Abstract
Many pregnant women see unorthodox medical providers in labour before presentation to the modern medical facilities after obstetric complications have arisen. This study evaluates the contribution of unorthodox medical facilities to the delays subsisting maternal mortality in a rural, poor and illiterate community. Data was collected prospectively on all referrals from outside the St. Vincent's hospital, over a three-year period. Seven hundred and fifty women were referred to the hospital and there were a total of thirty maternal deaths out of the 1268 live births, giving a maternal mortality ratio of 2366/100,000. Most of the referrals were patient-driven and verbal and came from traditional birth attendants (TBAs). The majority of the patients (86.7%) came in poor clinical conditions and some were moribund. The TBAs contributed most to maternal deaths. Prolongation of labour for more than 24 hours correlated positively with maternal mortality. Ruptured uterus complicating obstructed labour (34.8%) and haemorrhage (30.4%) were the leading causes of death in this series. The mortal delay suffered by pregnant women in accessing unorthodox medical attention deserves further attention in issues of maternal mortality in the underserved rural communities of Nigeria.
Introduction
There is a large amount of literature describing the peculiarities of maternal deaths in Nigeria that have resulted in an unacceptably high mortality rate. 1–4 These studies have all emphasized that unbooked pregnant women and women from the rural areas make an alarmingly high contribution to the number of maternal deaths. Paradoxically, although the rural communities constitute over 75% of the population of Nigeria, 5 these areas have very poor health-care facilities and considerably fewer trained medical personnel compared with the urban areas. 6 The vacuum that is created by the lack of health-care facilities has been exploited by spiritualists, herbalists and traditional birth attendants (TBAs).
The medically unqualified alternative care providers have usually acquired their skill through many years of apprenticeship or by family succession. They live among the people, practise their trade informally and have a flexible payment-reward system. Their practice is therefore popular and culturally acceptable, attracting an evergrowing clientele both from the rural and urban communities. 7,8 They form the primary entry into health care in the rural setting and any failure of their treatment is usually attributed to the machinations of the evil spirits and other extraneous powerful influences. In the event of such failures, patients may ultimately seek orthodox medical attention as a last resort.
The Safe Motherhood initiative that was launched in Nairobi, Kenya, in 1987, and subsequently in Nigeria in 1990, aimed to improve maternal health by reducing maternal mortality through the promotion of family planning, prenatal care, safe delivery and access to emergency obstetric care. The delay in the institution of effective medical intervention – the third arm of the classic model of the three levels of delay in obstetric practice as enunciated by Thaddeus and Maine 9 – has paid insufficient attention to the mortal delay caused to women in labour in the rural settings by the traditional obstetric care providers.
This study explores the contribution to maternal mortality of delays caused by the traditional birth attendants, spiritualists, herbalists and other informal traditional medical practitioners in a rural community in south eastern Nigeria.
Materials and methods
Study background
The St Vincent Hospital, located in the rural centre of Ndubia-Igbeagu, was established in the 1960s by the Catholic diocese of Abakaliki. It is staffed by nurses, midwives, medical officers and a specialist obstetrician and gynaecologist. It serves the local, mainly agrarian, population and its environs which have low literacy and high poverty rates. The majority of the inhabitants are Christians; the other belief systems are the traditional religions and animism. Numerous TBAs are integrated within the communities and serve as the first point of contact for health care. Referrals are not usually made to the mission hospital until late in the disease process. Efforts by the hospital to educate and train the TBAs were initially received with enthusiasm, but soon waned as the programme was interpreted as a means to divert clientele from the TBAs. The documented maternal mortality ratio of the region 10 is well above the Nigerian national average of 1000/100,000 live births. 6
Methods
Data was prospectively collected on mothers who presented from outside the facility with intrapartum or postpartum maternal morbidity over a three-year interval – January 2004 to December 2006. We sought information on the socio-demographics of the parturients, points from which they presented, referrals, delays in unorthodox care centres and outcome of management in terms of maternal mortalities. The medical officer, who was trained as a research assistant and was conversant with the local dialect, interviewed the parturients where their clinical condition permitted, using a standard format. When necessary, information was obtained from a close relative. Ethical clearance was obtained from the hospital management and consent sought and obtained from the subjects on enrolment into the survey. For this study the Kuti classification of the clinical condition of the patient at presentation was adopted as follows: stable – when women presented with stable vital signs and without added complications from previous mismanagement or delay in presentation; and poor – when the patients were admitted with abnormal vital signs or superimposed complications. They were also referred to as moribund when they presented with severely compromised vital signs and circulatory collapse.
Data was entered into and analysed using the Epi Info statistical software package version 3.3.2. The chi-square test was employed for statistical association and a P value of <0.05 taken as significant.
Results
Seven hundred and fifty women were referred to the St Vincent's Hospital during the study period; they comprised 487 referred intrapartum and 263 referred in the postpartum period. There were a total of 1268 live births. Of these, 1118 delivered at St Vincent's hospital while 263 delivered outside the facility. Three of the women who were referred intrapartum were dead on arrival. Of the 1018 deliveries at the centre, 531 (52.2%) presented to the hospital and their labour was entirely managed by the hospital. Of these, 516 (97.2%) were booked in to the hospital and the rest were unbooked. The remaining 487 (47.8%), 318 (65.3%) of whom had earlier accessed prenatal care at the St Vincent's hospital, were referred in labour from various centres. The rest (34.7%) were unbooked. The 263 referred mothers who presented with postpartum complications had all been treated by TBAs or by the spiritual homes. Over 52% of these had earlier sought and received antenatal care from the facility.
Table 1 shows that the bulk of the referrals (650, 86.7%) came from the TBAs, 52 (6.9%) from nurse-led centres, 28 (3.7%) from the primary health care centres, 11 (1.5%) from spiritual homes and nine (1.2%) from peripheral clinics. The modes of referral were also shown. The majority (63.3%) came on self-referral from the centres. Of those treated by TBAs, 69.1% were self-referrals and 30.9% were referred verbally. None of the women from the spiritual homes were officially referred. Verbal referrals were the predominant mode of referral from the nurse-led facilities (76.9%) and primary health-care centres (75.0%). All the cases from the peripheral medical clinics had written referrals. Of those treated by TBAs, 70.9% were in a state of circulatory shock at presentation, 25.7% were clinically stable, 19 (2.9%) were moribund and 0.5% were brought in dead.
Types of referral and clinical condition of patients at presentation at the Mission Hospital (n = 750)
Nearly 43% of the referring centres were located within 10 to 20 km of the mission hospital and 18.8% were sited over 20 km away (Table 2). Over half of the patients spent more than 10 hours between referral and/or decision to leave the referring centres and presentation at St Vincent's Hospital. Only 10.3% spent less than 5 hours getting to the hospital. The majority of the patients arrived at the centre in the morning (56.3%). The corresponding figures for women referred from the TBAs and the spiritual homes were 61.7% and 63.6%, respectively. Those from the nurse-led facilities, the primary health centres and the privately operated medical clinics presented mostly during the day (afternoon or evening; Table 3).
Distance and time from the referring facility to presentation at the Mission Hospital
Factors contributory to maternal death in 26 parturients
*Three patients were dead on arrival
From the 1268 live births, there were a total of 30 maternal deaths during the period giving a maternal mortality ratio (MMR) of 2366 per 100,000 live births. Only four of the women who presented at the Mission Hospital directly suffered maternal death. Two died as a result of anaesthetic complications and two from uncontrollable eclampsia. Twenty-six of the referred patients died, 22 from the TBAs and two each from the spiritual homes and the nurse-led facilities (Table 4). No deaths were recorded among those referred from the primary health-care centres or the peripheral clinics. Ruptured uterus complicating obstructed labour was the most common cause of maternal death in the referred cases accounting for 34.8% of the deaths. This was closely followed by obstetric haemorrhage (30.4%). Sepsis accounted for 17.4% of the deaths and there was one anaesthetic death (Table 4).
Sources of referral and major causes of maternal deaths
MMR, maternal mortality ratio; TBA, traditional birth attendant
Further analysis of the maternal deaths in Table 3 reveals that duration of labour longer than 24 hours was positively correlated with maternal mortality (P value = 0.0012, degrees of freedom [df] = 2). Even though more women who died presented to the referral centre later than 10 hours, the difference in duration between referral and presentation did not achieve statistical significance (P value = 0.899, df = 2). Significantly, however, more deaths occurred in patients who presented during the night (P value 0.000, df=2). Most of the patients (82.6%) died within 12 hours of presentation at the Mission Hospital.
In Table 5, the reasons for the mortal delay in accessing orthodox medical attention from the referral centre ab initio, or the delay in getting to the centre after referral or decision to leave the referring centre were evaluated. All the parturients suffered a delay by seeking help and spending substantial periods of time at the referring centres. Transportation difficulties were also a major source of delay (92.3%). Other factors included: the association of orthodox medical centres with operative delivery (69.2%) or blood transfusion (34.6%); financial considerations (57.7%); spouse's decision to seek attention from alternative sources or his refusal to take the decision to seek help (30.8% and 19.2%, respectively); and suggestions from peers to patronize unorthodox sources (30.8%). Six women (23.1%) had confidence in the TBAs, while in 19.2% the referring centres resisted early referrals. Among the patients 15.4% were unaware of the modern medical facility in the locality and 15.4% were deterred by the perceived negative attitude of the medical personnel at such centres.
Analysis of the background factors underlying the mortal delay
Discussion
The framework of delays subsisting maternal mortality as reported by Thaddeus and Maine 9 presupposes that a decision is taken by a parturient to seek medical attention from a competent source. The delays appertained to late decisions, transportation difficulties and the institution of definitive therapy. This profile may be broadly applied to all settings, including the rural population. However, in communities where poverty and illiteracy are rife, and charlatans offer a range of obstetric services, there may be another form of delay. This may be termed the ‘mortal’ delay and may be defined as delay caused by seeking medical attention from incompetent sources which result in a delay in referring the patient to an appropriate care facility and, consequently, in serious obstetric complications. This was clearly demonstrated in our study. Although it was not possible to determine the total number of parturients that sought attention from unorthodox facilities, referrals from those sources accounted for 59% of all cases handled at the referral centre. Obstetric danger signals went unrecognized by the unorthodox facilities, referrals were delayed or refused and, eventually, the patient presented at the orthodox medical centres in a clinical state. In this survey, the majority of the patients were in clinically poor conditions at presentation. Some were moribund on admission and died within hours of presentation. The long stay in the unorthodox centres greatly increased the total duration of labour which correlated very strongly with maternal death. Prompt access to emergency obstetric care is the key to reducing maternal mortality. 6
Prevention of prolonged labour has been identified as crucial to reducing maternal morbidity and mortality attending labour, hence the employment of the partograph in modern obstetric units. 11,12 The resistance of the TBAs and spiritualists to early referrals was clearly identified by the relatives of those who died as contributing to the mortalities. This was supported by the fact that there were no official referrals in more than 63% of the cases (Table 3). In such instances, the relatives of the patient often remove the parturient from the care of the TBA or spiritualist against their advice in order to seek alternative medical attention. The lack of knowledge and competence on the part of the TBAs and spiritualists was evident in this study, as all the postpartum referrals that were previously under their care had major complications and were mainly in poor clinical condition on admission. Intrapartum pointers and antenatal risk factors had clearly not been recognized. In an earlier study in Ilesa, southwestern Nigeria, Kuti and co-workers 1 noted similar findings. All the patients from the TBAs in their study were mismanaged and over 83% presented in a poor clinical condition. Correspondingly, the mission homes mismanaged 87.5% of their cases, 75% presenting in poor condition. Although a lack of, or delay in, referrals by such attendants may be attributed to ignorance, apprehension over possible scrutiny by, or reprisals from, the formal health sector may be contributory. In most states in Nigeria, maternal mortalities in the rural areas go unreported, undocumented and uninvestigated.
Resistance to referrals often implies that referrals occur in the night when transportation difficulties may further increase labour duration and adverse maternal outcome. Transportation in rural Nigeria is adversely affected by the poor road network, difficult terrain, especially during the rainy season, and poor communication network. It was not surprising that the majority of such parturients arrived at the referral centre in the morning after having waited throughout the night. Delays may also have been a result of the patient's arrival at the referral centre during the night, when the emergency obstetric and ancillary services may not have been optimal and may have been another reason for the poor outcome (Table 4).
The significant contribution of patients from the traditional birth homes underscores the importance of the TBAs in childbirth in rural Nigeria. 7,8 They draw their patronage from both the rural and urban population 8 and therefore they cannot be ignored. It should also be noted that the majority of the referred parturients had earlier booked and received prenatal care at the Mission Hospital but elected to deliver outside the modern facility. In Nigeria it is documented that only a proportion of the women who utilized professional care in the antenatal period, delivered under the care of skilled birth attendants, professionals or in modern health centres. 6 Nevertheless, this characteristic may serve as a platform for an educational programme for women in the prenatal period on the advantages of delivery by skilled carers and the attending complications that may follow delivery from other centres.
Most maternal deaths in Nigeria are preventable (Table 4). Recognition of obstetric complications is paramount in the prevention of maternal mortality. The source-specific MMR for the TBAs, nurse-led facilities and spiritualists (3385, 3846 and 18,182 per 100,000 live births, respectively) are sources of concern. Target 5 of the Millennium Development goals may remain unattainable if concerted effort is not made to tackle the mortal (type ‘M’) delay in economically deprived and ill-educated populations of the developing countries. Also contributing to the type ‘M’ delay is the negative impression of the orthodox facilities, including the presumed nonchalant attitude of health workers, which cause parturients to turn to alternative health systems. The role of the male in male-dominated populations was also evident in this study. The men should, therefore, become the targets for any interventional strategies in order to eliminate this mortal delay and ensure the timely uptake of the maternity services available at the modern health facilities. Financial constraints were also reasons given for patronizing the unorthodox health facilities.
Policy makers, donor agencies, governments and their agencies should ensure the fair distribution of health-care facilities, formulate strategies aimed at encouraging the uptake of modern maternity services among the rural population, discourage the patronage of charlatans, eliminate the type ‘M’ delay and, thus, ensure realization of the Millennium Development Goals.