The search engines PubMed, Medline, EMBASE, the Cochrane Library, and Google Scholar were initially searched up to Sep 16, 2016 and then searches were revised on September 27, 2017, but we found that data was saturated, and no new themes were emerged. Search terms were a combination of [(‘care seeking’ OR ‘care-seeking’ OR ‘health care’ OR ‘health care seeking’ OR ‘community based intervention*’ OR ‘community-based intervention*’) AND (mother* OR maternal OR women OR newborn* OR neonat*)] used as medical subject headings and keyword terms in the title/abstract. No language restrictions were applied. Grey literature and reference lists of included studies were also searched to identify studies. We considered studies from LMICs that assessed the factors associated with health care seeking for maternal and newborn illnesses in observational or qualitative studies. We did not consider studies on health care seeking for specific maternal and newborn illnesses such as jaundice etc. or for preterm babies. We considered recommendations for systematic reviewing of qualitative studies ¹² . We used the PRISMA checklist PRISMA statement in reporting systematic reviews from the observational studies ¹³ . The 22-items STROBE checklist was used to assess the methodological quality of the cross-sectional studies ¹⁴ . Studies that fulfilled the methodological criteria of more than 18 points were classified as high quality, between 12–18 as moderate quality and below 12 were classified as low quality.

ZSL and PM independently reviewed the retrieved articles in two stages. First, relevance was assessed from the title and abstract and if relevance was still unclear, the full text was read. Any disagreement was referred to a third reviewer (CC or ZAB). Factors responsible for health care seeking patterns for maternal and newborn health from observational studies and qualitative studies were separately analysed. Study design, country of study, setting, participants, and results were recorded for each study. We performed a narrative synthesis of the findings from the included studies, as included studies were observational and qualitative in nature.

Primary caregivers in all included studies were usually mothers; however, mothers-in-law, grandmothers, fathers, neighbours, traditional healers and opinion leaders in the community were among the many people involved as caregivers for mothers and babies. Across the studies, it was observed that mothers/families do seek care for neonatal illness ⁸⁰ ; however, complications during pregnancy are not considered as an illness and many signs are considered as normal, even when painful and constant ^{92– 102} . In certain studies, bleeding was not considered to be a complication ¹⁰³ , and in such situations decisions to seek health care were often delayed. Women who expressed pain verbally were considered as disobedient and therefore maintaining silence was considered appropriate ^{104, 105} . Missing antenatal care visits were reported to be due to heavy and unavoidable workloads at home ⁹¹ , and a few studies reported that mothers-in-law privileged household chores over women’s health ^{91, 99} . Some families perceived that some common neonatal symptoms should or cannot be treated at health facilities and therefore traditional care should be sought ¹⁰⁶ .

In India, women during pregnancy are usually advised to be cautious while eating “hot” or “cold” food, and to eat less otherwise the baby can grow too large and therefore lead to a difficult birth ⁹² . A qualitative study from Pakistan (Baluchistan) ¹⁰⁷ described that the dai (traditional birth attendant (TBA)) usually places mustard oil on her fingers and massages the vaginal walls to ease the birth, and inserts vaginal and anal pessaries after birth to help shrink the uterus and to provide support for the uterus and backbone. They also prefer women to eat Goandh (edible gum) combined with turmeric powder, and dried dates in milk to induce heavy vaginal bleeding so that all unclean blood is drained from the body, thus predisposing to postpartum haemorrhage. In situations when the placenta does not expel normally, the dai enters her bare hands in the uterus or puts hair into the mother’s mouth to induce vomiting ¹⁰³ . Eating vegetables rather than meat during pregnancy is preferred as it is considered to increase the production of breastmilk and freshens its taste ¹⁰⁸ . During infant illnesses, mothers prefer to give ‘ rabadi’ (prepared by cooking millet flour and yogurt), ‘ khichchadi’ (a semi-liquid rice and pulses mixture) and ‘mateera’ (watermelon curry) to their febrile children in conjunction with breastmilk ¹⁰⁹ .

While illnesses, particularly in women who are not pregnant, are considered unimportant, evil spirits and fate (Allah’s will) are reported to be the cause of these illnesses ^{103, 110} . Faith healing is important in many cultures. A study from Ghana ¹¹¹ named three major religions that practised faith healing and each religion has a specific healer. On the other hand, most of the communities in Asia and Africa believe that certain precautions during pregnancy or immediately after birth will ward off the evil eye (a gaze or stare superstitiously believed to cause harm) and will prevent the infant from getting sick ^{92, 112} . This includes isolating women and their baby in a room for a certain period of time after childbirth and lighting a fire at the entrance where they are confined ⁹² . In order to prevent them from evil eyes, people reported keeping the pregnancy secret from people outside of close relations ⁹⁶ .

Mothers may consult family and friends when the danger signs are not clear or unusually severe ^{107, 113} . However, in severe illness, decision-making power can be switched to more experienced members of the extended family, which can cause significant delays in decision-making. Many of the studies reported that in scenarios when women had money, they hurried to pursue treatment options from a health care facility despite several familial pressures. A study from Tanzania reported ⁹⁵ that having an option of home birth was found to be a hurdle in emphasizing the importance of skilled birth care ^{100, 104, 114– 116} . Trust for someone from the same community, sharing the same values and speaking the same language, was another factor that encouraged women to give birth at home and with a TBA ^{38, 117} . However, it was apparent from the studies that if women continued to suffer, then they do seek care from western-trained care providers ¹⁰⁷ .

Decision-making emerged as a complex issue. Decision-making power is less likely to be with the woman and mostly rests with their partners and mothers-in-law. Women who had no income source were usually those who had no rights for decision making ¹¹⁸ . Several studies reported that the major barrier for institutional care was gaining permission from husbands ^{92, 93, 104, 119– 124} . Women are considered inferior to men and their disobedience often results in physical and emotional violence ¹²³ . If husbands are absent, women face difficulties in receiving permission from her husband’s parents or other elders for seeking care and this results in even greater delays. Husbands and elders often have control over finances and women are mostly dependant on them ^{38, 93, 99, 108– 110, 117, 121, 122, 125– 131} . Deciding to seek care can incur transportation costs, user fees, cost of medicines, and possibly ensuing costs of misdiagnosis and treatment failures ⁹³ . Considering all these barriers, women often postpone seeking help, with the hope that the problem will subside on its own.

When a family is willing to seek care and arranges the money required, other challenges such as physical transference of mothers and newborns to health facilities becomes a problem. The situation is even worse if complications arise at night, when risk of being attacked by criminals’ increases or when transport providers raise their taxi or car-hire charges ^{96, 118, 121, 122, 125, 127, 130– 143} . Studies on people living in very remote areas reported factors such as distance to health facility and related transportation issues, lack of financial resources, encountering swollen rivers on the way, fear of encountering wild animals, shame about too many pregnancies or being of advanced age and pregnant as some of the critical reasons for not seeking care. Studies also reported other factors responsible for not seeking health care such as non-availability of staff at facility, rude behaviour of health care staff, and poor quality of care ⁹⁶ . Fear of operative procedures was reported as a factor hindering care-seeking ¹⁴⁴ . These were usually based on previous experience and contact with health care staff and the health care service received ^{93, 122, 125, 145, 146} .

Cost is another important barrier to seeking care from trained health professional. However a study from rural Mexico reported that cost of care from TBAs is sometimes higher than facility birth but women prefer them because they can give birth at home ¹⁴⁷ . Many women also preferred giving birth at home because they preferred a squatting position for giving birth that was also endorsed by TBAs ¹⁴⁷ . Relatives being not allowed at facilities during the childbirth was another factor expressed for giving birth at home ¹⁴⁷ .

Women’s previous encounters with health care staff and facilities were reported as a key factor for decision making ¹⁴⁸ . Further, many of the danger signs are not considered as pregnancy-related complications ^{38, 80, 109, 113, 117, 118, 149} , and thus families seek help from traditional healers, community health workers or drug sellers. Households often regard accessible and less expensive care such as herbal and home remedies or locally available drugs more highly ¹⁵⁰ . Workers from these types of care were often praised as they give time to patients and consider their social and cultural aspects as well.

Women and families usually opt for medically qualified birth attendants where women are perceived to have possible birth complications. Where TBAs detected a complication at home, women were provided with referrals. Women also preferred SBA when they wanted to have a tubal ligation performed ¹⁴⁷ . Studies reported that perceived fear of being torn in hospital, where Caesarean section was required ¹⁵¹ discourages women to seek institutional care for childbirth ¹⁰⁴ . Lack of privacy at care facilities and being examined in the open are other factors for not seeking care at clinics ¹⁵² .

Pregnant women or mothers with ill newborns usually experienced long waiting times when seeking hospital care ^{93, 104, 153, 154} . Most of the included studies cited that health professionals have poor attitudes towards poor or pregnant women, which are stigmatizing ^{118, 131, 144, 145, 155} . Studies pointed out that health care staff examine women in hurry and at many occasions did not clarify their concerns. Staff may stigmatize women, criticising them for their age and number of pregnancies and judge them on their practices for family planning ¹⁵² . Staff behaviour is therefore a major barrier for accessing care ¹⁵¹ .

Sometimes, women are referred to another facility due to lack of trained staff and functional equipment and supplies that lead to further delays. Women may be asked to pay for fuel for the ambulance to take them to the other facility. They then may be required to pay for medicines and other supplies, and when stocks of these run out, there are further delays in receiving care ¹¹⁸ .

All data underlying the results are available as part of the article and no additional source data are required.

Open Science Framework: Health care seeking. https://doi.org/10.17605/OSF.IO/5UT6X ¹⁵ . Supplementary Files contain information concerning characteristics of the studies included in this review.

Open Science Framework: PRISMA 2009 Checklist for this study. https://doi.org/10.17605/OSF.IO/5UT6X ¹⁵ .

Data are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).