Vermeersch et al. Malaria Journal 2014, 13:50 102462 http://www.malariajournal.com/content/13/1/50 RESEARCH Open Access Introducing the concept of a new pre-referral treatment for severely ill febrile children at community level: a sociological approach in Guinea-Bissau Audrey Vermeersch1ˆ, Anaëlle Libaud-Moal1, Amabelia Rodrigues2, Nicholas J White3,4, Piero Olliaro5,4, Melba Gomes5, Elizabeth A Ashley3,4 and Pascal Millet1* Abstract Background: Innovative strategies are needed to tackle childhood mortality in the rural tropics. Artesunate suppositories were developed to bring emergency treatment closer to severely ill children with malaria in rural areas where injectable treatment is not possible for several hours. Adding an antibacterial rectal drug would extend this strategy to treat non-malarial febrile illness as well. The objective of these studies was to assess acceptability of such a new pre-referral strategy by healthcare providers and likely uptake by the population. Methods: Two qualitative studies were conducted between May and July 2009. Study 1 investigated the acceptability of introducing a combined antimalarial-antibacterial suppository by interviewing 27 representatives of the three administrative levels (central government, regional, local) of the health sector; Study 2 investigated treatment-seeking behaviour and acceptability of this intervention at community level by interviewing 74 mothers in 2 villages. Results and Conclusions: Up to 92% of health representatives were in favour of introducing a new pre-referral strategy to tackle both malaria and non-malaria related severe illnesses in Guinea-Bissau, provided it was endorsed by international health authorities. The main obstacles to implementation were the very limited human and financial resources. In the two villages surveyed, 44% of the mothers associated severe illness with fever only, or fever plus one additional symptom. Mothers’ judgement of severity and ensuing decisions were not specific for serious illness, indicating that initial training to recognize signs of severe disease and treatment availability for non-severe, fever-associated symptoms will be required to prevent overuse of a new intervention designed as a pre-referral treatment for severely ill children. Level C health centres were the first resort in both villages (50% and 87% of respondents respectively). This information is encouraging for the implementation of a pre-referral treatment. Keywords: Severe febrile illnesses, Malaria, Artesunate, Antibiotics, Rectal administration, Pre-referral treatment, Rural areas * Correspondence: pascalmillet1@gmail.com ˆDeceased 1 EA 4575 Développements Analytiques et Pharmaceutiques appliqués aux Maladies Négligées et aux Contrefaçons, Université Bordeaux Segalen, Bordeaux, France Full list of author information is available at the end of the article © 2014 Vermeersch et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. Vermeersch et al. Malaria Journal 2014, 13:50 Page 2 of 10 http://www.malariajournal.com/content/13/1/50 Background antimalarial-antibacterial pre-referral treatment at all It has been estimated that Plasmodium falciparum mal- levels of the health sector; (ii) to evaluate the know- aria and acute respiratory infections (ARI) together ac- ledge, attitudes and practices of mothers and primary count for 25% of childhood deaths worldwide. The health care staff confronted with febrile illness in clinical manifestations of malaria and ARI are often children in rural areas in order to define implementa- similar, which tends to lead to frequent overdiagnosis of tion of home-based pre-referral treatment. malaria [1,2]. The health of a febrile child can deterior- Guinea-Bissau was chosen because it has a high mal- ate very rapidly and a large number of the deaths in aria prevalence and limited access to health structures rural areas occur before patients have reached a health for the majority of the population who live in rural areas care facility. In Tanzania, up to 40% of children who die (64%) [7]. The average life expectancy is 46 years and have never been seen in a health centre; in Bolivia this Guinea-Bissau ranked 164 out of 169 countries in the figure approaches 74% [3]. In severe falciparum malaria, 2010 Human Development Index [8]. the risk of death increases in line with the delay in ad- Infectious diseases account for most health care con- ministering an effective anti-malarial drug and is max- sultations, in particular malaria, ARI, diarrhoeal disease, imal in the first 24 hours. Therefore, the speed at which HIV and tuberculosis. The first three of these account appropriate management is instituted is the key factor for 65% of deaths in children less than five years of age. influencing the outcome of treatment. Administration of More than 50% of medical consultations are for malaria, rectal artesunate as a pre-referral treatment was evalu- and the country reported 50 deaths for 100 000 admissions ated in a large randomized, placebo-controlled trial to health care facilities in 2012. Artemether – lumefantrine of more than 12,000 malaria patients conducted in is the recommended treatment for uncomplicated malaria. Bangladesh, Ghana and Tanzania. Such treatment re- Intra-venous quinine is the recommended treatment for duced death or permanent disability in patients with se- severe malaria. ARI represents the second most common vere malaria who experienced delays greater than six cause of death in children aged less than five years in hours in reaching a health care facility [4]. Pre-referral Guinea-Bissau [9]. treatment with a single dose of rectal artesunate has been introduced within the Integrated Management of Methods Childhood Illnesses (IMCI) in remote areas. Two qualitative studies were performed: Knowing that a sizeable proportion of deaths in young children labelled as malaria are probably the result of Study 1 – Interviews with health sector representatives bacterial sepsis, in particular pneumonia, there is a Public health structure strong argument for extending the strategy of emergency The organization of the national public health service is pre-referral treatment to antibacterial as well as anti- pyramid-shaped with three levels, local, regional and malarial therapy by co-formulating an antibiotic with an central. The central level is represented by the Ministry artemisinin derivative in suppository form [5,6]. Such a of Public Health (MPH) and consists of the Office of product has the potential to reduce mortality by redu- Services, the National Programmes, Simão Mendes Na- cing the delays in antibiotic treatment among children tional Hospital and other national specialized hospitals who can no longer take oral antibiotics. and centres. The central level draws up policies and Effectiveness of this strategy under real-life conditions strategies, regulates health care activities and coordinates not only depends on demonstrating efficacy of the for- external aid. The regional level is represented by the Re- mulation but ensuring access to treatment among re- gional Hospital, Regional Health Office and regional mote populations who are likely to benefit most from health care teams. It is in charge of implementing the this kind of intervention is a major challenge. This annual programme. At the local level, the health struc- process starts with the incorporation of the new treat- tures provide primary care and consist of Levels A, B or ment into a country’s policy. Thereafter, a continuous C depending on geographical location, access to a re- supply of the drug in areas of need, should be estab- gional hospital, and size of population served: lished. In the rural tropics the nearest health centre is often several miles away from a village, thus the speed of  A: may serve also as secondary level of care, detecting serious illness and patterns of treatment- physicians, nurses, clinical, surgery and some seeking behavior that emphasize treatment and referral laboratory diagnosis, referral system for the regional will also be factors affecting how a policy performs in hospitals; practice.  B: secondary level of care, physicians, nurses, clinical The main objectives of this case study, which took and basic laboratory diagnosis; place between May and July 2009, were: (i) to evalu-  C: primary level of care, nurses only, first contact ate the acceptability of a proposed strategy of rectal between community members and other levels of Vermeersch et al. Malaria Journal 2014, 13:50 Page 3 of 10 http://www.malariajournal.com/content/13/1/50 health facility. Very basic clinical care and no Tchale village has around 1,000 inhabitants, mainly an- laboratory diagnosis. imists of the Balante ethnic group, and is situated 90 mi- nutes (by foot) from a Level C health centre and three These health care facilities are managed by a technical hours (by foot + car) from Bissora sector Level B health team and management committee. Representatives of centre. The nearest health care facility able to manage the three levels of the health sector were interviewed severe malaria or ARI is Bissau National Hospital, three (Table 1). The total number of respondents was 27. hours away from the village by road. Each village consists of 32 plots distinct from each Study 2- Interviews with mothers, community health other; the number of houses per plot may vary from workers, and health professionals from level C and B two to 21. At least one woman per plot and two women health care facilities (in rural Guinea-Bissau) in larger plots were interviewed in order to reach a tar- Selection of the community study sites and randomization get of 40 respondents. Most women interviewed had Tchale and Bantandjan villages in Oio region were se- never attended school. Apart from picking cashew nuts lected. They were two of the furthest villages from a during the harvest season, most had no income- health centre with laboratory diagnostic facilities and generating activities. The survey was conducted during ability to administer injectable treatment. The target the cashew nut harvest so many women were not at population of this survey was mothers of children under home. Thirty six housewives were interviewed in Tchale five years of age. Inclusion criteria were availability dur- village, on the assumption that field- and home- ing the study and informed consent of respondents. The working mothers shared responsibilities and income sample size was 40 mothers per village. from each plot equally. Table 1 Description of interviewees in Study 1 Category of respondent Organization/Department Position Number Representatives of the Health Departments Department of Family Health Director 1 at central/government level National Malaria Control Programme Director 1 National Programme For Health Development Director of national control programme 1 for malaria, HIV and tuberculosis National Programme For Health Development Responsible for cost recovery 1 Department of Primary Health Care Director 1 Department of Pharmacy and Medicines Assistant Director 1 CECOME* Director 1 Representatives of the Health Departments Regional departments of Health Oio Region Director and Assistant Director 2 at regional level CECOME* regional depot/distribution centre, Manager 1 Mansoa Representatives of International organizations WHO Guinea-Bissau Head of Department of Communicable 1 and donors Disease Control WHO Guinea-Bissau WHO representative in Guinea-Bissau 1 Médecins Du Monde Medical coordinators 2 UNFPA Guinea-Bissau Director of Department of Reproductive 1 health UNICEF Guinea-Bissau Programme Director 1 National Programme For Health Development Global Fund representative 1 Plan International Medical coordinator 1 WHO Guinea-Bissau Malaria programme director 1 Health professionals at different levels Simao Mendes National Hospital Paediatrician 1 Mansoa Regional Hospital Paediatrician 1 Farim Health Centre (Type A)** Nurse 1 Bissora Health Centre (Type B)** Nurse 1 Health Centre (Type C)** Nurses 4 *CECOME Central de Compra de Medicamentos (Central Office for Purchasing of Essential Medicines, Guinea-Bissau). **Health centre Type A, B or C: classification of primary health care centre according to activity. Level A has the most complete package of care including some surgery. Vermeersch et al. Malaria Journal 2014, 13:50 Page 4 of 10 http://www.malariajournal.com/content/13/1/50 Bantandjan village is seven hours by foot or 75 mi- and aims of the project. Verbal consent was obtained nutes by car from a Level C health centre. The nearest before each interview. For interviews that were tape- reference structure is Mansoa Regional Hospital, situated recorded, permission was requested at the start of between two hours and 15 minutes and eight hours the interview. from the village, depending on the type of transport used. Bantandjan has about 350 inhabitants who are Results mainly Muslims from different ethnic groups: Mandigoes, Study 1 Manjaques, Mansanques, Pepels, and Balantes. The village Acceptability of the proposed strategy of a combined is made up of 52 houses. One woman was interviewed antimalarial-antibacterial emergency pre-referral treatment in each of 40 households selected at random from a list All respondents judged that the proposed strategy was of all village homes. Most women interviewed had never suitable for the epidemiological context in Guinea- attended school. Apart from the palm oil harvest, they Bissau for children less than five years of age. Several in- had no income-generating activities. The survey took terviewees stressed their interest in having a combined place during palm oil season and 38 mothers were inter- medicine capable of treating both malaria and respira- viewed in Bantandjan village. tory infections as a strategy to combat high childhood In each health area, but not in each village, there is mortality rates. Some respondents remarked upon the community health services called Basic Health Struc- pertinence of proposing a treatment which contains tures (BHS), managed by Community Health Workers medicines already well known and conforming to the with very basic health training (no health related treatment protocols in place in the country, and in a diploma), and traditional midwifes. The health care facil- form more suitable for young children than tablets. Ac- ities selected for the study were close to Tchale and cording to one paediatrician, the idea of a co-formulated Bantandjan: four BHS, one Level C and one level B treatment was particularly suited to Guinea-Bissau since health centres surveyed and named by the mothers. The there was already a national move towards integrated respondents of the health centre survey were health pro- management of sick children for both malaria and re- fessionals in charge of these facilities. They were Com- spiratory infections in malaria-endemic regions. The munity health Workers in BHS, nurses in in level C, and current national health policy for rural areas has been one doctor in level B health centres. trying to revive and strengthen the most basic peripheral health posts in an attempt to bring emergency treatment Interview methodology to the village level. Several respondents stressed the diffi- For both studies, interviews were semi-structured and culties of geographical access that delay the arrival of conducted with the help of pre-prepared interview patients at health care facilities, particularly during guides. The language of the interviews was French at the the rainy season. According to one respondent, a pre- central level and Portuguese Creole at the regional and referral treatment at community level might also help to local levels, via translators. For study 1, open questions counteract the delay in seeking treatment related to cul- were used to evaluate acceptability of a pre-referral tural practices: treatment at all health care levels. For study 2, open questions were used to evaluate the knowledge of the ‘At community level there are cultural difficulties respondents of severely ill febrile children, and closed because it is up to the father to take the decision to questions were used to define their attitudes and prac- bring the child to a health centre. The mother waits tices for treatment. The data collection tools were pre- for her husband to decide; that can take some time. tested in Bissau rural areas. This situation would justify the use of a pre-referral treatment in the community.’ Data analysis Interviews were transcribed verbatim. Data collected was The interviewees were aware that such a strategy categorized by theme. Specific comments were selected would comply with WHO recommendations to use rec- either because they were repeated frequently or because tal artesunate as emergency treatment. they best reflected the overall opinion of the respondents For 23 out of 25 respondents (92%) administration of on a specific topic. a combined antibiotic-anti-malarial treatment would be an acceptable solution for the emergency pre-referral Ethical approval treatment of children aged less than five years. One sin- The study protocol was approved by the Ethical Com- gle reservation mentioned was the risk that the advice to mittee of the Department of Hygiene and Epidemiology seek further treatment at a health centre would not be of the Ministry of Health in Guinea-Bissau. Study partic- followed by parents once the suppository had been ipants were given a document outlining the background administered. Vermeersch et al. Malaria Journal 2014, 13:50 Page 5 of 10 http://www.malariajournal.com/content/13/1/50 Two respondents remarked upon the acceptability of health workers and basic health care units that are the strategy from a public health point of view in a con- closest to the population. text where there was no alternative. One of the paedia- tricians felt that a combination treatment containing In summary, the conditions necessary for the imple- an antibiotic would be an advance compared to giving mentation of a new treatment in the health sector iden- artesunate alone. In practice, certain medicines are tified were: already given by the rectal route (quinine, phenobarbit- one, paracetamol) without problems. The availability  Demonstration of efficacy and/or heat stability (five of a heat-stable suppository to treat children was con- respondents) sidered advantageous. Some health professionals under-  Integration into the national treatment protocols lined the importance of giving clear information to (six respondents) mothers and community health workers on the use  Validation of the treatment by WHO (two of this route of administration in order for it to be respondents) adopted easily.  Information, education and communication campaigns aimed at the population (18 respondents) Conditions for use of a combined antimalarial-antibacterial  Training, information and remuneration of the suppository formulation health technicians (18 respondents) According to the representatives of the various health-  Follow-up and supervision of the community health related institutions interviewed (six respondents), the workers (five respondents) introduction of such a treatment into the country  Population payment (token payment) towards the would necessitate a modification of the policy for the cost of the medicine (one respondent) management of malaria in children. This would imply an application to include the new medicine into the Objections, obstacles and difficulties anticipated for National List of Essential Medicines and the treatment introduction in the country and for its use at community policy documents of the National Malaria Control level Programme and Integrated Management of Childhood Interviewees postulated potential obstacles to the intro- Illness. This would be conditional on the identification duction of a medicine at the community level as follows: of partners with the necessary financial and technical expertise.  Poor availability of personnel at the community The interviewees also discussed factors affecting access level: community health workers are already to treatment by the target population. Three main as- overworked and very often they do not stay a long pects were highlighted: time in the same village (four respondents)  Financial difficulties of the population in rural areas  For a large majority of the respondents, cost was the which could prevent referral to health centres, major determinant of accessibility to medicines. because of transport, consultation and They referred to the strategy implemented to make accommodation costs (three respondents) Coartem® affordable to the population:  If the constituents of a new product were not already part of the national treatment policy for ‘The cost of the new medicine must not be higher than treatment of malaria and respiratory infections it the cost of Coartem® ..(…) We know how much the would be difficult to get approval (three population is able to pay and not all of them can respondents) afford Coartem®. We cannot give the medicine free of  The administrative and financial burdens incurred charge but the price must not be an obstacle.’ when a new medicine is introduced into the national policy (two respondents) Half of the participants interviewed believed that a  Cultural obstacles: reliance on traditional healers by subsidy of a new medicine would be necessary in the population in some areas (two respondents) order to ensure accessibility.  The difficulties of ensuring a regular supply of a new  Half of those interviewed cited availability of the medicine, particularly if the process involves medicine as an important condition to ensure access external donors, as seen with the introduction of to treatment. For them, the assurance of a regular artemether – lumefantrine (Coartem®) financed by supply from the manufacturer to the village was the Global Fund (one respondent) vital, with a reliable distribution system.  Problems of access to treatment in villages where  Two of 6 respondents felt that access to the there is no community health worker (one medicine should be guaranteed by the community respondent) Vermeersch et al. Malaria Journal 2014, 13:50 Page 6 of 10 http://www.malariajournal.com/content/13/1/50  The contradiction between emergency treatment Table 3 Treatment of first resort in Tchale and proposed with the planned introduction of rapid Bantandjan diagnosis tests for malaria in the peripheral health First healthcare % Tchale % Bantandjan posts (one respondent). choice (n = 36) (n = 38) Home treatment 5.6 2.7 Study 2 Traditional practitioner 27.8 10.5 Knowledge, attitudes and practices of mothers faced with a Level C health centre 50.0 86.8 seriously ill febrile child Tchale village (36 respondents) Level B health centre 11.0 0 Knowledge of mothers regarding serious conditions National Hospital 5.6 0 in children When asked to describe the signs and symptoms of a seriously ill child who may be dying, out of 36 mothers, 22.2% quoted fever only and 36% quoted fever plus one or more other symptoms Table 2). child had not improved after being treated in a Level C health centre. Attitudes and practices of mothers regarding serious illness in children Treatment of first resort When presented with a hypo- Time spent in the village before departing for a health thetical situation in which their child is in a serious structure and justification condition (febrile and cannot swallow) and asked to The delay before departing for a health facility varied describe their first care-seeking response, 50% of 36 from zero to seven days. The average lapse of time was mothers said they would go to the Level C health centre 2.13 days (standard deviation = 1.58). Reasons given for and 27.8% answered that they would consult a trad- the time delays were: itional practitioner first (Table 3).  To find money for consultation fees and drugs (83% of respondents) Various treatment-seeking behaviours described by the  To find money for transport (61% of respondents) mothers for the treatment of their child and justification  To find a means of transport (25% of respondents) When asked what they would do if their child was  To wait for the family decision-maker (2% of not cured after the first care-seeking response, most respondents) of the women who had opted for the traditional prac-  To be available to go (2% of respondents) titioner chose to go to a Level B or C health centre  Other (2% of respondents) (Figure 1 and Table 3). For these women, the fact that the treatment provided by the traditional practi- Two women stated they would not encounter any tioner did not cure the child meant that the disease problems and could leave the village whenever they was not of spiritual origin. wanted to. Four mothers mentioned the need for a spir- Out of the 22 women who opted for a Level B or itual ceremony as one of the reasons for a delayed de- C health centre for the initial treatment, 31.8% said parture to a health facility. they would choose to consult a traditional healer next, while 50% would go to a higher level facility if the Bantandjan village (38 respondents) Table 2 Signs and symptoms of serious illness reported Knowledge of mothers regarding serious conditions by mothers - Study 2 in children Of 38 mothers, 7.9% quoted fever only and Symptoms quoted % Tchale % Bantandjan 57.8% quoted fever plus one or more other symptoms. by mothers (n = 36) (n = 38) All mothers considered that a child unable to eat or Fever only 22.2 (8) 7.9 (3) drink was in a severe state (Table 3). Fever plus another symptom 22.2 (8) 36.8 (14) Fever plus two other symptoms 8.3 (3) 18.4 (7) Attitudes and practices of mothers regarding serious Fever plus three other symptoms 5.5 (2) 2.6 (1) conditions in children Other symptoms – without fever 30.6 (11) 15.9 (6) Selection of the treatment of first resort Out of 38 Don’t know 11.2 (4) 18.4 (7) mothers, 86.8% said they would go to the Level C health Total 100 (36) 100 (38) centre and 10.5% answered that they would consult a Other symptoms: fatigue, vomiting, diarrhoea, anorexia, respiratory distress, traditional practitioner as a treatment of first resort prostration, convulsion, headache, icterus, pale face. (Table 3). Vermeersch et al. Malaria Journal 2014, 13:50 Page 7 of 10 http://www.malariajournal.com/content/13/1/50 Figure 1 Patterns of treatment-seeking behaviour described by the mothers in Tchale. Various treatment-seeking behaviours described by the  To find money for transport (97 of respondents) mothers for the treatment of their child and justification  To find a means of transport (97 of respondents) The various itineraries described by the mothers with explanation are presented in Figure 2 and Table 3. The Only one respondent asserted that she would not en- four women (11%) who opted for the traditional healer counter any problem. as a first resort would go to a Level C health structure if the child had not been cured by traditional cere- Knowledge, attitudes and practice of staff of primary mony. Of the 87% of women who opted for a Level C healthcare facilities managing complicated febrile illness health centre as treatment of first resort, 6% would in children choose to consult a traditional practitioner and 67% In order to provide complementary information, BHS, would go to the regional or national hospital to have level C and one level B health care facilities were vis- their child treated. ited. In each of them, the person in charge was asked about his or her knowledge of serious conditions in Time spent in the village before departing for a health children. facility and justification The time delay in the village before departing for a Features of a serious condition health facility was shorter than in Tchale and ranged The staffs in charge of health facilities were asked to de- from zero to four days. The average lapse of time re- scribe the characteristic symptoms of a serious condition ported was 1.34 days (SD = 0.91). Reasons given for the in children, apart from high fever. The respondents of delay were: the Level C health centres quoted convulsions and two of them also mentioned loss of consciousness. One doc-  To find money for consultation fees and drugs (97% tor mentioned lethargy, facial oedema, vomiting, diar- of respondents) rhoea, loss of appetite, and weight loss. The community Vermeersch et al. Malaria Journal 2014, 13:50 Page 8 of 10 http://www.malariajournal.com/content/13/1/50 Figure 2 Treatment-seeking behaviour described by the mothers in Bantandjan. health officer interviewed from the BHS mentioned in most cases fever was associated with malaria while coughing, vomiting and diarrhoea. cough and abnormal respiratory rate were associated with ARI. Differential diagnosis for malaria and acute respiratory infections Description of health care provision BHS and level C health facilities did not have the cap- When a febrile child who cannot swallow consults a acity to carry out laboratory diagnosis and the commu- BHS, the community health officer systematically refers nity health officer had no knowledge of malaria rapid him to the nearest health centre. No treatment is admin- diagnostic tests (RDTs). Only one health centre had the istered to the child before referral. In the Level C health necessary equipment to perform a blood smear. The centres unable to perform laboratory diagnosis, basic re- diagnosis of ARI or malaria was based on clinical assess- suscitation is provided and a child is then sent to the ment in the three health facilities visited. In BHS, out of nearest reference health centre. At one of those centres four health professionals interviewed, two did not know the only pre-referral treatment received by a child is di- about RDTs. Indeed, RDTs were not part of the list of azepam given by intramuscular (IM) injection in case of medical equipment provided to BHS or first level C convulsions. The nurse then refers the child, which in health centres. Each health officer had a different de- most cases means the mothers have to walk carrying scription of ARI and malaria symptomatology, although their child to the referral facility. In the other two Vermeersch et al. Malaria Journal 2014, 13:50 Page 9 of 10 http://www.malariajournal.com/content/13/1/50 centres, the treatment of a febrile child unable to swal- In Guinea-Bissau, all mothers acknowledged that chil- low usually consists of a quinine drip with 5% glucose. dren who could not swallow were in a serious condition. In one of those centres, the staff will also administer On the other hand, some mothers described such a var- paracetamol by IM injection or as an addition to the iety of signs of illnesses that there would be a real risk parenteral solution. Referral is immediate, or within an that an emergency treatment might be too late and thus hour of receiving intravenous (IV) therapy depending on not guarantee a child’s survival during transportation to the centre. In health centres managed by a doctor, the the reference health centre if its use was restricted to child is treated on site. On arrival, the child receives very severe cases. If a new treatment does not prove its symptomatic presumptive care, while waiting for the re- efficacy, there is a possibility it might not be accepted by sult of blood smear test. The treatment consists of IM the community. injections of paracetamol and/or diazepam in case of As for the treatment-seeking behaviour, 66.7% of the convulsions. If the blood smear is positive, the child is mothers interviewed in Tchale and 86.8% of those inter- treated according to the treatment protocol for severe viewed in Bantandjan declared they would go to a health malaria, i e, IV quinine with 5% glucose. If the blood structure when their child was in a serious condition. smear test is negative, the doctor tries to diagnose alter- These women were not opposed to a pre-referral emer- native pathologies depending upon the clinical presenta- gency drug treatment readily for their child. However, tion. If the child is assessed as having acute respiratory 27.8% of the mothers interviewed in Tchale and 10.5% infection, treatment is normally ampicillin IV injection of those interviewed in Bantandjan said they would for seven days. The child is put under observation in the first opt to consult a traditional healer. Tchale mothers health centre for two days. If there is no improvement selected this option for cultural reasons despite the after two days, the family is referred to Simao Mendes proximity of a level C health centre, compared to Ban- National Hospital in Bissau. tandjan. It is possible that the mothers with strong spir- itual beliefs would be more reluctant to accept the Discussion administration of the drug as the treatment of first re- The proposed strategy of introducing a combined sort. Indeed, a study conducted in Tanzania on rectal antimicrobial-anti-malarial emergency pre-referral treat- artesunate has shown that the perception of the causes ment was acceptable by health sector representatives of the diseases might influence the implementation of an and was felt to be well adapted to the context in early treatment [14]. Bantandjan women chose to con- Guinea-Bissau. This approach was also considered to fit sult a traditional healer mainly because there was no well with the current move to revitalize the most periph- level C health facility in the village and access to the first eral health facilities and would be in line with WHO rec- health centre was difficult for geographical or financial ommendations for the use of pre-referral treatment with reasons. It is anticipated that these women would not be rectal artesunate [10]. Use of the rectal route to deliver so reluctant to consult a community health worker in medicines was not perceived to present a problem and charge of dispensing a life-saving drug in the village. in fact this approach was considered to fill a gap since, Most of the Tchale mothers and all of the Bantandjan currently, there is no emergency treatment option avail- mothers making the trip to the health centre chose a able for sick febrile children in remote areas of Guinea- Level C health centre. A survey in these facilities showed Bissau. that, in most cases, the Level C health centre does not In studying the knowledge, attitudes and practices of have the capacity to perform laboratory diagnosis, or to mothers in Tchale and Bantandjan villages and health initiate appropriate drug treatment. Furthermore, most professionals in the nearest primary health care facilities, of these centres refer seriously ill children to the next fa- the objective was to describe the real conditions in cility after dispensing emergency first aid. which any future pre-referral rectal treatment would be We did not probe the fact that, by definition, pre- used. The questionnaire did not intend to investigate the referral treatment would need to be followed by referral perception of intra-rectal drug administration at the vil- to a health facility in our study. Yet, provision of rectal lage level. Nevertheless, informal discussion with the artesunate or a combination-therapy that treats both health representatives indicated that rectal administra- malaria and ARI with emergency pre-referral combin- tion of drugs was not a problem in the country. Similar ation therapy treatment, means that a suppository dis- compliance to rectal administration has been described penser in the village would have to direct the mothers in other West African countries [11,12]. However, a towards a Level A or B health centre or the hospital. study performed in Laos indicated reluctance by the Lao Consequently, mothers of treated children would need community to use the rectal route, thus highlighting the to accept the need for referral and agree to facilitate need to take in account local behaviour before introdu- transit of the child to referral facilities which are located cing a new therapy [13]. further from the village than basic health units. A study Vermeersch et al. Malaria Journal 2014, 13:50 Page 10 of 10 http://www.malariajournal.com/content/13/1/50 of the factors influencing compliance with referral advice Author details 1 after pre-referral treatment in Tanzania has shown that EA 4575 Développements Analytiques et Pharmaceutiques appliqués aux Maladies Négligées et aux Contrefaçons, Université Bordeaux Segalen, if the child presents signs of serious illness such as coma Bordeaux, France. 2Projecto de Saúde de Bandim, Bissau, Guinea-Bissau. or convulsions, transit to a referral health facility was in- 3 Faculty of Tropical Medicine, Mahidol University, 420/6 Rajvithi Road, creased three-fold [15]. Our results here show that there Bangkok 10400, Thailand. 4Centre for Vaccinology and Tropical Medicine, Churchill Hospital, Oxford OX3 7LJ, UK. 5WHO Special Programme for can be a substantial time delay before departing to a Research and Training in Tropical Diseases (TDR), United Nations health facility (average of 2.13 days in Tchale and Development Programme (UNDP)/World Bank, Geneva, Switzerland. 1.34 days in Bantandjan) because of poor transport and Received: 11 October 2013 Accepted: 21 January 2014 lack of money. Even if the treatment was available in the Published: 6 February 2014 village, mothers of sick children would still need to com- ply with referral advice- and mechanisms for facilitating References 1. Punt J, Mwangi I, McHugh K, Marsh K: Clinical overlap between malaria referral were not explored in our study. and severe pneumonia in African children in hospital. Trans R Soc Trop Med Hyg 1996, 90:658–662. 2. Mwanziva C, Shekalaghe S, Ndaro A, Mengerink B, Megiroo S, Mosha F, Conclusion Sauerwein R, Drakeley C, Gosling R, Bousema T: Overuse of artemisinin- Guinea-Bissau has the infrastructure necessary to make a combination therapy in mto wa mbu (river of mosquitoes), an area misinterpreted as high endemic for malaria. Malar J 2008, 7:232. new medicine available at community level. The condi- 3. Gomes M, Ribeiro I, Warsame M, Karunajeewa H, Petzold M: Rectal tions for introducing a new treatment were clearly laid artemisinins for malaria: a review of efficacy and safety from individual down by study respondents and related to properties of patient data in clinical studies. BMC Infect Dis 2008, 8:39. 4. Gomes A, Faiz M, Gyapong J, Warsame M, Agbenyega T, Babiker A, Baiden the medicine itself (efficacy, heat stability, cost, availability, F, Yunus E, Binka F, Clerk C, Folb P, Hassan R, Hossain M, Kimbute O, Kitua A, validation by health authorities) and its integration into Krishna S, Makasi C, Mensah N, Mrango Z, Olliaro P, Peto R, Peto T, Rahman the health system (revitalization and expansion of the net- M, Ribeiro I, Samad R, White N: Pre-referral rectal artesunate to prevent death and disability in severe malaria: a placebo-controlled trial. Lancet work of health care providers at community level, avail- 2008, 8:41. ability and training of health care workers). The ability of 5. Kallander K, Tomson G, Nsabagasani X, Sabiiti JN, Pariyo G, Peterson S: Can mothers to recognize early signs of severe illness is a key community health workers and caretakers recognise pneumonia in children? Experiences from western Uganda. Trans R Soc Trop Med Hyg factor favouring use of pre-referral treatment. On the 2006, 100:956–963. other hand, this would imply providing, in parallel, oral 6. O’Dempsey TJ, McArdle TF, Laurence BE, Lamont AC, Todd JE, Greenwood artemisinin-combination treatment for mild malaria to re- BM: Overlap in the clinical features of pneumonia and malaria in African children. Trans R Soc Trop Med Hyg 1993, 87:662–665. duce the evolution of disease to severe malaria, and early 7. Rodrigues A, Schellenberg JA, Kofoed P, Aaby P, Greenwood B: Changing diagnosis and treatment of pneumonia via oral antibiotics pattern of malaria in Bissau, Guinea Bissau. Trop Med Int Health 2008, to prevent evolution to severe pneumonia. 13:410–417. 8. United Nations Development Programme: Global Report on Human Development 2005. Available at http://hdr.undp.org/en/media/ Competing interests HDR05_fr_complete.pdf. The authors declare that they have no competing interests. 9. World Health Organization: Country health system fact sheet Guinée Bissau. ; 2006. Available at http://www.afro.who.int/home/countries/fact_sheets/ guineabissau.pdf. Authors’ contributions 10. Artesunate for the treatment of severe malaria. Available at http://archives. AV passed away on December 3, 2013. This article was seen and approved who.int/eml/expcom/expcom15/applications/formulations/artesunate.pdf. by her for initial submission to the Malaria Journal. AV and AL carried out 11. Huyghens P, Konate B, Barennes H: Hygiène et socialisation du nourisson: the study in Guinea-Bissau; AR coordinated the study in Guinea-Bissau; NJW, le lavement rectal en milieu urbain à Bobo Dioulasso. Cahiers Santé 2002, EAA, MG, PO, and PM conceived the study, participated to its design and 12:357–362. coordination, and helped draft the manuscript. All authors read and 12. Ndiaye JLA, Tine RC, Faye B, Dieye HL, Diack PA, Lameyre V, Gaye O, Sow approved the final manuscript. HD: Pilot feasibility study of an emergency paediatric kit for intra-rectal quinine administration used by the personnel of community-based Acknowledgements health care units in Senegal. Malar J 2007, 6:152. We thank all the persons who agreed to take part in this study in Bissau and 13. Inthavilay S, Franchard T, Meimei Y, Ashley EA, Barennes H: Knowledge and the region of Oio; the Guinean Ministry of Health, the representatives of acceptability of the rectal treatment route in Laos and its application for Health Departments at national and regional level, the representatives of pre-referral emergency malaria treatment. Malar J 2010, 9:342. international organisations and donors and the health professionals in the 14. Warsame M, Kimbute O, Maochinda Z, Ruddy P, Melkisedick M, Peto T, Simao Mendes National Hospital, regional hospital and health centres in Oio, Ribeiro I, Kitua A, Tomson G, Gomes M: Recognition, perceptions and and the Bandim health project who provided invaluable logistical support treatment practices for severe malaria in rural Tanzania: implications for and contacts for our study. The study was supported financially by the accessing rectal artesunate as a pre-referral. PLoS One 2007, 2:149. Welcome Trust, Grant Ref: 085242/Z/08/Z. We thank you Hubert Barennes for 15. Simba DO, Warsame M, Kimbute O, Kakoko D, Petzold M, Tomson G, Premji reviewing the manuscript. Z, Gomes M: Factors influencing adherence to referral advice following pre-referral treatment with artesunate suppositories in children in rural Tanzania. Trop Med Int Health 2009, 14:775–783. Disclaimer The opinions expressed in this paper are those of the authors and may not doi:10.1186/1475-2875-13-50 reflect those of their employing organizations. PO and MG are staff members Cite this article as: Vermeersch et al.: Introducing the concept of a new of the WHO; the authors alone are responsible for the views expressed in pre-referral treatment for severely ill febrile children at community level: this publication and they do not necessarily represent the decisions, policy a sociological approach in Guinea-Bissau. Malaria Journal 2014 13:50. or views of the WHO.