Introduction

Asthma is the most common chronic childhood illness, with an estimated global prevalence of 14% in the general population1. The prevalence of asthma symptoms in children less than 5 years in low and middle-income countries (LMIC) ranges between 11-15%2,3. A study in Uganda showed that 41.2% of children less than 5 years with acute respiratory symptoms at a tertiary care hospital had asthma symptoms4.

Under-diagnosis of asthma in young children is still a big problem in LMICs5,6,7. Although asthma symptoms commonly start in infancy, many young children with asthma symptoms remain undiagnosed and/or misdiagnosed as pneumonia and are inappropriately treated with antibiotics8,9,10. A study in Uganda indicated that more than 90% of children with asthma symptoms were misdiagnosed as pneumonia and treated with antibiotics6. Consequently, the affected children do not get appropriate care, leading to repeated and costly hospital visits and contribute to antimicrobial resistance. Untreated asthma symptoms affect growth and development and is associated with poor quality of life of the affected children and their caregivers7,11.

Diagnosing asthma in young children can be difficult because recurrent coughing and wheezing due to viral upper respiratory tract infections (URTI) is common in this age group10,12. However, the probability of asthma is high if the child has 1) recurrent symptoms of cough, wheeze and heavy breathing for 10 days or more and often triggered by URTIs, 2) worse in the night, and 3) during exercise or laughing. Some have allergic rhinitis, conjunctivitis, dermatitis, food allergy or family history of asthma. A positive response to bronchodilators and inhaled steroids (ICS) supports the diagnosis of asthma1. To date, there are no objective tests for asthma in children less than five years.

The reasons for underdiagnosis of asthma in young children could be related to low awareness and understanding of the disease among health care professionals, partly due to old textbooks and guidelines, sub-optimal clinical assessment of the children with respiratory symptoms13,14,15, and over-diagnosis of pneumonia according to the guidelines on Integrated Management of Childhood Illnesses (IMCI)16. Poor access to relevant inhaled asthma medication and equipment like spacers and nebulizers17,18,19, which are important in the process of making a diagnosis, could be contributing to the under-diagnosis.

This study sought to explore the caregivers’ and health care providers’ understanding and practices around diagnosis and management of asthma in children less than 5 years, and how they contribute to diagnostic delays and inappropriate care of the affected children. We hypothesized that understanding and management of asthma symptoms in children less than 5 years by HCWs would be poor leading to substantial underdiagnosis of the disease.

Results

Demographic characteristics of study participants

A total of 25 respondents were interviewed (Table 1): eight (8) caregivers (CGs), nine (9) healthcare workers (HCWs), four (4) herbalists, two (2) drug shop attendants (DSA) and two (2) Village Health Team (VHTs).

Table 1 Description of study participants (N = 25).
Full size table

Themes and quotes

Four main themes were chosen: Symptoms and triggers, Diagnosis, Treatment and Health care System. Each theme is described first by the CGs, then by the healthcare providers.

The quotes are provided in the text and in Supplementary Table 1.

Symptoms and triggers

Several CGs indicated that the symptoms of their child started mainly during infancy, but in some cases, the symptoms started at around 2 years. The typical symptoms were recurrent and often lasting more than two weeks or long-term. The episodes often started as a runny nose and an irritating dry or wet cough, mainly at night, associated with a sound in the chest and difficulty in breathing. In many cases it worsened, and the children ended up in the health facilities or in hospital because they could hardly breathe. Common triggers for the respiratory symptoms were cold weather and a common cold. Wheezing was commonly described as ‘a sound made by a sick chick or sleeping cat’ or ´noise in the lungs´. This was referred to as ‘okukyolya’ in the local language. Some CGs thought that the symptoms were inherited from the parents.

Since childhood, about six months old, when he catches a mild flu that you would regard “little”, there’s a way he breathes that’s so desperately difficult. In fact, most of the time we have been admitted to the hospital. It’s been happening till now (CG 7, Female)

‘Whenever he gets flu, he also gets cough, and it worsens. Within a short time, he develops difficulty in breathing.’ (CG 4, Female)

He played in the rain while at his grandmother’s home, so he started getting cold in the evening and started coughing”. The coughing kept on getting worse and he developed difficulty in breathing and wheezing. (CG 2, Female)

The HCWs indicated that they recognized that some preschool children had acute but recurrent respiratory symptoms especially a running nose and coughs that were commonly associated with difficult breathing and wheezing. In addition to the respiratory symptoms, some children presented with fever, headache, vomiting and poor appetite. The HCWs recognized a pattern where children first developed a common cold, followed by breathing difficulties, without fever and that when they treated some of those children with oral bronchodilators and steroids, they would get better, only for the symptoms to recur after a few weeks.

HCWs also indicated that in some cases, parents with asthma would indicate that their children might have asthma, because they observed that the child’s symptoms were similar, including frequent cough, difficult breathing, and a sound in the chest.

According to HCWs, the most common triggers and causes of respiratory symptoms, as described by CGs, were cold weather, common colds, dust, overcrowding in homes, exposure to tobacco smoke, poor ventilation, malnutrition, and witchcraft.

There are those who come with cough below 5 days, there are those who come with cough that has been there for more than three weeks, and there are those who come with cough that has been on and off. (HCW 5, Female)

Some parents were diagnosed with asthma, so if they notice some of the symptoms in their children, they come saying that “health worker, I think my child is developing asthma like me, because the health workers told me that I have asthma”. (HCW 9, Female)

The herbalists were commonly consulted by CGs with young children with respiratory problems like cough, breathing difficulties and wheezing. They had encountered many children with chronic or recurrent respiratory symptoms, often due cold weather, a common cold, dust, malnutrition and inheritance from parents.

They wheeze and have shortness of breath. The child wheezes and reacts to dust and cold (Herbalist 2, Female)

Asthmatic children have a common symptom of reacting to cold weather. When it’s cold they cannot even go out of the house. And the child can´t go to school if they don’t have enough clothing or a thick jacket that keeps them warm. They can´t move in the rain. (Herbalist 2, Female)

The drug shop attendants (DSA) indicated that in most cases, the CGs did not go with the children when buying medicine. The DSA mainly focused on providing what the CG wanted or based on the prescription they presented. In a few cases, they asked about symptoms, and CGs mentioned cough, breathing difficulties, common cold and fever.

When the VHTs met children with chronic/recurrent respiratory symptoms, they referred them to the health facility for care. As such, they did not have much information to describe the scope and pattern of the symptoms.

Diagnoses

When asked about the name of the diseases their children suffered from, the CGs mentioned suffocation, pneumonia, tuberculosis, bronchitis, upper respiratory tract infections or malaria. The CGs were confused and frustrated because different HCW provided many different diagnoses from time to time, and most times, they were not given a diagnosis. Some CGs thought about the possibility that their children could be having asthma; however, the asthma diagnosis had been doubted by the HCWs at primary care and in paediatric care at hospital.

Some doctors would say the boy is suffering from bronchitis, others could diagnose asthma, others would say severe pneumonia. Each doctor would give a different diagnosis. (CG 1, Female)

Initially they used to say that he had malaria, and it was causing the dry cough until he developed difficulty in breathing, and they diagnosed asthma at four years old (CG 4, Female).

There was a time I accepted it as asthma because it was too much. All the signs were those of asthma, but when I go to hospital, many doctors still insist that it is not asthma, maybe with time, but as for now they say its pneumonia. But I thought the boy had asthma. (CG 1, Female)

The HCWs reported that the diagnoses they commonly made when children presented with respiratory symptoms included pneumonia, upper respiratory tract infection, tuberculosis and malaria. Several HCWs indicated that they did not make a diagnosis of asthma in children less than 5 years. Instead, they assigned a diagnosis of bronchiolitis, ‘reactive airways disease’ or ‘allergic cough’ and congestive respiratory illness. They also attributed the repetitive symptoms to parents’ failure to adhere to prescribed medicines.

Children who presented with chronic symptoms were clinically assessed for pulmonary tuberculosis and/or HIV and were examined with chest radiographs, and GeneXpert for sputum analysis for those children who could cough up the sputum and asked about the HIV status of the parents.

The assessment for possible asthma focused on the physical examination findings, especially auscultatory wheeze, and not on the history, and when such signs were absent, asthma was not considered as a possible diagnosis.

Some HCWs recognized the acute symptoms of asthma like wheezing and difficult breathing and would have preferred to give nebulized salbutamol, but this was not available in the health facilities and so they were forced to use oral salbutamol.

In most cases we don’t rush so quickly to diagnose asthma because these children are so prone to pneumonia. Usually, I first treat pneumonia and when the treatment fails, that’s when I resort to the asthma treatment like aminophylline. (HCW 9, Female).

There are those I diagnose bronchiolitis, but there are those two- to five-year-olds, in most cases I take it as congestive respiratory illness though I don’t classify it as asthma but in most cases when you give them a bronchodilator, they become better. (HCW7, Male)

The herbalists mainly received children who had long-standing coughs, and the most common diagnosis they thought about was tuberculosis or in some cases asthma. For children who had breathing difficulties, which they described as congestion of the lungs with mucus, they made a diagnosis of asthma and treated as such. For other presentations, they referred children to health facilities for investigations and care.

I tell them that based on the symptoms the child is having, the child might be having asthma or TB, so the child needs to be investigated. I tell them to take the child for testing, then we proceed with the treatment. (Herbalist 1, Female).

Although the drug shop attendants did not interfere with most of the children´s prescriptions, they thought that the most common diagnosis in relation to cough symptoms was pneumonia. For some of the children whose CGs reported repetitive symptoms, they told them that their children had asthma, and that it was incurable.

I tell them it’s pneumonia. It’s the most common around here. Asthma is there but most of the time its pneumonia and cough. (DSA 2, Female)

Treatment

The CGs consistently described two clear patterns of how their children were treated: starting with one type of medication, mainly antibiotics and cough syrups if the child had presented with acute or recurrent symptoms or antimalaria medicine if the child had fever. The children were repetitively given the same medicines. The CGs also frequently thought that the child had pneumonia and so would go to pharmacies or drug shops and buy antibiotics. However, the medicines were often ineffective. They expected to get medicines to cure the disease and were disappointed when they were told that the medicine only relieves symptoms.

The third time we took him to the government health facility, he did not completely improve on the medicines…which were mainly cough syrups and antibiotics like Co-trimoxazole and Amoxicillin. (CG 4, Female)

They were still treating malaria because of the fevers, particularly he was given ACT, then after discovering that it was pneumonia, he was given injections of X-pen and Gentamycin and sometimes they would add Panadol and amoxyl. (CG 3, Male)

In two cases, the children were prescribed inhalers, but these were not available in the health facilities, so they were expected to buy them from private pharmacies, which they could not afford. One CG resisted in giving the inhaler to the child, because she doubted that the child had asthma.

Some CGs were told by some HCWs to go to the herbalist and have herbal medicines to treat the symptoms because the children were still too young to use asthma medicines.

We were given some inhalers but me I have never used them because I didn’t accept that it is asthma because in the family background, we don’t have anybody having asthma. (CG 1, Female)

All HCWs reported that they prescribed antibiotics for children who presented with acute symptoms of cough and difficult breathing because they had made a diagnosis of pneumonia. Some of them also prescribed antimalaria medication when children presented with fever in addition to the antibiotics. Some HCWs prescribed antibiotics even when they noted that the children had no signs of pneumonia, and had clearly been classified as ‘No pneumonia, cough or cold’ according to IMCI guidelines. A few HCWs adhered to the guidelines and did not prescribe antibiotics. They would prescribe cough syrup, Vitamin C and antihistamines if the children had a common cold and advised the CGs to give the child fruits. On the other hand, many mothers complained about not getting the antibiotics as they expected and the HCWs explained to them why the child did not need antibiotics.

HCWs used the diagnosis of allergic cough when the cough was triggered by cold weather and gave oral prednisolone which resulted in a quick recovery. Another HCWs reported that when children presented with breathing difficulties associated with wheezing, she would prescribe intravenous medications especially aminophylline and hydrocortisone and oral salbutamol and prednisolone to use at home.

Sometimes, HCW gave oral bronchodilators, but when the children were referred to the regional referral hospitals, some children received inhaled medicines.

While some HCWs indicated that they would prescribe inhalers to children with wheeze if available, others thought that inhaled medicines were not suitable for young children and would cause damage to the airways, and that if they started using them during early childhood, they would need to take the inhalers until adulthood. The HCWs also indicated that when they wanted to give the child inhaled medicine, the CGs did not want inhalers at all and would ask if there were no tablets instead.

If it is a simple cough, we initiate them on Septrin and if they respond, then that’s ok. We are also giving cough mixture called Beta-linctus. if it’s a severe cough, we give IV treatment. (HCW 3, Male)

I have realized that many of the children suffer from allergic coughs… and when you give a bronchodilator and at times a steroid, they tend to get better very fast. (HCW 9, Female)

If its due to either weather that it’s an allergic cough so we advise the CG on how to handle the kid, how to give some steroids (tablets). (HCW 3, Male)

They don’t like it (inhaled medicine). If you tell them to buy it, they say that “if you use that one (inhaled medicine), your child may never heal. (HCW 9, Female)

The herbalists explained that in most cases, CGs went to them to seek care for their children at a point when they felt that the health system had failed them. They indicated confidence that they could treat children with asthma, and that they got cured within a short time. They perceived themselves as problem solvers. The herbalists used herbs from different parts of the country and used foods and believed that when the foods boosted the child’s immunity, the asthma symptoms disappeared. They recognized the importance of also taking the children for an assessment by trained HCWs, although realizing that hospital care often doesn’t make a difference.

Their parents take them to the hospital but no difference. Then they ask me if I had medicine that can cure asthma. I tell them that I do, give it to them and they are cured. They cure completely. (Herbalist 2, Female)

The drug shop attendants commonly provided what the CGs asked for or followed prescriptions. They sometimes advised use of antibiotics when they (CGs) described symptoms which they (DSAs) thought were due to pneumonia.

The healthcare system

The CGs were frustrated with the healthcare system. They repeatedly visited the health facilities and were unsatisfied and tired of visiting multiple HCWs. The different HCWs provided many different diagnoses from time to time.

Moreover, they were unhappy because the drugs were unavailable at the health facilities, and they had to buy them. Several could not afford to buy the medicine. In addition, instead of spending money for transport to the health facilities, they resorted to buying medicines at local drug shops using previous prescriptions, whenever the children developed respiratory symptoms.

She is fed up going to the health facility. She is fed up. Every time she is in the hospital. She is tired of moving up and down due to the child’s sickness. (CG 3, Male)

The HCWs expressed similar sentiments of lack of medicines in the health facilities and indicated that CGs spent a long time at the facilities but often ended up with no drugs.

The HCWs were frustrated when caregivers sought care late, when the symptoms had progressed, often due to poverty. The herbalists agreed.

Although the HCWs indicated that they knew about inhaled salbutamol, most had never used it because it was not available at their health facilities. Due to lack of drugs and limited equipment like nebulizers and inhaled medicines as well as oxygen, their ability to care for their patients were reduced. The lack of medicine in their health facilities resulted in self-medication from the nearest drug stores. HCWs were challenged by little continuity and follow ups.

Then another issue is lack of drugs at government facilities. Our mothers come in, they pass through the system, someone comes in at 8 or 9am, goes through the line of laboratory, through the line of drugs, by 2 pm, the dispensary is telling her that they are out of stock. This is an embarrassment. (HCW 1, Male)

Then often with severe pneumonia, our mothers come in late. They wait for the situations to worsen because you ask ‘ how long has the child had fast breathing’ and they say four days and the chest indrawing, two days. (HCW 1; Male)

Self-medication, sometimes with the wrong drugs. Most CGs having children with these complications are frustrated with the system and don’t use qualified health workers. (HCW 1, Male).

Discussion

We aimed to explore the understanding, practices, and perspectives regarding diagnosis and management of chronic or recurrent respiratory symptoms in children less than 5 years of age. Qualitative interviews with caregivers (CGs), healthcare workers (HCWs), herbalists, drug shop attendants (DSAs) and Village Health Team (VHTs) in Jinja district in Uganda were triangulated.

All CGs described that their child had recurrent and/or long-term cough, wheeze and respiratory distress, which is suggestive of asthma1. The CGs experienced that their child either got no diagnosis or frequently were diagnosed with pneumonia, bronchiolitis or bronchitis, and sometimes with malaria by the HCWs. The children were repeatedly treated with antibiotics, often combined with other drugs. This is in accordance with the HCWs interviews, informing that children with respiratory symptoms often were diagnosed with pneumonia or bronchitis and treated with antibiotics, even when children didn´t present with fever. Children were examined for tuberculosis if they had long term or recurrent cough. Malaria was expected when a child had high fever.

The HCWs indicated that they did not make a diagnosis of asthma in children less than five years. Instead, they assigned a diagnosis of bronchiolitis, ‘allergic cough’, congestive respiratory illness or allergic cough. In contrast, the herbalist did use the asthma diagnosis in young children and provided them with herbal medicine and dietary advice.

The study strongly indicated an underdiagnosis of asthma and an overdiagnosis of pneumonia. These findings are similar to those from other studies5,6,7,13 and highlight the urgent need for interventions to improve clinical care practices and health system capacity strengthening, to support early and correct assessment of children with respiratory problems, and optimal care.

Antibiotics were prescribed for children with cough, even in cases where the HCWs diagnosed obstructive airways disease and prescribed oral bronchodilators and oral systemic steroids. Studies in rural Kyrgyzstan, Vietnam and Uganda also indicated the frequent misdiagnosis of childhood asthma as pneumonia and the irrational antibiotic use6,7,15. The inappropriate prescription of antibiotics by the HCWs, and the CGs frequent purchase of the same antibiotics over the counter is a vast problem, expensive for the CGs and society and contributing to the antibiotic resistance crisis20. Antibiotic resistance is a growing public health challenge especially in low-income settings4,7,9,20,21,22.

The presence of wheeze led the HCWs to think about an obstructive airways disease, but the HCWs doubted the asthma diagnosis before 5 years. However, even when typical recurrent or long-term asthma symptoms were told, but without auscultatory wheeze noticed in the consultation, or when CGs strongly and rightly asked if their child could suffer from asthmatic symptoms, HCWs doubted the possibility of asthma. Terminology gaps probably extended the diagnostic difficulties. In many languages in Uganda, there is no single term for wheezing. Moreover, there is no direct term for wheeze in many languages and usually the assessment is based on descriptions such as a ‘whistling sound’ or ‘noise made by a sleeping cat or water boiling in a kettle’ 23, with some caregivers describing it as a sound arising from the throat24,25, leading to misdiagnosis or diagnostic delays.

The HCWs and some CGs thought that asthma is a serious disease and cannot be diagnosed in young children, which also has been shown in other studies25. A study in Uganda indicated that many HCWs are uncomfortable with assessing patients with chronic respiratory symptoms [28]. Probably the lack of clarity on asthma in young children in textbooks and clinical guidelines on management of respiratory illnesses in children16,26 could be contributing to the limited knowledge among the HCWs.

In some cases, children with chronic cough were assessed for possible pulmonary TB, which is an appropriate practice in areas with high TB burden like Uganda. However, other possible causes of chronic cough such as asthma were not assessed. This could be due to low awareness and lack of knowledge and skills in management of childhood asthma14.

Both HCWs and CGs thought that when a child with recurrent respiratory symptoms had high fever it could be due to malaria. In such cases, children were given anti-malaria medication, often in addition to antibiotics. This is probably related to symptom overlap between malaria and pneumonia. Studies in malaria-endemic countries in Africa have shown symptom overlap between malaria and pneumonia in up to 30% of cases27,28,29. It is therefore not surprising that the CGs and HCWs thought about the presence of malaria in children presenting with fever and cough, although this could lead to over-diagnosis of malaria and/or pneumonia and under-diagnosis of asthma. Careful clinical assessment for possible asthma and testing for malaria is recommended.

Healthcare system challenges

The study highlighted numerous health system challenges for the CGs and HCWs including the children’s multiple visits to the health facilities, lack of continuity of care, lack of diagnostics and time for explaining the treatment. This is in accordance with a recent Ugandan study showing that clinical consultations of young children presenting with respiratory symptoms were typically very brief, lasting 2–4 min13, much like a study in low-income settings in Greece, Vietnam, and Kyrgyzstan15. Furthermore, the lack of continuity of care made it difficult for the HCWs to learn from the outcomes of their consultations. Similarly, the lack of medicine including inhaled asthma medicines at the government facilities made it difficult for HCWs to become familiar with inhaled medicine and confirm the diagnosis of asthma, as recommended in guidelines1. These challenges probably contributed to delayed diagnosis, misdiagnosis and repetitive consultations and hospitalisations and high healthcare costs to the caregiver and for the health system, similar to findings in other studies in Sub-Saharan Africa5,30,31.

The health system challenges were partly responsible for CGs seeking care elsewhere including local herbalists. Interestingly, some herbalists acknowledged and used the asthma diagnosis in the young children and treated with herbs and advice on lifestyle and food supplements. Little is known about the herbs used by the herbalist in this study, and whether they may contain bronchodilators and/or steroids or other active ingredients which are responsible for the symptom’s relief. The use of supplements is potentially a good practice especially in Uganda where many children suffer chronic undernutrition32, but requires further research.

Methodological considerations

The main strengths of this study were that it involved the scope of stakeholders that are commonly involved in the care of children with asthma symptoms; CGs, HCWs, local herbalists, drug shop attendants and community health workers. This approach helped to understand the similarities and divergence in understanding and perspectives. Furthermore, this knowledge is helpful in designing strategic interventions for diagnosing and managing children with asthma symptoms.

The study was conducted in rural primary care settings where most of the children present for care. These findings are therefore generalizable to most of the settings in Uganda and similar settings globally.

Even though the study was conducted nine years ago, the findings are still very relevant because the challenges around asthma care still exist, given that there have not been any interventions to address the challenges discussed in this paper. Additionally, there is still very little information about the practices, experiences and perspectives of management of young children with recurrent respiratory symptoms in low-resource settings particularly in Africa. These findings can be useful in similar settings to inform further research on interventions for improvements.

The major limitation for this study was that we could not get the views of the CGs who did not seek any care for their children’s respiratory symptoms or sought care elsewhere, because the CGs who interviewed were identified through the health facilities as having young children with recurrent and/or long-term respiratory illnesses.

Conclusions and recommendations

Under-diagnosis of asthma in children less than 5 years is still a challenging problem in LMICs. This qualitative study explained how misdiagnosis of asthma as pneumonia and malaria in young children leads to overuse of antibiotics and antimalarial medicines and leads to prolonged illness and suffering to the child and caregiver, and high health care costs.

The study indicated, that HCWs had insufficient knowledge about asthma epidemiology, symptoms and treatment and the lack of inhaled asthma medicine as standard prevented the diagnostic evidence. The high pressure on the healthcare system, with multiple visits to the health facilities due to sub-optimal care and lack of continuity of care, diminished the opportunity for relevant diagnostic steps, treatment and follow ups.

The study underscores the need for a comprehensive strategy that supports appropriate asthma care including updated textbooks and guidelines and an urgent need to strengthen the capacity of the HCWs and health system to appropriately diagnose and manage children with respiratory symptoms. Availability of inhaled asthma medicines and delivery devices, and a chronic care model is crucial. Research on the potential role of local herbs and nutrition in management of asthma are also recommended.

Methods

This was a cross-sectional qualitative study conducted between June and August 2016. We adopted the Consolidated criteria for reporting qualitative research (COREQ) checklist to promote complete and transparent reporting of our findings33. The study was conducted in Health Centre IVs (HC IVs) and communities in Jinja district, Eastern Uganda. HC IVs are primary care facilities that offer both out- and in-patient medical care as well as emergency obstetric care.

The study participants were caregivers (CGs) of young children with recurrent and/or chronic respiratory illnesses and health care providers. The CGs were identified in two ways; 1) health facility records were checked to identify children who visited the health facility (at least once a month) for respiratory symptoms, 2) CGs who sought care for their children with recurrent respiratory symptoms during the study period. Children with tuberculosis were excluded.

The health care providers were divided into four categories: Healthcare workers (HCWs) who provide clinical care at HC IVs (doctors, clinical officers, nurses), herbalists, drug shop attendants (DSA) and Village health Teams (VHTs). The clinicians were selected based on their experience in clinical care of children (at least 6 months). We aimed to have a mixture of medical doctors, nurses and clinical officers. The VHTs were identified by the health unit staff. The DSAs and herbalists were identified by the VHTs. For all categories of respondents, women and men were purposively included.

Clinical officers undergo a three-year diploma course in clinical medicine and are the major clinical care providers in primary care facilities in Uganda. Although nurses are primarily trained and expected to provide nursing care, they are also involved in clinical consultations in Uganda34.

The HC IVs are linked to the communities through community health workers referred to as VHTs who receive basic training on health issues of public health interest like immunization, recognition of danger signs in children and pregnant women, and management of common diseases. Drug shops are small privately-owned drug stores with a limited range of medications, mainly oral antimalarials, antibiotics, cough syrups and painkillers. They are managed by nurses, midwives or pharmacy technicians that dispense medications either based on the prescriptions or symptoms presented by patients. In this study, they are collectively referred to as DSAs. The research assistants contacted the identified potential participants either physically (if they were in the health facilities) or through phone calls and provided information about the study. For those who accepted to participate, further arrangements for the interviews including date, time and venue that were private, feasible to conduct interviews and convenient to the participants, were discussed and agreed upon.

The interview guides were developed by two authors (Nantanda and Østergaard), based on the study objectives. The interview guides were pilot tested on 2 CGs and 2 HCWs and revised. During the pilot testing, the research assistants summarised the findings using a debrief matrix, which was then used to discuss the findings and the necessary changes made. All interviews were conducted after obtaining written informed consent from the participants. The interviews lasted an average of 60 min and were audio-recorded. The research assistants summarized their findings in a standardized matrix, and any emerging issues identified, were further explored in subsequent interviews. This process also helped to identify the point of saturation35. The research assistants were fluent in English and Lusoga (the local language) and the interviews with the caretakers, herbalists, VHTs and DSAs were conducted in Lusoga and later translated to English.

Data management, analysis, and interpretation

A hybrid content-thematic analysis approach36 combining inductive and deductive methods, was employed. To facilitate this, all interviews were audio recorded and transcribed verbatim, with those conducted in Lusoga being translated into English for consistency. All the transcripts and audio recorders were kept securely at Makerere University Lung Institute. To complete the analysis, we thoroughly read the transcripts to familiarize ourselves with the data. We then initiated the coding process with a subset of four transcripts, aiming to uncover and identify prominent and recurring patterns and themes. The initial codes and coding framework was then reviewed and refined by the qualitative research team to ensure consistency, accuracy, and clarity, resulting in a revised codebook.

We indexed the data by assigning specific codes to relevant segments, and then through a process of iterative analysis, we identified and developed higher-level categories that helped to clarify and refine the meaning of the data in relation to each research question. Coding was done by grouping data in line with the themes and sub-themes and these were discussed among the co-authors. Key quotes were collaboratively reviewed and discussed by the co-authors. Subsequently selected quotes were used to illustrate and present the findings in the results section.

The study was approved by Mulago Hospital Research and Ethics Committee (MHREC), the Uganda National Council for Science and Technology (UNCST), the Leiden University Medical Centre ethical review board in The Netherlands, and The Danish Data Protection Agency (J.nr. 2017-41- 5051). Administrative clearance to collect the data from the health facilities was provided by the Jinja District Health Officer (DHO) who directly supervises the health facilities in the district.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.