Perspectives on child diarrhoea management and health service use among ethnic minority caregivers in Vietnam

Coverage of government primary health care (PHC) facilities has greatly improved in the past two decades in Vietnam. By the end of 2008, 99% of all communes (the local administrative unit in Vietnam) in the country had a communal health station (CHS), including the northern mountainous communes inhabited by ethnic minorities [1]. Communal health stations provide primary health care including free immunization for children, treatment for minor diseases and common drugs (including treatment of child diarrhoea and respiratory infections), assistance during deliveries and take part in health promotion activities in the communes including mother and child health education, and hygiene and sanitation education. This study was conducted in the Northern mountainous regions of Vietnam where 32% of CHSs having doctors and 86% having midwives or assistant doctors specialized in paediatric and obstetric care [2]. CHSs are small health stations situated in each commune, which refer seriously ill patients to larger inter-communal clinics and district hospitals, which provide facilities for surgeries and in-patient treatment as well as more diagnostic services (e.g. simple lab-tests etc.).

Most (97%) villages in the Northern regions of Vietnam also have active Village Health Worker (VHWs) [2], who form a crucial part of the PHC system. They are laymen with short-course training in health and human diseases, typically three months of training plus occasional upgrade courses. They work part-time and have five official tasks; perform health education, disease prevention including hygiene promotion, mother and child health care including advocacy of family planning, first aid and disease treatment, and to participate in health programs with authorities and CHSs (e.g. inform caregivers about child immunization and give education to caregivers about child nutrition, care and hygiene) [3]. Private health providers are also a crucial part of the wider health system in Vietnam including a large numbers of traditional medical and herbal experts, spiritualists, and private bio-medical practitioners [4]. It has also become normal practice among Vietnamese to purchase and self-medicate with over-the-counter drugs, many sold without prescriptions, from the large number of private drugstores and pharmacies [5‐7].

Improving child health indicators is a major priority for the Vietnamese government, who have initiated several child health programs and policies under the National Target Programs on Health since the early 1990'ties. Prevention of communicable diseases, improving coverage of the expanded child immunization program and improving child nutrition status are some of the main child health objectives. These programs all aim at providing basic child health services and promotion to all parts of the country. To ensure affordability and increase coverage of services, the National Health Care Fund for the Poor was established in 2002 to provide insurance for ethnic minorities and poor communities, making all treatment and drugs at CHSs and district hospitals free of charge for those insured [8]. However, child health inequalities still exist between rural and urban, and between majority and minority population groups.

Diarrhoeal diseases are the number two leading cause of morbidity and is endemic in the northern mountainous regions of the country inhabited mainly by EMGs [9], as are parasitic infections of children [10]. It was recently estimated that 12% of children in a mountainous rural district of Vietnam had experienced diarrhoea within two weeks [11]. Children living in the northern and central highlands also experience higher risks of malnutrition compared with children living in urban and rural lowland [11, 12]. A recent analysis of official statistics also highlights that far more ethnic minorities and children from the northern regions live in poverty, have poor shelter and limited coverage of water, sanitation and immunization compared with urban regions and Kinh children [13].

In principle, all caregivers in Vietnam have access to a well-established and free PHC system for children. And several studies have shown that self-medication of child illnesses is also common in urban as well as rural Vietnam [14] and that antibiotics are commonly prescribed, and overprescribed, to children [14‐16]. One study in a Kinh populated rural district found that only 54% of caregivers used antibiotics and 36% used analgesic and antipyretics to treat the child's diarrhoea, while only 9.7% gave the child ORS [15]. Compared with the majority of Vietnamese (Kinh) parents, quantitative studies have found that ethnic minority caregivers are even less likely to give ORS and seek treatment at CHSs when children have diarrhoea [17, 18]. One of these studies also found that ethnic minority caregivers perceived diarrhoea episodes as less severe compared with Kinh caregivers [18]. Indirect costs of treatment, user fees and long distances to health facilities have also been mentioned as important factors that hinder the use of health facilities among poor and the EMGs living in highland areas [19, 20].

However, the perspectives of Vietnamese ethnic minority caregivers on child illness management and treatment preferences and perceptions are not well described in the literature. This study explores treatment seeking strategies among ethnic minority caregivers of children with diarrhoea in Northern Vietnam, in order to suggest improved health services for EMGs and other vulnerable groups in Vietnam.

Six months of fieldwork was conducted in two adjacent rural communes in the Northern Province of Lao Cai in Vietnam in 2008, investigating the general living and working conditions, perceptions and practices of health, hygiene and sanitation among ethnic minorities [21]. For the current study, and building on previous findings, two additional months of fieldwork was conducted in 2009 focusing on caregivers' management and perceptions of child diarrhoea.

The total population of the two communes was 10,000 with 85% being ethnic minorities. Fieldwork was conducted among the four most populated minority groups in the area; Giáy, Tày, Xa Phó (living in lowland villages) and Red Dao (highland people). About 40% of households were below the government poverty line of 2006 - 2010 for rural areas (defined by a monthly income below 200,000 Vietnamese Dong per person, approximating 10.5 USD), with a higher concentration of poor households among the Xa Phó (living on the edge of the forested hills) and the Red Dao in the highland [22]. All ethnic minorities depend on subsistence rice farming, animal husbandry and forestry with all adults doing hard physical fieldwork daily. All four ethnic groups have traditionally lived in multigenerational households and are all patrilineal, with men being the heads of households and owners of all property, and patrilocal with daughters moving to their groom's families when married.

Further details on the study area and the minority groups have been reported earlier [21].

Two CHSs, one also hosting the Preventive Medicine Office in charge of conducting preventive activities in both communes, were located in the lowland part of each commune. Both CHSs were run by assistant doctors and had nurses, pharmacists and nurse midwives employed. Both CHSs had a selection of common drugs and remedies available including ORS, antibiotics, other anti-diarrhoeals and allopathic medicines. Village Health Workers had been appointed by the CHSs for almost all of the 39 villages in the communes. Government district and private hospitals and drug stores were located in the district town 3-5 kilometres away from the clinics. No private bio-medical practitioners, clinics or drug stores were found in the two communes. But several spiritualists (Thày Mo's) were identified as important health providers in highland villages, performing fortune telling and warding rituals to strengthen health. The lowland villages (Tày and Giáy) were located closest and within five kilometres to the CHSs. The Xa Phó villages were located within eight kilometres, while highland villages were scattered and located on steep dirt roads up in the mountains 10-15 kilometres away from the CHS. No public transportation exists and people walk, bike or drive motorbikes to the CHSs and district towns.

The main source of data was 43 semi-structured in-depth interviews with caregivers of pre-school children (six years and below) who had experienced a case of diarrhoea in the past month. In addition, two weeks of observations of patient-staff interactions were performed at the two CHSs and three focus group discussions (FGDs) were conducted with men. Data from participant observations in four different study communities conducted in 2008 was included as general background information of the living conditions in highland and lowland settings (methodology has been described in details elsewhere [23]).

Information about diarrhoea cases in 10 lowland and two highland villages (Table 1) of the four ethnic groups were collected by asking caregivers if their child had experienced an incidence of loose or watery stools at least three times per 24 hours during the past month, using the diarrhoea definition of WHO [24]. Several sampling methods were used to ensure that cases included mild and severe cases, and cases treated at home and by various health providers. These included two weeks of observations at the two CHSs (eight cases identified), a project community diarrhoea survey covering both communes (seven cases), by references from VHW's (nine cases), and house-to-house visits by the first author and assistants in villages (19 cases).

Interviews included questions on perceived causes of diarrhoea, decision making about treatment, treatment choices, and use of health services (See additional file 1: Caregiver Interview Guide). The main informant was the adult caregiver who had attended to the child during sickness, but during seven of the interviews more than one caregiver was present and answering questions for their children. Attempts were made to avoid having grandparents in the same interviews as mothers to allow mothers to talk openly about family decision making. To gain in-sight into influences on treatment seeking and choices of fathers and grandfathers, three FGD's were conducted among men from two lowland villages (Tày and Xa Phó groups) and a highland village (Red Dao). Characteristics of participants are shown in Table 1.

Interviews and FGDs were conducted in Vietnamese by two Vietnamese assistants with the first author being present during half of the interviews and all of the FGDs to ask elaborating and clarifying questions. Young local women assisted with translation during interviews in the highland and Xa Phó villages if the interviewees did not speak Vietnamese.

Caregivers with sick children encountered at the CHSs and in households were not interviewed, but asked if they were willing to be interviewed at a later time once the child had recovered. Two caregivers with sick children met at the CHS declined to be contacted again. One caregiver with a sick child undergoing treatment was met in a Xa Phó village and interviewed at a later stage. All interviews were audio recorded with informed verbal consent and promised full anonymity and confidentiality. Recordings were transcribed, cross-checked and translated ad verbatim into English by the two assistants. Data was coded for thematic content analysis [25] by the first author using pre-set and emerging empiric and theoretical categories by the NVivo 8 software program [26].

The study was approved by the National Institute of Hygiene and Epidemiology in Hanoi, The Department of Health in Lao Cai province, and Peoples' Committees (government authorities) in the two study communes.

Some clear differences in treatment seeking and health care choices were found between the poorer ethnic groups living and working in the highland (Dao and Xa Phó) compared with richer groups living in the lowland setting (Tày and Giáy). As described in case 1 (Table 2), greater difficulties in assessing treatment at health facilities were observed for highland caregivers living furthest away from CHSs and hospitals, compared with the richer and more centrally located communities in the lowland.

Even though health insurance would cover the direct costs of treatment and admissions, highland caregivers (e.g. case 1, Table 2) in particular worried a lot about the substantial indirect costs such as transport, buying food, arranging home-sitting, and paying for extra drugs and services at the CHS or hospitals. Distance and time was also a constraint for highland caregivers from poor families without means of transportation, who would have to spend half or a full day if walking to the CHS. Caregivers acknowledged that this was too time consuming if a child was seriously sick. In focus group discussion and interviews, caregivers all said that they would borrow or spend their money to pay for transport to get to the CHS in time, despite having very low incomes. The cost of petrol or renting a motorbike driver for a round-trip from a highland village and the CHS was 20-25,000 Vietnamese Dong, equalling several days of work for a highland farmer. In comparison, lowland caregivers (Giáy and Tày) lived a maximum of five kilometres away from the CHS and all had bicycles or motorbikes accessible within their households.

Observations in households (conducted during fieldwork in 2008) and interviews with caregivers confirmed that all four groups of ethnic minorities live according to patriarchal gender roles, with the oldest men taking all major decisions and the youngest women being the main domestic labour force. This was most strongly expressed in highland families, where young mothers were also the main agricultural labour force of the families, and who expressed low decision taking powers about their children's health, which complicated seeking the government health facilities (Table 2). In the highland, six out of 19 caregivers of children during sickness were grandmothers, and mothers and grandmothers indicated that elders were the primary persons in the household in charge of deciding, preparing and administering the treatment for a sick child. Mothers were not allowed to travel unaccompanied or with male non-family members to health facilities and needed permission from elders to stay at home from field work to care for the sick child (Table 2). A group of highland men explained: "when the child gets fever, falls ill, belly pain or something similar, the grandparents have to take care of the child and prepare the medicines". The men continued: "Everything must be discussed with the grandparents [...]. After discussing, if all agree, we let the child go (to the CHS)" (FGD, Dao men).

Despite the fact that all caregivers, from highland as well as lowland communities, highly appreciated access to Government health services and drugs free of charge, many caregivers sought these services with some apprehension. None of the interviewed caregivers sought the advice or were visited regularly by the VHWs working in their villages. Caregivers said they normally only had contact with the VHWs when attending child vaccinations or child nutrition surveillance on a monthly or bi-monthly basis, which were also perceived to be the main tasks of a VHW; "He just weighs the small children once per month. I only saw him doing that"(Xa Phó mother). As expressed in both cases (Tables 2 and 3), caregivers also said that VHWs did not have important health services to offer sick children such as drugs, first aid, ORS, or good medical knowledge. All caregivers had therefore experienced receiving a standard advice of going to the CHS from the VHW. This made them go directly to the CHS without consulting the VHW first.

Health Station staff, on the other hand, were perceived to be qualified to diagnose diseases and prescribe drugs. But as exemplified in case 2 (Table 3), the majority of caregivers had bad experiences when consulting the CHS. Their illiteracy and 'backward' lifestyles were often commented on negatively by different types of health staff. Thirteen caregivers told stories of being scolded, ignored, or blamed for seeking treatment too late or too often. A caregiver from the highland explained; "Most of the people here are scolded by the doctors. Doctors often blame people for not taking good care of their children [...] They should try to understand the situation here. It is really hard (to live) in a place like this" (Dao woman). A young lowland mother added: "I just want the doctors to be more enthusiastic with patients [...] Only when people fall sick, they come to the clinic, so the doctors have to behave softly with them so people can be satisfied" (Tày mother). Language barriers were also mentioned by caregivers during observations and in all interviews with highland women. They felt they could not communicate accurately in Kinh language with staff at the CHS. Health staff also expressed frustration about not being able to communicate with patients. They therefore used very simple language or, in rare cases, used other waiting patients as translators or asked patients to have someone in their own community read and explain the prescriptions to them (from a prescription book).

To avoid being misunderstood or perceived as backwards, caregivers uniformly said that they never shared ideas about causes of diseases, asked clarifying questions about the prescribed drugs or told any health staff including VHWs, about the home-made treatments they had used before coming to the CHS. This was verified during observations at the CHS. Here, no caregivers were observed to share any information on the child disease or ask about prescriptions, diagnose, or the guidelines given by health staff. Staffs were observed to only ask few direct questions on disease symptoms and personal details.

As illustrated in the two cases (Table 2 and 3), caregivers sought and chose treatment according to a pragmatic logic of medical pluralism, applying and combining a range of treatments from different medical systems to increase chances of effect and recovery. This strategy was described with the Vietnamese proverb 'praying in four directions' (Vái tứ Phương): "We call it 'praying in four directions'; when we get sick; we do whatever we have to" (FGD, men, Highland). A highland mother added: "It is safer when doing all. Because sometimes doing 'mo' will (make him) recover, sometimes going to ask for drugs at the health clinic will" (highland mother).

As described in the two cases (Table 2 and 3), caregivers commonly shifted between folk, popular, and professional bio-medical treatment systems as a disease developed and choices were often based on the perceived severity of the disease. Home-administered remedies, including herbal mixtures, concoctions and massage were perceived sufficient for 'light' or 'minor' (nhẹ) episodes lasting two - four days with irregular defecation of liquid stool, belly pains and general uneasiness. This type of diarrhoea was also described as a 'normal disease' (bệnh bình thường) common during childhood, easy to treat and not severely affecting the child: "The old people said that the children have diarrhoea when they grow up [...] It's not a special thing if a child gets diarrhoea" (Xa Phó mother, lowland). A lowland mother added: "I saw the diarrhoea as normal; he had diarrhoea for three-four days and then recovered. [...] He didn't lie down and he went out to play all day as usual" (Tày mother). Among all caregivers, diarrhoea was perceived 'heavy' or 'serious' (nặng) when lasting more than 2-4 days, re-emerged shortly after recovery or included high fevers, sudden onset of very frequent defecation, and the child becoming very lethargic, or losing weight. In these cases all caregivers mentioned consulting the CHS and taking 'western medicines' (thuốc tây) as a necessary treatment. Referral to a hospital was also generally agreed as a necessary action if treatment with western drugs did not stop the diarrhoea. This was also acknowledged among older caregivers from highland communities with longer distances to the CHS; "When he kept being very sick, we couldn't let him stay at home. We went down to the health clinic to ask for drugs... He was sick at home for two-three days and it was getting too serious... The longer we stay at home, the more serious it gets" (Dao Grandparents). Nevertheless, a total of six children (four from the highland) had been sick with diarrhoea for several weeks and were seriously weakened before being taken to the CHS and admitted up to 12 days at the inter-communal clinic or the district hospitals receiving frequent intravenous sodium glucose solution and drugs.

Respondents frequently used the expression of a treatment or method being 'compatible' (hợp) or 'suitable' (phù hợp) for a child with diarrhoea. Compatibility turned out to be a central and complex concept, functioning at various levels and having implications on all aspects of diarrhoea management, including the perceived cause, the chosen treatment and evaluations of its outcome.

Low usage of government health providers among EMGs compared to Kinh have previously been ascribed to failures in insurance coverage [19] and logistical constraints to access health services from the highlands [19, 28]. But as pointed out by London [8], health service fee exemptions only cover one component of health service costs, while indirect costs may present bigger challenges for the poor. This was also expressed by highland caregivers in this study, who worried a lot about the substantial related costs of admittance, which potentially delayed seeking treatment for seriously sick children.

Interestingly, despite having insurance with access to free treatment for their children at government CHSs, higher costs and longer distances, this and one other study [19] have identified a willingness among ethnic minority caregivers to buy drugs from private drug stores. The common practice among the rural population of Vietnam of seeking private health services and buying over-the counter drugs for self-medication [5, 7] apparently extend to poor EMGs.

The study also showed that VHWs, who are the closest health providers, were not considered qualified by caregivers, who instead bypassed VHWs and travelled long distances to access private or public health facilities. Problems of distance to health services might therefore be lessened by upgrading services at community level, either by increasing skills and competences of VHWs in remote communities to perform effective health promotion and basic treatment, or by increasing the frequency of outreach visits by CHS staff to remote communities.

The study also identified a local explanatory model of disease management among ethnic minority caregivers, clearly differing from a bio-medical treatment system. Simultaneous resort, local medical cosmologies of obtaining bodily balances, concepts of compatibility of treatment and humoral qualities of medicines (e.g. hot/cold and heavy/light) as identified in this study are all commonly found in the Asian context of medicine and well described in minority as well as majority populations in Asia and Vietnam [7, 29‐31]. However, our study also found that health staff ascribed such health seeking patterns as 'ethnic' and 'backwards', and that government health services are not always the first and most well-liked choice of health service for ethnic minority caregivers.

Health is a key objective of state policies and governance in Vietnam and health sciences in modern Vietnam have become symbolically associated with socialist modernity, rationality, and progress [8, 32]. Further, ethnic minorities have commonly been described as 'marginal' and 'developmental backwards' and therefore targeted by the socialist state to 'assimilate' to mainstream developmental standards [33, 34] - while also being encouraged by the state to preserve cultural traditions not posing any threats to progress of the socialist state [33, 35]. 'Superstitious beliefs and practices' such as shamanistic rituals and animal sacrifice have been deemed as 'backward' health practices and a constraint for progress and modernity [32, 35]. Hence our study provide further evidence that ethnic minorities seem to balance between practicing, to them, meaningful health rituals, while also interacting with a modern government health system. This might explain why ethnic minority caregivers do not reveal local explanatory models of disease to health staff and opt for drugs from the private sector, where they are not met with demands to change health behaviours.

This study agrees with previous findings on health service use in Vietnam, that low trust and dissatisfaction with medical staff [17], and discouraging communication with health personnel affects disadvantaged women's and ethnic minorities' treatment seeking behaviours and use of the public health sector [28, 36]. This is underlined in a recent World Bank report stressing that female ethnic minority caregivers in Vietnam are particular vulnerable and in special need of appropriate health services [34]. The public health system of Vietnam has been described as particularly 'provider-centred', not anticipating patients to speak up their needs or questioning health providers [37, 38]. This is possibly worsened for illiterate EMGs who also live in areas with poor health service coverage and where it is difficult to attract and retain qualified health staff [2]. Hence, this study argues for a need for the Vietnamese health system to focus on service levels in remote areas of the country and to change the approach to marginal population groups in Vietnam to increase their trust in and use of government health services.

Cross-cultural competence models [39, 40] have been developed to address communication challenges in the health sector. They stress that health staff working in ethnically diverse populations must understand patients' actions and perceptions of disease as expressions of their social realities, including constraining socio-economic factors and local explanatory models of disease. For health staff in daily contact with EMGs in Vietnam this clearly implies acknowledging the flexible 'praying in four directions' approach to choosing treatment and local cosmologies of disease management, as well as avoiding the persisting stereotyping of EMGs as 'culturally backwards'. To address these needs, Vietnamese Medical Universities have recently strengthened community health training of medical students [41], while other staff groups remain to be considered. This study primarily focused on understanding the perspective of ethnic minority caregivers' on use of child health services. Further studies into the constraints of delivering effective government health care in mountainous areas of the country, would add important information to understanding the full complexity of providing effective health care for all children in Vietnam.

Our findings also corroborate other studies highlighting the potential problem of self-medication and over-prescription of drugs to children in Vietnam. Training of health staff should therefore focus on improved prescription practices and guidance to patients on how to administer drugs and ORS. They should encourage open discussions with patients on when to expect effects from drugs, and when to shift treatment and health care system. Small-scale interventions have managed to improve drug prescription and purchasing practices in urban sites in Vietnam [42‐44]. It is strongly recommended that such interventions be scaled up to include rural and mountainous areas, where ethnic minorities are in need of good guidance.

Empowering and encouraging Vietnamese caregivers with good care practices can enhance and sustain better child-care and health care practices in rural communities [45‐47]. The current study highlights that most caregivers from all four EMGs recognize symptoms and evaluate the severity of diarrhoea within few days of onset and respond caringly and rapidly by applying various treatments. However, caregivers felt that hardly any encouragements or preventive advice were given to support these care giving efforts during contact with VHWs or CHS staffs. We highly recommend that all staff at government health facilities use every opportunity to praise and encourage such good care giving behaviours. This can form a fruitful starting point for further integrating preventive messages and child health promotion into the contact with caregivers. This may strengthen caregivers' experience of being in contact with a caring and trustworthy health system.

In the present study we also saw strong indications of patriarchal gender roles creating difficulties for young highland mothers in particular to follow health messages and seek CHSs, depending on male and older family members' permissions. These findings are in agreement with other studies in the rural population of Vietnam [30, 48], and other ethnic minorities in Asia [29], which show that older community members and grandmothers, in particular, play core roles in child disease management and care. However, current child health promotion programs in Vietnam and health promoters mainly target and educate mothers as sole caregivers. It is therefore strongly recommended that health promotion in Vietnam include other family members as key people in children's 'care groups' and their resources in child care. These people can be reached through different grassroots organisations (e.g. Farmers' Unions and Women's Union) or community stakeholders in daily contact with men and elders, e.g. village heads, diviners and CHS staff.

Finally, we acknowledge the beneficial practices of rehydrating children with diarrhoea using continued breastfeeding and home-made oral rehydration solutions of green teas, water or rice gruels with added salt and meats. These practices should be known to all health providers including spiritualists, pharmacists and VHWs as beneficial and included as advice during contact with caregivers. This might be more effective than prescribing ORS to less severe cases, since it is perceived 'compatible' with children, while ORS is perceived a non-compatible western drug and also administered incorrectly by many caregivers.

This study has provided new information on the qualitative aspects of why and how ethnic minority caregivers in Vietnam use and choose health services for child diarrhoea. However, despite triangulation of methods and a thorough sampling strategy the study faces some limitations:

Firstly, as also pointed out by Toan et al [19], the geographical inequity in access and use of government and private health services, as well as language barriers and poor economical conditions, are anticipated to be stronger for EMGs living in less developed areas of the country e.g. in high mountains with longer distances to health facilities and urban environments. Secondly, difficulties of gaining access to 'the insider perspective' as outsiders to a research community can present serious limitations to qualitative research [49]. Since some caregivers expressed being marginalised and disrespected by Kinh people due to their ethnicity, the researchers' and assistants' backgrounds (a white westerner and of Kinh origin from urban developed environments) might have caused reluctance from respondents to disclose actions which could be perceived as 'wrong', 'traditional' or 'backward'. This problem was anticipated by the first author having good prior knowledge of the living conditions and having socialized with many community members and their children during a previous six month long fieldwork in the same study area. Being outsiders to the area was also mentioned as a benefit by some respondents, who felt safe to disclose bad experiences with health staffs and family members since the researcher and assistants were not affiliated with any local authorities or organisations. However, it did constitute a special challenge to gain the trust of young highland mothers living in patriarchal families and making them speak up about sensitive issues of gender and decision making. Research assistants therefore aimed at interviewing mothers without the presence of elders and men and encouraged trusted friends and sisters to take part. Thirdly, some information is likely to be lost or transformed during translations with the first author not speaking Kinh, the research assistants not speaking ethnic minority languages and the need for double translation in Xa Phó and Dao communities where some women did not speak Kinh. The quality of translation was validated as research assistant's cross-checked and discussed recordings and each other's transcriptions of all interviews and FGDs with the first author.