Model Inputs \(\rightarrow\) Health System Parameters \(\rightarrow\) Recognition/Referral
In order to be treated, a complication must first be recognized, and perhaps referred to an appropriate facility. Recognition/referral of a complication is conceptualized as the “first delay” in the framework developed by Thaddeus and Maine.[1] We model different probabilities of recognition/referral by starting delivery site and complication type/severity.
The broader concept of recognition/referral encapsulates different aspects of complication management and health-seeking behavior such as: technical knowledge of complication danger signs, cultural acceptability of referral (ex: whether the husband is present and gives permission for the wife to go), financial barriers, childcare arrangements, etc. A systematic review found five main sub-themes related to the first delay: socio-economic factors, community perceptions about obstetric complications, women’s autonomy and insufficient male partner involvement in the management of obstetric emergency, maternal obstetric history and health service utilization, and women’s knowledge about obstetric danger signs.[2] These factors contributed to delays at home to seeking appropriate obstetric care in a timely manner when needed.
We undertook a targeted literature review of recognition/referral pathways and women’s knowledge of obstetric danger signs, summarized by country below.
Bangladesh:
A study describing the process of recognition and response to symptoms during potential prolonged labour found that the most frequent and usually first action was to seek care from untrained attendants at home, followed from professional attendants outside the home, with care-seeking outside the home occurring a median of 19 hours after perceived labor onset.[3] Another study in rural Bangladesh of women who reported a ‘serious or very serious’ complication during their last pregnancy/delivery found that most women sought care for ‘serious’ complications (86%), with 42% seeking multiple sources of care.[4] A community-based cross-sectional survey in Matlab of women who delivered in 2014 found that 26% of women could recognize three or more pregnancy danger signs, and 23% could recognize three or more delivery danger signs.[5]
Ethiopia:
A systematic review and meta-analysis found that 32% (95% CI: 25.6-38.5) of pregnant women were prepared for birth and its complications, and only 26.33% of pregnant women were aware of danger signs during pregnancy.[6] Another systematic review and meta-analysis found that the level of women’s knowledge about obstetric danger signs during pregnancy, delivery, and postpartum was 48%, 43%, and 32%, respectively. Maternal age, education, income, health service use, distance from facility, and women’s autonomy were reported in several studies as determinants of women’s knowledge of obstetric danger signs.[7]
Ghana:
A household survey in East Mamprusi District found that 51.1%, 29.4%, and 34.6% of women knew at least three key danger signs during pregnancy, delivery, and postpartum period, respectively.[8]
Haiti, Malawi, Senegal:
A study of ANC clients based on service provision and assessment (SPA) surveys in Haiti, Malawi, and Senegal found that in all three countries more than half of clients knew at least one danger sign, but most clients could report only one, while almost no clients knew all seven.[9]
India:
A study of pregnant women attending the outpatient department of an urban tertiary care hospital for the first time found that 20%, 15.8%, and 12% had knowledge of at least 3 danger signs of pregnancy, labor, and severe illness in newborns, respectively.[10] A cross-sectional study of antenatal and postnatal mothers registered in a Rural Health Training Center in Kancheepruam district, Tamil Nadu found that 21% of women had adequate knowledge (correctly answering at least 6/12 questions) regarding the danger signs of pregnancy, with maternal literacy and adequate prenatal care major determinants of respondent knowledge.[11] Another study in Dakshina Kannada district, Karnataka found that 53.8% of women were aware of at least 6 danger signs.[12] A study in a rural block of Haryana found that >8 danger signs were known to only 9.5% of participants.[13] Studies have found that knowledge of danger signs is likely somewhat higher among trained birth attendants. A study of Accredited Social Health Activists (ASHAs) in a rural district in Karnataka found that 1%, 4.8%, and 7.2% were aware of all key danger signs for labor and child birth, postpartum period, and pregnancy period, with a majority (71%) scoring 4-7 out of 8.[14] A study in three states (Maharashtra, Rajasthan, and Odisha) exploring the adequacy of knowledge and clinical skills of ayurvedic and homeopathic practitioners engaged by the state governments to provide maternity care services found that nearly 76% had adequate theoretical knowledge of essential obstetric care and identification and management of complications of pregnancy.[15] Looking at actual responses to complications (not just general knowledge of danger signs), studies have found higher proportions of women who seek care. Using a broader definition of “complication”, the Coverage Evaluation Survey (CES) 2009-10 found that 32.8% of women who experienced a complication did not consult anybody, implying that the remaining ~70% of women did seek some form of care.[16] Similar findings were observed by Montgomery (2014): a study based on a nationally representative survey found that two-thirds of women died seeking some form of healthcare, with most seeking care in a critical medical condition - it also found that rural areas of poorer states had proportionately lower access to and utilization of healthcare services than in urban areas, but that this rural-urban disparity was not seen in richer states.[[17][Montogmery]] Another study found that out of 153 maternal deaths identified through community workers, verbal autopsies indicated that 23% occurred at home and that 30% occurred on the way to a health facility.[18]
Indonesia:
A study in Jayawijaya district, Papua province in eastern Indonesia, found that the decision to seek care lies predominantly with the husband. There was no significant lag time between the decision-making stage and the process of seeking care for the maternal cases. On average, the interval between symptom recognition and the decision to seek care in the home was less than 6 hours. Rapid decision-making usually occurred when visible symptoms of excessive bleeding began to appear.[19]
Kenya:
A study of women attending antenatal care at Kenyatta National hospital found that 67% of respondents knew at least one danger sign in pregnancy, while only 6.9% knew of three or more danger signs, with level of education positively influencing birth preparedness.[20] A CEmOC facility-based qualitative study of 30 women who experience obstetric “near miss” in Malindi found that 16/30 women experienced delays in making the decision to seek care, compared to 8 women who experienced delays in reaching the care facility, and 6 experienced delays in receiving care at the facility.[21] A community-based cross-sectional survey of community health workers/volunteers in North and East Kamagambo, Migori County found that 60% of participants in North Kamagambo knew 3 or more danger signs in 3 or more categories, compared to 24% of participants in East Kamagambo, revealing varying levels of knowledge among the two populations of lay health workers.[22]
Madagascar:
A convenience sample of 372 women in their first year postpartum between April-October 2015 found that knowledge of at least one danger sign varied from 80.9% of women knowing danger signs in pregnancy, to 51.9% at delivery and 50.8% postpartum, with women with higher education more likely to know danger signs.[23]
Malawi:
A study of primigravidae women in a rural health center in Blantyre found that 82% (95% CI 67-96) of women had some knowledge and could make an informed decision to go to a health facility with pregnancy complications, with 61% (95% CI 42-79) able to do so with complications after delivery.[24] A following study of primigravidae women in an urban health center found that 60% of the participants were knowledgeable about obstetric complications in pregnancy, however, 73% and 82.2% did not know of any problems that could occur during and after delivery, respectively.[25] Participants had limited knowledge of complications that may need immediate treatment during all three periods: 58% (95% CI 43-73) had some knowledge and could make an informed decision to go to a health facility with pregnancy complications, however only 24% (95% CI 11-38) could do the same for complications after delivery.[25]
Morocco & Netherlands:
A qualitative study to describe knowledge of hypertensive disorders in pregnancy among pregnant Moroccan women in Morocco and the Netherlands found that half of them had never heard about hypertension in pregnancy and had no knowledge of symptoms or alarm signals.[26]
Myanmar:
A cross-sectional survey of 262 auxiliary midwives from July 2015 to July 2016 found that only 8% were able to identify at least 80% of 20 critical danger signs, and 57.6% knew three or more critical danger signs during pregnancy, and 54.2% knew four or more critical signs during birth and postpartum.[27]
Nepal:
A study of group antenatal care in six villages in Achham found that among the control cohort, 8-10% of women could identify danger signs during pregnancy, labor and childbirth, or postpartum period.[28]
Nicaragua:
A study examining the impact of a radio-education intervention on knowledge of pregnancy danger signs found that the total number of signs identified by study participants increased by 53.8%, and found that urban setting and more than sixth-grade education were factors making it significantly more likely to score higher on post-test related to knowledge of danger signs.[29]
Nigeria:
A community-based survey in Osogbo metropolis found that 70.8% of women were aware of danger signs in pregnancy, with vaginal bleeding the commonest danger sign mentioned.[30] A cross-sectional study at the postnatal ward at Federal Teaching Hospital in Ebonyi found that most of the women knew about birth preparedness (87.7%) and complication readiness (79.5%).[31] A cross-sectional study of recently delivered women at selected health facilities in Ikenne found that 34.6% of women had knowledge of 5+ danger signs of pregnancy.[32] A cross-sectional study of pregnant women attending the prenatal clinic at a tertiary hospital found that 81.5% were “well prepared” for birth and its complications, however, only 27.4% knew key danger signs during labor/delivery and 24.9% knew those in the first 2 days after delivery.[33] A study of women in Zaria found that 18.31%, 9.89%, and 9.24% knew at least four danger signs during pregnancy, labor and delivery, and postpartum period, respectively.[34] A study evaluating the Maternal and Child Health Integrated Program (MCHIP) in Kano and Zamfara states found that mothers’ knowledge of birth preparedness/complication readiness was 32.2% among those who received counseling compared with 11.2% among those who did not.[35]
Pakistan:
A study in three rural areas around Islamabad found that the majority of women (75%) were unaware of key obstetric danger signs and symptoms, with traditional practices and home remedies commonly used to manage complications in pregnancy before seeking medical care, which was only done when the condition became unmanageable at home.[36]
Rwanda:
A study in 2015 at University Teach Hospital of Kigali of all admitted pregnancy-related complications found that the majority of first delays resulted from mothers failing to recognize the problem (22.3%). Women with hypertensive disorders represented 9.1% who failed to recognized signs of severity, followed by women who experienced abortive outcomes (abortion or ectopic pregnancy).[37] A cross-sectional study among pregnant women who were referred to Ruhengeri hospital between July and November 2015, found that out of 350 women, 296 (84.6%), 271 (77.4%) and 288 (82.3%) could mention at least one key danger sign during pregnancy, labor and postpartum respectively, but only 23 (6.6%) could mention three or more key danger signs during all three periods.[38]
South Africa:
A study of 340 pregnant women in KwaZulu-Natal found that although most of the study population (92%) attended health care facilities, only half (52%) of them knew about some of the “danger signs” of pregnancy, and 39% of them knew about their HIV status.[39]
Tanzania:
A community-based cross-sectional study of women who delivered in the previous two years in Chamwino found that 58.2% of respondents were considered to be prepared for birth and its complications, with maternal education a significant determinant of preparedness.[40] However, 68.6% of women knew 0 danger signs, with 6.1% knowing 1-4 and 25.2$ knowing 5+.[40] A follow-up study in Chamwino District found that 25.2% of respondents could mention at least five danger signs in any of the three phases of childbirth (pregnancy, childbirth, postpartum) with at least one in each phase, while 68.7% did not mention any danger signs.[41] A study of women who became pregnant or gave birth in the two preceding years in Mpwapwa found that 14.8% of women knew three or more obstetric danger signs, with the most commonly known signs including vaginal bleeding during pregnancy (19%), foul smelling vaginal discharge (15%), and baby stops moving (14.3%).[42] A study to evaluate the effectiveness of Home Based Life Saving Skills education found that the proportion of women who knew three or more danger signs improved substantially: 15.2% vs 48.1% during pregnancy, 15.3% vs 43.1% during childbirth, and 8.8% vs 19.8% in the postpartum period.[43] A study of women how had been pregnant in the past two years found that 51.1% knew at least one obstetric danger sign, with 26% knowing at least one danger sign during pregnancy, 23% during delivery, and 40% after delivery. However, few women knew three or more danger signs. Having secondary education was found to increase the likelihood of awareness of obstetric danger signs six-fold (OR=5.8, 95% CI 1.8-19.0) in comparison with no education at all.[44] A study of randomly selected pregnant women attending antenatal care in Dodoma municipality found that only 46.0% of women in the intervention and 44.7% of participants had knowledge on obstetric and newborn danger signs, rising to 77.3% after the intervention, compared to 48.0% in the control group.[45] A study in Kinondoni Municipality, Dar es Salam found that 57.8% of women were able to spontaneously mention only one to three danger signs, 31% had correct knowledge of at least four signs, and 2.7% were not able to mention any item.[46] However, among women who reported actually experiencing danger signs during their pregnancy (17.4% of respondents), 91% visited a healthcare facility.[46] The most commonly known pregnancy signs were vaginal bleeding (81%), swelling of the fingers, face, and legs (46%), and severe headache (44%), and older women were more likely to have knowledge of danger signs than young women (OR 1.61, 95% CI 1.05-2.46).[46]
Turkey:
A study of 125 pregnant women who had been admitted to the emergency department for antenatal bleeding found that advanced age, high level of education, lack of health insurance, receiving antenatal care, nuclear family structure, and knowledge of the danger signs during pregnancy were found to affect the use of emergency obstetric care services.[47]
Uganda:
A study of 810 women admitted in the antepartum period to Mulago hospital, found that only about 1 in 3 women were able to mention at least three of the five basic components of birth preparedness/complication readiness, and 1 in every 4 women could not mention any of the five components.[48] A study of 764 recently delivered women from 112 villages in Mbarara district found that 52% of women knew at least one key danger sign during pregnancy, 72% during delivery, and 72% during postpartum. Only 19% had knowledge of 3 or more key danger signs during the three periods. Overall 35% of the respondents were birth prepared. Young age and high levels of education had a synergistic effect on the relationship between knowledge and birth preparedness.[49] A study 0f 80 pregnant women attending antenatal clinic in Adjumani found that 76.25% of the respondents mentioned vaginal bleeding and 62.5% vomiting as danger signs in pregnancy, while 12.5% did not know any danger sign in pregnancy.[50]
We used hierarchical logistic regression models to model the probability of recognition and referral. Based on findings in the literature, we assume that women with higher education are more likely to recognize and have the autonomy and necessary support to be referred. We also assume that more severe complications are more likely to be recognized and referred, and that recognition/referral probabilities increase with more advanced delivery site, with Home being the lowest, followed by Home-SBA, non-EmOC, BEmOC, and CEmOC. These constraints were enforced when sampling probabilities in the model. Based on the literature, we also assume that ANC visits improve the probability of recognizing complications for women delivering at Home (see ANC).
If a delivery complication occurs, the probability of recognition/referral is drawn. A complication has to be recognized in order to be treated in the model, and depending on the woman’s current delivery site and the severity of the complication, she may be referred to another facility. Specifically, we assume that any complication that occurs at Home or non-EmOC will be referred, and any ‘severe’ complication that occurs at a BEmOC will be referred. Recognized complications that occur at a CEmOC facility are not referred. We currently assume that no false positives occur (i.e. 100% specificity).
GMatH (Global Maternal Health) Model - Last updated: 28 November 2022
© Copyright 2020-2022 Zachary J. Ward
zward@hsph.harvard.edu