Study setting

The study was carried out in Kibaale district, a remote, rural area in mid-western Uganda. The estimated population of 613,200 is predominantly located in dispersed villages and engaged in subsistence agriculture. In Uganda, all primary healthcare services (including malaria diagnosis and treatment) are provided free-of-charge at public health facilities. Artemisinin-based combination therapy (ACT) was introduced as treatment for malaria in 2006, with artemether-lumefantrine (AL) (Coartem^®) as the first-line treatment for uncomplicated malaria [32]. In 2011, a histidine-rich protein 2 based RDT (SD BIOLINE malaria Ag-Pf, SD 05FK60), was introduced into 30 public health facilities which lacked functional microscopy, ahead of a planned nationwide roll out. (The process for introducing RDTs in the district has been described elsewhere [2].) At a similar time, RDTs were also introduced as part of an integrated community case management (ICCM) programme with community health workers, known in Uganda as Village Health Teams (VHTs), which contributed to raising community awareness of testing.

Sampling

This study drew on the same sampling approach and patient sample as the main study [2]. Using prescription data from outpatient registers over a two month period, lower level health facilities were purposively sampled according to ‘prescribing performance’ (proportion of RDT-negative patients prescribed anti-malarials), with the aim of exploring a range of clinician prescribing behaviours [2]. This resulted in a selection of 12 facilities, including seven Health Centre (HC) IIs and five HC IIIs. All prescribing clinicians at the 12 health facilities were targeted for observation, with the aim of observing three fever cases per provider. Patients were selected for observation on a rolling basis at the point of care. All observed patients (or caretakers of child patients) were targeted for semi-structured interview.

Observation and patient/caregiver interviews

The main study reported findings from observation (related to case management) and healthcare provider interviews; this paper reports findings from observation (related to communication) and patient/caregiver interviews. Data collection was carried out in November and December 2011, approximately ten months following the introduction of RDTs, a cross-sectional point when their use was expected to have been integrated into routine practice and there would be some awareness and expectation of routine testing for malaria among patients presenting with fever [2]. Observation and semi-structured interview guides were pre-tested at health centres in two neighbouring districts, as under the main study. The data was collected by four social scientist research assistants with experience in qualitative research and a range of language skills relevant to the setting. All research assistants received two days training on the specific technical scope of the study and use of the data collection tools.

Observation was conducted by two of the four research assistants. During the preparatory phase, the two observers simultaneously observed three cases and then systematically compared transcripts for discrepancies, with the aim of reducing inter-observer variability [2]. The observation guide sought to capture exchanges and interaction between the healthcare provider and the patient throughout the care process. Provider-recorded diagnosis or patient complaint, RDT result, and prescription information were also abstracted from patients’ record books. Observation procedures have been described in detail in relation to the main study [2].

Patient interviews were audio-recorded and conducted in the language of the respondent. The main languages spoken in the area are Runyoro-Rutooro, Rukiga, Rufumbira, Runyankole, Rukonjo, and Runyarwanda-Kinyarwanda, though Luganda and Kiswahili are also used. Given the range of languages, research assistants were trained to conduct oral translation of the patient interview guide at the point of interview. Once an observation was completed, the observed patient (or caregiver) was invited for interview with the observer, accompanied by a second research assistant. For observed child patients, the interview was conducted with the accompanying caregiver. Scope of enquiry included demographic characteristics, reason and expectations for visit, choice of health facility and prior care-seeking experience there, familiarity and prior experience with RDTs, and present testing and treatment experience (communication of test purpose, perceptions of testing, experience with receiving and understanding test results, perceptions of test results, understanding and acceptance of treatment provided, and overall satisfaction with visit).

Observation notes and verbatim interview transcripts were prepared in English in the field throughout data collection. To avoid loss of meaning and interpretation bias, key terms in local languages were retained alongside the English translations. Sections of the transcripts were compared with audio-recordings for review of translation accuracy.

Analysis

Analysis followed the ‘framework’ approach [33], whereby a pre-existing coding frame was developed based on the scope of enquiry to which codes were added on review of the data. All data were coded and indexed in Excel (Microsoft) and analysed according to the most salient themes by RA (MPH), with regular review and discussion with CS (MPH) and AN (MD, MPH). Content analysis of observation data also used a summative approach in order to assess relative frequency of certain aspects of communication (e.g. linguistic phrases and keywords used to communicate the purpose of testing, RDT result, diagnosis or treatment plan). Summative findings were then interpreted contextually. Triangulation of individual-level data (observed interaction and linked patient/caregiver report) was also performed for validation and deeper exploration of key themes.

Ethics, consent and permissions

Ethical approval was granted by the Uganda National Council for Science and Technology (UNCST, HS 1009). Informed oral consent was obtained from all healthcare providers prior to observation. To avoid altering routine clinical practice, clinicians were provided with consent statements and asked to inform and consent patients to the presence of the observer during consultation. If any patient had not consented to the presence of the observer at any time, the observer would have simply left the room and returned at the start of the next consultation, though this situation did not arise. Oral consent for interview, audio-recording and record book review for both RDT result and prescription information was subsequently obtained from the patient or caregiver. In this case, oral consent was requested for reasons of comfort given the low literacy among the general patient population and associated challenges with reading and signing a consent form. Participant consent (from both providers and patients/ caregivers) was witnessed by the note taker and documented in the interview notes: a series of tick marks were made as each point of consent was explained and the final consent for interview or observation noted. Consent statements also included information on the study’s broad aims, confidentiality, respondent rights and uses of the data. Unaccompanied minors were not observed.

Study participants

As reported in the findings from the main study, a total of 55 patients were observed across 22 healthcare providers [2]. No observed patients or caregivers refused to be interviewed. All patients were tested for malaria with an RDT with slightly more than two-thirds (n = 38) testing negative. Key demographic characteristics of participating patients are provided in Table 1; characteristics of the observed clinicians have been described elsewhere [2].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic characteristics of observed patients.

https://doi.org/10.1371/journal.pone.0159525.t001

Overall care process

In all of the observed health facilities, RDTs were systematically performed on patients complaining of fever. Generally, the RDT was performed after patient registration and/or consultation. There were three providers, however, who employed the RDT as a de facto screening test, rather than a confirmatory diagnostic. At most of the observed health facilities (10), RDTs were performed in the same room as patient registration and consultation. However, the RDT was performed by a different health worker from the consulting provider in nearly half of observed cases. In these cases, patients were either sent to a testing space within the same room or told to leave the room and wait to be called back in for testing. At health facilities with only one provider, some opted to ‘batch’ fever cases requiring testing and others tested patients during the consultation. Following testing, almost all patients and caregivers were told to “go out and wait to be called” or were asked to wait in the same room where they were tested (often remaining in the consultation room as health workers proceeded to register and/or consult other patients). Frequently, several activities were occurring at once in the same room, making the space congested and effectively eliminating the possibility of a confidential patient-provider exchange.

One of the health workers was on the clerking table, another one on RDTs and transferring data to the outpatient register, while the third one was dispensing the drugs to the patients. Everything is done from the same room (consultation, RDT, prescription and dispensing of drugs). The table used for clerking (registration) was also used for dispensing the drugs and the RDTs were being done on the examination bed just behind the clerking table. There is no privacy in this health facility because there are three health workers all seated in the same room, with each attending to a patient(s) and also the other patients are seated very close to the table where consultation is done. At one point, there were eight people in the room.

[Observation notes, HCII].

On average, patients encountered 1.9 health workers during the course of their visit; this was the same at both HCIIs and HCIIIs, although a greater proportion of observed patients at HCIIIs saw two or more health workers (70% vs. 55%). Many patients did not re-see the health worker who conducted their initial consultation and test. In some facilities, task shifting was also common during the course of the day, with the health worker who conducted patient registration in the morning subsequently shifting to pharmaceutical dispensing or recording patient outcomes in the outpatient register later in the day. Although the average length of patient visit (time from beginning of consultation to treatment dispensing) was 1 hour 17 minutes, the duration of the clinical encounter (time spent directly interacting with a health worker for registration, consultation, testing and dispensing) ranged from three to 35 minutes [2].

Talking about fever and malaria

Patient-provider communication surrounding malaria testing was complicated by linguistic issues. In many of the languages spoken in the study area, the concepts of fever and malaria are frequently expressed using the same word, “omuswijja”. Although “omuswijja” is a non-specific term which may refer to any febrile illness, it may also comprise other symptoms that overlap with the clinical presentation of malaria [34]. Malaria can be more precisely expressed as “fever/malaria caused by mosquitoes” (“omuswijja gw’emibu”, taking one example in Runyoro-Rutooro). However, few providers or patients seemed to make this distinction, at times employing the word “omuswijja” with the intent of conveying malaria, and at other times, overlooking any difference in meaning.

Patient agendas and acceptance of testing

When asked about expectations for their visit, patients and caregivers mostly talked about receiving medicine to treat their problem or illness as identified by the health worker. About half of all patients specifically mentioned expecting to receive “drugs for fever/malaria” or specific drugs. Just a few patients mentioned an expectation of being tested.

The majority of patients had had some prior exposure to RDTs, through having been tested themselves, having had a family member tested, or having seen the test performed by VHTs in their village. Just three of the 55 patients reported no experience with or awareness of RDTs prior to their visit. There was wide reporting of the principle of “testing before treating”, with reference to messages from radio shows, health workers or key figures in the community.

“I have never had this test done. I always bring my children to test but not myself. I have heard about this test being advertised on radio, even the VHTs in the village carry out these tests among children below five years. Even health workers encourage people to test first before getting medicine…The [radio] programme was about the importance of testing before taking drugs and they also said that it is a standard procedure to test before getting malaria tablets.”

[PT01, RDT-negative male adult, seen by HW01, nurse in charge, HCIII]

Patient acceptance of testing appeared to be high. Most patients appeared to welcome the test in order to know the illness underlying their symptoms and to enable receipt of appropriate treatment, rather than medication based on “guessing”.

“I felt happy because I knew that I would be given the right treatment basing on what the results would show…Before the introduction of this test, we were given drugs without testing and most people would either take a long time to improve or they would totally fail to heal. But now these tests guide health workers in administering the right drugs hence making their work easier.”

[PT08, RDT-positive female adult, seen by HW03, nursing assistant, HCIII]

Test purpose

Both observation and patient interviews highlighted little explanation by the healthcare provider about the purpose or rationale for testing. Prior to testing, less than half of all observed patients were told something about the test purpose that specifically related to seeking confirmation of malaria infection. Healthcare providers more commonly mentioned non-specific aims such as “testing” or “checking blood”. When malaria was mentioned, a non-specific term for “fever” was often used interchangeably with “malaria”. In two cases, it was suggested by providers that the test could potentially enable diagnosis of multiple infections (“find the cause of the sickness”). In nine cases (across seven different providers), patients received no communication prior to testing; the health worker simply initiated a blood draw without speaking to the patient. For some patients, the lack of communication seemed of little consequence given they already felt familiar with the testing process, but others reported that more information on the reason for testing would have been useful.

“I don’t think the health worker explained to me clearly what the test was for. He just told me that the child was going to be tested for malaria caused by mosquitoes. I would like him to explain what would happen if the result was positive or if it was negative and which treatment he would give him.”

[PT21, Caregiver of RDT-negative male child, seen by HW07, nursing assistant, HCIII]

Vague or limited explanations of test purpose appeared to contribute to confusion over what the test was able to detect, with many patients suggesting it could identify all febrile illnesses, differentiate between types of malaria/fever, or identify any disease. Two patients believed they were being tested for HIV and two further patients reported that the provider mentioned differentiating malaria and typhoid. A clear overlap between patients reporting insufficient explanations of test purpose and those who were unclear on the purpose of testing was observed.

“The health worker told me that if am complaining of malaria, I would be tested first to determine the type of malaria and I understood that I would get the right treatment for the type of malaria that I have.”

[PT55, RDT-negative female adult, seen by HW22, nursing assistant in charge, HCII]

“I did not know they were testing for what. I thought they were testing for HIV or what. This is the first time to hear that this thing [points at the RDT in the interviewer’s hand] tests for malaria. They tested me without telling me anything, wrote in the book and gave me medicine.”

[PT42, RDT-positive female adult, seen by HW18, nursing assistant, HCII]

Few patients were observed to, or reported, asking questions at the time of testing. Most patients reported they had no questions to ask, given their familiarity with the testing process, though some expressed reticence to ask questions, mostly due to a lack of opportunity or comfortable environment. Providers often used an instructional tone when introducing the testing process (e.g. “Wait here to get tested”); this emphasised authority may have discouraged patients from asking questions.

“I am going to check your blood for fever (omuswijja).”

[HW19, nursing assistant acting in charge, speaking to PT46, RDT-negative female adult, HCII]

“I had some questions but I didn’t ask because there were many people around and I thought they might hear what I had to say and since the health worker was alone working on all patients, I thought she might not have enough time to explain all that I wanted to ask. So I decided to keep quiet and reserve my questions for another visit.”

[PT53, RDT-negative adult, seen by HW22, nursing assistant in charge, HCII]

Communication of test result

The majority of patients (around three-quarters) were either explicitly informed of their test result or were told that they did or did not have malaria or fever. Half of these communications were specific about the test: the provider reported a negative or positive test result or mentioned something related to “testing” when informing the patient that he did or not have malaria or fever. Among RDT-positive patients, the specific test result was mentioned in slightly less than half of encounters; however, all but four of the 17 patients who tested positive were told that they had malaria. The specific test result tended to be communicated more frequently to patients who tested negative: more than two-thirds of the 38 RDT-negative cases were informed of the result or an outcome associated with testing.

[No specific mention of test results, but confirmation of malaria diagnosis:]

“Your child has malaria and she is anaemic so am referring you to [name of] hospital.”

[HW02, nurse-midwife at HCIII, speaking to caregiver of PT06, RDT-positive child]

[Test result explicitly associated with outcome:]

“Blood tests are out, and they indicate that the child has fever caused by mosquitoes, so we are going to give you some medicines to give this child.”

[HW11, nurse at HCIII, speaking to caregiver of PT27, RDT-positive child]

“RDT results are out and they are showing that the child is negative and the child does not have fever. Now we are going to treat cough and give you medicine.”

[HW14, nurse at HCIII, speaking to caregiver of PT33, RDT-negative child]

The use of non-specific illness classification terms seemed to contribute to misunderstandings about the meaning of the test result.

The health worker informs the patient that he does not have “malaria” and the patient refuses the results saying, “What?”

H: “I have told you that the test results show that you do not have malaria fever.”

P (angrily): “But I have fever.”

H: “No you don’t have fever. So we have these tablets here for cough since it was the one which came first.”

P: “For you the only tablets you give are for malaria, what if I have other types of fever?”

H: “Here we do not test for typhoid fever and if you think you have typhoid you have to go and seek treatment from elsewhere.”

[HW17, nursing assistant, HCII, and PT40, RDT-negative male adult]

Gaps in communication of test results occurred across health facilities and providers: the 13 patients who were not informed of test results were distributed across eight health facilities and 12 providers, suggesting broad inconsistencies in routine practice. In some cases, absence of communication about results appeared to be related to organisation of service delivery and clinician workload. At health facilities where multiple health workers attended patients, gaps in responsibility with regards to communication were apparent, with documentation in the patient’s record book facilitating communication between health workers, perhaps obscuring the absence of verbal communication to the patient. In some cases, test results were communicated by the dispensing health worker, and in other cases, patients never learned their results.

Patient is called in the room at 13:05 and she enters and stands next to the health worker. The health worker does not tell her the results of the test but instead asks for her book: “Give me your book?”

[Observation of PT51, RDT-negative female adult, seen by HW21, nurse in charge, HCII]

Some patients reported reaching their own conclusions about the test outcome, based on the treatment received and their prior knowledge and experience with testing and taking anti-malarials, or from reviewing what had been recorded in their record book. However, in one case, a lack of direct communication on the test outcome led to complete misinterpretation by the patient; the patient was observed to have tested negative, but he believed that he had actually tested positive and had malaria. As with the initial testing process, patients rarely raised concerns or sought clarification on the test outcome.

“[Frowning]. I knew the child had malaria from the drugs that were given to me but the health worker did not tell me anything. Whenever I am given these drugs [she looks at the AL] after the child has been tested I just know the child has malaria. But that health worker did not tell me anything.”

[PT30, caregiver of RDT-positive female child, seen by HW13, nurse in charge, HCIII]

“Can I argue with the health workers who are trained? I don’t trust the results but I cannot challenge health workers who are trained.”

[PT16, RDT-negative female adult, seen by HW06, nurse, HCIII]

Diagnosis—told what don’t have, but not what do have

Although more than two-thirds of patients who tested negative were informed of the test result, they were not always provided with an alternative diagnosis or an explanation of the rationale for treatment. Among the 38 patients who tested negative, clinicians were observed to provide three levels of diagnostic information:

1. Half (19) were told they did not have malaria and provided with a minimal or unspecific explanation about another possible cause of illness. In some cases, the provider just noted that he would therefore treat other symptoms or patient complaints.

“We are going to give medicine for fever and cough, so do not get worried, so you can now follow me so that I give you drugs.”

[HW09, nursing assistant, HCII speaking to PT23, RDT-negative male adult]

One patient found out what he was being treated for only after asking:

P: “So the tablets you have given me are for what?”

H: “For cough and flu.”

[PT40, RDT-negative male adult and HW17, nursing assistant, HCII]

2. More than a third (14) of patients who tested negative were told they did not have malaria, but were not given any alternative explanation for their illness. In many of these cases, drugs were simply dispensed.

“Now the RDT results are out and you do not have fever caused by malaria, so take this medicine [chlorphenamine], one tablet three times a day. Then this [paracetamol], two tablets three times a day, then this [trimethoprim-sulfamethoxazole], two tablets two times a day.” [Provider does not explain the different names of the drugs to the patient.]

[HW05, nursing assistant, HCIII speaking to PT15, RDT-negative female adult]

3. The remaining patients (5) were not provided with any diagnostic information.

Among patients who tested negative, a main point of frustration was the provider’s failure to reconcile the negative result with the patient’s symptoms: patients expected to be told more about their illness and expected the provider to explain why they had symptoms of malaria yet the results were negative. Some of these patients explained that they understood their symptoms could be caused by “another disease” or “something else” and that they were being given medicine for other symptoms or complaints, but remained dissatisfied at not being told more about their own diagnostic outcome.

“I was told that the results indicate that [name of child] doesn’t have malaria. I didn’t understand the results very well because my son develops high temperature especially in the evening and he also lost his appetite. So if the health worker says that he doesn’t have malaria, then I get confused and I don’t know what he is suffering from. The health worker told me that it could be the cough that is causing the high temperature and she added that she would give me medicine that will cure him. I needed to know what could be the cause of loss of appetite but I didn’t ask the health worker because there were many patients waiting to be treated.”

[PT21, caregiver of RDT-negative male child, seen by HW07, nursing assistant, HCIII]

“The nurse did not explain. She just told me that the child does not have malaria. That information was not enough, the nurse told me that the child does not have malaria and yet he convulsed last night and even has a high temperature now. What is the cause of that?”

[PT18, caregiver of RDT-negative male child, seen by HW06, nurse, HCIII]

Trusting the test result

Overall, patients reported fairly high levels of trust in the test results, particularly among patients who tested positive, who commonly observed that the results “confirmed expectations”. Across all patients, factors that appeared to be important in fostering trust of results included prior positive personal or peer experience with testing and treating (“other people come and get better”) and a belief in technology (“tests don’t lie”). Among those who tested negative, visualisation of the test results (seeing the lines on the cassette, observing other sick-looking patients who also tested negative), trusting health worker competency, and receiving a satisfactory alternative explanation for the cause of fever also appeared important in fostering patient trust in the result.

Generally, doubts about test accuracy were confined to patients who tested negative. Slightly less than a third of patients testing negative expressed doubts; this was particularly apparent among those who demonstrated confusion over what the test can detect and who felt a disconnect between feeling symptoms of malaria and not receiving an alternative diagnosis.

“I don’t know whether to trust the test or not because there are times especially in the evening when I feel feverish and get signs of malaria but if the test says that I am negative, then I don’t know whether to believe it or not.”

[PT26, RDT-negative female adult, seen by HW10, nurse in charge, HCIII]

Benefit of testing

All patients who tested positive reported that there was some benefit to testing. Most patients who tested negative also perceived some benefits from testing, such as enabling the provider to “treat what he knows” or “treat the right illness”, as well as learning that they did not have malaria. However, some RDT-negative patients reported no benefit to testing, linking this closely to the absence of an alternative diagnosis or the receipt of treatment; a few patients indicated that there would have been a benefit to testing if they had tested positive.

“I didn’t benefit anything because I have been given medicine without knowing what I am exactly suffering from.”

[PT51, RDT-negative female adult, seen by HW21, nurse in charge, HCII]

Treatment plan

Although many patients were not informed of the rationale for treatment, the dispensing clinician generally provided some explanation of how to administer prescribed medicines. However, healthcare providers rarely mentioned actions to be taken in case the patient did not improve. Referral was made in four of the 55 observed encounters, however there were no discussions relating to the execution of referral (which patients don’t always have the means to do).

Most patients expressed confidence that they would improve with the treatment received or that they were uncertain and would have to wait and see. Confidence in the treatment plan appeared to be fostered by belief in the ‘testing before treatment’ approach, a familiarity with and belief in the effectiveness of the medication received, and previous health recoveries following visits to the same health centre.

Patient dissatisfaction with treatment appeared to result primarily from the absence of an alternative diagnosis and perceptions of not receiving adequate treatment. These perceptions were influenced by patient expectations (desired treatment or expectation of receiving anti-malarials), patient understanding of treatment purpose, the adequacy (quantity and type) of drugs prescribed, and the availability of prescribed drugs at the health facility. While some patients expressed dissatisfaction over not receiving drugs, in part this appeared to reflect poor comprehension of the treatment plan, rather than not receiving AL or an alternative treatment. For example, one adult patient [PT54, RDT-negative female adult seen by HW22, nursing assistant in charge, HCII], who was told that she did not have malaria but was not informed of a diagnosis or the purpose of prescribed medicines, expressed her frustration despite receiving trimethoprim-sulfamethoxazole, mebendazole, paracetamol and chlorphenamine (during interview, she was able to identify all of the drugs except trimethoprim-sulfamethoxazole). Similarly, two caretakers of RDT-positive children also expressed concerns about the treatment plan: one because she was not informed of the test result and the second because she did not receive all prescribed drugs and was not informed about the purpose of the drugs.

“I don’t think I received the right treatment because [name of child patient] was not given drugs. I was told to buy three types of drugs and am not even sure that [child’s name] will improve because she was not given anti-malaria drugs yet she has a high temperature and she is generally weak. I think it is not the right treatment because she was not given [AL] yet they have it in stock.”

[PT48, caregiver of RDT-negative female child, seen by HW20, nurse in charge, HCII]

“When they test you and find you with malaria, you are given drugs but when they test you and tell you that you do not have malaria, you go away without drugs. The health worker did not explain the results for me to understand so as not to go away with doubt.”

[PT20, caregiver of RDT-negative male child, seen by HW07, nursing assistant, HCIII]

In spite of reported doubts about proposed treatment plans, few patients were observed or reported raising questions or requesting alternative treatments. Some patients implicated themselves as passive recipients of care, appearing either intimidated by, or respectful of, the knowledge or authority of healthcare providers. Just one patient reported explicitly asking for AL after being informed that his results were negative.

“I don’t know whether I received the right treatment because the health worker tells me that the child will get better yet she has not given me drugs for malaria. So I am not sure the child will get better. Maybe after improving, I will know that the treatment was right…I didn’t ask for any other drugs because I feared the health workers since all of them were in the room. Maybe if I was alone with one health worker in the room, then I would ask for other drugs.”

[PT36, caregiver of RDT-negative male child, seen by HW15, nurse in charge, HCII]

“I don’t know whether I got the right treatment. I will only know that it was the right treatment after I totally cure. I asked for [AL] because it’s the only anti-malaria drug that doesn’t bring dizziness and body weakness but the health worker told me that she can’t give [AL] because the results indicate that I don’t have malaria.”

[PT28, RDT-negative female adult, seen by HW12, nursing assistant, HCIII]

Limitations

Assessment of, and ability to, interpret non-verbal communication could have been aided by video recordings of clinical observations though this was not done. However, the observed exchanges between healthcare providers and patients were frequently discussed between the two observers to support consistent interpretation and documentation of the patient-provider exchange. Although all attempts were made to ensure that healthcare providers continued ‘normal, everyday practice’, some observer-expectancy effect may have occurred, in that providers may have been more inclined to perform to a higher standard than usual. Finally, negative test results were not evenly distributed by age, limiting our ability to interpret any potential differences between adult and child patients, or between patients with positive and negative test results.

Recommendations

The process through which RDTs are utilised to reach a decision about patients’ care should feature as more of a dialogue between healthcare provider and patient. This would allow providers the opportunity to understand better the patient’s condition, thus supporting the overall diagnostic process (through active and responsive—not just instructive—problem solving), as well as building patient trust and understanding of his/her own condition. A dialogical approach would also allow patients the opportunity to understand the clinical rationale behind care decisions, important if patients are to fully engage in the treatment plan, and which will ultimately drive perceptions of quality of care and future care seeking decisions [43, 44]. Improved healthcare provider communication skills are required. A stronger emphasis should be given to communication in clinical (including nursing) curricula, ad-hoc trainings, clinical mentoring and improved support supervision. Evaluations of supporting interventions introduced alongside RDTs to improve inter-personal communication skills and practice among health workers could support the prioritisation and design of appropriate strategies. The challenge faced by healthcare providers will be to accept change in an existing system of ‘opaque diagnosis’, embedded in a system of clinician authority and patient trust [20].

The absence of specific patient-provider communication standards should be addressed through the development of context-relevant communication protocols and job aids, drawing on case scenarios, expected to provide cues for clinicians to more systematically integrate information gathering and sharing into their interactions. Recent research in Uganda has identified patient-centredness as an aspiration for both health workers and community members for public health facilities [45], a useful foundation for the introduction of such initiatives. Protocols and communication guides should also provide clarification on what RDTs are able to detect.

Initiatives, which recognise and are suitably adapted to the local socio-cultural context, should be developed to generate demand for information among patients, who should be supported in understanding their health care rights and empowered to ask questions; their evaluation will also be critical to furthering understanding of local barriers and constraints. This study confirms that a shift in the patient-provider relationship from its former emphasis on paternalism, to a recognition of the importance of an informed and actively participating patient, is emerging as a growing need and desire. However, in low-resource settings where health workers are already stretched, it may not be realistic, nor desirable, to expect health workers to assume sole responsibility for actively engaging patients in the care process. Complementary interventions on the demand side should be explored to better inform patients and the general public on the care process, diagnostic tools and treatment options, with the aim of enabling patients to assume greater responsibility for the prevention, detection and treatment of health problems in a manner that supplements professional service. The strategic use of social and behaviour change communication, applying targeted messaging and tailored approaches, to promote a range of supportive preventive, care-seeking and disease management behaviours is vital for creating demand for testing and for building trust in results [46, 47], particularly when patients receive RDT-negative results and are unsure of what to do next. Communication interventions that support the adoption of diagnostic testing by both patients and providers is also a necessary step for improved treatment and surveillance of malaria, and increases the likelihood of a good return on investment for malaria programmes [48].

There are also implications for organisation of care. Clarifications on task management may be required, such as relating to who communicates test results (prescribing clinician or dispensing health worker). Where possible, physical spaces should be configured to allow for private, detailed patient history taking and assessment, and healthcare providers encouraged to respect patient confidentiality. Given RDTs were the first point-of-care diagnostic introduced at many lower level health facilities in Uganda and similar contexts, experience in their use may also have valuable implications for wider practice.