Utility of the three-delays model and its potential for supporting a solution-based approach to accessing intrapartum care in low- and middle-income countries. A qualitative evidence synthesis

Figures & data

Table 1. Search terms.

Sample	Woman OR mother OR pregnant OR parturient OR female
Phenomenon of Interest	Delays = delay OR wait OR time OR 3-delays AND Intrapartum care = Intrapartum OR delivery OR labour OR childbirth OR birth OR obstetric AND Low income settings = low income OR developing country OR LMIC OR LMICs OR sub-Saharan Africa OR Asia OR middle-income countries OR Latina America
Design	NA
Evaluation	Views OR opinions OR perceptions OR beliefs OR attitudes
Research type	Qualitative OR Mixed-method OR Phenomenology OR Grounded theory

Figure 1. PRISMA flow chart.

Figure 2. New contributing factors and re-grouping proposed to the three delays model.

Figure 3. Changes proposed to the structure of the three delays model.

Figure 4. The women’s health empowerment model, developed by the authors.

Table A1. Studies using the 3-delays model without any modification (N = 5).

					Factors contributing to the delays
No	Author and (ref) Year of publication Study aim	Country	Type of study Sample population	Method of data collection Key informants	1^st Delay	2^nd Delay	3^rd Delay	Individual plans of the woman
1.	Urassa et al.,1997 To identify operational factors contributing to maternal death	Tanzania Ilala district Dar es Salaam Urban	Mixed-method 117 Maternal deaths	Verbal autopsy, including interviews Relatives and health staff, senior administrators and health institutions	Women’s status (decision-making by mother-in-law, husband, nurse) Tradition (use of traditional healer first)	Distance Long transfer (travel time)	Poorly staffed facilities (limited number of nurses and doctors, lack of motivation) Poorly equipped facilities (blood, lack of drugs and theatre not always working) Inadequate management (poor decision, wrong management)
2.	Barnes-Josiah et al., 1998 To analyse the complex interplay between the personal decision of Haitian women and their families to seek care when faced with an obstetric emergency	Haiti	National prospective study with qualitative component 12 Maternal death	Verbal and social autopsies Creation of case histories Husband, mothers, friends, neighbours, TBAs, physician and clinic personnel	Indirect cost of seeking care Perceived quality of care (known lack of personnel at nearest hospital)	Distance Lack of means of transport Cost	Poorly staffed facilities (inability to perform a c/section) Inadequate management (inability to diagnose and treat acute conditions, being given wrong treatment and early discharge, inability/unwillingness to arrange referrals)
3.	D’Ambruoso et al., 2010 To extend the standard VA to gather and present information on experiences of emergency care-seeking, from the perspective of family members who experience some or all of the relevant event.	Burkina Faso (Ourgaye: remote rural setting) Indonesia (Serang and Pandelang: rural settings)	Mixed-method study 70 Maternal deaths (Burkina Faso) 104 Maternal deaths (Indonesia)	Verbal autopsy Structed questionnaires with open-ended questions Relatives of the deceased women	Illness factor (no knowledge danger signs) Women’s status (decision-making by family members) Perceived accessibility (transport/cost) Previous experience of care (dissatisfaction with midwives) Tradition: belief in magic and God not requiring medical intervention	Distance Lack of means of transport (use of non-motorised vehicles, limited a night) Cost for transport (loans)	Poorly staffed facilities (lack of staff) Poorly equipped facilities Inadequate management (delayed treatment, care contingent to payment)
4.	Khan and Pradhan, 2013 *Indirect use of the model To understand the missed opportunities to save maternal lives and explore social dimensions contributing to maternal mortality	India Five districts: Palamu, West Singhbhum, Giridih, Godda and Gumla	Cross-sectional study with qualitative data collection 833 Maternal deaths	Verbal autopsy Interviews with husbands and mothers in law	Illness factor (+ knowledge of danger signs; perceived severity of illness but no action) Socio-legal issues (unsafe abortion) Economic status (poverty) Tradition (use of local remedies to treat danger sign before formal care)	Distance Lack of means of transport (not always motorised) Cost of transport	Poorly staffed facility (lack of competences to identify complications, provision of wrong treatment) Inadequate management (care not available at the first facility needing further referral)
5.	Sharma et al.,2017 To investigate recognition, decision-making and care-seeking among facilities who experienced a maternal death, a reported postpartum haemorrhage; sequence of care-seeking actions; the role of husbands in these processes; and how perception of risk influence decision-making	Nigeria Jigawa state: 24 districts	Qualitative study included in an RCT 10 Maternal death 10 Women with PPH	Illness narrative and qualitative group interviews Women with PPH. For MD: husband, family members, neighbours, TBAs	Illness factor (no symptoms recognition by family members, different perception of severity of conditions based on previous experience) Women’s status (decision-making mainly by husband, mother-in-law, sister, traditional birth attendants) Perceived accessibility (cost) Previous experience with care (fear of health workers, absence of providers) Tradition (fatalism, use of spiritual care before formal care)	Lack of transportation

Table A2. Studies using the three delays model but adding new factors explaining the delays (N = 10). New factors are underlined.

					Factors contributing to the delays
No	Author and (ref) Year of publication Study aim	Country	Type of study Sample population	Method of data collection Key informants	1^st Delay	2^nd Delay	3^rd Delay	Individual plans of the woman
1.	Essendi et al., 2011 *Indirect use of the model To investigate poor urban Kenyan men’s and women’s views on the factors that hinder the uptake of formal obstetric care services	Kenya Nairobi slum of Viwandani and Korogocho	Qualitative study Women with obstetric complication (sample size not reported)	Interviews and focus group discussion (16 in total) Women, male partners, Traditional Birth Attendants; Female opinion leaders; Male opinion leaders.	Illness factor (+ knowledge of danger sign)History of uncomplicated delivered limited cause for alarm Women’s status (decision-making by mother-in-law and husband or TBAs) Economic status (unaffordable care) Tradition (TBA’s choice as affordable) Perceived accessibility (cost of transport at night, insecurity at night) Satisfaction with service (unfriendly staff/procedures)	Distance (insecurity at night) Lack of means of transport at night Cost for referrals Poor road network	Poorly equipped health facilities (blood, drugs) Inadequate management (poor referrals) Need to show proof of ANC attendance to be admitted Care contingent to payment (additional amount for complications) and detention without bills been cleared Disrespectful care: poor attitudes, preference for male birth attendant, fear of ill treatments
2.	Jammeh et al.,2011 To provide a better understanding of the barriers to timely access to emergency obstetric care resulting in perinatal deaths and in survivors of severe obstetric complications	The Gambia Bansang hospital Central river region rural	Qualitative study 20 women surviving severe obstetric complications	Individual in-depth interviews Women, family members and TBAs	Illness factor (+ knowledge danger signs/perceived severity of illness) Women’s status (decision-making by mother-in-law and husband) Indirect cost of seeking care (transport and food at hospital) Tradition (use of local remedies to treat danger sign before formal care)	Distance (remote rural areas) and dependence from ferry service Poor road network (rainy season) Lack of means of transport Lack of ambulances for referrals	Poorly equipped facilities (blood)	Care seeking decisions and procedure Process of getting care
3.	Kabali et al., 2011 To report the circumstances around the occurrence of complications that lead to death or near-miss, drawing on the testimonies of women who survived serious complication, and of families of women who passed away.	Democratic Republic of Congo Five District of Kinshasa city, urban	Qualitative study 208 Near-miss women 103 Maternal deaths	Semi-structured interviews and verbal autopsy Near-miss women For deceased women: husbands, mothers, step mothers, sister, sister in law, friends, neighbours	Illness factor (knowledge danger sign but neglecting them, severity of illness not threatening) Perceived accessibility (cost for C-section not affordable) Tradition (traditional remedies when formal care failed)		Poorly equipped facilities (blood, theatre not available) Inadequate management (students not doctors for inability to pay) Payment before treatment Poor attitude of staff (negligence)
4.	Echoka et al., 2014 To explore the barriers to EmOC services by women who experienced life-threatening obstetric complications or near miss	Kenya Malindi district, costal urban	Qualitative study 30 Near-miss women	In-depth interviews Women	Illness factor (fair knowledge of danger signs) Economic status (home delivery to contain cost)	Poor road infrastructure Lack of ambulance (need to organise own transport) Multiple referrals	Poorly staffed facilities (unavailability of doctors) Poorly equipped facilities Inadequate management (long queue before being treated)	Info collected on birth preparedness and place of birth Results: women made plans with TBAs and organise their own transport
5.	Gebrehiwot et al., 2014 To explore health service providers’ perceptions of facilitators and barriers to the utilisation of institutional delivery	Ethiopia Tigray, two rural districts: Ganta-afeshum, Kilte-awlaelo	Qualitative study 12 Health extension workers 4 Midwives	In-depth interviews	Illness factor (+ knowledge of danger signs) Perceived accessibility (family’s reluctance to ask for transport to non-relative) Tradition: home delivery preferred/beliefs around going to health facility while pregnant	Distribution and location of facilities (mountain setting) Poor road network Poor network connection to call an ambulance	Poorly staffed facilities (lack of competences to detect complications and organise referrals) Poorly equipped facilities (equipment, electricity, water Relatives on allowed in the delivery room Negative attitude of providers
6.	Lori and Starke, 2012 To explore the cause and circumstances surrounding maternal mortality and severe morbidity	Liberia One rural county	Secondary analysis of maternal death and near-miss audits with interviews 120 Near-misses’ women 28 Maternal deaths	Interview with open-ended questions Women (NM) For deceased women: family members, community members and health workers	Illness factor (+ recognition of obstetric complications) Culture (hiding the problem/self-treatment, use traditional remedies before formal care) Women’s status (decision-making by family)	Distribution and location of facilities (remoteness) Lack of means of transport Non-motorised transport Cost for transport Lack of assistance at night (safety) Seeking care at multiple levels before tertiary	Poorly equipped facilities (delay receiving blood)
7.	Cham et al., 2005 To describe the socio-cultural and health service factors associated with maternal death to create an operative understanding of the concept of access	The Gambia Central and upper river division, rural	Qualitative study 42 Maternal deaths	Verbal autopsy and group interviews Family members, in-laws, husbands Health staff: midwives, ambulance drivers, generator operators, lab personnel, taxi drivers, ferry captain	Women’s status (following advice of older women about when to seek care) Economic status (cost of care) Perceived quality of care (fear of punishment from providers, maternal care only available when clinics opens, poor attitude of staff) Belief (progress of labour follows times linked to Muslim prayer) Underestimation of severity of complications due to previous uneventful home birth	Distance (travel time) Poor road network Lack of motorised vehicles Prolonged transportation Lack of ambulances/fuel Multiple referrals (Seeking care at inappropriate level)	Poorly staffed facilities (doctors, lack of competences) Poorly equipped facilities (blood, cost of supplies) Inadequate management (delay in providing care) Information from providers about where to give birth not clearly understood
8.	Mgawadere et al., 2017 To used the 3DM to analyse the delays associated with all maternal deaths occurred in one year.	Malawi Mangochi, semi-urban	Mixed-method study 151 Maternal deaths	RAMOS survey, including verbal autopsy Creation of case histories Household members, relatives, TBAs, medical personnel	Illness factor (lack of knowledge of obstetric complications) Women’s status (no financial empowerment if husband not available) Economic status (low family income) Perceived accessibility (distance) Previous experience with care (bad, poor quality) Tradition: visit TBA or traditional healer first Not wanting to stay in labour ward for long period Uneventful previous deliveries at home	Poor road network Distribution and location of facilities Distance (travel time) Cost for transport Slow means of transports	Poorly staff facilities (limited number, lack competences) Poorly equipped facilities (drugs, supplies, equipment, blood) Inadequate management (wrong assessment, wrong diagnosis, wrong treatment, lack of treatment guidelines) Inadequate referral system (transport not arriving for referral)
9.	Pagalday-Olivares et al., 2017 To assess the feasibility, in terms of potential of and requirements, of eHealth solutions to improve maternal healthcare in remote areas of Kpando, Ghana	Ghana Kpando: 6 Island and 5 peninsulas of Lake Volta, rural	Mixed-methods study Pregnant women (sample size not indicated) 15 Health workers 12 Maternal care stakeholders	Semi-structured interviews and focus group discussion Women, TBAs, men, Midwives, Representative of education services, national health insurance, representative of private facilities	Education (lack) Tradition (TBA as preferred choice of care)	Poor road network Distribution and location of facilities (geographical setting) Lack of means of transport Cost for transport	Poorly staffed facilities Poorly equipped facilities Inadequate management (inconsistent communication) Healthcare providers lack of knowledge about clients’ previous complications Poor referral system
10.	Sk et al., 2019 To identify facility and community level factors that contribute to maternal deaths.	India West Bengal, urban	Mixed-methods study (use of primary and secondary data) 317 Maternal deaths	Verbal autopsies (40) + hospital records Family members, neighbours and relatives	Illness factor (lack knowledge of danger sign of pregnancy and severity of illness) Education (illiteracy and ignorance) Tradition and beliefs No birth preparedness No antenatal care Previous uncomplicated pregnancies	Lack of means of transport Poor condition of roads Cost for transport and treatment (delay to mobilise funds, borrowing money) Mean of transport (not motorised) Multiple referrals	Poorly staffed facilities (lack medical staff) Poorly equipped facilities (blood and medical supplies) Inadequate management (delay in starting treatment)	Birth preparedness included in the assessment but not discussed in the results

Table A3. Studies proposing a change in the definition of delay (N = 3).

					Factors contributing to the delays
No	Author and (ref) Year of publication Study aim	Country	Type of study Sample population	Method of data collection Key informants	1^st Delay	2^nd Delay	3^rd Delay	Changes to the definition of delays
1.	Rodríguez Villamizar et al., 2011 To define critical points of change in the maternal care process of care, guide the decision-making around the issue and support the strengthening of policies for the provision of services with a view of achieving the MDG of improving maternal death.	Colombia Bucaramanga	Retrospective descriptive study 10 Maternal deaths	Epidemiologic reporting cards, clinical histories, field visits with interviews Family members and obstetricians	Illness factors (no knowledge of danger sign) Late access to ANC thus late learning of danger signs Women’s unawareness of their health rights	Poor referral between facilities (difficulties between lower and higher-level facilities)	Poorly staffed facilities (limited number of staff, lack of competencies) Poorly equipped facilities (blood) Inadequate management (incorrect diagnosis and treatment)	First Delay separated in 2 phases Recognition of danger sign of pregnancies Decision making process around seeking care and action
2.	Jithesh and Ravindran, 2016 To explore social and health system factors contributing to the higher proportion of maternal deaths in the district.	India Wayanad, Kerala, less developed district (rural)	Qualitative study 14 maternal deaths	Social autopsy with In-depth interviews Household members, gynaecologists, PHC medical officers, junior public health nurses	Women’s status (concealing their status for fear of being blamed by family members) Economic status (poverty) Perceived accessibility (remoteness) Perceived quality of care (fear of C-section) Domestic violence Previous birth at home	Distribution and location of facilities (remote location) Poor road condition Difficult accessibility during rainy season Multiple referrals (facilities in opposite directions)	Poorly staff facilities (poor competencies and accountability, fear for decision taken) Poor equipped facilities (blood) Inadequate management (poor stabilization of women before referral, sub-standard care)	Delay’s definition to capture delays due to multiple referrals 1st delay: delay of > 1 h from the occurrence of complications during pregnancy or delivery to actually starting the journey to appropriate health facility 2^nd delay: any delay of > 1 h from the start of the journey (where complication occurred) to reaching a facility where the woman receives definitive treatment.
3.	Chalet et al., 2017 To reaffirm and challenge assumptions around the drivers of the first two delays and discuss implications for policy makers and programme implementers working at the household, community and health system levels.	Ethiopia, India Indonesia Nigeria Tanzania Uganda Nepal	Qualitative multi-country study 45 Maternal deaths 84 Maternal ill cases (PPH only) Events narrative Ethiopia: 51 India: 32 Indonesia: 16 Nepal: 32 Nigeria: 40 Tanzania: 48 Uganda: 48	Reports, facility-based records, discussions, in-dept interviews and focus group discussions Women groups, villages leaders, community health workers	Illness factors (partial recognition of danger signs) Perceived severity of illness influences the choice of care (biomedical vs traditional) Women’s status (decision-making by older women and husband) Perceived quality of care (bypassing lower level facilities due to low quality) Tradition (traditional healers before formal care)	Lack of means of transport	Poorly staffed facilities Poorly equipped facilities (medications and supplies) Inadequate management (poor quality care) Long waiting time because of referrals	First Delay separated in 3 phases: Recognition of potentially life-threatening maternal and new-born illness Understanding of illness severity Decision making process around care seekingCombined the 3-Delays Model to Pathway to survival to better explore the drivers of the first two delays.

Table A4. Studies adding a fourth delay (N = 4).

					Factors determining the delays
No	Author (ref) Year of publication Study aim	Country	Type of study Sample population	Method of data collection Key informants	1^st Delay	2^nd Delay	3^rd Delay	Definition of 4^th delay
1	Combs Thorsen et al., 2012 To identify the socio-cultural and facility-based factors that contributed to maternal death	Malawi Lilongwe, 2 urban hospitals	Qualitative study 32 Maternal deaths	Case notes reviews (VA) and in-depth interviews Healthcare staff, family members, neighbour’s, traditional birth attendants	Illness factor (no recognition of danger signs and severity of illness by woman and husband) Perceived accessibility (remoteness) Tradition (preference for home delivery/TBA instead of institutional care)	Distance Lack of motorised vehicles in rural settings Lack of means of transport (at night)	Poorly staffed facility (lack of staff, lack of competences and communication skills) Poorly equipped facility (medicine and blood) Inadequate management (incorrect diagnosis, inadequate clinical work, lack of monitoring, attentiveness)	The patients conceal the HIV status and religion to the provider thus delaying treatment
2	MacDonald et al., 2018 To explore determinants of maternal mortality from the perspectives of women of near-miss maternal experiences and community members and their solutions to reduce maternal mortality in the community	Haiti Rural communities close to the mountains	Qualitative study (Participatory Action Research) 5 near-miss women	Semi-structured interviews Focus group discussions Women, men community leaders and traditional birth attendants	Illness factor (lack of awareness, denial of danger signs, waiting too long to seek care) Socio-legal issues (hiding pregnancy to partner for fear of financial worries) Women’s status (reduced ability to make informed choices, decision-making by partner) Economic status Perceived accessibility (cost of care and transportation) Tradition (local medicine before formal care) Negligence from the woman/ignorance of young women	Distance (time) Cost for transport	Poorly staffed facility (lack of competences) Disrespectful care (feeling unwelcomed, humiliated by HCW, lack of respect toward accompanying TBA)	Delays from community accountability for maternal death due to Lack of a system to support transports Trusting traditional practice endangering women (pressure) and use precious resources
3	Wallace et al., 2018 To gain insights into what influences people’s decision to seek care antenatally and during labour and birth.	Timor-Leste 4 municipalities: rural, peri-urban and urban settings	Qualitative study 17 women with various reproductive history	Semi-structured interviews (17) and focus group discussion (9) Women and men/partners	Women’s status (husband’s decision of place of birth and reliance on mother’s in law and TBA to resolve issues at home) Economic status Perceived quality of care (facility birth perceived good but not allowing to perform cultural blessing, fear of not having privacy and family support) Short labour Limited birth preparedness ANC provide confidence of healthy pregnancy not requiring institutional birth	Distance Lack of means of transport Cost for transport Poor road conditions Lack of ambulance/late arrival	Disrespectful care: staff ignoring cultural practices and perceptions	Delay from perceptions of respectful quality care Misgivings about staff Perceptions of hospital environment and policiesIndividual plans of the womanMinimal exploration of birth preparedness
4	Påfs et al., 2016 To explore care-seeking and experiences of maternity care among women who suffered a near-miss event in the early or late stage of pregnancy, and to identify potential barriers and health system limitations to maternal survival	Rwanda Kigali, 3 hospitals, urban setting	Qualitative study 47 women	Naturalistic inquiry using open ended questions Women	Illness factor (fail to recognise the need for care) Socio-legal issues (keep pregnancy secret if outside marriage, fear of stigmatisation) Economic status (no support from partner, lack of national insurance) Perceived accessibility (unaffordability of health services) Perceived quality care (fear of being mistreated if coming too early) Belief (late disclosure of pregnancy because of fear of witchcraft) Tradition (traditional medicine because of delayed payments)	Problems in referrals (lack of transports and constrains communication between facilities)	Poorly staffed facilities (lack of staff) Poorly equipped facilities Inadequate management (incorrect diagnosis and treatment, wrong referral, cost not covered by insurance, receiving hospital unprepared for to handle an emergency) Disrespectful care: Poor patient-provider interaction and communication Fourth delay Disrespectful care (distrust in health staff) Self- treatment	Third Delays split in 2 parts: Third delay: Delay in receiving care Fourth Delay: Delay from perception of respectful quality care Splitting the identification of factors affecting the delays by distinguish between formal care seeking and informal care seeking

Table A5. Studies proposing other changes (N = 3).

					Factors determining the delays
No	Author (ref) Year of publication	Country	Type of study Sample population	Method of data collection Key informants	1^st Delay	2^nd Delay	3^rd Delay	Other changes
1	Castro et al., 2000 To seek a comprehensive knowledge of the characteristics of maternal mortality in this setting, and identify factors that can be modified through concrete interventions	Mexico 3 states: Queretaro, San Luis Potosi, Guerrero urban	Qualitative study 145 maternal deaths	Verbal autopsy with open ended questions Relatives: mother, sister, acquaintance, husband	Illness factor (not actions from women despite presence of danger signs) Women’s status (required partner approval before seeking care) Perceived accessibility (cost for transport, service and medicine) Perceived quality of care (previous negative experience) Beliefs (women’s needs to ensure complications, interpretation of sign and symptoms through a non-medical paradigm addressed via local remedies) Domestic violence	Distance (remoteness, marginality) Lack of means of transport Prolonged transportation time Seeking care at more than 2 facilities	Poorly staffed facilities (lack of staff, limited competencies and training) Inadequate management (incorrect management, early withdraw medications) Long waiting despite serious complications	Classification of factors contributing to the delays into: Subjective Interactional StructuralThese are used to formulate solutions
2	Gabrysch and Campbell, 2009 Explore the scope of determinants of skilled birth attendance, including preventive care seeking for delivery in LMICs.	Multiple countries	Literature review 2 reviews 80 original studies		Socio-cultural factors (maternal age, marital status, ethnicity, religion, traditional beliefs, family composition, mother’s education, women’s autonomy) Perceived need of care (information availability, health knowledge, pregnancy wanted, perceived quality of care, ANC use, previous facility delivery, birth order, complications) Economic accessibility (mother’s occupation, Husband’s occupation, ability to pay) Physical accessibility (region – urban/rural, distance, transport, roads) Perceived quality of care	Economic accessibility (mother’s occupation, Husband’s occupation, ability to pay) Physical accessibility (region – urban/rural, distance, transport, roads)	Quality of care (preventive or emergency)	Distinguish between quality of emergency care from quality of preventive care
3	Sorensen et al., 2011 To analyse the main dynamics and conflicts in attending and providing good quality delivery care in a rural setting in Tanzania	Tanzania Kagera region, nortwest, rural	Qualitative study 31 mothers 32 relatives 19 healthcare providers	Semi-structured interviews and Questionnaire Women, relatives, TBAs	Women’s status (decision-making by husband) Perceived quality of care (perceived incapability of local facilities to manage birth complications, home birth preferred because of closeness to family support) Tradition (trust in TBA to be safe) Fail to follow healthcare providers’ advices due to ignorance and culture	Distance Lack of means of transport Cost for transport for referral Travel at night challenging (fear of thieves and wild animals)	Poorly equipped facilities (limited availability of supplies) Inadequate management (staff not available at night or long wait before arrival)	Suggest a new model (Actantial model) with 4 components: Subject, Aim, Helpers, Obstacles To facilitate the identification of responsible agents and strategies of action to improve access to EmOC.