UBC Faculty Research and Publications

Patterns of treatment-seeking behaviors among caregivers of febrile young children: a Ugandan multiple… Kassam, Rosemin; Sekiwunga, Richard; MacLeod, Duncan; Tembe, Juliet; Liow, Eric Feb 16, 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12889_2016_Article_2813.pdf [ 623.95kB ]
JSON: 52383-1.0307462.json
JSON-LD: 52383-1.0307462-ld.json
RDF/XML (Pretty): 52383-1.0307462-rdf.xml
RDF/JSON: 52383-1.0307462-rdf.json
Turtle: 52383-1.0307462-turtle.txt
N-Triples: 52383-1.0307462-rdf-ntriples.txt
Original Record: 52383-1.0307462-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessPatterns of treatment-seeking behaviorsamong caregivers of febrile young children:a Ugandan multiple case studyRosemin Kassam1*, Richard Sekiwunga2, Duncan MacLeod1, Juliet Tembe3 and Eric Liow1AbstractBackground: The vast majority of malaria deaths in Uganda occur in children five and under and in rural areas.This study’s exploratory case study approach captured unique situations to illustrate special attributes and aspectsof treatment-seeking during a malaria episode.Methods: During August 2010, a qualitative exploratory study was conducted in seven of Butaleja District’s 12 sub-counties. Multiple case study methodology consisting of loosely-structured interviews were carried out with eightcaregivers of children five and under in the local dialect. Caregivers were geographically distant and not known toeach other. Interviews were translated into English and transcribed the same day. Data were analyzed usingcontent analysis.Results: Of the eight cases, children recovered fully in three instances, survived but with deficits in three, and diedin two. Common to all outcomes were (1) triggers to illness recognition, (2) similar treatment sequences and practices,(3) factors which influenced caregivers’ treatment-seeking decisions, (4) challenges encountered while seeking care atpublic health facilities, (5) cost burdens associated with managing malaria, (6) life burdens resulting from negativeoutcomes from malaria, (7) variations in caregiver knowledge about artemisinin combination therapy, and (8) varyingperspectives how malaria management could be improved.Conclusions: Despite the reality that caregivers in Butaleja District generally share similar practices, experiences andchallenges, very few children ever receive treatment in accordance with the Uganda’s national guidelines. To bringnational practice into conformance with policy, three advances must occur: (1) All key stakeholders (those affiliatedwith the formal health system - public facilities and licensed private outlets, unlicensed drug vendors, and caregivers ofyoung children) must concur on the need and the means to improve malaria management, (2) all health providers(formal and unlicensed) need to be engaged in training and certification to improve timely access to affordabletreatment irrespective of a region’s remoteness or low population density, and (3) future public health interventionsneed to improve caregivers’ capacity to take the necessary actions to best manage malaria in young children.Keywords: Malaria, Treatment-seeking, Behavior, Child, Caregiver, Uganda, Case Study, Experiences* Correspondence: rosemin.kassam@ubc.ca1School of Population and Public Health, Faculty of Medicine, University ofBritish Columbia, 2206 East Mall, Vancouver, BC V6T 1Z3, CanadaFull list of author information is available at the end of the article© 2016 Kassam et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Kassam et al. BMC Public Health  (2016) 16:160 DOI 10.1186/s12889-016-2813-7BackgroundIn 2010, there were an estimated 216 million episodes ofmalaria worldwide and 655,000 deaths [1]. Approxi-mately 81 % of these episodes and 91 % of deaths werereported in the African Region, and nearly 86 % of thedeaths were in children under five. Globally, malaria hasbeen reported as the third largest cause of death in2010, after pneumonia and diarrhoea [2]. Among thosewho survive, many are left with persistent anemia, im-paired brain function, or paralysis [3]. The highest bur-den is in children because of their lack of immunity tothe parasite [4–6]. Reducing child mortality from treat-able diseases has therefore become a global priority, withAfrican leaders at a summit in Abuja, Nigeria in 2000pledging to halve malaria mortality by 2010 and two ofthe United Nation’s Development Goals focusing on re-ducing this unacceptable high level of mortality [5, 7–9].One proposed declaration and plan of action from theAfrican summit was to ensure by 2005 at least 60 % ofthose suffering from malaria receive prompt access to af-fordable and appropriate treatment within 24 h of theonset of symptoms [8]. However, despite several large-scale efforts to disseminate effective case management,the Abuja target continues to be elusive, with over39,000 children each year dying of malaria in Uganda[10]. A large proportion of caregivers continue to delayseeking appropriate care for their children, many chil-dren are treated presumptively, and many others receiveineffective antimalarials [7, 11–15].Several key impediments have been proposed in theliterature to explain the gap between policy and practicewithin rural settings. At the service level - a large num-ber of Ugandans have limited access to publicly fundedhealth facilities, regulated pharmacies and trained healthprofessionals, thereby relying on the unregulated privatedrug delivery sector that is untrained and unlicensed[16]. The popularity of the unregulated sector is furtherheightened because of their easy access, they do not re-quire patients to wait in long queues or travel long dis-tances, and they offer a range of drugs that can bepurchased in any amount and without the need for pre-scription [15, 17, 18]. However, given that a large per-centage of drugs and/or drug dosages dispensed by suchvendors are inappropriate, their prominence presents amajor challenge [12, 19]. At the household level - thereare inappropriate self-management practices and anover-reliance on less effective antimalarials [15]. Suchpractices also influence what drug vendors sell, who oftensuccumb to selling what the customer demands [18, 20].At the community level - there are limited resources tosupport informed decision-making and demand for ap-propriate case management by caregivers [15, 16, 19].Lastly, at the government level - resources are unavailableto enforce existing drug vendor regulations, which iscompounded by the predicament that such enforcementwould create in compounding the suffering of millionswho do not have access to regulated health facilities andtrained health professionals [16, 21].A comprehensive strategy to improve caregivers’ man-agement of malaria for young children will require morethan policy changes. It will require an understanding ofsteps taken by caregivers when treating their children’smalaria, an explication of factors that predispose, enableand reinforce desirable treatment-seeking behaviors, andan understanding of elements which need to be en-hanced and supported. This in turn, will require infor-mation on caregivers’ current treatment seeking patternsand their experiences with the health system. Accord-ingly, this investigation’s case study approach capturesunique situations to illustrate special attributes and as-pects of treatment-seeking during a malaria episode,including both good practices and weaknesses in care-giving. The exploratory case studies also revealunrecognized factors influencing caregiver practices,help explain factors already identified, and determinefactors that can be used as leverage points for defining ahealth promotion program.This study’s objectives were to examine treatment-seeking patterns and to describe experiences of eightcaregivers while managing the fever episode (presumedto be malaria by the caregiver) in their child of five yearsand under. Specifically, this research explored (1) the se-quence of treatment steps taken by caregivers, (2) treat-ment options available to them when visiting sourcesexternal to their home, (3) challenges they encounteredwhen seeking treatment, (4) financial and social burdensassociated with malaria, and (5) caregivers’ knowledgeabout what is the best antimalarial for young children.The study sought to generate a broad perspective ontreatment-seeking behavior by caregivers who treatedtheir children and whose child experienced one of threedifferent outcomes: (1) the child was cured (positive out-come), (2) the child survived but experienced a perman-ent disability (negative outcome), or (3) the child died(negative outcome). This study focused on the child’smost recent febrile episode resulting in the outcome ofinterest.MethodsThis qualitative exploratory study was conducted inButaleja District, Uganda as part of a baseline assess-ment for a larger study to examine caregivers’treatment-seeking behaviors for children of five yearsand under with presumed malaria. A multiple case studymethodology was implemented over a period of twoweeks in August 2010, to understand caregivers’treatment-seeking practices – whether anticipated oremergent – and to determine factors influencing theseKassam et al. BMC Public Health  (2016) 16:160 Page 2 of 24practices. Ethics approval for the project had been previ-ously obtained from the Uganda National Council forScience and Technology, the Office of Research EthicsSimon Fraser University and the University of BritishColumbia British Columbia’s Behavioral and ResearchEthics Board.SettingThe study investigated caregivers in Butaleja District lo-cated in rural eastern Uganda approximately 38 kmsouthwest of the nearest large city Mbale and 210 kmnortheast of the capital Kampala [22]. Butaleja’s adminis-trative structure consists of 10 sub-counties (mostlyrural) and two town councils (designated urban centres).Based on the 2002 national census, the population ofButaleja District for 2010 was estimated at 206,200 [23].The predominant ethnic group is the Banyole tribe andthe predominant spoken language is Lunyole [22]. Thedistrict normally experiences two major rainfall periodsbetween May and October, although it commonly expe-riences unpredictable rainfall patterns that result in se-vere flooding creating swamps, submerging gardens,destroying roads and leaving many families homeless[24]. The district’s economy is chiefly subsistence farm-ing, with almost four-fifths of the population deriving itslivelihood from crop production [24]. Poverty is gener-ally a society-wide phenomenon, although women arelikely to be poorer than men because they lack inde-pendent sources of income and so have less access to re-sources [23].Malaria was the highest ranked cause of morbidity inthe District in the period 2007–2009, with about eight inevery 10 persons experiencing malaria/fever symptoms[23]. The public health infrastructure in Uganda is strati-fied into four levels: hospital at the district level, HealthCentre (HC) III at the sub-county level, Health Centre(HC) II at the parish level, and Health Centre (HC) Iconstituting of a network of volunteer community healthworkers (CHWs) scattered in villages across the district[25]. At the time of this study the District had one pub-lic hospital located in Busolwe town council. While itwas national policy to make cost free ACT availablefrom all levels of the health system (levels I to publichospitals), supplied as the Coartem® brand, at the timeof the study no CHWs were provided with ACTs [26].Other ACT brands could also be purchased from privateoutlets, with subsidized pricing offered only at licensedprivate outlets through the Affordable Medicines Facility– Malaria (AMFm) scheme [27]. The District howeverhad no pharmacies and only a few licensed drug shopslocated mostly in town centres and market areas. Thelargest fraction of private vendors included unlicensedprivate vendors distributed across the district who didnot have formal training in the management of malaria[28]. Table 1 summarizes public and private outletsavailable at the time of the study in the sub-counties andparishes where study participants resided.Study sampleThe target population consisted of caregivers who re-sided within Butaleja District, had at least one child fiveyears and under, who provided most of the children’sday-to-day care, were able to understand and speak thelocal dialect of Lunyole, and who were willing and ableto provide consent to participate in the study. Using pur-posive sampling, caregivers were recruited who hadunique experiences managing malaria in a child of fiveyears and under, resided in different regions of the Dis-trict, and who were willing to discuss their experiences.For all eight children, malaria as the cause for the feverepisode was established by their caregivers and/or healthproviders based on presumptive diagnosis. As part of anearlier household survey process, the research team hadidentified households with caregivers who appearedeager to share their treatment-seeking experiences alongwith an annotation of whether their child had recovereduneventfully from a febrile episode or had experienced anegative outcome. At the time of the survey, these care-givers had also been asked if: (1) they would agree to becontacted again for a more in-depth discussion abouttheir experiences with managing malaria in one of theiryoungest children, and (2) for their permission to sharetheir name and contact information with the study’s re-cruitment team. A list of potential caregivers for the casestudy was then created with caregivers classified accord-ing to their experiences: (1) those who sought treatmentexternal to their home and the child improved or thosewho did not seek external care and the child survived,(2) those who sought treatment external to their homebut the febrile illness persisted resulting in some form ofdisability, (3) those who sought treatment external totheir home but the treatment failed and the child died.Names of potential caregivers from each of the categor-ies were written on slips of paper, folded, put it in a con-tainer and drawn randomly using a ballot process.Caregivers were subsequently contacted by the studyteam and invited to participate. The first eight consent-ing caregivers became the case study participants.Data collectionCase study interviews were carried out by two trainedresearch assistants recruited from the District and fluentin verbal and written English as well as Lunyole. Eachresearch assistant was responsible for conducting and re-cording the interview. The study team traveled to the in-dividual caregivers’ homes to conduct the interviews.While an interview guide was provided to the researchassistants for areas to explore, the interviews wereKassam et al. BMC Public Health  (2016) 16:160 Page 3 of 24Table 1 Distribution of public health and private outlets in sub-counties (parishes) where cases residedCases (CS):sub-county (parish)→Outlet type ↓CS 01:Himutu(Kaiti)CS 02:Busaba(Buwihula)CS 03:Nawanjofu(Bingo)CS 04:Naweyo(Kaiti)CS 05:Butaleja(Nakwasi)CS 06:Naweyo(Kaiti)CS 07:Naweyo(Kaiti)CS 08:Busabi(Malangha)Health centre III – – 1 – 1a – – –Health centre II – – 1 – 1 – – 1aPrivate clinic – – – – – – – –Pharmacies – – – – – – – –Drug shop (Licensed) – 2 – – – – – –Drug shop (Unlicensed) – 4a 3 5b – 5a 5b –General shop (Unlicensed) – – – 1 – 1 1 –Market stall (Unlicensed) – – – – – – – –Kiosk (Unlicensed) – – – – – – – –Mobile vendor (Unlicensed) – – 1 – – – – –Total outlets 6 6 6 2 6 6 1aOne drug shop located in the village where caregiver residedbThree drug shops located in the village where caregiver residedKassametal.BMCPublicHealth (2016) 16:160 Page4of24largely unstructured and pursued information providedby caregivers rather than adhering to a rigid protocol.Research assistants offered occasional prompts to keepthe conversations flowing about the child’s history pre-ceding the fever episode, signs and symptoms during thefever episode, treatment they sought – where and why,challenges they faced, length of the episode, child’s statusfollowing the episode, possible improvements to caregiv-ing; nevertheless, each case study was a conversation ra-ther than a structured interview. Data from case studieswere collected in the local language over two to fourvisits, each lasting no more than one hour. BecauseLunyole is not a written language, conversation was ver-batim translated into English by the research assistantsas they took interview notes. To ensure quality and ac-curacy, all conversations were audiotaped with the per-mission of all participants. Components of theconversation dealing with medications used during thechild’s treatment were verified by asking caregivers toidentify photos of each drug or medicine from a poster-sized collection of images of known drugs and their vari-ous brand names thus minimizing problems of recall orunknown drug names. At the end of each day, the re-search assistants met with a senior member of the re-search team to debrief and share critical perspectives,and all transcriptions were reviewed by the team within24 h of completion.AnalysisQualitative data analysis techniques were used to analyzethe data. All caregivers’ responses were transcribed ontoMicrosoft Word documents, formatted, and entered intoa qualitative data analysis computer program (QDAMiner) for subsequent coding, thematic analysis, and in-terpretation. All transcripts were reviewed and codedusing a constant comparative method. As a first step,three of the authors independently reviewed all tran-scripts line-by-line identifying both anticipated andemergent themes. The themes were then discussed, de-bated, and where necessary relabeled or regrouped intonew themes using open, axial, and select coding tech-niques [29]. Once the labeling was completed, thethemes were grouped into categories and analyzed con-sistent with the study’s overall objectives for specific in-dicators, patterns, and trends in both anticipated andemergent outcomes. Qualitative research issues of trust-worthiness and replicability (validity and reliability) wereverified by examining the number of times each themewas mentioned by different caregivers – either in the af-firmative or negative. Furthermore, issues of trustworthi-ness and confirmability were assured by using multipleinvestigators in the analysis process and with repeatedcycles of theme identification, definition, and reconcili-ation until all research team members concurred on thecoding system and ultimately agreed on which utterancesrepresented which codes [30]. Subsequently, to enablecomparison across cases, select data was rendered intotables to illustrate the sequence of treatment steps takenby caregivers, their experiences with external sources, andtheir knowledge about antimalarial medicines.ResultsEight caregivers (seven female and one male) partici-pated in the case studies (CS). Children generally experi-enced one of three outcomes from their febrile illnessand treatment received: (1) a positive outcome wherethe child improved (CS 06, 07, 08), (2) a negative out-come where the child survived but experienced an irre-versible disability (CS: 01, 02, 03), or (3) a negativeoutcome where the child died (CS: 04, 05). Caregivers’and their index children’s demographic characteristicsare summarized in Table 2. Caregivers’ ages ranged from22 to 45 years, and their index children’s ages rangedfrom two months to four years at the time of the febrileepisode. All caregivers were peasant farmers living inmonogamous households. Two of the caregivers hadcompleted some level of post-secondary education, fivehad some level of primary schooling, and one had noformal education. Five of the caregivers reported beingprotestant, one was Muslim, and two made no mentionof their religious affiliations. The eight caregivers repre-sented seven of the 10 sub-counties in Butaleja District.Case study summaries (Additional files: Table S1 andTable S2)Case study (CS) 01Date of episode not discussed). At 10 pm, the caregivernoticed that her child was feverish and convulsing. Sheinitiated home management with Coartem®, Panadol®and physical supportive therapy using a wet cloth to de-crease the fever before taking the child to a public healthfacility the following morning. The public facility(Kangalaba HC III) offered no treatment but instead re-ferred the child to a hospital in the neighboring districtof Mbale for blood transfusion. At the hospital, the childwas admitted and given six quinine injections over threedays and two-week worth of quinine syrup. Post-discharge the child’s vision worsened and the left handbecame lame. Over the next month, the caregiver tookthe child to visit optometrists and doctors in the districtof Tororo and was told the child had received too muchquinine and nothing could be done to help the child.Case study (CS) 02(Episode two years prior interview). At 4 am, the care-giver noticed her child was ill with fever and convulsing.The caregiver initiated home management with trad-itional remedy (smearing pounded onions of the child’sKassam et al. BMC Public Health  (2016) 16:160 Page 5 of 24body and giving the child to drink some), and then at10 am took the child to a public health facility (BusabaHC III). At the public facility the child was given a bloodtest, Septrin®, Panadol®, and quinine tablets, and was toldto buy additional quinine syrup from a private drugvendor. Because the child continued to convulse, thecaregiver continued to give quinine syrup daily for abouta year, while periodically returning to the same public fa-cility as well as visiting the local drug shop to find somerelief for her child’s convulsions. However, no furthertests or different treatments were offered. When thechild suddenly stopped speaking she stopped the quinineand reverted back to using traditional remedies (bathingin mululusa and tomato leaves) for a period of onemonth. She eventually stopped using the traditionalremedies when the child did not respond and, a yearlater, at the suggestion of a friend, she inquired aboutanticonvulsant tablets at a HC III and got referred to themain hospital where the child received additional quin-ine and convulsion tablets. The child now receives anti-convulsant tablets daily which the parents obtain fromthe hospital on a weekly basis. While the frequency ofconvulsions decreased, the child lost the ability to speak.Case study (CS) 03(Episode two months prior interview). At 5 pm, thecaregiver noticed her child had a swollen stomach andwas feverish, vomiting and convulsing. She initiatedhome management with tradition remedy (smearingpounded onions on the child’s body) to lower the child’sfever, and at 10 am the next day visited a private drugshop in Busolwe. The drug vendor assessed the child tohave cerebral malaria and gave the child an injectionand various tablets including Panadol®. The caregiverreturned the next day for the same medicines. Thechild’s fever gradually improved after two-weeks, but tothis day continues to experience convulsions and a swol-len stomach.Case study (CS) 04(Episode five years prior interview). At 4 pm, the care-giver noticed her child was ill with fever and short ofbreath. She initiated home management with Panadol®,Septrin® and supportive therapy using a wet cloth. Thefollowing day at 7 am when she saw the child’s conditionhad worsened, she took the child to a public health facil-ity (Nabiganda HC II). No treatment was given at the fa-cility, instead the child was referred to a hospital in theneighboring district of Mbale for blood transfusionwhich took two and a half hours to reach. When thechild arrived at Mbale the doctors ordered blood tests,but during the three hours while the tests were beingrun the child died.Case study (CS) 05(Episode five months prior interview). At 5 pm, the care-giver noticed her child was ill with fever, vomiting,coughing and experiencing diarrhea. She initiated homemanagement with supportive therapy using a wet clothto cool the child’s fever before visiting a public health fa-cility at 6 pm (Nakwasi HC III). At the health facility thechild received Coartem® and was told to buy quinineTable 2 Caregivers’ and index child’s demographic characteristicsCases→Demographics ↓CS 01 CS 02 CS 03 CS 04 CS 05 CS 06 CS 07 CS 08Household InformationSub-county Himutu Busaba Nawanjofu Naweyo Butaleja Naweyo Naweyo BusabiParish Kaiti Buwihula Bingo Kaiti Nakwasi Kaiti Kaiti MalanghaCaregivers’ Demographic CharacteristicsAge 30 years 28 years 42 years 43 years 22 years 30 years 45 years 25 yearsEducation PartialprimaryPartialsecondaryPartialprimaryPartialprimaryPartialprimaryNone PartialsecondaryPartialprimaryOccupation Farmer Farmer Farmer Farmer Farmer Farmer Farmer FarmerReligion Protestant Protestant Protestant Not Available Not Available Muslim Protestant ProtestantMarriage status YesMonogamyYesMonogamyYesMonogamyYesMonogamyYesMonogamyYesPolygamyYesMonogamyYesMonogamySpouse’soccupationFarmer Mechanic Farmer Farmer Farmer Farmer Farmer TeacherNo. Children 5 2 9 7 4 6 6 5Index Child’s Demographic InformationGender Male Female Female Male Female Male Male MaleAge 3 years 4 years 3 years 1 year 2 years 2 months 1 year 1 yearKassam et al. BMC Public Health  (2016) 16:160 Page 6 of 24syrup from a drug shop. Despite treatment, the childcontinued to worsen. The caregiver returned to the HCtwice over the subsequent two months without any im-provement. Three months later, the caregiver took thechild to a private drug vendor because the public facilitywas closed. There the child was given tablets to treatworms, the child appeared to improve after the initialdose but quickly turned for the worst and died.Case study (CS) 06(Episode two months prior interview). In the morningthe caregiver noticed that her child was short of breathbut decided to do nothing until the next morning whenthe child’s condition worsened, at which time she visiteda private drug shop. The drug vendor gave the childAspirin© for three days which the caregiver adminis-tered. No additional medicines were purchased becausethe household could not afford to buy more medicines.The caregiver said it took two weeks for the fever to re-solve on its own.Case study (CS) 07(Episode two months prior interview). At 9 am, whenthe caregiver noticed the child was ill with fever, chills,and vomiting, he initiated home management with Pana-dol®. Later the same day when there was no improve-ment the child was taken to a private drug shop. Thedrug vendor took a blood test, gave the child two quin-ine injections, and the child gradually improved over afew days eventually returning to normal.Case study (CS) 08(Episode two months prior interview). At 7 am, thecaregiver noticed her child was feverish and not beingher usual self. She initiated home management withCoartem®, Panadol® and mululusa, and used supportivetherapy with wet cloth to reduce the fever, but the child’sconditioned continued to worsen. At 10 am the nextday, the child was taken to a public health facility(Muhuyu HC II) where the child was given moreCoartem®, Panadol®, some kind of injection and ReneColdease®. The child’s illness was cured in 3–4 days.Case study findingsEight broad themes summarize caregivers’ treatmentseeking patterns for the reported febrile episodes, theirexperiences, what they knew and what they learned.These include: (1) how caregivers’ recognized theirchild’s illness, (2) the sequence of treatments sourcesand treatment practices involved, (3) factors which influ-enced caregivers’ treatment-seeking decisions, (4) chal-lenges they encountered with seeking care at publichealth facilities, (5) the burden of cost associated withmanaging malaria, (6) the burden of living with anegative outcome from malaria, (7) caregivers’ know-ledge about ACT, and (8) caregivers’ perspectives of howmanagement of malaria could be improved. Quotes thatbest illustrate the above themes and represent viewsfrom all three outcomes of interest are presented.Additional file 1: Table S1 illustrates the sequence ofcare received by the child, the initial symptoms and thetime of day when symptoms were first noticed, the dif-ferent external sources visited and treatments receivedduring these occurrences, and outcome(s) at the end ofeach treatment phase. Additional file 2: Table S2 outlinescaregivers’ treatment-seeking experiences when seekingtreatment from an external source for the episode offever in question. Tables provide a “running description”of history, treatments, caregiver experiences, and subse-quent reflection. Entries in the tables are arranged inorder of the episode’s outcome: child recovered, childsurvived but with deficits, or the child died.Recognition of malaria“Recognition” examines how caregivers identified their childwas suffering from malaria. It captures the most commonsymptoms recognized, the time of day when symptomswere first noticed, and how this affected caregivers’ treat-ment–seeking decisions (Additional file 1: Table S1).Symptoms and time of recognitionWith the exception of one caregiver, “fever” or “hot body”was the most common symptom mentioned, followed bycoughing/congestion observed in cases that had positive out-comes as well as ones where the child died (CS: 04, 05, 06,08). Convulsion was mentioned only by those caregiverswhose child had survived with a negative outcome (CS: 01,02 and 03), and it commonly accompanied fever. Some ofthe other primary symptoms mentioned included “red eyes”,“diarrhea”, “swelling of the body”, “pain within body”, “thechild not eating”, “vomiting”, and “making excessive noise”.The responses from the case studies showed that for allthree children with positive outcomes (CS: 06, 07, 08), theirsymptoms were noticed early in the morning. Conversely,for those children who died or survived with a negative out-come, their caregiver first noticed their symptoms at night.Treatment actionsThe steps taken by caregivers to treat their children oncethe initial symptoms were noticed were termed TreatmentActions. Included in this theme are the different steps takenby caregivers, treatment sources accessed, types of treat-ments received, and outcome at the end of each treatmentoccurrence (Additional file 1: Table S1). The multiple stepstaken by caregivers represent treatment failures and at-tempts to seek additional care to resolve the febrile episode,as well as manage adverse outcomes from the illness. Thetreatment sources accessed were classified as either homeKassam et al. BMC Public Health  (2016) 16:160 Page 7 of 24management or external treatment which typically in-cluded a public health facility or a private drugvendor. Home management was defined as any careinitiated from the home setting with resources withinthe household, such as physical supportive care, west-ern medicines, and/or traditional remedies. Time toseeking external treatment or where treatment wassought appears to have had little influence on theoutcome of the illness. Treatment at public health fa-cilities was rarely successful, with four of the fivechildren who visited a public health facility as part oftheir first external care experiencing a negative out-come. Of the two children who died, one died at areferral hospital and the other at home after severalvisits to a public health facility with repeat treatmentfailure.First actionAll but one caregiver (CS: 06) initiated treatment withsome form of home management soon after noticingfever. The one caregiver (CS: 06) who did not initiatehome management, opted to simply wait and observethe child for the first day, and initiated the first treat-ment on the next day with Junior Aspirin® obtainedfrom a private drug vendor when the child’s symp-toms worsened. Across the seven cases, three types ofhome management treatments were seen: westernmedicines, physical supportive therapy, and traditionalremedies. Examples of physical supportive care in-cluded bathing the child in cool water, washing thechild with a cool cloth or sponging the child to re-lieve fever. Western remedies used in home manage-ment included Coartem®, Panadol®, Septrin®, andquinine syrup. Traditional remedies included smearingonion paste over the ill child or sitting the child in abath of herbs such as mululusa.Western medicines were the most popular homemanagement, with all positive outcome cases (CS: 07,08) and half of the negative outcome cases (CS: 01,02, 04) receiving some type of medicine. There is nostrong pattern separating the cases with positive andnegative outcomes in terms of which western medi-cines were given. Of the three children given an anti-malarial, two received Coartem® (one with a positiveone with a negative outcome) and one with a negativeoutcome received quinine. On the other hand, sup-portive care only was used mostly by those who expe-rienced a negative outcome (CS: 01, 03, 04, 05), withonly one positive outcome case reported to have usedsupportive care (CS: 08). Children were given sup-portive care with/out western medicines. Traditionalremedies were only used by three caregivers (CS: 02,03, 08) whose children later went on to experience anegative outcome, only one used it in combinationwith a western medicine (quinine).Just over half of the caregivers representing childrenwith a negative outcome (CS: 01, 02, 03, 05) reportedinitiating home management only as a “first aid”measure, with the intent to seek subsequent treat-ment from an external source first thing in the morn-ing. Whereas the remaining caregivers (CS: 04, 07,08) indicated they decided to seek additional carefrom an external source only after noticing theirchild’s condition had worsened.I had Panadol® and I gave her, and I still saw that hercondition was still bad. I tried to take her to the healthunit. (CS: 07)I did not go to the health unit on that very daybecause I had trusted that bathing him with mululusaand giving him Coartem® and Panadol® would help.When I got to know that the fever had worsened, whenit dawned to morning, I took him to the health unit.(CS: 08)Caregivers started treatment at home because theyhad western medicines at home and trusted that theywould cure the child:I knew that this [Panadol®] and this [Septrin®] treats… fever.… In the health unit, the health providersteach us [to] use [Panadol® and Septrin®]. (CS: 04)I know that Panadol® could reduce a bit the pain inthe body…. The health provider told me, that this one[Panadol®] reduces painful body. (CS: 07)Second and subsequent actionsAdditional file 1: Table S1 shows that all second andmost subsequent actions took place externally andoutside the home for all caregivers. Case 02 was theexception, as she reverted from visiting externalsources and obtaining quinine treatment during sub-sequent actions to using traditional remedies fromhome when her child stopped speaking and she foundwestern medicines were not helping. Three of thecaregivers (CS: 02, 05, 07) reported visiting an exter-nal source within the same day, whereas theremaining caregivers went the next day. However, allcaregivers whose child experienced a negative out-come reported seeking care within 24 h of noticinginitial symptoms. The most common second actionKassam et al. BMC Public Health  (2016) 16:160 Page 8 of 24(and first external source) for five of the caregiversinvolved going to a public health facility, with fouropting for the nearest facility (CS: 02, 04, 05, 08). Ofthe cases with a positive outcome, only one (CS: 08)was taken to a public health facility, the other two re-ceived treatment at private drug shops. On the otherhand, all but one case with a negative outcome (CS:01, 02, 04, 05) sought their first external treatmentfrom a public health facility. All children who weresubsequently referred to higher level health facility orthose who required further treatment experienced anegative outcome. The most common medicines givenor prescribed by external outlets were Panadol® tabletsand/or some form of quinine. Only two children re-ceived Coartem® and both were at a public facility.Additional file 2: Table S2 illustrates caregivers’ expe-riences when seeking external care and whether ornot they accessed their nearest public health facility.Treatment in an ensuing malaria episodeWhile the intent of this study was to focus principallyon the single febrile episode leading to the outcomein question, some of the illnesses never subsided andchildren either continued to experience mild fever orexperienced frequent recurrent infections, which madeit difficult to separate the original episode from newinfections. As such, half of the cases (CS: 03, 06, 07,08) mentioned their treatment-seeking experienceswith a subsequent febrile episode as well. Two care-givers (CS: 03, 07) sought external care from a drugshop as they had done in the first episode, but onecaregiver (CS: 08) opted for a drug shop over publichealth facility in the subsequent episode. One case(CS: 06) continued to take no action since the house-hold “had no money to buy medicines”.Factors affecting decisionsDecision factors examined how caregivers’ previousexperiences and advice from family members influ-enced their decision about drug used and sourcesaccessed. All cases shared some examples of howtheir previous experiences had influenced their prefer-ence for certain drugs and/or treatment sources.Past experiences influencing drug usePrevious experience(s) with a western medicine wasan important factor influencing caregivers’ decisionsfor the current febrile episode, as were experienceswith current episodes at influencing future decisions.We found cases with positive outcomes (CS: 06, 07,08) were more likely to share both good and badpast experiences with drug use influencing currentdecisions, whereas those who had experienced anegative outcome (CS: 02, 03, 04) shared only goodexperiences with drug use. The following quotes il-lustrate why caregivers favored specific medicines.Okay I get him those medicines [Panadol® and Septrin®]because … they are the ones that I had [in the home]because I normally use and they work. (CS: 04)I normally buy Panadol®, it's the one that I usuallygive them and they swallow…. At times I give themPanadol® and they get cured well. (CS: 06)The fever [malaria] can match with quinine [usesquinine to treat]. (CS: 07)I knew it because whenever I go with my child tothe health unit, I am first given Panadol® andCoartem®. So, if I would pack some [Panadol® andCoartem®] … I can use them in case of a malariaattack…. From personal experience, whenever I usePanadol®, it relieves me of the fever and hightemperature. (CS: 08)Similarly, caregivers avoided the use of certainmedicines because of previous undesirable experi-ences with certain medicines. The following quotesare examples of some of these experiences:I used to see, if the child was just deteriorating … Iwould give [Coartem®]. Then I saw … it [Coartem®]… did not cure, did not cure at all, then Iabandoned it [Coartem®] … without seeing anychange…. (CS: 06)We used to use chloroquine long time ago butwhenever I would use it, it would itch my body, mywhole body. So I no longer use that chloroquine.(CS: 08)Past experiences influencing source of careCaregivers’ previous experiences with public healthfacilities and private outlets also contributed to theirdecision about which external source to use. Withrespect to negative experiences at public health facil-ities, caregivers discussed examples such as stock-outs or treatment failure influencing their decision to notKassam et al. BMC Public Health  (2016) 16:160 Page 9 of 24return to that facility. Such poor experiences were not exclu-sive to cases with positive or negative outcomes (CS: 01, 03,04, 06, and 07). As illustrated in Additional file 1: Table S1and Additional file 2: Table S2, caregivers still visited publicfacilities despite their frustrations with services at public facil-ities (CS: 01, 04).[Speaking of an earlier fever episode with anotherchild] On that particular day medicines had gotfinished from public health facilities [Namulo HC IIand Kangalaba HC III], so that is when I took thechild to the drug shop. But even when she treated,the child did not cure. So, I wait till they broughtmedicines to Nalusaga [Kangalaba HC III]. That iswhy I took [this] child to Nalusaga [Kangalaba HCIII] for treatment since he [other child] had failedto cure with the treatment given to him by the drugshop. At Nalusaga [Kangalaba HC III], that iswhere the child got cured.… [When] I took my childto … Namulo HC II … he failed to improve so Istopped going there. (CS: 01)The reason why I decided to go to the private [drugshop] is because I lost one of my children from thatBusolwe hospital. (CS: 03)They [drug vendors] provide for us with medicine,because at times we go to the government health units[and] medicines are not there, that's why we go to theclinics [drug shops]. (CS: 04)In Naweyo [HC] when you go, when you go at times,you find that there is no health provider, and at timeswhen you find the health providers, he writes for you[prescribes] and at times tells you that there is nomedicines then you come back empty-handed or withnothing. (CS: 07)Advice influencing treatment sourcesAdvice from family or friends also influenced care-givers’ decision of where to seek external treatment.All caregivers cited receiving such advice either dur-ing the initial phase of the illness or during subse-quent actions when the child had not improved.Most caregivers, three of the five with negative out-comes (CS: 02, 03, and 05) and all with positivecases (CS: 06, 07, 08), mentioned taking the advicegiven to them. Comparing the quotations with care-givers’ actions in Additional file 1: Table S1 confirmsthat such advice was followed.… [in] 2010 … we were told [by friends] about tabletsthat treat convulsions [from fever], that those tabletswere found in Busolwe [hospital]. So, that is when wewent to Busolwe to obtain those tablets. (CS: 02)My mother told me not to dress up the child in anyclothes [because of fever] but to wrap the child in ababy shawl and proceed directly to the health unit[drug shop]. When he convulsed, I ran very fast and[wrapped] wet cloth over his body. I then looked for abicycle and rushed him to a [drug shop] in Busolwe.(CS: 03)My husband used to tell me that we take the child tothe clinic [HC]. Now, when the illness attacks the childand it was severe … [I] discuss with my husband and… go to the clinic [HC]. (CS: 05)Now, the person [caregivers’ sister] tells me that takethe child to the health unit [drug shop] and the childwill cure of fever [malaria]. (CS: 06)So when … I knew that it was fever and I called thechild’s mother [the wife], the mother also touched,touched her and saw that she was badly off and I toldher we take the child to the health unit [drug shop]maybe the health provider can help us. (CS: 07)He [the husband] tells me that take the child to healthunit [HC]. (CS: 08)Advice given to caregivers influencing treatment choicesAdvice from family and friends also influenced caregivers’initial treatment choices for their child. Three of the cases,representing both positive and negative outcomes (CS: 02,03, 08), who were given advice about which source to ac-cess also spoke about receiving advice on how to treattheir child. The only three cases to use traditional remed-ies (CS: 02, 03, 08) stated that they had done so becauseon advice received by a friend, mother, or grandmother.Kassam et al. BMC Public Health  (2016) 16:160 Page 10 of 24Only one caregiver (CS: 02) reported using a westernmedicine (quinine) at the advice of people she met on theroad and at boreholes. Others used western medicineswhich had previously been recommended to them byhealth providers.There is a friend of mine who had a child who also usedto convulse [from fever], so that friend of mine is the onethat advised me to use the onion [so] I pounded theonions, had her drink some and smeared the rest on herbody…. I would narrate [tell] the child's illness to her[my friend] and she would tell you what to do. (CS: 02)Others [people from the village who she met on theroad and at boreholes] would tell me that “I also hada child that used to convulse like yours, I continued togive her that quinine syrup until she got fine/cured”….So, we used to give her the quinine.… We continuedbuying and giving her that quinine syrup. They [peoplefrom the village] gave us advice to continue giving herthat quinine syrup so whenever the syrup would getfinished, we would still buy more quinine syrup andgive it to her. (CS: 02)I got a wet cloth and pass it over his body again andagain. [Then] my mother and mother-in-law said “ifyou pound onion and smear it all over his body, be-cause it could be evil spirits [causing the convulsions]and the evil spirits can smell the onions and they runoff. [So] I got onions and smear them all over his body.(CS: 03)I was at my parents' home when the child fell ill. I hadgone to mourn a dead relative when the child was ill.So at the funeral, my grandmother advised me to usemululusa. (CS: 08)Caregivers also relied on advice given by health pro-viders during previous interactions. Two of the cases withpositive results (CS: 07, 08) and one whose child had anegative outcome (CS: 04) initially managed their childusing home management with a western medicine basedon advice received from health professions during a previ-ous interactions. Although only one caregiver (CS: 08) in-dicated she had been told been told by a health providerto use Coartem® when she herself got sick from malaria,so she believed it to be appropriate for her child. The ad-vice for cases 04 and 07 did not follow the governmentrecommendations for treating malaria (Additional file 1:Table S1).In the health unit, the health providers teach us to usethis [Panadol®] and this [Septrin®]. (CS: 04)The health provider [Drug Shop Vendor] told me, thatthis one [Panadol®] reduces painful body. (CS: 07)I learnt [about Coartem®] because I also normally fallsick. So, when I fall sick and go to a health providerand explain to her about my illness, she tells me to goand swallow Coartem®, Panadol® and the other tabletsfor coughing whose name I do not remember. She tellsyou that “for every tablet … you will notice the bodyheat and fever being relieved”. So on that first daywhen I noticed [child had fever] I gave him theCoartem®. (CS: 08)Challenges while seeking care at public health facilitiesTreatment-seeking at public health facilities was asso-ciated with several challenges, with all caregiversmentioning one or more of these challenges. Chal-lenges made up the largest part of the case study dis-cussions and were not always specific to the currentfebrile episode in question. Despite the many chal-lenges expressed by caregivers, a large majority tooktheir child to public facilities on multiple occasionswhen their child did not improve. The most fre-quently voiced challenges included getting to a publichealth facility and the quality of service once at thepublic facility. Examples of challenges encounteredonce at public facilities included regular medicationstock-outs, getting the runaround by the health staff,incivility of the staff, and the cost of medicines(Additional file 2: Table S2).Access to public health facilitiesLack of access to public health facilities resultingfrom long geographic distances was a major concernfor all caregivers as most often had to walk on footto get there. Caregivers’ frustration with access is il-lustrated in the following three quotes:The challenge I face was to be referred to Businghu[Busiu Hospital in Mbale] yet [because] I did nothave the means to get there. So I kept thinkingabout what to do. So then, I asked my husbandabout what we could do…. Transportation fromKassam et al. BMC Public Health  (2016) 16:160 Page 11 of 24Nalusaga [Kangalaba HC III] to Businghu costs5,000 UGX using a taxi…. So it was my husbandthat found a way out. I guess he borrowed moneyfrom a friend. So with that money we went toBusinghu. (CS: 01)We walk [7 miles to Busolwe], we walk with our feet,that is when we have failed to get money. At times, weride on the bicycle but when we fail to use the bicycle,we walk on foot to the hospital to get the medicines.(CS: 02)It is there the clinic [drug shop] that was near, theother one [the HC] is far. And also I was challenged inform of money to go to the health unit. (CS: 07)Availability of health providersFurther complicating access to treatment was the un-availability of health providers when caregivers arrived atpublic facilities. Challenges of low staffing or absentee-ism were experienced by several of the caregivers (CS:02, 03, 04, 05, 07, 08).The challenge I faced is that I got there before thehealth providers. The health providers were not atthe health unit by the time I got there. It was openand there was an askari [watchman] but the healthproviders had not yet come. That is the challengethat I faced. But as soon as they [the healthproviders] got there [30 minutes later], they workedon my case. (CS: 02)Our health unit has only one health provider, I wentthere around 10:00 am and by 2:00 pm my child hadnot received any medical attention…. [Eventually] Shetouched and felt the boy's body heat with the back ofher hand so with that, I guess she recognized what sheshould do with the child and the medicines to give fortreatment. (CS: 08)Access to medicineOnce at a public health facility, a primary complaintvoiced by six of the caregivers was unavailability of med-ications (commonly referred to as “stock-outs”) at theLevel II and level III public facilities (CS: 01, 02, 04, 05,06, 07). Additional file 2: Table S2 shows that blood testswere rarely taken, but when performed at all, weremostly available at higher level public health facilities,thus limiting the spectrum of care available at lowerlevel public facilities.There are those [health facilities] that are nearby but Idon’t go … because at times the medicines are finished… so you say let me obtain help from the other HC.[Health providers at Nalusaga: Kangalaba HC III]told me that for us here, we do not have medicines foryour child but you go to Businghu [Busiu hospital inneighbouring Mbale District], so then she wrote for mea reference letter and I went to Businghu. (CS: 01)I found challenges in mostly when I went and foundno medicines there. I used to take her to the clinic[HC] when the illness is severe…. (CS: 05)… at times when you find the health providers, hewrites for you [prescribes] and at times tells you thatthere is no medicines then you come back empty-handed with nothing. (CS: 07)Getting the runaroundParallel to the issue of stock-outs is the “runaround”given to caregivers when a public health facility is out ofmedicines. In circumstances when public facilities areexperiencing stock-outs, caregivers are either referred toanother public facility to manage the child or they aresent to a private drug vendor with a prescription to pur-chase their medicines, resulting in further delay in treat-ment. Caregivers whose child experienced a negativeoutcome were those who most often described gettingthe runaround. (CS: 01, 02, 03, 04, 05). Case 01 providesa good instance of a caregiver being sent to multiple lo-cations, first because of stock-out at a nearby facility,then because of suffering an irreversible adverse out-come from over-medication. In the end the caregiver istold there was nothing that could be done for her child.It started with fever and then the child started toconvulse, so I took him to health unit at Nalusaaga[Kangalaba HC III]…. They referred me to Busiuhospital [in neighbouring Mbale District], so I tookhim to Busiu hospital … and they told me to go buywater for the drip. After they discharged us, the childshowed improvement but then the child’s sight becamepoor, child could no longer see property so they told usthat "maybe the quinine that was given to the childwas too much"…. So then we took him to Tororothinking that maybe the eyes had a problem, thatKassam et al. BMC Public Health  (2016) 16:160 Page 12 of 24maybe the eyes were sick. We went to Tororo to check/examine the eyes and the eyes did not have anyinfection and they told us that maybe the medicinewas too much, the quinine that they injected the childwith was too much…. [As] the child was curing … thishand [holds the boys left hand that is lame] becamelame. And even now, something attacks the brain andthe brain stops working well … and the child fallsdown.... So, then, that is how he is..... (CS: 01)I went to Budumba HC III people told me that "thereare health providers who occasionally go there. Theytreat children that convulse. They have medicines thatthey give out freely". So with that, I went there. I wentthere with my daughter who is ill and I spoke to thehealth provider who told me that those have providersthat treat convulsing children no longer went toBudumba HC III so he told me to go to Busolwe. Hetold me to go to Busolwe hospital so that is when Istarted to go to Busolwe. (CS: 02)At Nabiganda [HC II], we did not get any treatment.When we reached, they [the health providers] told us that"we are not going to treat the child". Take him to Mbale[hospital]…. They said that … the blood is finished [childis anemic] so you go to Mbale. When I reached the healthunit [Mbale hospital] they told us to buy a … set and acannula … [so can draw] the child’s blood … theyremoved blood from him and they told us we take it andtest it [get blood cross-matched at laboratory] so that theyget [right] blood [and] they can give him. (CS: 04)Negative experiences at public health facilitiesAll caregivers expressed some level of discontent withthe level of services they received, as well as with howthey were treated by health providers at public health fa-cilities. Once again, these complaints were more pro-nounced among caregivers whose child experienced anegative outcome compared to those whose child had apositive outcome. The complaints commonly involvedfailure to receive appropriate services and lack of atten-tion at health facilities.When they tested [the child’s blood], they did not tell methe illness that my child was suffering from. (CS: 01)That unit at Bingo HC II, when I go there, I am givenonly Panadol®, they do not examine the child, they donot even measure the child's temperature…. They[health professionals at Bingo HC II] do not haveenough capacity to handle such a condition of a childso that is why they refer you to Busolwe mainhospital…. Since we smeared him with onions, wecould not dress him up, but still the truth is that hehad no clothes of his own…. The health provider saidthat "why have you brought the child when he is naked”[the health provider at the government facility wasangry that the child was not clothed]. (CS: 03)On reaching [Mbale, since] 11 o'clock they had notgiven us [the child] blood, the child at 1 PM he died.(CS: 04)The health provider just looks at you and he feelsproud … [like] he own things, goes and comes back.Even if you go, there is nothing you get, the healthproviders are proud.… So there is nothing you get, theway you have gone, is the way you come back [child isignored and not treated]. (CS: 06)The health provider at times is busy with his work,can take long or he is having other patients. So there,we face also a challenge there. (CS: 07)Burden of costThe cost burden associated with treating malaria in theiryoung children was a significant concern for all caregiversin this study. Caregivers reported incurring a variety ofcosts during the process of seeking treatment, includingcosts associated with: having to purchase medicationsfrom private outlets during periods of stock-outs, takingtransportation to reach a public facility, managing a childfebrile episode when there is treatment failure or treat-ment (or lack of) results in negative outcomes.The cost of medicinesMedication stock-outs at public facilities inevitably re-quired caregivers to purchase medications from privatevendors for the health provider at the facility to adminis-ter or the caregiver to give to their child at home. Thisresulted in significant medication costs to households,thus negating the existing national treatment policy re-quiring children five and under seen at public facilitiesbe treated for free. All caregivers in this study mentionedexperiencing such medication costs.Kassam et al. BMC Public Health  (2016) 16:160 Page 13 of 24Because there was no quinine syrup at Busaba HC III,they told me to go and buy from anywhere. So I wentto that drug shop and I asked if they had the syrupand they told me that they had it and that it costs2500UGX. (CS: 02)I face challenges when I could go to Jabusiba [NakwasiHC III] and there is no treatment and they could tellme to go and buy. [But] malaria could not improveand it came back, she [the child] started swelling … Ihad to take her back again to the HC. (CS: 05)There are times when you go to the health unit and donot find any tablets there. So, if there are no medicinesat the health unit and are lucky enough you find thehealth provider there, she can ask for a book [write aprescription] in which she write the tablets which youhave to buy for your child's illness. (CS: 08)Costs associated with finding time and takingtransportationMaking time and finding transportation to reach publichealth facilities or private clinics was another factor add-ing to the burden of cost. All caregivers mentioned thecost of transportation as a burden, except for one casewith a positive outcome (CS: 08) who was given Coar-tem® from a public health facility to keep at home for fu-ture use. Getting the runaround and being sent toseveral public health facilities and drug shops in order toreceive treatment adds another layer of transportationcosts. In some cases, such costs deterred people fromseeking care at public facilities.The challenges that I face were about moneybecause transportation to Nalusaga [Kangalaba HCIII] on a motorcycle costs 1500UGX …. The [next]challenge I face was to be referred to Businghu[Busiu hospital in Mbale] yet [because] I did nothave the means to get there. Transportation fromNalusaga to Businghu, costs 5000UGX using ataxi…. I asked my husband about what we coulddo … he borrowed money from a friend. So withthat money we went to Businghu. (CS: 01)I had also footed, walked on foot up to the healthunit .… With the transportation problem, if I havemoney, the motorcycle is easy to use but the problem ismoney. You may want to use the bodaboda[motorcycle for hire] but when you do not have themoney. Don't you see?.... The bodaboda guys[motorcycles] tell you to pay 4000UGX to and fromBusolwe hospital. (CS: 02)The child had malaria when he was still very young,just one week old. We were reluctant to take him tothe health unit … [because] we failed to get money fortreatment.… So that is how this cerebral malariastarted. (CS: 03)Transport [is a challenge] in the way that you may notbe having a bicycle. And so the children fall sick, youneed to find a way to the HC. (CS: 04)I lacked transport to reach me to other HC … in termsof money for getting a car, motorcycle, like that. (CS: 05)I have never gone to Busolwe hospital … because it isfar, and also for transport the money is too much [fortransportation to Busolwe]. (CS: 07)Cost of managing negative outcomesAnother important contributor to the cost burden werecosts associated with managing negative outcomes of amalaria episode. Children who survived with a disabilitypresented further costs to their families by requiring medi-cines to treat impairments resulting from the severity ofthe illness or because of inappropriate management of themalaria episode. Three of these children (CS: 01, 02, and03) survived but with a negative outcome, all required on-going treatment and time.So, when the child was curing in the process, this hand[holds the boys left hand that is lame] became lame.And even now, something … like epilepsy … attacksthe brain and the brain stops working well … and thechild falls down…. So, then, that is how he is, I lookafter him in such state. (CS: 01)Those [convulsion] tablets are found in Busolwehospital, they are distributed every Friday, they giveyou tablets to last a whole week. So every Friday, weget those tablets for her…. We walk [7 miles toKassam et al. BMC Public Health  (2016) 16:160 Page 14 of 24Busolwe], we walk with our feet, that is when we havefailed to get money. At times, we ride on the bicyclebut when we fail to use the bicycle, we walk on foot tothe hospital to get the medicines. (CS: 02)The child's illness has caused us a lot of povertybecause when it [convulsion] catches him, we do nothave money. We have to work in other people'sgardens to pay for the cost [medicines andtransportation]. So now, I do not know what to do tomake the child's illness to get cured for good. (CS: 03)Sources of moneyA further primary hurdle faced by most caregivers whoobtained treatment for their child was the difficulty inobtaining money. Caregivers stated they had to seekcredit from private vendors, borrow money from friends,obtain money from their spouse and family, deposithousehold goods, do extra work in exchange for moneyor sell household goods to pay for treatment.[Sometimes] I deposited the gomesi [dress] as securityand obtained money … [from] a woman at a bar inthe trading centre … when I wanted to buy themedicine for injection and tablets…. [Another time] Iborrowed a bicycle which my husband deposited [as]security…and then he used the money to pay for thetablets and other medicines and then he also paid forthe bodaboda. After that is when I go to the healthunit or the other private clinic at Busolwe [to buy themedicine]. (CS: 03)I take her to a clinic in Lelesi [drug shop] and theytreat her with quinine … and tell me to come backwith the money later. Then afterwards I look formoney. You go, then someone gives you a garden, youdig, then in exchange she gives you money [and youpay the shop]. (CS: 05)I am challenged in a way of money…. If am to go tothe health provider [public health facility], they willtell me that they need money [before give treatment]yet I do not have [so] I [have to] go to borrow and tohelp me treat the child. (CS: 06)Burden of living with a negative outcomeLiving with a negative outcomes resulting in irreversiblephysical and/or mental disabilities generates uniqueburdens for both the child and the household. Unlikethe burden of cost which can be discharged in a seasonor so, physical and mental defects last for the lifetime ofboth child and household.Effect on the child and the householdNegative physical effects on the child were the most com-mon burden mentioned and included decreased mobility,sight, and mental capacity. Cases 01, 02, 03, and 05 all hadchildren who experienced prolonged episodes of malaria.As illustrated in these caregiver reports, these resulted inconsiderable suffering for the child, and impacted care-givers by adding to the burden of cost, requiring constantand vigilant care, dealing with socially inappropriate be-haviors from the child, and living with enduring sadness.So when the child was curing in the process, this hand[refers to boy’s left hand] became lame. And even now,something … like epilepsy … attacks the brain and thebrain stops working well it throws him down and thechild falls down. I guess you can see the scars due tothat [child to several wounds and sores on the foreheadsand at the chin]. So then that is how he is, I look afterhim and such state. (CS: 01)Convulsions fever caused her to stop talking and thebrain became affected, she became mentally disturbed,affected and ill. She used to speak everything, even sing.If she could still speak, she would have by now alreadywelcomed you, greeted you. She used to even thank mefor the cooking after she had a meal [caregiver is nowvery emotional, close to tears]. But from the time shestopped talking, that is the same time during which hermental capacity was lost, and now she is mentallyunstable…. Before getting mentally incapacitated, shewould go and defecate in the latrine, but right now shejust defecates anywhere that she finds. It can even behere right now, she can come and defecates here…. Shecan even go and defecate in the house. (CS: 02)Whenever he catches the malaria, I do not want toleave him all by himself, I need to move with him allaround [wherever] I go…. I go with him. When he gets asevere attack, and yet I have got no money, [I] still andmove with him wherever I go [even when gardening]….He [still] suffers the convulsions. Stomach swells verymuch. Whenever he convulses, [stomach] too swells upand it can continue to swell up to three months nonstop.He also does not eat … as for the stomach that onefailed to get back to normal. In fact, it only continues toKassam et al. BMC Public Health  (2016) 16:160 Page 15 of 24swell and swell even more. [The child's stomach is veryswollen like a balloon]. That child's illness has causedus a lot of poverty because when it catches him, we donot have money. (CS: 03)Body swells, both her [the child’s] legs and her hands.The cheeks would also swell, she would have diarrhea,she would cough, mucus would come out of the noseand her body would get very hot…. She continuouslyfell sick. God himself decided to take her. But I tried alot to give her treatment and I was defeated, then thechild died. (CS: 05)Knowledge of government recommended medicineCaregivers reported a variety of beliefs about usingwestern medicines or traditional remedies to treatmalaria in young children. Preference for westernmedicine over traditional medicines to treat malariawas expressed by all caregivers, although some didmention traditional remedies also had role in someinstances. Seven of the cases (CS: 01, 03, 04, 05, 06,07, 08) acknowledged hearing about ACTs (or Coar-tem®), but only one (CS: 04) of the four cases askedto name the government recommended antimalarialcorrectly stated ACT/Coartem® (Table 3). Thus, notall caregivers who had heard of ACT or Coartem®recognized that Coartem® was an example of ACT,and not everyone was familiar with ACT or Coartem®being the government recommended antimalarial(Table 3). Additionally, only two sought out Coartem®to keep at home for future use (CS: 01, 08) whichthey initiated as part of home management, and onlyone caregiver (CS: 08) requested that their child’scurrent episode be treated with an ACT (or Coar-tem®) when visiting an external source.I know that the Western [medicines] work better thanthe traditional [because] I know that [Coartem®] workson malaria. (CS: 01)[Western medicine] works speedily/fast and traditionaltreats but not much…. I know also that traditional onesalso work … here in the village, we have our local ortraditional medicine … if she vomits, they help her innot vomiting … it works [the traditional]. (CS: 07)Myself. I knew that Coartem® also would play animportant role in child's body…. I had got theCoartem® from the health provider at Muhuyu HCII because she is my husband’s friend. Previously,the child had been ill and had been given Coartem®to give to him. But on top of that, the healthprovider gave me other [additional] Coartem® to uselater on in the future just in case any of my childgot sick with malaria…. I learnt because I alsonormally fall sick … when I fall sick and go to ahealth provider and explain to her about thoseconditions, she tells me to go and swallowCoartem®. (CS: 08)Table 3 summarizes caregivers’ knowledge aboutACT (and Coartem®), whether the child received it,and where they normally obtained it. We found thatno febrile illness was treated in accordance with gov-ernment malarial policy: receiving a confirmatorydiagnosis for malaria and initiating a first-line anti-malarial (quinine for those less than four-months andACT for those greater than four-months) within 24 hof fever onset. While seven of these eight childrenwere older than 4 months at the time of the febrileepisode, and therefore eligible to receive an ACT,only three ever received it during the course of theirillness.Lessons learned and suggestions for futureimprovementsCaregivers reported they had learned from this ex-perience, and discussed how this will change their fu-ture treatment-seeking behaviors. They also madesuggestions about what they believed need to occurwithin the health system to improve malarial treat-ment for children five and under.How this experience changed caregivers’ treatment seekingbehaviorsThe experiences from this febrile episode influencedtreatment-seeking behavior for all caregivers whosechild had experienced a negative outcome (CS: 01, 02,03, 04, 05), and all had several things to say duringthis portion of the interviews. Only one caregiverwhose child had a positive outcome (CS: 08) partici-pated in this discussion. Only one caregiver whosechild survived with a disability (CS: 03) said shewould not treat her children with quinine. One care-giver (CS: 01) stated that while she herself would notgive quinine she would give it if it was recommendedby a health provider, even though she now believesCoartem® is the best.I know that it [Coartem®] works best on malaria … Icannot use quinine again because I got him lame. NowI take him to a health provider who would know howKassam et al. BMC Public Health  (2016) 16:160 Page 16 of 24to handle my son’s malaria…. [However] If it is ahealth provider [who recommends it], I do not refuseto use the quinine because it is the rule from thehealth provider. (CS: 01)I have learnt that I should not use lots of quinine. I wastold about the effects of quinine so now I also know thatquinine damages the brain and mental capacity if youuse it in large amounts. [Now] when she gets malaria,we give her these very [convulsion] tablets that weobtain from Busolwe hospital on Fridays. (CS: 02)Nowadays when he gets sick did normally do not givehim [quinine] injections…. I bathe him. If I havemoney, I get him some Panadol® so that the fever isrelieved a bit. (CS: 03)Both caregivers whose child died said they would initi-ate treatment more promptly, one (CS: 05) by makingsure she had Coartem® at home to initiate as home man-agement, the other (CS: 04) still remained uninformedabout the first-line antimalarial treatment.I [have learned to quickly] take her to the healthunit. Okay I [get] him those tablets [Panadol® andSeptrin®] because one time I had gone and theyare the ones that I normally get for treatment. Ieven use these for myself for treatment.(CS: 04)Okay when it comes [fever], I usually haveCoartem® in the house, so when the sickness comes,it is the one that I usually give before I rush her tothe health unit. (CS: 05)Two of the cases, one with a positive outcome (CS:08) and one with a negative outcome (CS: 02), statedthat they had learned from their experience to notuse traditional remedies.Table 3 Caregivers’ knowledge of government recommended medicineaKnowledgequestions→Cases (CS) ↓Does CGknow thefirst-line gov’tantimalarial?Has CGheard ofACTs?Has CGheard ofCoartem®?Does CGknowCoartem® isan ACT?Where did CGhear aboutACTs?Was child given an ACTduringthis malaria episode?Where doesCG usuallyobtain ACTs?Fromhome(Yes/No)From 1st externalsource (Yes/No)Positive Outcome: Child RecoveredCS 06 No No Yes No N/A No No (DS) PHF for free(not this episode)CS 07 No Yes(But gives wrongexamplesa)No No • Drug shops• Gov’t health providers• RadioNo No (DS) No evidenceACT was everobtainedCS 08 N/A N/A Yes N/A N/A Yes(leftover)Yes (PHF) PHF for freeNegative Outcome: Child Survived but with DeficitsCS 01 N/A Yes Yes No • Gov’t health providers• RadioYes(leftover)No (PHF) PHF for freeCS 02 No No No N/A • Not heard No No (PHF) Not for this indexchildCS 03 N/A Yes No No • Radio No No (DS) PHF for free(not this episode)Negative Outcome: Child DiedCS 04 Yes(Says ACTs& Coartem®)Yes Yes Yes • Gov’t health providers• RadioNo No (PHF) PHF for free(not this episode)CS 05 N/A N/A Yes N/A N/A No Yes (PHF) PHF for freeAbbreviations: Artemisinin Combination Therapy (ACT); Caregiver (CG); Government (gov’t); Drug Shop (DS); Not Applicable (N/A) - In some of the interviews theseissues were not discussed; Public Health Facility (PHF)aExamples given: Antibiotics; Aspirin®; quinine; chloroquineKassam et al. BMC Public Health  (2016) 16:160 Page 17 of 24I used [traditional meds] but they failed so I give upon them. (CS: 02)I used it but it [mululusa] did not heal my child. Sonow, I do not use it. I realized that it [mululusa]wastes time. (CS: 08)Three of the five cases with a negative outcome (CS:02, 03, 04) indicated they would in the future seek carefrom a public health facility.Now, what I have learnt is that if any other child has orsuffers a convulsion, I do not follow anyone's advice ofwhat to use to treat the child. I know that I have to go tothe health provider whom I explained to him he advisesme accordingly. (CS: 02)I [would] take her [immediately] to the health unit[previously treated the child at home and only tookthe child to a health unit when symptoms got worse].(CS: 04)In addition to treatment-seeking behavior, most care-givers (CS: 01, 02, 03, 07, 08) discussed the need to im-prove preventive measures and overall well-being of thechild.To sleep in a clean place. You should sweep wherethey sleep even if they only sleep on a papyrus mat.(CS: 02)Not giving them cold food in the morning since it iscold, it has parasites. So it is better that youprepare and give children warm food like porridgeto eat. (CS: 03)If child has a poor appetite, always try to give himvariety. If child fails to eat something, try and givehim something else to eat. Do not just ignore thechild without giving him something to eat. Try yourbest to give him various foods to eat until child getssomething that he desires to eat. In that way, I amhelping to give care to child who has malaria. Sothat the child doesn't stay the whole day hungryand yet he is suffering from the sickness as well.(CS: 08)Four of the five cases with negative outcomes (CS:01, 02, 03, 04) mentioned that nets were an import-ant way to improve caregiving. While only two ofthese four cases (CS: 01, 04) reported actually usingnets in their home, one (CS: 01) qualified the use ofnets by saying that there was no way to avoidmalaria.They [children five and under] sleep under the net….You cannot avoid or prevent yourself from a mosquito.At times, you can be seated and it comes and bites you.So in case you have medicine it helps us. (CS: 01)Suggestions for improving health delivery at the systemslevelCaregivers’ suggestions for improving health deliverywere directed toward alleviating challenges they facedwhen seeking treatment from public health facilities.Two suggestions made by half of the cases with bothpositive and negative outcomes (CS: 01, 02, 05, 08)were to bring health services closer to their homeand to have more personnel at public facilities.I think that the government should help us and buildhealth units that are nearby so that even us who stayfor we can start moving shorter distances to the healthunits. (CS: 01)So, at least, various roles can be shared up [have morestaff to do the different tasks]. One health provider canbe dispensing the tablets and the other health providercan be injecting, rather than just one person giving outtablets and at the same time injecting and at thesame time receiving the patients. You can imagine, ofabout 30 people or 25 people, how can one healthprovider attend to all those people at once.(CS: 08)Four other caregivers (CS: 03, 04, 05, 07) suggestedthey were interested in being given more medicine bythe public health facilities. Both cases 05 and 07 saidthat the types of medicine they want are ACTs. Whilecase 05 reported she had heard of Coartem® and herchild had received Coartem® at the public health facil-ity; case 07 was less familiar with Coartem® but hadbeen told it was the first-line government recom-mended antimalarial by the research assistant con-ducting the household survey as part of the largerstudy. Two caregivers (CS: 05, 06) also indicated thatpublic health facilities need to provide them with in-formation on how to better care for their child.Kassam et al. BMC Public Health  (2016) 16:160 Page 18 of 24They [public health facilities] should immunize formalaria. (CS: 03)They [public health facilities] should bring for us helpby giving and providing us with tablets and medicines… specifically Coartem® … in village…. [Additionally] Iwould love to be told when the child fall sick anddoesn't want to eat what do I do. (CS: 05)We want government to bring for us mosquito nets,and also those [ACT] medicines they give us and be inthe house as first aid so that we know what to do forthe child very fast. (CS: 07)DiscussionWhile case studies do not allow generalizability to alarge population, they do provide greater insight into in-dividual behaviors. These case studies offer valuable tes-timony of eight caregivers’ circumstances, experiencesand challenges with treatment-seeking in Butaleja Dis-trict in rural Uganda. In our study we examined the se-quential steps taken by caregivers to manage their child’sfebrile illness, their experiences with the various sourcesthey visited and challenges they encountered during thisprocess. Overall, caregivers had a pluralistic approach totreatment, with many taking multiple steps in an at-tempt to resolve their child’s febrile episode and to avoidadverse outcomes. Despite the variability in outcomes –illness improved, negative outcome with disability andnegative outcome resulting in death - caregivers gener-ally shared similar practices, experiences and challenges.Home managementSimilar to other regions in Uganda and sub-SaharanAfrican countries, most caregivers in our study initiatedtreatment with home management either as a primaryapproach to treat the febrile illness or as a temporarymeasure before seeking treatment from an externalsource [17, 31]. Home management commonly involvedthe use of supportive treatment, traditional herbs, and/or western medicines kept in the home for future use.Home medicines were mainly sourced from private drugshops or public health facilities, and some instances wereleftover from previous treatments. Only one caregiveropted to do nothing other than observe the child as theirinitial act. While just under half of the caregivers re-ported that they did not intend to seek external carewhen they initiated home management, all caregiverswhose child did not improve sought treatment from anexternal source within the same/next day. The strategyto start with home management allows caregivers to waitin hope the less expensive home medicines will suffice,postponing direct and indirect expenses associated withseeking external treatment [32]. This approach also en-ables those wishing to seek further treatment to securemoney, liquid assets, or find alternate source of labor topay for these expenses [17, 31, 33–36]. The practice toseek help from external sources only when home man-agement is perceived to have failed or if the child is pre-senting with symptoms reflecting severe illness such asvomiting or diarrhea, has also been observed in severalother regions of Uganda, sub-Saharan African countries,as well as south Asia [17, 33–37]. Thus, despite care-givers reporting a preference for public health facilities,direct and indirect costs frequently compel them to startwith home management [35, 37]. As evident in ourstudy, the delay in seeking external care often results inthe child’s condition progressing from mild to severe,resulting in negative outcomes such as irreversible dis-ability or death.Use of multiple treatment sourcesGetting the runaround at public health facilities was acommon occurrence in our study. A greater proportionof caregivers in our study chose public health facilitiesover private outlets (ration of 5:3) for their first externalsource, although three of the five were later referred toprivate drug shops to purchase medications for the child.Consequently, more caregivers obtained their treatmentsfrom private outlets than public facilities. Furthermore,half of the caregivers required multiple visits to the sameor different sources because their child either did notimprove, the child went on to suffer an adverse out-come, or public facilities where they were visiting wereexperiencing stock-outs of medicines and/or blood sup-ply. In a few instances, caregivers were referred to higherlevel public health facilities for more invasive treatmentwithout any treatment initiated at the primary externalsource. Thus, similar to what other researchers in sub-Sahara Africa have observed, caregivers in our studyvacillated between multiple sources, travelling long dis-tances and incurring numerous costs in an attempt tofind a cure [32, 38, 39].Use of multiple treatmentsThe use of multiple treatments to manage children wascommon in our study. This practice of polypharmacy bycaregivers likely occurs in hope that at least one ap-proach or synergistic effects from the combination oftreatments will help cure malaria in their child [17, 32].“Multiple treatments” included the use of multiple west-ern medicines together, and sometimes in combinationwith supportive care and traditional remedies. Whilepolypharmacy was common with home management,Kassam et al. BMC Public Health  (2016) 16:160 Page 19 of 24this practice was also observed with public health facil-ities and private drug shops. Our results are in line withother studies that have found traditional and moderntreatments, both effective and ineffective, to coexist[17, 32, 40]. A rural study conducted in Mali foundcaregivers often sought care from HCs when homeremedies did not alleviate symptoms, but then theyreturned to traditional therapy if treatment from theHC failed. In our study, one caregiver reported reinstat-ing a traditional remedy when the child failed to im-prove, supplementing the treatment prescribed by thehealth facility.Delays in treatmentWhile most caregivers responded to their child’s symp-toms on the same day with home management, someversions of home management no doubt caused delay inseeking treatment from external sources to later in theday or the next day. For many children this wait was toolong, with some children’s condition progressing frommild to severe. In some cases the wait may have beenlonger as caregiver waited for home remedies to take ef-fect. Appearance of illness at night was another deter-rent to seeking immediate treatment at an externalsource, since caregivers waited for the sun to rise. Onceat a public facility, medication stock-outs caused stillfurther delays by requiring caregivers to secure add-itional funds and to travel further distances to a privateoutlet where medicines could perhaps be purchased.Other barriers encountered at public facilities by care-givers in our study included long wait times and unavail-ability of staff. Prompt treatment with antimalarial isdefined by the World Health Organization (WHO) andthe Uganda National Malaria Control Strategy as ad-ministration of treatment within 24 h of onset of fever[7, 41]. Research exploring health system practices inUganda and other sub-Saharan African countries hasconfirmed delay in treatment to be a common occur-rence, usually associated with lack of sufficient and ap-propriate staffing, long wait times, and unavailability ofmedications [33–36, 42–44]. Thus, despite the fact thathalf our study’s caregivers reported seeking care from apublic facility within 24 h, treatment may not have beeninitiated within those same 24 h.Prompt diagnosis and treatment with appropriateantimalarialOnly a small proportion of eligible children in our studyreceived an ACT from an external source (25 %), despiterevised guidelines prompting its use, recent initiatives toimprove its access from private licensed facilities, and moreambitious national targets for treating malaria. The UgandaNational Malaria Control Strategy recommends that allsuspected cases of malaria be confirmed diagnostically, andthose testing positive receive a first-line antimalarial within24 h of fever onset [7]. The target set by the UgandaMalaria Control Strategic Plan was to have 85 % of chil-dren under five receive correct treatment according to thenational treatment guidelines with 24 h of the onset ofsymptoms by year 2010 [7]. For uncomplicated malaria,the first-line antimalarial consists of oral quinine for chil-dren younger than four months and ACT for those olderthan four months. For severe malaria, treatment is initiatedwith injectable quinine followed by oral quinine or ACTdepending on the child’s age. Additionally, in circum-stances where a child needs to be referred to anotherhealth facility, both the WHO 2010 guideline and UgandaNational Malaria Control Strategy recommend initiatingtreatment with rectal or intramuscular artesunate, arte-mether injectable or intravenous quinine prior to referral[7, 41]. For most children in our study, diagnosis was pri-marily presumptive with only two children receiving a con-firmatory blood test for malaria, one at a public facility andanother at a private outlet. In two of the three instanceswhere ACT was given, it was initiated as part of homemanagement. However, only one child was later main-tained on the ACT when seen at the public health facility,the other was switched to oral quinine. Overall, one care-giver reported initiating ACT after visiting a public facility.Thus, in this study, public facilities were not dependablesources for receiving treatment as outlined by the nationalmalaria policy. Children were no more likely to receive anappropriate antimalarial treatment if they were first seen ata public facility as opposed to a private outlet. The low useof ACTs in our study is in line with national and regionalfindings in Uganda which show the usage of ACT to be at21 % in 2009 and 44 % in 2012, with an even lower propor-tion demonstrated to receive an ACT within the same/nextday (18 % in 2009 and 36 % in 2012) [45, 46]. For thosechildren in this study who died from a protractedcase of fever, one died in spite of receiving an ACT.One possible explanation for the death may be thelack of proper clinical assessment and overall man-agement by the public facility. Since many childhoodillnesses share common symptomatology such asfever, fatigue, weakness, nausea and diarrhea, diagnos-tic testing for malaria is integral to a child receivingappropriate treatment. As observed in this study andreported by other, in circumstances where quality ofservices are poor and public health facilities are notadequately resourced, experiencing a negative out-come is not only inevitable but unfortunately predict-able, suggesting the need to strengthen casemanagement [39, 43].Stock-outs of ACT at public health facilitiesFor children in our study, it is possible that public healthfacilities may not have prescribed an ACT because theyKassam et al. BMC Public Health  (2016) 16:160 Page 20 of 24were already aware that few private drug shops – whichare mostly unlicensed in Butaleja District – carry it.With funding from the Global Fund, Uganda imple-mented a national program in Spring of 2011 to dissem-inate ACTs through licensed private outlets at asubsidized cost [27]. However, purchase of ACTs from alicensed private outlet was beyond the reach of manycaregivers, as most are poor and cannot afford to travellong distances to licensed private outlets. Half of thecaregivers in our study relied on public health facilitiesto obtain an appropriate antimalarial, and the use ofunlicensed private outlets was in large part due to un-availability of medicines at public facilities. In fact twocaregivers in our study reported obtaining ACTs frompublic health facilities well in advance of an illness andkeeping it at home for future use because of frequentstock-outs. Thus, in Butaleja District public facilitieswere not reliable sources of ACTs; and while unlicensedprivate vendors were an important source of antimalarialtreatment, they were not licensed to sell ACTs.Caregivers prefer public health facilities despitechallengesDistances to health facilities were a note of concern byall caregivers, as many lacked resources such as moneyand transportation to travel to these facilities. Amongthose who were able to reach a health facility, they werefurther burdened by long wait times, medication andsupply stock-outs, staff shortages, inattentive staff, staffthat exhibited poor attitudes, and staff who delivered anoverall poor quality of care. Despite such unpleasant ex-periences, many caregivers still expressed greater prefer-ence for public health facilities, with just over a halfchoosing public facilities as their first external sourcewhen their personal attempts to resolve the illness failed.As has been suggested by others, caregivers’ preferencefor public facilities may be associated with the belief thatproviders at these facilities are more qualified and expe-rienced than those at private outlets [17, 34, 47]. Know-ledge of such obstacles impeding access to prompt andappropriate treatment at public facilities in Uganda isnot new and have been reported by many others [17, 44,48, 49]. Our study however points to the fact that suchchallenges continue in spite implementation of nationalmalaria control policies and training of public providersto improve case management. The literature suggeststhat interventions more far-reaching than health pro-vider training are needed to improve care at public facil-ities [17, 48, 49]. These include implementing regularand adequate supervision, improving worker motivationsand perceptions, improving customer service and com-munication between staff and caregivers, changingpolitical and economic environment and incentives,mitigating misappropriation of drugs to ensurecontinuous stock of ACT, and removal of ineffective an-timalarials from facilities [17, 48, 49]. Until such time aspractices at public health facilities meet guidelines, pro-moting the use of public facilities will remainproblematic. Public health researchers and program im-plementers need to consider multifaceted approaches in-volving all key stakeholders – including both caregiversand unlicensed drug vendors, to improve prompt and ef-fective management of malaria in young children.RecommendationsA number of interventions have been implemented inUganda to strengthen case management within the reg-ulated public and private settings. However, consideringthe central role of caregivers and of the unlicensed pri-vate sector in the management of fever for young chil-dren in rural settings, future initiatives need to bemultifaceted by including caregivers and all servicesaccessed by caregivers [15, 17, 45, 48, 50, 51].In Butaleja District where a large proportion of thepopulation relies on unlicensed outlets, the current pol-icy to limit dissemination of AMFm subsidized ACTs toonly licensed private outlets has not resulted in accept-able malaria management practices [27]. One option toaddress the current gap between policy and practice isto extend the ACT subsidy program along with the ne-cessary training to the unlicensed private sector [15, 51].An additional strategy to engage the unlicensed sectormay be through the integrated community case manage-ment (iCCM) program. There have been recent nationalinitiatives to bring public services closer to the commu-nity through the introduction of iCCM, although at thetime of this study this service had not yet been intro-duced in Butaleja District. This program is aimed at de-livering prompt and free treatment to children underfive for malaria, pneumonia, and diarrhea, through theuse of volunteer CHWs who are recruited by the com-munity but supported and supervised by the publichealth system [7, 52]. While short-term studies havedemonstrated care delivered by CHWs to be effectiveunder study conditions – where close supervision anduninterrupted supply of ACT is assured, long-term ef-fectiveness still needs to be demonstrated [53, 54]. Someof the challenges levelled against the original HomeBased Management of Fever Program (HBMF) uponwhich the iCCM is modelled include frequentmedication stock-outs and absenteeism of CHWs [55].Given that iCCM is governed under the public healthstructure and CHW model continues to rely on volun-teers, it also runs the risk of being plagued by similarproblems [54, 56]. Current data from national and re-gional studies indicate that to-date there has been a lessthan optimal uptake of CHW services through theiCCM program [45, 48, 50]. Whether this low uptake isKassam et al. BMC Public Health  (2016) 16:160 Page 21 of 24a consequence of earlier negative experiences with theHBMF program still needs to be investigated. What wedo know from previous treatment seeking behavior stud-ies is that poor quality of care and/or stock-outs canundermine the potential impact of providing free healthcare [33]. Given that most unlicensed vendors are firstand foremost part of the communities they service, onepossibility would be to consider a CHW model that in-cludes them. Such a model may mitigate the concern ofmedicine stock-outs afflicting public outlets – since pri-vate vendors can offer a constant supply of ACTs; and itmay alleviate concerns with absenteeism because of theiron-going presence.Shaping caregivers’ expectations and decisions may beanother important element to meet national case man-agement objectives [18, 20, 48]. For years social scien-tists have advocated that acceptance of any healthcareintervention will ultimately depend on individuals’accepting and demanding such interventions [57].Accordingly, in circumstances where caregivers lack theappropriate knowledge or conviction, the literature indi-cates that they will often limit the demand for effectiveinterventions even when it is available [20, 58]. Whilemost caregivers in this study reported hearing of ACT,there was a low level of awareness that ACT was thefirst-line antimalarial. It is therefore not surprising thatmost caregivers did not request an ACT when obtainingtreatment. Other studies in rural eastern Uganda havealso found caregiver to commonly confuse non-antimalarials with antimalarials, with many opting forless effective therapies [17, 48]. Similarly, the low use ofdiagnostic tests to confirm suspected case of malaria inthis study was also a concern. While we did not explorecaregivers’ views of diagnostic tests, studies exploringcommunity acceptance of routine diagnostic proceduresto confirm malaria have identified disturbing findings.For instance, in Uganda, it was reported that communitymembers generally feared getting their children’s bloodtested for malaria because of concern that their childcould get infected with HIV in the process, the bloodwould be used to test for HIV rather than malaria, orthat the blood could end up in the wrong hands andbe used for witchcraft [59]. Another study in Nigeriareported that while community members acknowl-edged the importance of testing for malaria, manyremained doubtful about the reliability of the tests,especially when results are negative [60]. Given thatthose who are informed about the benefits of ACTsand of diagnostic testing are more likely to advocatefor their use when seeking treatment for their child,this study clearly highlights the need to developtargeted education for caregivers around diagnostictesting and prompt treatment with first-line antima-larials [18, 59, 60].LimitationsCase study methodologies offer unique opportunities toexplore and examine complex phenomena within naturalcontexts, and – in the case of this study – throughmultiple lenses [61]. The intent of this study was todocument individual cases to capture thick descriptionsof unique situations and to illustrate special attributesand aspects of care giving during a malaria episode, aswell as to determine factors influencing caregiver prac-tices as leverage points for defining a health promotionprogram. Despite their advantages, case study limitationsneed to be recognized and findings from this study needto be interpreted accordingly. First, case study method-ologies generally use smaller sample sizes than found inmost quantitative studies, thus this study makes noclaim to generalizability. Second, case studies are subjectto selection bias because each case is purposively se-lected by the research team to explore specific experi-ences. In our study, results represent insights from eightcaregivers, thus the study findings are not necessarilyrepresentative of all caregiver experiences in ButalejaDistrict. Third, the research was conducted retrospect-ively and therefore was subject to recall bias. While itwas possible to minimize recall bias for those caregiverswhose children had a positive outcome by limiting thediscussion to the very last fever episode, it was difficultto minimize it from a methodological perspective forcases that experienced negative outcomes, as it requiredcaregivers to reflect back to an episode that may haveoccurred over a year ago. Thus, recall bias may have re-sulted in omission or over-emphasis of certain discretebut relevant steps. Lastly, cases were recruited based oncaregivers’ presumptive diagnosis of malaria. This couldhave resulted in sampling of some cases which may nothave been malaria. However, given that fever commonlyserves as a proxy for malaria in high endemic-regionswhere use of routine diagnostic procedures remains low,the sampling process followed real life practices.ConclusionsFindings from our study propose that these eightcaregivers in Butaleja District generally shared similarpractices, experiences and challenges when seekingtreatment for children five and under, with few childrenever receiving treatment in accordance with the Ugandanational guidelines. To improve timely access to ACT,our results support the need to include all key stake-holders in future public health interventions aimed atimproving malaria management in young children. Trad-itionally, training and certification has been limited toproviders affiliated with the formal health system - pub-lic health facilities and licensed private outlets. However,given the weak infrastructure and limitations of the for-mal health system, focusing on these groups alone is notKassam et al. BMC Public Health  (2016) 16:160 Page 22 of 24sufficient. As noted in this study and acknowledged byother researchers, unlicensed drug vendors as well ascaregivers continue to play an essential role in the man-agement of fever in young children in rural settings. Ac-cordingly, future health delivery models need toconsider how unlicensed vendors can be leveraged tohelp attain the national target – to have 85 % of childrenunder five receive antimalarial treatment as per guide-lines. Additionally, future public health interventionsneed to improve caregivers’ capacity to take the neces-sary actions for their children by better informing house-holds on how best to manage malaria in young childrenand to advocate for appropriate treatment.Additional filesAdditional file 1: Table S1. Caregivers’ treatment seeking actions andsequence from point of initial symptomsa. (PDF 194 kb)Additional file 2: Table S2. Caregivers’ experiences with externalsourcesa. (PDF 171 kb)Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsRK contributed to the conception and design of the study. RK, RS, and JTcontributed to acquisition of the data. RK analyzed and interpreted the datawith contribution from DM, EL, and RS. RK drafted the manuscript articlewith input from DM, EL, JT, and RS. All authors have read and approved thefinal manuscript.AcknowledgementsThis research was made possible by the contribution of eight caregivers whoshared their personal encounters and experiences – we appreciate theirparticipation. While this study did not receive direct funding, it was part of alarger malaria exploratory study funded by the Canadian Institute of HealthResearch and the Hampton Fund Research Grant to define sustainablecommunity-based interventions for improving access and use of antimalarialdrugs in Uganda. As such, the study benefited from the organizationalinfrastructure of the larger study. We therefore recognize the Butaleja Districtofficers and staff, Makerere University, and the entire Ugandan and Canadianteam members for their support. We want to thank Ms. Josephine Nyamonyaand Ms. Irene Tabo for their diligent work as research assistants. We arespecifically grateful to Dr. Kenneth Mweru (then District Health Officer forButaleja District) for supervision of the field researchers, and to Mr. DanielHashasha for facilitating the initial contact with local village leaders. Wethank Ms. Nicole Chaudhari, Ms. Nabeela Rasool, and Mr. Maichael Thejoe fortheir assistance with formatting and proofreading of the manuscript.Author details1School of Population and Public Health, Faculty of Medicine, University ofBritish Columbia, 2206 East Mall, Vancouver, BC V6T 1Z3, Canada. 2ChildHealth and Development Centre, School of Medicine, Makerere University,KampalaP.O.Box 7062Uganda. 3The Islamic University, Mbale, Uganda.Received: 27 September 2015 Accepted: 2 February 2016References1. World Health Organization (WHO). World malaria report 2011. http://www.who.int/malaria/world_malaria_report_2011/9789241564403_eng.pdf.Accessed 1 Dec 2014.2. Liu L, Johnson HL, Cousens S, Perin J, Scott S, Lawn JE, et al. Global,regional, and national causes of child mortality: an updated systematicanalysis for 2010 with time trends since 2000. Lancet. 2012. doi:10.1016/S0140-6736(12)60560-1.3. ter Kuile FO, Parise ME, Verhoeff FH, Udhayakumar V, Newman RD, van EijkAM, et al. The burden of co-infection with human immunodeficiency virustype 1 and malaria in pregnant women in sub-Saharan Africa. Am J TropMed Hyg. 2004;71 Suppl 2:41–54.4. United Nations Children's Fund (UNICEF). Malaria and Children: progress inintervention coverage. http://www.unicef.org/health/files/Malaria_Oct6_for_web(1).pdf. Accessed 1 Dec 2014.5. World Health Organization (WHO). The global burden of disease: 2004update. http://www.who.int/healthinfo/global_burden_disease/GBD_report_2004update_full.pdf. Accessed 1 Dec 2014.6. Steketee RW, Nahlen BL, Parise ME, Menendez C. The burden of malariain pregnancy in malaria-endemic areas. Am J Trop Med Hyg. 2001;64(Suppl 1–2):28–35.7. Ministry of Health. Uganda malaria control strategic plan: 2005/06 - 2009/10.http://archiverbm.rollbackmalaria.org/countryaction/nsp/uganda.pdf.Accessed 1 Dec 2014.8. World Health Organization (WHO). The African summit on Roll Back Malaria.http://www.rollbackmalaria.org/microsites/wmd2011/abuja_declaration_final.html. Accessed 1 Dec 2014.9. United Nations. The Millennium Development Goals report 2008. http://www.un.org/millenniumgoals/2008highlevel/pdf/newsroom/mdg%20reports/MDG_Report_2008_ENGLISH.pdf. Accessed 1 Dec 2014.10. World Health Organization (WHO). World malaria report 2008. www.who.int/malaria/wmr2008/malaria2008.pdf. Accessed 1 Dec 2014.11. World Health Organization (WHO). Scaling up home-based management ofmalaria: from research to implementation. http://www.who.int/tdr/publications/documents/scaling-malaria.pdf. Accessed 1 Dec 2014.12. Greer G, Akinpelumi A, Madueke L, Plowman B, Fapohunda B, Tawfik Y,et al. Improving management of childhood malaria in Nigeria and Ugandaby improving practices of patent medicine vendors. http://www.basics.org/documents/pdf/ImprovingMalariaMgmtPMVs.pdf. Accessed 1 Dec 2014.13. Hall S. People first: African solutions to the health worker crisis. http://www.chwcentral.org/sites/default/files/People%20First-%20African%20solutions%20to%20the%20health%20worker%20crisis.pdf. Accessed 1 Dec 2014.14. Zurovac D, Tibenderana JK, Nankabirwa J, Ssekitooleko J, Njogu JN, RwakimariJB, et al. Malaria case-management under artemether-lumefantrine treatmentpolicy in Uganda. Malar J. 2008. doi:10.1186/1475-2875-7-181.15. Rutebemberwa E, Pariyo G, Peterson S, Tomson G, Kallander K. Utilization ofpublic or private health care providers by febrile children after user feeremoval in Uganda. Malar J. 2009. doi:10.1186/1475-2875-8-45.16. Ministry of Health Uganda, USAID, Makerere University School of PublicHealth, Health Systems 20/20. Uganda health system assessment 2011.http://health.go.ug/docs/hsa.pdf. Accessed 1 Dec 2014.17. Rutebemberwa E, Nsabagasani X, Pariyo G, Tomson G, Peterson S, KallanderK. Use of drugs, perceived drug efficacy and preferred providers for febrilechildren: implications for home management of fever. Malar J. 2009. doi:10.1186/1475-2875-8-131.18. Goodman C, Brieger W, Unwin A, Mills A, Meek S, Greer G. Medicine sellersand malaria treatment in sub-Saharan Africa: what do they do and how cantheir practice be improved? Am J Trop Med Hyg. 2007;77 Suppl 6:203–18.19. Goodman C, Kachur SP, Abdulla S, Bloland P, Mills A. Drug shop regulationand malaria treatment in Tanzania–why do shops break the rules, and doesit matter? Health Policy Plan. 2007. doi:10.1093/heapol/czm033.20. World Health Organization (WHO). The Roll Back Malaria strategy forimproving access to treatment through home management of malaria.http://whqlibdoc.who.int/hq/2005/WHO_HTM_MAL_2005.1101.pdf.Accessed 1 Dec 2014.21. Tawfik Y, Northrup R, Prysor-Jones S. Utilizing the potential of formal andinformal private practitioners in child survival: situation analysis andsummary of promising interventions. http://pdf.usaid.gov/pdf_docs/PNACP202.pdf. Accessed 1 Dec 2014.22. Wikipedia contributors. Butaleja District. http://en.wikipedia.org/wiki/Butaleja_District. Accessed 1 Dec 2014.23. Uganda Bureau of Statistics (UBOS). Higher local government statisticalabstract: Butaleja District. http://www.ubos.org/onlinefiles/uploads/ubos/2009_HLG_%20Abstract_printed/Butaleja%20DLG%20Abstract-Final.pdf.Accessed 1 Dec 2014.24. Butaleja District Local Government NEMAN, United Nations DevelopmentProgramme (UNDP)/Poverty and Environment Initiatives (PEI). DistrictKassam et al. BMC Public Health  (2016) 16:160 Page 23 of 24environmental policy. http://www.nemaug.org/district_policies/Butaleja_District_Environment_Policy.pdf. Accessed 1 Dec 2014.25. Uganda Bureau of Statistics (UBOS). 2010 Statistical abstract. http://www.ubos.org/onlinefiles/uploads/ubos/pdf%20documents/2010StatAbstract.pdf.Accessed 1 Dec 2014.26. Nanyunja M, Nabyonga-Orem J, Kato F, Kaggwa M, Katureebe C, Saweka J.Malaria treatment policy change and implementation: the case of Uganda.Malar Res Treat. 2011. doi:10.4061/2011/683167.27. AMFm Independent Evaluation Team. Independent evaluation of phase 1 ofthe Affordable Medicines Facility - Malaria (AMFm), multi-countryindependent evaluation report: Final report. http://www.theglobalfund.org/documents/amfm/AMFm_2012IEPhase1FinalReportWithoutAppendices_Report_en/. Accessed 1 Dec 2014.28. Presentation to Global Health and Innovation Conference (GHIC). Access toeffective antimalarial therapy remains low in rural Uganda - assessingunlicensed drug vendor contributions, presented on 12 April 2014 inNewhaven, Connecticut. http://www.uniteforsight.org/conference/speaker-schedule-2014.29. Strauss A, Corbin JN. Basics of qualitative research: techniques andprocedures for developing grounded theory. Thousand Oaks: SagePublications Ltd; 1998.30. Golafshani N. Understanding reliability and validity in qualitative research.Qual Rep. 2003;8:597–607.31. Williams HA, Jones CO. A critical review of behavioral issues related tomalaria control in sub-Saharan Africa: what contributions have socialscientists made? Soc Sci Med. 2004. doi:10.1016/j.socscimed.2003.11.010.32. Ellis AA, Traore S, Doumbia S, Dalglish SL, Winch PJ. Treatment actionsand treatment failure: case studies in the response to severe childhoodfebrile illness in Mali. BMC Public Health. 2012. doi:10.1186/1471-2458-12-946.33. Diaz T, George AS, Rao SR, Bangura PS, Baimba JB, McMahon SA, et al.Healthcare seeking for diarrhoea, malaria and pneumonia among childrenin four poor rural districts in Sierra Leone in the context of free health care:results of a cross-sectional survey. BMC Public Health. 2013. doi:10.1186/1471-2458-13-157.34. Uzochukwu BS, Onwujekwe EO, Onoka CA, Ughasoro MD. Rural–urbandifferences in maternal responses to childhood fever in South East Nigeria.PLoS ONE. 2008. doi:10.1371/journal.pone.0001788.35. Malik EM, Hanafi K, Ali SH, Ahmed ES, Mohamed KA. Treatment-seekingbehaviour for malaria in children under five years of age: implication forhome management in rural areas with high seasonal transmission in Sudan.Malar J. 2006. doi:10.1186/1475-2875-5-60.36. Kemble SK, Davis JC, Nalugwa T, Njama-Meya D, Hopkins H, Dorsey G,et al. Prevention and treatment strategies used for the communitymanagement of childhood fever in Kampala. Uganda Am J Trop MedHyg. 2006;74:999–1007.37. Mohan P, Iyengar SD, Agarwal K, Martines JC, Sen K. Care-seeking practicesin rural Rajasthan: barriers and facilitating factors. J Perinatol. 2008. doi:10.1038/jp.2008.167.38. Lowassa A, Mazigo HD, Mahande AM, Mwang'onde BJ, Msangi S, MahandeMJ, et al. Social economic factors and malaria transmission in Lower Moshi.Northern Tanzania Parasit Vectors. 2012. doi:10.1186/1756-3305-5-129.39. Deressa W. Treatment-seeking behaviour for febrile illness in an area ofseasonal malaria transmission in rural Ethiopia. Malar J. 2007. doi:10.1186/1475-2875-6-49.40. Hildenwall H, Rutebemberwa E, Nsabagasani X, Pariyo G, Tomson G,Peterson S. Local illness concepts: implications for management ofchildhood pneumonia in Eastern Uganda. Acta Trop. 2007. doi:10.1016/j.actatropica.2007.02.003.41. World Health Organization (WHO). Guidelines for the treatment of malaria.http://www.ncbi.nlm.nih.gov/books/NBK254223/. Accessed 1 Dec 2014.42. Nuwaha F. People's perception of malaria in Mbarara. Uganda Trop Med IntHealth. 2002. doi:10.1046/j.1365-3156.2002.00877.x.43. Kahabuka C, Kvale G, Hinderaker SG. Factors associated with severe diseasefrom malaria, pneumonia and diarrhea among children in rural Tanzania - ahospital-based cross-sectional study. BMC Infect Dis. 2012. doi:10.1186/1471-2334-12-219.44. Kiwanuka SN, Ekirapa EK, Peterson S, Okui O, Rahman MH, Peters D, et al.Access to and utilisation of health services for the poor in Uganda: asystematic review of available evidence. Trans R Soc Trop Med Hyg. 2008.doi:10.1016/j.trstmh.2008.04.023.45. ACTwatch Group. Household survey Uganda 2012 survey report. http://www.actwatch.info/sites/default/files/content/publications/attachments/ACTwatch%20HH%20Report%20Uganda%202012.pdf. Accessed 1 Dec 2014.46. ACTwatch Group. Household survey report (baseline) Republic of Uganda03/09-04/09. http://www.actwatch.info/sites/default/files/content/publications/attachments/Uganda%20Household%20Baseline%2C%20ACTwatch%202009.pdf.Accessed 1 Dec 2014.47. Chuma J, Gilson L, Molyneux C. Treatment-seeking behaviour, costburdens and coping strategies among rural and urban households inCoastal Kenya: an equity analysis. Trop Med Int Health. 2007.doi:10.1111/j.1365-3156.2007.01825.x.48. Littrell M, Gatakaa H, Evance I, Poyer S, Njogu J, Solomon T, et al.Monitoring fever treatment behaviour and equitable access to effectivemedicines in the context of initiatives to improve ACT access: baselineresults and implications for programming in six African countries. Malar J.2011. doi:10.1186/1475-2875-10-327.49. Moszynski P. Disappearance of drugs undermines Uganda's fight againstmalaria. BMJ. 2010. doi:10.1136/bmj.c2611.50. Rutebemberwa E, Kadobera D, Katureebe S, Kalyango JN, Mworozi E, PariyoG. Use of community health workers for management of malaria andpneumonia in urban and rural areas in Eastern Uganda. Am J Trop MedHyg. 2012. doi:10.4269/ajtmh.2012.11-0732.51. Konde-Lule J, Gitta SN, Lindfors A, Okuonzi S, Onama VO, Forsberg BC.Private and public health care in rural areas of Uganda. BMC Int HealthHum Rights. 2010. doi:10.1186/1472-698X-10-29.52. World Health Organization (WHO), United Nations Children's Fund (UNICEF).WHO/UNICEF Joint Statement: integrated community case management(iCCM). An equity-focused strategy to improve access to essential treatmentservices for children. http://www.who.int/maternal_child_adolescent/documents/statement_child_services_access_whounicef.pdf. Accessed 1Dec 2014.53. Ajayi IO, Browne EN, Garshong B, Bateganya F, Yusuf B, Agyei-Baffour P,et al. Feasibility and acceptability of artemisinin-based combination therapyfor the home management of malaria in four African sites. Malar J. 2008.doi:10.1186/1475-2875-7-6.54. Marsh DR, Hamer DH, Pagnoni F, Peterson S. Introduction to a specialsupplement: evidence for the implementation, effects, and impact of theintegrated community case management strategy to treat childhoodinfection. Am J Trop Med Hyg. 2012. doi:10.4269/ajtmh.2012.12-0504.55. Malimbo M, Mugisha E, Kato F, Karamagi C, Talisuna AO. Caregivers'perceived treatment failure in home-based management of fever amongUgandan children aged less than five years. Malar J. 2006. doi:10.1186/1475-2875-5-124.56. Hamer DH, Marsh DR, Peterson S, Pagnoni F. Integrated community casemanagement: next steps in addressing the implementation researchagenda. Am J Trop Med Hyg. 2012. doi:10.4269/ajtmh.2012.12-0505.57. Edberg M. Essentials of health behavior: social and behavioral theory in publichealth (texts in the essential public). Sudbury: Jones and Bartlett; 2007.58. Goodman C, Kachur SP, Abdulla S, Mwageni E, Nyoni J, Schellenberg JA, et al.Retail supply of malaria-related drugs in rural Tanzania: risks and opportunities.Trop Med Int Health. 2004. doi:10.1111/j.1365-3156.2004.01245.x.59. Mukanga D, Tibenderana JK, Kiguli J, Pariyo GW, Waiswa P, Bajunirwe F,et al. Community acceptability of use of rapid diagnostic tests for malariaby community health workers in Uganda. Malar J. 2010. doi:10.1186/1475-2875-9-203.60. Ezeoke OP, Ezumah NN, Chandler CC, Mangham-Jefferies LJ, OnwujekweOE, Wiseman V, et al. Exploring health providers' and communityperceptions and experiences with malaria tests in South-East Nigeria: acritical step towards appropriate treatment. Malar J. 2012. doi:10.1186/1475-2875-11-368.61. Baxter P, Jack S. Qualitative case study methodology: study design andimplementation for novice researchers. Qual Rep. 2008;13:544–59.Kassam et al. BMC Public Health  (2016) 16:160 Page 24 of 24


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items