attendance, and improved the health information system.


The maternal mortality ratio among Afghan refugees in the area improved from 291 per 100 000 live births in 2000 to 102 per 100 000 live births in 2004. The proportion of refugee births attended by skilled staff increased from 5% in 1996 to 67% in 2007. Complete prenatal care coverage increased from 49% in 2000 to 90% in 2006, and postnatal coverage more than trebled from 27% in 2000 to 85% in 2006.


Improved services, community involvement and education, good coordination, and effective systems succeeded in reducing maternal mortality in a traditionally conservative environment.

1 Introduction and background

After more than 20 years of civil war, conflict, drought, and famine, Afghans live with some of the poorest health indicators in the world. In its annual State of the World's Population report for 2007, UNFPA documented Afghanistan's maternal mortality ratio (MMR) as one of the highest anywhere—1900 deaths of women from pregnancy— or childbirth—related complications per 100 000 live births [1]. The infant mortality rate is 143 per 1000 live births [1]. The total fertility rate is an average of 7.11 children per woman and only 14% of births are assisted by a skilled attendant [1]. On average, an Afghan woman dies every 30 minutes from pregnancy-related causes [2].

Afghan refugee women fleeing to neighboring Pakistan after the 1979 Soviet invasion of their country did not find themselves in a much better position. In rural Pakistan, the outlook for women's health continues to be poor. The MMR in rural areas around Peshawar was calculated at 473 per 100 000 live births in a 2005 survey [3], and in the North West Frontier Province (NWFP) only 20.4% of women who give birth have the assistance of a skilled attendant [4]. The average fertility rate in rural areas is 5.4 children and the infant mortality rate is 102 per 1000 live births [4].

There is evidence that, until recently, conditions for Afghan refugee women living in NWFP were also poor. A reproductive age mortality survey (RAMOS) covering the period of January 20, 1999 to August 31, 2000 in 12 Afghan refugee settlements in Hangu district of NWFP receiving assistance from the International Rescue Committee (IRC) documented the poor conditions for women in this area [5]. It indicated that 41% of deaths among women of reproductive age were due to maternal causes; this compares with less than 1% in the United States in 1999 [6], [7]. The survey also found that 92% of these maternal deaths were preventable. Further, the MMR in Hangu district was estimated at 291 per 100 000 live births, and only 18% of all deliveries were attended by skilled personnel. The study revealed that 81% of women who died of maternal causes had at least one barrier to care. Examples included delays in recognizing complications, delays in reaching health facilities, and delays in receiving care after arriving at the facility [5].

A reproductive health knowledge, attitudes and practices survey (KAP) conducted by the IRC and the Centers for Disease Control and Prevention (CDC) in the same population in 2000 found low levels of literacy and education (59% of men had never attended school, and 99% of the women stated that they had never attended school and could neither read nor write). It also found a high birthrate (the average fertility rate of the refugee women was measured at 9 lifetime births) and low rate of family planning use, at 9.5% [8]. All of these factors contribute to poor maternal outcomes, as does the limited mobility experienced by these women, which constrains their access to health care [5]. These findings highlight the critical need among Afghan refugee women in Hangu district for access to good quality emergency obstetric care (EmOC) services.

The global Safe Motherhood Initiative has come to a consensus on 3 elements that are key to reducing maternal mortality: family planning, skilled care at all deliveries, and access to emergency obstetric care for all women with life-threatening complications [9]. There was an obvious need among this refugee population for services in these 3 key areas to reduce maternal morbidity and mortality and to improve the health of newborns and their siblings.

1.1 Background of the program

The IRC began its Primary Health Care (PHC) program for Afghan refugees living in Hangu district in 1980. At this time, IRC established Basic Health Units (BHUs), which provide both curative and preventive care, including prenatal and postnatal care, as well as referral to emergency obstetric services. There are currently 8 BHUs—one per camp, except for 3 BHUs which each serve 2 camps. The current (2007) refugee population in the Hangu and Thal areas of the Hangu district is approximately 96 300 in 11 camps along a 45 km stretch (Fig. 1). A host community of approximately the same size population also receives services through these facilities.

Details are in the caption following the image

Map of Afghanistan/Pakistan border area, including Hangu District.

Following the International Conference on Population and Development in Cairo in 1994, the IRC in Hangu expanded its focus to include reproductive health services as an integral component of PHC. At that time, according to IRC's health information system, more than 96% of all deliveries among this refugee population took place at home, about 30% of which were assisted by women trained as female health workers (FHWs)—community health workers who are not skilled attendants. Community traditions were very conservative, restricting women to their own homes or homes of close relatives. Women were not seen in the market and they were allowed only limited access to health services; for example, to take a sick child to the BHU. Pregnancy was not spoken about because of the fear that a miscarriage could be induced through a curse for any of a number of reasons, such as jealousy or vendetta.

In 1996, the IRC established the first facility in the area for managing basic obstetric emergencies. The facility began as a small building in the IRC health office compound called a minor operating theater (MOT), rather than a delivery room or basic EmOC center. It was named an MOT to avoid a backlash from the community against a facility designed to serve women during the very private act of giving birth. The MOT was staffed with Pakistani female doctors and nurse midwives as well as dayas (traditional birth attendants) from among the refugee population. Knowledge of the MOT spread throughout the area by word of mouth and through the positive reports of women who had used it. Because the IRC is a US-based agency, in September 2001 the IRC compound, including the MOT, was attacked and destroyed following the US bombing of Afghanistan. The facility was subsequently re-established in a nearby building, and this time renamed as an EmOC center, illustrating that over time the community had begun to accept this women's health facility.

However, insecurity and distance still prohibited women living in the farthest Thal camps from using this EmOC facility. As early as 2000, the community representatives of the distant camps cited the availability of an accessible birth center as one of their most urgent needs. In response, IRC opened a second facility to serve those camps in 2003. Today the two facilities provide basic EmOC services to the 11 camps (one for 6 camps and another for 5 camps). In addition, there is one comprehensive EmOC facility at Kohat Hospital, approximately 45 minutes away.

This paper details the strategies used by IRC to reduce maternal mortality among the women it serves in Hangu district, and reports the outcomes of those strategies.

2 Materials and methods

Although numerous factors contribute to a woman's health, the IRC staff identified safe motherhood as an essential service in 1996, and worked to develop a plan that was appropriate to the setting. Recognizing the unmet need for treatment of obstetric complications and birth spacing, the program focused on enhancing the availability and accessibility of these services. As the focus of this paper is on access to EmOC services, information on family planning aspects is not included here.

It has come to be understood that program management is as important as clinical care in the implementation of safe motherhood programs [9] and IRC's experience in Pakistan reflects this. IRC has reduced maternal mortality in Hangu district through a combination of provision of EmOC services, community education, health staff coordination, and improved health information systems (HIS).

IRC's program focused on the need to improve care during pregnancy, delivery, and postpartum, and to detect complications in time to provide prompt treatment when they occur. The program worked to improve knowledge throughout the community regarding the need for skilled delivery care and recognition of danger signs during pregnancy and childbirth. BHU staff use prenatal care visits as an opportunity to educate pregnant women on the need for delivery by skilled staff in the EmOC facility, as a complication may occur before, during or after delivery, and such complications cannot always be predicted. Furthermore, during prenatal care visits, when staff recognize risk factors for obstetric complications, they give the woman a referral form for the EmOC facility in advance. BHU staff also share this information with family members so they will be prepared to help the woman find timely transport to the EmOC facility at the onset of labor. This is important because it is often male family members who make the decision to bring a woman to the health facility. This community involvement and awareness-raising has led to increased service utilization and more timely decision-making by family members regarding EmOC.

Health staff also use prenatal care visits as an opportunity to advise women to report for postnatal care, especially for a check-up during the first 72 hours after giving birth.

EmOC staff are available or on-call 24 hours a day, 7 days a week. Each of the two EmOC facilities has labor and delivery rooms and a separate room for general examinations, running water and toilets, electricity with a backup generator, and an ambulance for emergency referrals.

A unique part of the health program is the network of volunteer community health workers (CHWs) and FHWs who focus on health education in the refugee community. Currently about 330 CHWs and 325 FHWs are active in raising awareness regarding PHC, including reproductive health. In addition, each camp has a health committee that represents their neighbor families through bi-monthly meetings with BHU staff.

BHU and EmOC staff, volunteer health workers, and an additional team of 4 IRC staff (2 male and 2 female) provide regular training to CHWs, FHWs, health committee members, teachers, and religious leaders about reproductive health. Training on safe motherhood was provided to key community representatives including CHWs (299), FHWs (244), women of reproductive age excluding FHWs (545), men (518), health committee members, all of whom are men (100), teachers (149), religious leaders (72), and private practitioners (6). These numbers reflect activities in 2007, but similar trainings have been conducted among these groups since 2003. The trainings, together with health education in health facilities, helped to increase the demand for prenatal care, delivery by skilled staff, and postnatal care.

Over time, the HIS was improved to measure the UN process indicators for EmOC services, which include the number of EmOC services available, the geographical distribution of EmOC facilities, the proportion of all births in EmOC facilities, the met need for EmOC services, cesarean deliveries as a percentage of all births, and the case fatality rate [10]. IRC also instituted a maternal death audit, checklists to monitor supplies, skills and procedures, and midterm evaluations. Staff were trained in the use of new HIS and processing of the relevant data. This led to improvement in staff skills in documenting cases in the facilities and to improvements in services based on findings from data analysis.

3 Results

3.1 Maternal mortality

The program has had a positive impact on maternal mortality among the Afghan refugees residing in Hangu district. The maternal mortality ratio in 2004 was 102 per 100 000 live births according to the HIS, compared with 291 per 100 000 live births documented in the 2000 survey (95% CI, 181–400).

3.2 Utilization

The EmOC services are now well accepted by the community and utilization rates have increased consistently since 1996 (Fig. 2). The proportion of refugee births in an EmOC facility increased from 4.8% in 1996 to 67.2% in 2007. According to a follow-up KAP survey done in December 2006, 40.6% of women knew 6 danger signs during pregnancy, and 55.9% knew at least 2 [11].

Details are in the caption following the image

Proportion of refugee deliveries by skilled staff in EmOC facilities, 1996–2007.

As the skills of the staff in the EmOC facility improved, along with community demand, the met need for treatment of complications increased considerably to a high of 86.9% in 2005 (Fig. 3). This proportion fell in 2006 when one EmOC facility was closed for 8 weeks due to insecurity, but rose again in 2007, despite the facility closing for another 2 weeks due to poor security.

Details are in the caption following the image

Met need for EmOC services, 2004–2007.

3.3 Cesarean deliveries

Among the total cases referred from the IRC-run Basic EmOC facilities between 2004 and 2007 (n = 492), 49% (239) were managed by cesarean delivery. The total cesarean rate for the refugee population was 1.4%, notably short of the UN process indicator recommendation of 5%–15%. One could expect that if a higher coverage of facility-based deliveries had been achieved, a better detection of complications and a higher cesarean rate might result. It must be kept in mind that as successful as the program has been, the estimated MMR remains above 100 maternal deaths per 100 000 live births. It is conceivable that a greater proportion of cesarean deliveries could be the missing factor in better control of maternal death.

3.4 Case fatality rate

The case fatality rate for pregnancy complications among the refugee population seen either in the basic or comprehensive EmOC facilities, was 0.2%, well below the UN target of less than 1%.

3.5 Coverage

Table 1 shows data from the HIS from 2004—the first full year of operations of the 2 basic EmOC facilities—to 2007, the most recent full year of operations. These are the numbers of all live births in the population during the year, not only births in health facilities. The refugee population decreased during these years, as some families returned to Afghanistan, yielding fewer births per year.

Table 1. Births to Afghan refugees in Hangu district, Pakistan
Year 2004 2005 2006 2007 Total
Live births 3918 3497 2950 2870 13 235
Refugee births in IRC BEmOC facilities 2390 2413 1829 1866 8497
Referrals from BEmOC facilities to hospital 120 126 105 141 492
Cesarean deliveries 52 77 50 60 239
Maternal deaths 4 4 5 4 17

Simultaneously, complete prenatal care coverage (defined as 3 or more visits) increased from 49% in 2000 to 90% in 2006, and postnatal coverage within 72 hours of birth increased from 27.2% in 2000 to 84.5% in 2006 (Fig. 4). An apparent plateauing of the rate of increase in prenatal and postnatal coverage during 2005 and 2006 may reflect the effect of the insecurity that closed the health facilities for periods of time. It also could be an illustration of the need for alternative behavior change approaches among the remaining 10% of women and their families not yet motivated to seek prenatal and postnatal care.

Details are in the caption following the image

Prenatal and postnatal coverage, 2000–2006.

According to HIS data, the neonatal mortality rate fell from 25 per 1000 live births in 2000 to 20.7 per 1000 live births in 2006. The HIS within these camps has undergone extensive review and refinement over the years and has been shown to be quite accurate with regard to vital statistics among the refugee population. All pregnancies are followed; all pregnancy outcomes, complications, births, and deaths are documented.

4 Discussion

It is evident that IRC's programs serving the health needs of Afghan refugees in Pakistan's Hangu District were effective in reducing pregnancy-related death. The cost per delivery in IRC BEmOC facilities was approximately US $61. Our discussion will explore the factors that contributed to the success of this endeavor.

The most important factor to which we can attribute the availability and utilization of EmOC services is the consistent presence of dedicated staff. The effort has been led since its inception by people determined to learn from global efforts and to implement evidence-based good practice to address maternal mortality. The steps they followed in improving obstetric services in this setting reflect the lessons described in the work of the Averting Maternal Death and Disability program, in particular the building blocks framework [12]. First facilities were established; staff were trained and services were available 24/7. Once services were up and running and utilization began to pick up, linkages were strengthened with basic primary and community health activities; community members were engaged and utilization was promoted through many approaches.

An HIS had been in place to monitor the primary health care system prior to the establishment of EmOC services. However, as EmOC services became better established, the HIS was refined to be able to capture the UN process indicators. Data from the HIS were analyzed to guide improvements in the EmOC program, for example, to demonstrate the need for a second EmOC facility near Thal to serve the distant communities that could not reach the Hangu center.

Unusual in relief settings is the number of staff who have worked long-term in this program. Both the health coordinator and the health manager have been on the job for more than 20 years. This allowed an incremental effort at instituting good practice and a consistent voice communicating new concepts to conservative community leadership. Staff are available around-the-clock in the EmOC facilities and most of them have also worked with the program for many years. In fact, many of the EmOC staff are themselves members of the community they serve. This has the effect of building community trust and enhancing continuity of systems and services. Privacy and a friendly environment have also been key in increasing utilization and client satisfaction.

The role of highly skilled staff is even more important in EmOC services than in general PHC units, as the number of serious cases that require immediate, lifesaving treatment is higher in EmOC facilities. Basic EmOC is more effective in reducing mortality if supported by efficient comprehensive EmOC, and IRC's staff are able to refer and transport patients to the area hospital when needed. Over time, the referral network has been strengthened with a communication network, an equipped ambulance, and a feedback mechanism between the IRC-run basic EmOC centers and the government-run referral hospital.

PHC services in the camps are also strong, and the strategy of linking EmOC with other PHC components has proved effective in heightening awareness of the importance of skilled attendance and postnatal care. Feedback from the community through meetings with health committees, CHWs, and home visits provided insights that led to improvements in quality of care.

Educating BHU staff, as well as male and female CHWs, women of reproductive age, men in the community, health committee members, teachers, religious leaders, and private practitioners in the danger signs related to pregnancy increased awareness and recognition of complications and motivated women to seek timely support by skilled attendants.

An essential lesson of success is available from this program. Despite the myriad challenges of providing basic EmOC in a traditional, conservative refugee community in a relatively insecure, resource-poor area, services were delivered effectively, data were collected accurately, and the program has demonstrated that the health status of the population improved.

5 Epilogue

Tragically in the past few months, as we prepared this article for publication, Taliban influence has penetrated the district increasing insecurity, preventing women healthcare providers from working, and laboring women from using EmOC services. The result can only be more maternal deaths. IRC is working to again re-establish EmOC services outside the Taliban controlled area, but this will mean women must travel longer distances. The possibility of a maternity waiting home is under discussion among the community leaders as they try to assure women's access to the life-saving obstetric care that they now value so highly.


The authors wish to thank the European Commission Programme of Aid for Policies and Actions on Reproductive and Sexual Health and Rights in Developing Countries which, from 2004 to 2007, provided critical support for IRC's delivery of basic emergency obstetric care for Afghan refugees in the Hangu District of Pakistan's North West Frontier Province.