Pneumonia Diagnostics Project a Qualitative Study to Explore

Pneumonia Diagnostics Project A qualitative study to explore experiences of Community Health Workers when managing pneumonia in children under five in Mpigi district, Uganda

December 2014

Dr. Diana Nanyumba, Research Officer

Dr. Akasiima Mucunguzi, Project Manager.

Table of contents

List of Abbreviations ...... 4 Acknowledgements ...... 5 Abstract ...... 6 Objective: ...... 6 Methods...... 6 Findings...... 6 Conclusion...... 6 Introduction ...... 7 Background ...... 7 Study aims and objectives ...... 8 Methods ...... 9 Study design ...... 9 Study population ...... 9 Data collection ...... 11 1. Location...... 11 2. Research assistants ...... 11 3. Consenting of participants...... 11 4. Focus Group discussions...... 12 5. Data Management...... 12 6. Data analysis ...... 12 Ethical approval ...... 13 Commentary ...... 14 Participants ...... 14 VHT knowledge and management of pneumonia...... 14 Enablers and constraints in pneumonia care at the community level ...... 16 Characteristics of an ideal diagnostic tool...... 20 Discussion ...... 22 22

Study limitations ...... 23 Conclusion: ...... 23 References ...... 24 Appendices...... 25 Appendix i ...... 25 ppendix ii ...... 29 Appendix iii ...... 31 Appendix iv...... 32 Appendix v ...... 33 Appendix vi...... 34 Appendix vii ...... 35

List of Abbreviations

ARI Acute Respiratory Infection CHW Community Health Worker DHO District Health Officer DHT District Health Team FGD Focus Group Discussion HC Health Center HSD Health Sub-district iCCM Integrated Community Case Management IMCI Integrated Management of Childhood Illnesses MC Malaria Consortium MOH Ministry of Health ORS Oral Rehydration Solution POx Pulse oximetry RAs Research Assistants RR Respiratory rate UNCST Uganda National Council for Science & Technology UNICEF United Nations Children’s Fund VHT Village Health Team VHTs Village Health Team Members. WHO World Health Organization

Acknowledgements

This study is made possible by funding from the Bill and Melinda Gates Foundation. Thank you to the VHTs in Mpigi district who volunteered their experiences caring for children with pneumonia in their communities. The research assistants for their help collecting the data, the Mpigi district health team and Ministry of Health for their unwavering support throughout this stage of the research.

A team assisted in the preparation of this report. Thank you to the following individuals for their contributions: Kevin Baker, Diana Nanyumba, Akasiima Mucunguzi.

Abstract

Objective: To capture the experiences of 24 ICCM trained VHTs in Mpigi district managing pneumonia at the community level specifically around the enablers and constraints they face as well as their preferences around the ideal pneumonia diagnostic tool that could improve the accuracy of pneumonia diagnosis which is acceptable to caretakers.

Methods. Three focus group discussions were conducted with a total 24 VHTs in Mpigi district. A discussion was conducted every third day with a group of 8 VHTS. Audio recordings of the discussions were translated from the local language to English and transcribed. The transcripts were uploaded into QSR NVIVO 10 and analysed according to predesigned thematic framework.

Findings. Most caretakers have confidence in the VHTs and the services they provide as well as the ARI timer. The VHTs demonstrated an understanding of the signs and symptoms, diagnosis and treatment of pneumonia and have a willingness to serve their communities. However they listed a number of challenges related to the ARI timer include: durability (46%) of which 64% thought it breaks down fast and 27% alluded to a short battery life; time consuming (38%); 31% thought the ticking sound every second causes confusion to the VHT and discomfort to the sick child. Low community awareness in sections of the community means that some children are not taken to the VHTS for care. 46% expressed concern that sections of the community had low confidence in them as health care workers and the services they provide with some caregivers questioning the efficacy of the medicine. Supply chain management was a major setback for all of them. The key ideal tool characteristics include: automation (88%), easy to use (71%) and durable (63%).

Conclusion. There is need for an improved pneumonia diagnostic tool for community health workers, which is easy to use, durable and also acceptable to caretakers. More work needs to be done to improve community awareness of pneumonia and the availability of community based care. An in-depth analysis of the amoxicillin supply chain bottle necks is recommended to address the underlying causes of the lack of effective access to amoxicillin.

Introduction

Background Pneumonia is a leading cause of morbidity and mortality in children less than 5 years in sub- Saharan Africa, South Asia and South East Asia. In 2013 pneumonia caused the death of an estimated 905,059 children in the said regions. In 2013, 30 countries including Uganda contributed to 85% of child pneumonia deaths experienced worldwide.

In Uganda MoH has implemented the Village Health Team (VHT) concept since 2006 and in 2010 the ICCM strategy was added to the VHT responsibilities. In the VHT concept, every village (the lowest administrative unit in Uganda) is supposed to have five community health volunteers.

The selection criteria for the volunteers are as follows: maturity (over 18); village resident; literacy in local language; good community mobiliser and communicator; dependable and trustworthy; interested in health and development; with a demonstrated level of volunteerism. The intension is to include members from different backgrounds (community medicine distributors, extension workers, peer educators, traditional birth attendants etc.)

These volunteers are trained in a five day basic package following which they are supposed to mobilize their communities for health action across a wide range of disease prevention and promotion interventions including malaria prevention, water and sanitation and family health. The primary role of Uganda’s VHTs is to prevent disease through health promotion and to refer seriously ill patients to the health facility.

Two of the five VHT members (VHTs) are further selected against other criteria and trained in a six day ICCM package. The ICCM strategy adds treatment by VHT members to the scope of VHT activities so that preventative and curative activities will complement each other. Following the training, they are supplied with a kit of pre-packaged medicines (colour coded coartem, rectal arteseunate, colour coded Amoxicillin, zinc and ORS), commodities and supplies including diagnostic tools. This should enable them to manage children with signs of uncomplicated malaria, pneumonia and diarrhoea and refer children with signs of severe illness to a health facility.

The ARI timer is the recommended timing device for assessment of fast breathing to classify pneumonia but it is not without shortcomings and challenges. Several research groups have developed applications and devices that use different approaches to measure and classify

respiratory rate but these haven’t been comprehensively evaluated for clinical performance, usability and acceptability in an endemic setting.

The pneumonia diagnostics project titled “Improved tools for the measurement of respiratory rate and oxygen saturation for the diagnosis of pneumonia” is being conducted in four countries including; Uganda, South Sudan, Ethiopia and Cambodia.

The project goal is to identify the most accurate, acceptable, scalable and user-friendly respiratory rate timers (RR) and pulse oximeters (POx) for the detection of the signs of pneumonia in children by community health workers and front-line health facility workers in the four countries.

The project objectives are;

1. To systematically review the landscape for existing RR mobile phone applications (apps), automated RR timing tools and POx devices appropriate for low income settings 2. To identify using predefined criteria the most promising and appropriate devices for field testing in sub-Saharan Africa and South East Asia. 3. To establish the accuracy of the RR timing / classification device to diagnose symptoms of pneumonia and the POx devices to measure oxygen saturation, respectively, when used by community and first level health workers in Sub-Saharan African and South Est Asia. 4. To explore the acceptability and usability of excising RR mobile phone apps, automated RR timing tools and POx devices as perceived by parents, community and first level health facility workers in Sub-Saharan Africa and South East Asia. 5. To disseminate the findings nationally and internationally and advocate for the manufacturing of selected devices for introduction to scale.

To meet the objectives the project will be implemented in six stages.

Study aims and objectives

Stage one of the project was aimed at capturing the experiences of ICCM trained VHTs in Mpigi district managing pneumonia at the community level specifically around the enablers and constraints they face as well as their preferences around the ideal pneumonia diagnostic tool that could improve the accuracy of pneumonia diagnosis which is acceptable to caretakers.

Methods

Study design

This qualitative research employed FGDs to capture information on the current constraints in diagnosing pneumonia at the community level. Three FGDs were conducted every third day to allow for data collation and compilation before the next. A group of 8 VHTs participated in a single FGD each lasting 1 to 2 hours.

Study population Mpigi district is located in central Uganda (see Fig 1) and is one of the districts that (until Dec. 2014) are supported by Malaria Consortium to implement ICCM with funding from UNICEF. It is largely rural and has implemented the VHT strategy the longest.

Mpigi district consists of two health sub- districts (HSD); Mawokota North and Mawokota South, seven sub counties and 332 villages. It has a total population of approximately 210,000 with 40,000 children less than 5 years. Mpigi has one referral hospital (HC4), 18 sub-county health centers (HC3), 412 first level health facilities (HC2) and 656 VHTs (HC1) who are all trained in ICCM and actively supported by Malaria Consortium

It is largely rural and has implemented the VHT strategy longer than all the other districts. Mpigi district has 650 VHTS who are all trained in ICCM of which up to 40% are male and 60 % are female. The participants were drawn from 6 of the 7 sub counties including; Buwama, Kituntu and Nkozi in Mawokota South and Mpigi town council, Kiringente, Kamengo in Mawokota North. Four participants were purposefully selected from each sub county to make a total of 24 participants; age and gender were put into consideration. These were representative of the wider VHT population in Mpigi district.

Of the 24 VHTs selected, 58% were female and males constituted the remaining 42%. The average age was 46 years ranging from 25 to 58.

Figure 1: Map of Mpigi district

Data collection

1. Location. Three focus group discussions (FGDs) were conducted in Mpigi district with 24 ICCM trained VHTs in three groups of eight.

Table 1. FGDs conducted in Mpigi district

FGD Location Date Number of Participating sub HSD VHTS counties 1. Mpigi Health 01/04/2014 8 Mpigi town council Mawokota North Centre IV Kamengo Mawokota North

2 Buwama Health 04/04/2014 8 Kiringente Mawokota North Centre III Buwama Mawokota South 3 Bukasa Health 09/04/2014 8 Kituntu Mawokota South centre II Nkozi Mawokota South

2. Research assistants. For this exercise four research assistants were recruited. In a two day workshop held at Malaria Consortium, they received training covering the project background, qualitative research methods specifically round about the conduct of FGDs, data collection and management and the appropriate 5trfuse of the data collection tools.

3. Consenting of participants. Before the start of each discussion, the moderator introduced the project to the VHTs. They were each handed a translated project information sheet which they read. The research officer and the project manager responded to queries that the participants raised following which the research assistants guided the participants through the consenting process. Each participant completed a pair of consent forms of which one was kept by the project team and the other by the participant for record purposes.

4. Focus Group discussions. All communication with the VHTs was in Luganda. The FGDs were conducted in quiet rooms with minimal interruption. Two of the four research assistants took notes during the discussion (field notes), one moderated and the other ensured audio recording of the discussions was done. The discussions each lasted between 1 to 2 hours.

5. Data Management. Following each FGD the team convened to debrief and share data collected. The data were verified by the research officer to ensure completeness and relevance. The field notes and audio recordings were backed up on the Malaria Consortium server, external hard drives and project laptops. The consent forms and attendance sheets are stored away safely by the research officer.

Table 3. Data collected at the end of each FGD. Principal output Specific out puts Measurement Verified by Consent forms Number of consent forms correctly 8 forms Research Officer completed Attendance sheets Attendance sheets correctly 2 attendance Research Officer completed sheets Field notes Copies of field notes handed to RO 2 copies Research Officer upon return from the field Fair notes Fair notes complied and handed to RO 1 copy Research Officer within 24 hours of FGD Audio recordings A copy stored on RO’s laptop and EHD 2 audio Research Officer upon return from the field recordings A transcript in English handed to RO 1 transcript Research Officer within 72 hours of the FGD

6. Data analysis The transcripts were uploaded into QSR NVIVO 10 and analysed according to a predesigned thematic framework.

Ethical approval

The pneumonia diagnostics project proposal was approved by the Makerere University School of Medicine Research and Ethics Committee for one year; from 07 March 2014 to 06 March 2015 (see Appendix vii). Final approval was granted by the Uganda National Council for Science and Technology on 24 March 2014 (Appendix viii)

Commentary

Participants The participating VHTs were purposefully selected from 6 of the seven sub- counties in Mpigi district. Age and gender were factored in to obtain a representative sample of participants

Mpigi HC IV Buwama HC III Bukasa HC II (09/04/2014) Interviewee Age Years of Interviewee Age Years of Interviewee Age Years of Experience experience experience M1 52 3 B1 52 10 N1 41 7 M2 44 3 B2 58 10 N2 47 7 M3 49 7 B3 38 10 N3 39 7 M4 52 3 B4 25 4 N4 44 7 M5 44 3 B5 47 10 N5 42 7 M6 45 7 B6 56 10 N6 45 8 M7 52 7 B7 56 10 N7 50 8 M8 45 8 B7 38 10 N8 39 8 Table 2: Participants of the three FGDs held in Mpigi district.

VHT knowledge and management of pneumonia.

For 25% of the VHTs pneumonia was the most common disease affecting children under the 5 years in their communities with the rest indicating that malaria affected children under five years the most.

All the participants demonstrated a good understanding of the basic signs and symptoms, diagnosis and treatment of pneumonia in children less than 5 years.

“My timer is functional and I use it all the time to examine children with cough, before I start to examine the children I normally explain to their parents that I am going to look at how the child is breathing. If they are breathing below the threshold for their age it means that they have regular cough and I will not give them medicine. They always come to me saying; my child is coughing, assess them and see if they will need medicine or not. After that I examine the child, so they trust my word.” Ugandan VHT.

“I have a timer and I use it to determine the respiratory rate of a child with cough. I expose the child’s chest and count the breaths they take in one minute, which is the respiratory rate. If it is above the upper limit for their age, then they have pneumonia.” Ugandan VHT

“In addition to what she has said, after establishing the respiratory rate, I use the sick child job aid to establish where they fall – whether they have pneumonia or not - and the dose of amoxyl to dispense given their age.” Ugandan VHT.

“Their chest falls in deep and in addition you will sometimes hear a little noise/ sound in their chest as they breathe.” Ugandan VHT.

“The child looks so weak and they labour to breath. They look like they haven’t got strength left in them.” Ugandan VHT

They acknowledged that there are varying severities of pneumonia some of which require care at a higher level necessitating referral to government health facilities and/ or private clinics.

“Some children also wheeze, I normally refer these because I cannot manage them appropriately.”

Ugandan VHT

“In addition we visit our patients in their homes to find out if they’re getting better or not. If we find that they are not improving or getting worse, we refer them to the nearest health facility.” Ugandan

VHT

“There is a stage of the disease that is not ours; those that are very ill. We have er, er referral forms.

We fill them out and refer them to hospitals.” Ugandan VHT

Enablers and constraints in pneumonia care at the community level

All the VHTs were trained to diagnose and treat pneumonia in their communities. They were supplied with ARI timers and treatment job aids to guide their management decision making processes. In addition most VHTs expressed a willingness to provide care.

“We ought to treat them, we are all the same.”- Ugandan VHT

However, 58% (14) indicated that they would benefit from regular support supervision. They expressed a need for regular refresher training in the diagnosis and treatment of pneumonia with 17 %( 4) stating that they had forgotten how to use the ARI timer.

“We also need refresher trainings to remind ourselves of things we learnt in the past… as well as new things. The more knowledge you have the better you get at what you do.”- Ugandan VHT

“The timer disturbs us I have to admit; at least from my perspective we need to be re- trained on how to use a timer.” -Ugandan VHT

“I want to agree a little with what the previous speaker said about the timer. We did not all grasp the basics of how to use it. You find that some VHTs run out of amoxyl faster than other. This suggests that they have difficulties using the timer and they dispense medicine indiscriminately” -Ugandan VHT

They attend to most of the children with pneumonia in their communities and refer the severely ill to government health facilities and private clinics.

“What I wanted to add is that most of them come to us.”- Ugandan VHT

“They usually come to us the VHTs because we live amongst them”.- Ugandan VHT

“I see most of the children in my community who suffer from pneumonia.”- Ugandan VHT

“I see them all, parents do appreciate our work. The private clinics in my village are running out of business.”- Ugandan VHT

“There is a stage of the disease that is not ours; those that are very ill. We have er, er referral forms. We fill them out and refer them to hospitals.”- Ugandan VHT

Most caretakers have confidence in the ARI timer and trust the treatment decisions taken by the VHTs when they use it to evaluate children with cough.

“Parents are always happy when you use the timer, they are confident that you are doing the right thing. They say, “They are skilled, they even have special tools you see, when she pressed it, it went tick tack” so they go home happy. When you say to them you know what, I am not going to treat this child, the children that I treat are like this and that, they go home satisfied and will not say that you simply got rid of them”- Ugandan VHT

“Parents like the timer and they trust what I say to them only if I have used the timer to assess their child.”- Ugandan VHT

They are also very confident in their roles when they use the ARI timer to diagnose pneumonia; they say it increases their credibility as healthcare providers in their communities.

“What I can say is that it is helpful to us. You see when you use it to examine a child and find that they do not have pneumonia, you can confidently send a parent home without medicine.”- Ugandan VHT.

However, they expressed challenges using the ARI timer. Forty six percent (11) VHTs suggested the ARI timer is not durable; with 64% alluding to the fact that they break down quickly, while 27% suggested that a short battery life was the reason why they it was not durable. Up to 80% of the participants had at one time or the other used a clock/ watch or mobile phone stop clock as alternative tools to determine a child’s respiratory rate.

“They break down really fast. For many of the VHTs, the timers broke down after a short while.”- Ugandan VHT

“The timers have a short battery life, they drain quickly.”- Ugandan VHT

“My timer broke down so I use my mobile phone instead.” Ugandan VHT

“We have tried it with a disco watch RN [digital watch] in my village. It does the same job. Many VHTs use this method. I have to add that this only improvisation, we need proper tools.” Ugandan VHT

The ARI timer was reported to be time consuming by 38% (9) of the participants, inconsistent by 2 and confusing by 1 respondent. Thirteen percent (13%) said the ticking sound from ARI timer was a source of discomfort to sick children.

“For me, the timer is time consuming.”- Ugandan VHT

“When you are going to use it you commit your time. It is not like fever, you have to commit time to really find out if a child has pneumonia. Many times children start to cry when you are counting and so you have to stop and wait for them to calm down. It is until they calm down that you carry on which normally means that you have to start counting all over again. You waste a lot of time with the timer.”- Ugandan VHT

“You see I repeat the process up to four times just to be sure I have the right count.”- Ugandan VHT

“If they could give us a timer which when you place on a child gives results in a minute or so rather than us having to count one, two, three... It’s confusing sometimes.”- Ugandan VHT

“The problem I have found is that children are afraid of it, they cry and this wastes time.”- Ugandan VHT

“Yes the young ones say a year old and less fear the timer, especially the ticking sound.”- Ugandan VHT

The VHTs were acutely aware that the communities they serve play a significant role in the management of pneumonia in the community. Forty two (42%) were concerned that sections of the community had little confidence in them and the services that they provide often times questioning their capabilities and the efficacy of the medicines they dispense.

“There are those who say the VHTs are not qualified. They ask “where did they go to school?” They despise us. They say we do not know anything.”- Ugandan VHT

“Some parents have doubts about the efficacy of the medicine.”- Ugandan VHT

“ ..you know, some parents don’t believe in the timer, they wonder how you get result, some think you’re making things up.”- Ugandan VHT

Twenty five percent indicated that some caretakers seek alternative sources of care including; divine intervention, local witch doctors, herbalists and homemade remedies to treat pneumonia.

“Some stay away and make their own herbal concoctions to treat pneumonia.”- Ugandan VHT 18

“Yes, I had a case recently that sought divine intervention from a Pentecostal church."- Ugandan VHT

“Depending on their reasoning, if they fail to treat the cough, they consult witch doctors.”- Ugandan VHT

Seventeen percent expressed that one of the major setbacks in community management of pneumonia is the lack of awareness about pneumonia in sections of the community and called for a scale up in pneumonia awareness campaigns.

“I attended the last world pneumonia day celebrations. It would be nice if pneumonia awareness was increased in the communities. You see with malaria, there are many awareness campaigns including distribution of free mosquito nets. This is not so for pneumonia, there is little awareness about it which affects our work, it holds us back. There is a lot out there about malaria; campaigns, nets but nothing about pneumonia. There was world pneumonia day but not many people knew about it. We would like more activities that increase pneumonia awareness; it will make our work a lot easier.”- Ugandan VHT

On a quarterly basis VHTs are supplied with medicine to treat malaria, diarrhoea and colour coded amoxicillin to treat pneumonia in children less than five years. However, medicines supplies chain management ranked as one of the major setbacks in the management of pneumonia at community level with 58% of the VHTs stating that the amoxicillin supplied did not meet the need, often times running low on stocks soon after replenishment.

“I want to add that we need more medicine. We get low stocks compared to the demand. The number of people that visit us has increased from when we first started but the stocks haven’t. Children are born every day.”- Ugandan VHT

The VHTs reported that personal protective equipment including; gloves, aprons, hand washing soaps would improve infection control at their homes. Supplies including gumboots, back packs, bicycles could improve the quality of services they provide to the community.

“We lack protection for ourselves and our children. We too have children at home and I fear that they are at risk of contracting diseases from the children that are brought to my home for care. You see our homes are like hospitals. In addition, I fear that I may be a source of infection for my family, if there are things like soaps that we could use to wash our hands after attending to these patients, I think they will help.”- Ugandan VHT

“We need aprons, children vomit on us during assessments all the time.”- Ugandan VHT 19

Characteristics of an ideal diagnostic tool. Eighty eight percent (21) of the participants described an automated diagnostic device of which 95% expressed a desire for a placed device with a digital (57%) or colour display (38%) of results. One participant described a breathalyser – like device.

“If they could give us a timer which when you place on a child gives results in a minute or so rather than us having to count one, two, three…”- Ugandan VHT

“I would make a tool that has a colour display. To get a result I would place it on the child’s chest and it would display colours to indicate whether or not they have pneumonia.”- Ugandan VHT

Usability was highlighted by 71% (17) of the participants as an important characteristic with 41% suggesting that an ideal tool should be easy to use and 58% proposing that it give quick results.

“It should be easy to use so that I can easily determine whether a child has pneumonia or not.”- Ugandan VHT

“If it were up to me I would make a tool that wouldn’t waste my time. A tool that gives results quickly.”- Ugandan VHT

Sixty three percent (15) of the participants indicated that durability is an important characteristic for a diagnostic device of which 27% (4) alluded to a long lasting tool, 4 a rechargeable tool, 3 alluded to a long battery life, one participant was referring to a mechanical robustness and still one suggested that spare batteries be provided.

“I would like for the battery life to be looked into. If we can get something that is rechargeable; it doesn’t have to be the timer.”- Ugandan VHT

“We want something long lasting and easy to store.”- Ugandan VHT

“…it would not break easily even if it was dropped accidentally on the floor.”- Ugandan VHT

“I would like for the battery life to be looked into. If we can get something that is rechargeable; it doesn’t have to be the timer.” Ugandan VHT -

25% of the respondents considered tool accuracy as an important characteristic and an equal number suggested that it should be acceptable to caregivers as it builds their confidence.

“We need one that’s more accurate than the timer because it saves time as well as medicines.”

“A tool that parents are happy with and whose results they trust.”- Ugandan VHT

Only 33% (8) of the respondents gave their thoughts on the ideal tool design; 3 thought it should like a digital thermometer, 2 a mobile phone, and three more each suggested a pen, medallion and BP apparatus.

Discussion

Most VHTS thought malaria was the most common disease in their communities however their diagnosis and treatment of malaria is presumptive based largely on fever. The sensitivity of fever to detect malaria is 100% but specificity barely reaches 10%. Kallander et al (2005) found that the practice of treating children with respiratory symptoms with antimalarial drugs was common. Given the symptom overlap between pneumonia and malaria (Kallander et al 2004), there is a suggestion that pneumonia is often misdiagnosed as malaria and therefore under reported. Kallander et al (2008), found that most children who died from pneumonia were repeatedly first treated at home with antimalarials.

The VHTs were each supplied with an ARI timer to diagnose pneumonia. However all of them have experienced one challenge or the other with it in the dispensation of their responsibilities. Most of them suggested it was not durable and often broke down while others pointed out that it has a short battery life. This is a key point as a short battery life and breaking down easily in effect means there is nothing available as a diagnostic tool Up to 80% of the participants had at one time or the other used a clock/ watch or mobile phone stop clock as alternative tools to determine a child’s respiratory rate. It was deemed time consuming requiring them to count breaths for a full minute and in most cases repeating the count to confirm accuracy. Another shortcoming of the ARI timer is the ticking sound which they said was a source of discomfort to the children. This highlights the need for an improved diagnostic tool. In addition the sound is a source of distraction for the VHTs and for this reason some of them described the ARI timer as confusing. According to Nanyonjo et al (2012) Counting RR is challenging and misclassification of observed rate is common. The ticking sound every second can distract health workers and lead to counting of the sound instead of the chest movements. This was demonstrated by Svanberg M (2012).

Most sick children are taken to VHTs, however sections of the community still have low confidence in the VHTs while others have little confidence in the services they provide including the efficacy of the medicine they dispense. The VHTs indicated that some caretakers still seek alternative sources of care including; witch doctors, local herbalists, divine intervention while some resort to homemade remedies. All this seems to suggest that increasing community awareness about pneumonia is necessary. This is in agreement with Lye et al (1996) who found that increasing community awareness about pneumonia is important in reducing the incidence of severe

pneumonia. They also found that locally adapted behaviour change communication may reduce the incidence of severe pneumonia.

It was clear in all three discussions that supply chain management of amoxicillin was a major setback in the community management of pneumonia. All VHTs had experienced spells of medicine stock outs and indicated that they did not always have it available to dispense to sick children who needed it. This undermines effective access to amoxicillin for childhood pneumonia treatment. This is agreement with the WHO/ GHWA Report (2010) which notes that consistent and regular supplies of drugs and equipment are important to to the success of CHW programs. Without these, not only are CHWs unable to perform their duties, their standing within the community can be undermined. The specific underlying supply chain management barriers (i.e. forecasting, procurement, financing) were not elucidated.

The VHTs implied that an automated tool is the most ideal as it would take the need for counting away which would be less time consuming on their part. Usability, durability and tool accuracy were the other tool characteristics they deemed important.

Study limitations Whilst the project set out to gain an understanding of VHTs’ views on RR timers and POx devices, very little information was gathered on Pox devices.

Conclusion: There is need for an improved pneumonia diagnostic tool for community health workers. More work needs to be done to improve community awareness of pneumonia and the ability of VHTs to treat pneumonia cases in children. An in-depth analysis of the amoxicillin supply chain bottle necks is recommended to address the underlying causes of the lack of effective access to amoxicillin

References

1. Nanyonjo A, Källander K. Counting breaths in children: A comparative study of the accuracy of the ARI timer alone, the ARI timer with beads and the Malaria Consortium mobile respiratory timer. Kampala: Malaria consortium; 2012.

2. Svanberg M. Can community health workers in Mozambique diagnose and treat diarrhea, malaria and pneumonia in children? Evaluation of performance one year after training in Massinga district, Mozambique. Stockholm: Karolinska Institutet; 2012.

3. Kallander K, Nsungwa-Sabiiti J, Peterson S. Symptom overlap for malaria and pneumonia – policy implications for home management strategies. Acta Trop. 2004;90:211–4. doi: 10.1016/j.actatropica.2003.11.013. [PubMed] [Cross Ref]

4. Källander K, Tomson G, Nsabagasani X, Sabiiti JN, Pariyo G, Peterson S. Can community health workers and caretakers recognise pneumonia in children? Experiences from western Uganda. Trans R Soc Trop Med Hyg. 2006;100:956–63. doi: 10.1016/j.trstmh.2005.11.004. [PubMed] [Cross Ref]

5. Lye MS, Nair RC, Choo KE, Kaur H, Lai KP. Acute respiratory tract infection: a community- based intervention study in Malaysia. J Trop Pediatr. 1996;42:138–43. doi: 10.1093/tropej/42.3.138. [PubMed] [Cross Ref]

6. Kallander K, Hildenwall H, Waiswa P, Galiwango E, Peterson S, Pariyo G. Delayed care seeking for fatal pneumonia in children aged under five years in Uganda: a case-series study. Bulletin of the World Health Organization. 2008; 86(5): 332-8.

7. World Health Organization/Global Health Workforce Alliance, Global Experience of Community Health Workers for Delivery of Health-Related Millenium Development Goals: A Systematic Review, Country Case Studies, and Recommendations for Integration into National Health Systems. 2010.

Appendices.

Appendix i Discussion Guide 1: A study to capture current constraints in diagnosing pneumonia at the community level and suggestions on characteristics of a tool that could improve the accuracy of pneumonia diagnosis which is acceptable to the caretakers

FOCUS GROUP DISCUSSION

For Community Health Workers

DISCUSSION MODERATOR: ______

COMMUNITY/Village: ______

FOCUS GROUP DISCUSSION NUMBER: _

DATE OF FGD (dd-mm-yy):

TIME STARTED (hh-mm):

TIME ENDED (hh-mm):

FGD participant name Age Sex Length of time as CHW 1 2 3 4 5 6 7 8

Introduction: Welcome everyone; thank you so much for agreeing to be part of this discussion. Today we are going to discuss your experiences as a community health worker and also your opinions on community case management of pneumonia and the tools you use to diagnose pneumonia in your community.

(Please select from the following questions to generate and guide the discussion, depending on the direction of the conversation. Make sure time is allowed for the participants to respond to each other’s comments.)

CHW knowledge and management of pneumonia in the local health situation • Of malaria, pneumonia and diarrhoea, what do you think is the most common disease? (Probe: Why? Which disease brings the most people seeking services from you?) • How do you manage children with cough? • What are the common symptoms of pneumonia? • How would you diagnose pneumonia? • How would you treat pneumonia?

Current constraints in pneumonia care at the community level • Do you think you see and attend to all the children in your community who suffer from pneumonia? (Probe: If not, why not?) • From your experience, where do parents or carers of sick children usually go to get treatment for cough or pneumonia? (Probe: Why?) • Do you feel you are generally able to provide the care you would like to provide? (Probe: If not, why not?) • Are there any areas of your work where you feel you need to learn more skills? • Are there any areas of your work where you feel you need better tools?

Tool characteristics • Do you currently use any tools to diagnose pneumonia? If so, describe them. • What do you like about the tools you currently use? How do they help you? • What do you dislike or find difficult about the tools that you use? How could this be improved on?

• Which elements of a pneumonia diagnostic tool are important to you and why? (Prompt if necessary to cover each point. Select elements randomly in each discussion) o Easy to use/not confusing o Long lasting/durable o Reliable (i.e. doesn’t break) o Accurate o Simple (i.e. not technical – probe about mobile phone versus simple counting devices) o Acceptable (i.e. parents and patients believe the diagnosis).

• If you could design your own pneumonia diagnostic tool, what would it look like? How would it work?

We have come to the end of our discussion. Is there anything that you would like to add in relation to diagnosing pneumonia in your community?

I would like to THANK YOU again for finding the time to discuss this important topic with us and sharing your knowledge and experience.

Appendix ii Pneumonia Diagnostics Study Information Sheet – Community Health Worker Focus Group Discussions

We would like you to help with a research study. This information sheet will tell you what the research involves. Please take your time reading it. It can be read out to you if you prefer. Please ask questions and you can talk it over with others if you wish.

Overall, the study aims to improve the diagnosis of pneumonia in children under five years of age by community health workers. This is the first stage of the research and will look to capture current constraints in diagnosing pneumonia at the community level and suggestions on characteristics of a tool that could improve the accuracy of pneumonia diagnosis, which is acceptable to the caretakers.

Why have I been chosen for the study? You are a community health worker trained in the care and treatment of malaria, pneumonia and diarrhoea in under-five children in a number of locations including in the area where you live. Approximately 24 community health workers will participate in this stage of the study.

What happens if I agree to take part? You will be involved in this focus group discussion lasting approximately two hours. Whether you take part is your choice. Participation is completely voluntary; you may choose not to take part or to stop at any time. You will continue to receive the same diagnostic tests and medicines as usual if you do or do not agree to participate. If you participate you will help us find out the best way to diagnose pneumonia in the community. The Focus Group will be led by a trained member of our research team and will allow you and your colleagues to discuss a number of important issues concerning the diagnosis of pneumonia in your community.

What are the benefits of taking part? If you take part, it is possible that you will only participate in this element of the study; however we cannot be sure at this stage. You will only find out later. Otherwise, there are no direct benefits to you, but this study hopes to improve the care of children with pneumonia 29

like those you provide care for in the future. Your travel costs and light refreshments will be provided.

What are the possible disadvantages and risks of taking part? There are no added risks involved in participating in this study. Whether you agree or not to take part you will continue to receive same medicines and tests from the project.

Will my participation in the study be kept confidential? Yes. The information will be stored by a number, not by your name. The information will only be available to the researchers working on the study.

What will happen to the results of the study? They will be used to improve pneumonia treatment in under-five children care and specifically help in the development of better diagnostic tools for pneumonia. The results will also be published in medical journals. You will not be identified or identifiable in any reports of publications.

What happens if the research study stops earlier than expected? If it does, we will provide you with clear information as to why.

Who is doing and paying for the research? Malaria Consortium is carrying out the study, with support from the Bill and Melinda Gates Foundation, who are funding the study.

If you have any questions at any time, please ask a member of the research team or you can contact: xxxxxxxxx

Appendix iii

Participant Consent Form. Research study: A study to capture current constraints in diagnosing pneumonia at the community level and suggestions on characteristics of a tool that could improve the accuracy of pneumonia diagnosis which is acceptable to the caretakers. 1. I confirm that I have read and understood the information sheet dated ______, explaining the above research project and I have had the opportunity to ask questions about the study. 2. I understand that my participation is voluntary and that I am free to withdraw at any time without giving any reason and without any negative consequences. In addition, should I not wish to answer any particular questions, I am free to decline. Contact number of researcher: +XXXXXXXXXXXX 3. I understand that my name will not be linked to the research materials and any personal information that could identify me will be kept strictly confidential. I understand that my responses will be anonymised and that I will not be identified or identifiable in any report, publications or presentations that result from this research. 4. I agree for the data collected from me to be used in future research. 5. I give permission for this interview to be recorded, to be used only for analysis. 6. I agree to take part in the above research project.

Name of participant Date Signature/ Thumb print

Name of person taking consent Date Signature/ Thumb print

Appendix iv

Appendix v

Appendix vi

Appendix vii