https://orcid.org/0000-0002-0282-544X
Abstract
Individuals in rural communities in Madagascar must frequently travel long distances, over difficult terrain, to obtain basic care. The quality of care is often inconsistent and inadequate.
An exploratory mixed-methods study was conducted in select coastal communities in the Maroantsetra region of Madagascar to generate a more robust understanding of community and health care provider perceptions and how patients decide to seek health care, including the decision to use traditional medicine vs allopathic medicine. A total of 69 free-listing exercises and 21 facility assessments were conducted in eight communities.
Symptoms most commonly reported as reasons to seek health care included headaches, fever and cough. Decisions to access allopathic health care facilities depended on the intersection of geographic and financial access to health care facilities, perceived severity of the illness and the availability and confidence in traditional plant-based medications. Traveling salespeople, staff at local stores and pharmacy workers very often lacked formal training.
The decision to use allopathic medicine was determined by the perception of disease severity and when vulnerable populations, such as children and the elderly, were involved. Results provide insight into patterns, motivations and obstacles to health care utilization and decision making in the Maroantsetra region of Madagascar.
Introduction
Madagascar is the 10th poorest country in the world, with 76% of its population of 24.2 million living below the international poverty line of US $1.90/day.1,2 Roughly 60% of Madagascar’s population live in rural communities thinly scattered across the island nation.3 Patients must frequently travel long distances, over difficult terrain, to obtain basic care. As with many low- and middle-income countries, much of the population lacks access to consistent and high-quality health care.
As of 2013, available health care facilities in Madagascar included 2536 Centres de Santé de Base (CSBs) I and II, 86 district hospitals, 18 referral hospitals and 12 university hospitals with 17 established specialties.4 The health care structure is complex (Figure 1). CSB Is provide basic primary care and vaccinations. CSB IIs provide basic obstetric care in addition to primary care. District hospitals are the first level of referral centers for CSBs. The number of specialists at district hospitals depends on the population it serves as well as local resource availability. Regional referral hospitals are usually comprised of additional specialties and emergency care, while university hospitals encompass multiple subspecialties, including trauma care and comprehensive laboratory equipment. University hospitals receive referrals from all levels.5
Flow of patient referrals in the Madagascar health care structure.5
The majority of physicians in Madagascar are concentrated in urban areas. CSBs serve the dispersed populations in rural areas, referring complex cases to urban and university hospitals. However, traditional medicine is a core part of health care delivery in this island nation, due to both the integral role traditional healers play within the Malagasy culture and the service it provides in granting access to health care in remote regions of the country.6 Defined by the World Health Organization (WHO) as ‘The sum total of the knowledge, skill, and practices based on the theories, beliefs, and experiences indigenous to different cultures, whether explicable or not, used in the maintenance of health as well as in the prevention, diagnosis, improvement or treatment of physical and mental illness’, traditional medicine is used by 80% of the population in Africa to cover the gap in health care needs.7,8 Globally, the physician:patient ratio is 1.4 physicians per 1000 persons, but on the African continent that ratio is 0.27 physicians per 1000 persons. In Madagascar, there are only 0.14 physicians per 1000 persons.9,10
The unique biodiversity of Madagascar provides a rich substrate for the use of local flora for traditional medicine.11 A quarter of Madagascar’s botanical species are reported to have medical properties. Since 1996, the government of Madagascar has shown a commitment to explore the use of traditional medicines to improve access to health care in rural populations.12 Furthermore, since 2004, Madagascar Health and Environment Research (MAHERY), a consortium of Malagasy and American researchers, has focused on examining the intersection between human health and the environment. Its research activities have included evaluation of the use of traditional medicine by local people, the barriers in accessing health care and adequate nutrition, and the impact of overhunting and terrestrial wildlife declines on food security and nutrition.13 In 2012, Golden et al.11 conducted a study calculating the value of botanical ethnomedicines in a rainforest region of Madagascar, the Makira Protected Area. They found that the benefit value of botanical ethnomedicines was equivalent to 43–63% of the median annual household income in Madagascar, showing the perceived efficacy of traditional medicine over allopathic (conventional) medicine for certain diseases.
Currently, limited literature exists describing how patients in Madagascar decide to access care and whether to use traditional or allopathic medicines. Previous research has focused on febrile illness, demands of the urban population for access to health care and the socioeconomic factors that affect maternal mortality at the commune level in large urban sites.14–16 Thus the applicability of prior studies in the regions with the poorest access to health care is limited, a major challenge in Madagascar.17 A more robust understanding of community and health care provider perceptions, especially the decision to use allopathic vs traditional medicines in rural communities of Madagascar, may inform policy decisions, advance health systems development and generate innovative programming to improve access to health care across the country. We conducted an exploratory mixed-methods study to better understand the common sources of health care and patterns of health resource utilization in rural communities of the Maroantsetra region of Madagascar.
Materials and methods
This exploratory mixed-methods study employed free-listing exercises and facility assessments to identify common access points for health care resources in rural communities, conduct limited facility-based assessments and characterize the factors considered when making decisions about accessing care and choices between allopathic and traditional medicines.
Study site
This study was conducted in eight rural communities of the Maroantsetra region of northeastern Madagascar. Study sites were chosen due to their distance from urban populations and proximity to MAHERY field sites given the established relationship and rapport with the local population.13
Data collection
The study team collected data between August and December 2016. All research members were trained in using the interview instruments created (Figures 2 and 3). A Malagasy research assistant fluent in English, French and Malagasy conducted the majority of the interviews, given his fluency in the local language and familiarity with the culture. Data transcription occurred in real time. Data transcribed in Malagasy was subsequently translated into English. A limited number of assessments conducted by researchers in English were simultaneously interpreted by a Malagasy research assistant.
Key informant interview questions.
Facility assessment interview questions.
Subject selection and interview methods
In each community, the chief was approached for permission to perform the study. The first component of the study involved a free-listing exercise querying community members through key informant interviews on the main illnesses affecting the community, disaggregating responses specific to children, adolescents, adult women, adult men and the elderly.18 Subsequently, community members were asked to list the types of facilities and locations where community members sought care for those conditions identified in the free-listing interviews. Criteria for participation in the free-listing exercise included being a person >18 y of age capable of giving full informed consent and characterized by one of the following: a community member or leader, a traditional or allopathic health worker or a woman with children <5 y of age. The study team conducted free-listing exercises through individual key informant interviews with village chiefs, when available, traditional and allopathic health workers, adult men and women, and women with children <5 y of age. Given the small size of communities, data on the specific role participants played in the community were not recorded given their potential to provide identifiable information.
Free-listing exercises were designed to get a range of answers relevant to a domain of inquiry, not to provide frequencies of illnesses. All respondents were asked to report on diseases perceived as commonly found among all demographics of the population (i.e. women were asked about diseases among the elderly, children, men, etc.) to generate a broad view of diseases affecting the community.
Facility assessments identified sources of health care in surveyed study sites. ‘Health facilities’ were defined as any location where community residents with illnesses seek care (as identified by the free-listing exercises). At each site, at least one health worker provided a key informant interview. A ‘health worker’ was broadly defined as anyone involved in the care of patients or dispensing/selling treatment options to persons seeking care. Facility assessments focused on the type and number of health care workers available, qualifications (if any) of each health care worker, hours of facility operation, medical supplies available at the time of the assessment of each facility, the approximate number of people seen per day and per week at each facility and any additional comments regarding the function of each facility. All health facilities identified by community members in target communities were included in the health facility assessments.
Data analysis
The authors analyzed free-listing key informant interviews as well as facility assessment data using a deductive approach through interview questions. Simple tabulation and summative content analysis was achieved using key themes, which included common diseases experienced by the community, locations where treatment was sought and patterns of illness and care-seeking with regards to reproductive health care. The emergent framework was subsequently used to group data and identify relationships.
Results
Free-listing exercises
Sixty-nine participants were interviewed using free-listing exercises (Table 1). The median and average age of participants was 43 and 42 y, respectively (Table 2). Participation was distributed relatively equally among genders, with 33 (48%) males and 36 (52%) females.
Number of study participants by community
| Community | Free-listing exercises | Facility assessments |
|---|---|---|
| Antsirabe-Sahatany | 10 | 3 |
| Vinanibe | 10 | 3 |
| Ankofabe | 10 | 4 |
| Voloina | 10 | 3 |
| Rantabe Ambanizana | 10 | 2 |
| Mahalevona | 10 | 4 |
| Iharaka | 6 | 1 |
| Marofototra | 3 | 1 |
| Total | 69 | 21 |
| Community | Free-listing exercises | Facility assessments |
|---|---|---|
| Antsirabe-Sahatany | 10 | 3 |
| Vinanibe | 10 | 3 |
| Ankofabe | 10 | 4 |
| Voloina | 10 | 3 |
| Rantabe Ambanizana | 10 | 2 |
| Mahalevona | 10 | 4 |
| Iharaka | 6 | 1 |
| Marofototra | 3 | 1 |
| Total | 69 | 21 |
Number of study participants by community
| Community | Free-listing exercises | Facility assessments |
|---|---|---|
| Antsirabe-Sahatany | 10 | 3 |
| Vinanibe | 10 | 3 |
| Ankofabe | 10 | 4 |
| Voloina | 10 | 3 |
| Rantabe Ambanizana | 10 | 2 |
| Mahalevona | 10 | 4 |
| Iharaka | 6 | 1 |
| Marofototra | 3 | 1 |
| Total | 69 | 21 |
| Community | Free-listing exercises | Facility assessments |
|---|---|---|
| Antsirabe-Sahatany | 10 | 3 |
| Vinanibe | 10 | 3 |
| Ankofabe | 10 | 4 |
| Voloina | 10 | 3 |
| Rantabe Ambanizana | 10 | 2 |
| Mahalevona | 10 | 4 |
| Iharaka | 6 | 1 |
| Marofototra | 3 | 1 |
| Total | 69 | 21 |
Age distribution of free-listing exercise participants (N=69)
| Age (y) | n | % |
|---|---|---|
| Range | 19–73 | |
| 18–30 | 11 | 16 |
| 31–40 | 19 | 28 |
| 41–50 | 26 | 38 |
| 51–60 | 9 | 13 |
| >60 | 4 | 6 |
| Median | 43 | |
| Average | 42 | |
| Gender | n | % |
| Male | 33 | 48 |
| Female | 36 | 52 |
| Age (y) | n | % |
|---|---|---|
| Range | 19–73 | |
| 18–30 | 11 | 16 |
| 31–40 | 19 | 28 |
| 41–50 | 26 | 38 |
| 51–60 | 9 | 13 |
| >60 | 4 | 6 |
| Median | 43 | |
| Average | 42 | |
| Gender | n | % |
| Male | 33 | 48 |
| Female | 36 | 52 |
Age distribution of free-listing exercise participants (N=69)
| Age (y) | n | % |
|---|---|---|
| Range | 19–73 | |
| 18–30 | 11 | 16 |
| 31–40 | 19 | 28 |
| 41–50 | 26 | 38 |
| 51–60 | 9 | 13 |
| >60 | 4 | 6 |
| Median | 43 | |
| Average | 42 | |
| Gender | n | % |
| Male | 33 | 48 |
| Female | 36 | 52 |
| Age (y) | n | % |
|---|---|---|
| Range | 19–73 | |
| 18–30 | 11 | 16 |
| 31–40 | 19 | 28 |
| 41–50 | 26 | 38 |
| 51–60 | 9 | 13 |
| >60 | 4 | 6 |
| Median | 43 | |
| Average | 42 | |
| Gender | n | % |
| Male | 33 | 48 |
| Female | 36 | 52 |
Common diseases and care-seeking patterns
Respondents perceived headaches, fever and cough as the most common health complaints in their communities, as a whole. Respondents stated that cough was the most common disease among children, followed by fever and headaches. In general, participants reported seeking care at locations providing allopathic care (either the nearest hospital or the local dried goods stands, called epiceries, where allopathic medicines are sold alongside household items) only after traditional medicines had not worked. However, certain populations were thought to access allopathic care more quickly than others. For example, respondents reported that parents most often take their children to the nearest hospital when ill rather than waiting for a failed traditional medicine treatment. In fact, respondents reported that parents would, as a second choice, take children to the local epicerie in order to purchase allopathic medications and, only as a final option, might choose to use traditional medicine.
Respondents reported that adolescent women most commonly sought care for dental problems, fever and headaches. Life-threatening complications such as hemorrhage and severe infections from self-induced abortions using foreign objects and traditional plant-based medicines were also reported among this population. Adolescent men were thought to seek care most often for fiandry, back pain and fever. Fiandry is a local term that was mentioned often as a common disease, but it was difficult to translate directly into English. Local populations used this broad term often when referring to symptoms ranging from general fatigue and dehydration to conditions affecting the genitourinary system.11 The majority of respondents suggested that adolescent males and females initially sought care using either traditional medicines or by going to a CSB or hospital. Almost all respondents felt that the hospital was the most common location where the elderly sought care—more often than local epiceries or traditional medication.
Factors affecting care-seeking behavior
Participants stated that, in general, members of their community first try traditional medicine or allopathic medications from a local epicerie, with the exception of children and the elderly as noted above. If the illness was severe or progressed, they then contemplated whether they had the finances and available family to seek care at the nearest health facility. One participant stated:
They try to buy medications here in the community or go to the forest. If the illness gets worse, then they try to go to the hospital. It is very difficult to get [to the hospital], so that is only for extreme cases.
In the pediatric population, there is a perceived lower threshold to seek allopathic care initially at a local epicerie or hospital. Traditional medicines were used for simple complaints, when serious consequences of a disease were deemed to be less likely. A participant summarized care-seeking behavior for children by stating:
Since the doctors are very far, parents usually go to the epicerie first. If [the epicerie owner] has medication, he will give it to them. If not, he will send the child to [a nearby town with a higher level of care].
Adolescent women and men usually treat their symptoms first by using traditional medicine or by going to the local epicerie to purchase allopathic medications. Care-seeking behavior depended on the type of symptoms and the gender of the ill person. For example, community respondents reported that women might be less likely to seek care in a hospital or clinic setting where only male providers are present.
As a female participant reported:
(Adolescent women) always use traditional medication first. If they have a problem with this, they go to the epicerie.
Commenting on self-induced abortions, one participant stated:
Young women use local medication for abortion, and this can make them sick. Women don’t go to the hospital for this…People often use objects to get rid of them, causing hemorrhage and infections.
When discussing care-seeking behavior of adolescent men, a participant stated:
(Adolescent men) try to get medications from the epicerie first. If it is difficult, then they go to the hospital…The men usually go to the epicerie first since the owner is male. They sometimes show him their problem. Sometimes he doesn’t have the medication to treat them. If this happens, he will write down the name and the patient will go to the CSB to buy it. The epicerie owner will administer it after.
Discussing the elderly population, respondents suggested that community members prioritize seeking allopathic care, at least in part because of a higher perceived risk of poor outcomes resulting from an illness.
They always go to the hospital directly. Everyone in the community tries to help watch their house and children so that they can take the elderly to clinic. Sometimes they have to carry them for two hours if the sea is bad. If the sea is good, they go by boat.
Challenges to seeking care
Although many prefer care provided by government facilities, particularly for individuals perceived to be at highest risk, barriers exist. The most commonly cited barrier to seeking health care from an allopathic professional was distance to the nearest CSB or medical facility. Because most of the communities included in this study were in the rural regions of Maroantsetra, reaching a CSB I or II could take 1–8 h by foot or 1–4 h by river or sea. Travel by sea was less difficult but one had to own, borrow or rent a boat and take weather and sea conditions into consideration. For this reason, one participant stated:
First thing they do is go to the epicerie or use traditional medications. Hospital is the last choice.
Financial barriers present another significant obstacle. As is true in many contexts, the sick member of the household is not the only member who is affected by a particular illness. If care is sought at the nearest CSB, patients are usually perceived to be quite ill and often must be accompanied by another member of the household. Thus families have to take into consideration the estimated cost of allopathic care and daily income lost to the household by the sick patient and the member accompanying him or her. In contrast, traditional medicines are typically made from plants found in and around the community and thus are a fraction of the cost compared with even highly subsidized Malagasy allopathic medicine.11
Facility assessments
The research team assessed a total of 21 facilities (Table 1) in eight communities (Figure 4). ‘Health facilities’ were broadly defined and included CSB I and II (n=6), local dried food stands (epiceries) selling allopathic medications (n=7), traveling salespeople (n=5) and local pharmacies (n=3). Traveling salespeople were described as merchants who pass through communities to sell medications. All CSBs had at least one health care worker. Of all the facilities, one CSB was staffed by a physician and two facilities had one nurse each. An independent nurse was found to be selling medications on foot. All staff encountered in epiceries and pharmacies, as well as traveling salespeople (with the exception of the nurse identified above), had no formal health training. Epiceries were staffed by owners or family members of the owner. Five epiceries had staff who had received a brevet d'études du premier cycle (BEPC), equivalent to completing lower secondary education, and one epicerie had one staff member who had received a certificat d’etudes primaires èlèmentaires (CEPE), or a primary education; however, none had health-related training. Most facilities were open at least 12 h/d (CSBs), while the epiceries were generally available 24 h/d, 7 d/week. The travelling salespeople were present on an occasional basis, as they pass through one community to the next before returning to cities to restock. Most facilities had no waiting times. None reported expired medications. In 14 of the 21 facilities, health care workers reported providing some form of nutritional and maternal information, though the content of this counseling was not assessed. Health facility workers confirmed that the most common presenting complaints included fever, cough and headaches, mirroring reports from community members.
Facility locations where assessments were made.
Discussion
These findings aim to contribute to a better understanding of the common sources of health care and patterns of health resource utilization in rural communities of the Maroantsetra region of Madagascar.
The most common health conditions reported were headaches, fever, cough, fiandry and back pain. It is important to note that these are symptoms and they reflect the ways in which patients perceive illnesses in the Maroantsetra region. Because trained providers and point-of-care technology are sparse in rural regions of Madagascar, medical conditions are frequently identified and treated based solely on symptoms and a cultural understanding of the disease. Although only symptoms were reported, those reported by these communities overlap with some of the most common diseases reported in Madagascar, including acute upper respiratory infections, malaria and diarrhea. Combined, these three diseases contribute to almost 30% of the total years of life lost in Madagascar.19 Community-based health care protocols centered around these common conditions and their perceived symptoms may lead to improved provision of health care in these remote regions and improved patient outcomes. Input from local health care providers, researchers and community members may be used to develop locally relevant protocols to address presenting complaints at the community level and incorporate community norms regarding allopathic and traditional medicine.
Traditional medicine and local healers have always been central to health care in remote parts of Madagascar. In our study region in northeastern Madagascar, researchers found that the use of botanical ethnomedical treatments was well developed, with the use of 241 locally recognized plant species to treat more than 82 categories of illness in that region.11 This is unlikely to change given the strong cultural beliefs and a relative scarcity of allopathically trained professionals, including physicians, throughout the country.6 A representative household survey based on the Demographic Health Survey conducted in the Ifanadiana district of southwestern Madagascar in 2017 found low utilization of the formal health system.20 Research designed to identify safe and culturally meaningful methods of including these preferences and geographic realities in a formalized government health care program would help to engage a broad constituency in Madagascar, as access to allopathic facilities continues to improve.
Additionally, its extensive coastline and location in the Indian Ocean gives Madagascar a distinct susceptibility to climate change and natural disasters.21 Climate change affects health in multiple ways, including direct trauma from disasters, heat-related mortality and infectious and vector-borne diseases. These realities compound challenges faced by the Malagasy health system with regards to consistent staffing, supply chains and infrastructure.6,22 While these broader health system issues are being addressed, local healers and traditional medical treatments may provide cost-efficient, safe and more immediately accessible means of providing health care to remote populations.6 Finding ways of incorporating essential allopathic training into the traditional health care system may help improve access to lifesaving care.
In general, the decision to access care was thought to depend on a combination of geographic and financial barriers, the perceived severity of the illness and the availability and confidence in traditional plant-based medications. These findings were consistent with a 2013 study in Madagascar concluding that improved access to health care and shorter travel times would help reduce maternal mortality at the commune level.15 Notably, the threshold to seek allopathic care was lower for those perceived to be more vulnerable to illness, including pediatric and elderly populations. Confidence in traditional medicine was evident given the preference to seek this treatment modality first when dealing with illness in nonvulnerable populations.
Our findings are consistent with a prior study that found that infrastructure, training/staffing and the supply chain are key to strengthening health care in Madagascar.4 Although community health agents are found in some areas of Madagascar, our findings did not reveal a very active presence.23 We suggest developing and field testing a program that trains informal health care providers in protocol-driven community health worker skills, based on the common symptoms identified above. This would be best combined with community-based epidemiological surveillance systems, such as common tablet-based mobile health surveillance systems.24,25 In fact, the MAHERY team has begun pilot testing these initiatives on a local scale in the Maroantsetra region.
The Malagasy government is intending to roll out universal health care and a structure for community health workers in the near future. We agree that these steps would greatly enhance health care access throughout the island nation. Recent studies have documented the feasibility, effectiveness and acceptability of community-based distribution services of injectable contraceptives and delivery of prepackaged chloroquine, reducing the risk of dosage errors and compliance issues.22,26 Furthermore, recent literature reviews have shown that using community health workers for prevention and control of noncommunicable diseases, screening for cervical cancer and management of cardiovascular disease are both feasible and effective in low-resource settings.27–29 This further supports the training for community health providers in rural Madagascar as a viable approach in the region. Incorporating traditional medicine may serve as a means of increasing the uptake of services provided by trained health providers and help identify potentially dangerous conditions for early referral to government facilities.
Limitations
Our study was exploratory and qualitative, using purposive, nonrepresentative sampling. This very likely introduced some bias and may have excluded the perspective of hidden or particularly vulnerable populations. Furthermore, the perceived role of the research team as health care providers could have provided skewed results when discussing common diseases in the community. In some cases, female participants only consented to be interviewed if their husband was present. This may have influenced women’s responses or prevented them from stating an opinion that may have been against or different from the husband’s opinion. Lastly, because our study was conducted in Malagasy and subsequently translated or transcribed to English, translation errors may have been introduced. Finally, these results represent findings from a limited set of communities in the Maroantsetra region and cannot be generalized to the broader rural and urban Malagasy context.
Conclusions
Limited data currently exist on patterns, motivations and obstacles to health care utilization and decision making in rural Madagascar. The decision to use allopathic medicine was determined by the perception of disease severity and when vulnerable populations, like children and the elderly, were involved. These preliminary findings provide additional insight into community and health care provider perceptions, especially the decision to use allopathic vs traditional medicines in rural communities. These findings could help inform more community-based, locally relevant programming that incorporates informal providers and traditional medicine in safe and effective ways, assist in creating community-based health care protocols centered around common conditions and their perceived symptoms with the integral involvement of local health providers and community members, and aid in establishing avenues to incorporate essential allopathic training into the traditional health care system. Further study on community perceptions of disease, effectiveness of traditional medicine and the means by which informal health care providers can be safely integrated into rural Malagasy health care is warranted and should be conducted in close partnership with the communities themselves.
Authors’ contributions
CDG and PP conceived the study and designed the study protocol. NDB, PP and JFR conducted interviews, with JFR conducting all interviews in Malagasy. NBD analyzed and summarized data as well as drafted the manuscript. NBD, CDG and PP critically revised the manuscript for intellectual content. All authors read and approved the final manuscript. NDB and PP are guarantors of the paper.
Acknowledgements
The map in this article was created using ArcGIS software (Esri, Redlands, CA, USA). ArcGIS and ArcMap are the intellectual property of Esri and are used herein under license. Copyright © Esri. All rights reserved. For more information about Esri software, please visit www.esri.com.
Funding
None.
Competing interests
None declared.
Ethical approval
The research protocol was approved by the Institutional Review Board of the University of Southern California Health Sciences Campus. Researchers obtained verbal consent from all participants in Malagasy in a private area prior to all interviews. To protect confidentiality, researchers did not obtain identifying information on any study participant. The research team conducted all interviews in private settings to ensure the confidentiality of all responses. All data were stored on secure, password-protected laptops. All respondents were asked to provide information regarding health conditions in their communities in general, not their individual experiences with health issues, in order to protect confidentiality.
References
Statista. Physicians density worldwide as of 2013, by region (per 10,000 population). https://www.statista.com/statistics/280158/physicians-density-worldwide-by-region/ [accessed 6 August 2017].
Comments