Data collection.

We consulted institutions within the study countries and online sources for national policy documents that describe characteristics relating to CHW education, duration of training, selection, and scope of practice. To understand CHW challenges, we conducted focus group discussions (FGDs) and key informant interviews (KIIs) with purposively sampled community and formal health system level stakeholders in each study country between August 2015 and March 2016. Invitation of study participants was done through email, phone calls, and, where possible, face-to-face discussion.

A topic guide was developed to explore CHWs’ activities, including challenges that affect their performance. Relevant revisions were made after pretesting the guide and obtaining inputs from in-country partners. Data collection continued until data saturation was realised [21]. Data saturation was established through an iterative process of preliminary data analysis during data collection. It was noted that data from additional discussions and interviews continued to confirm emerging themes.

“Not knowing enough”.

Various stakeholders, especially nurse-midwives, expressed concerns over an inadequate breadth and depth of the CHW training curriculum, which constrained the acquisition of new knowledge and, consequently, affected their proficiency in assigned roles. Despite the relatively long training period of Level 2 paraprofessional CHWs to enable them to take on complex roles, including clinical procedures [25, 28, 34], there were concerns about their capacity to comprehend and apply the principles underpinning these clinical procedures. For example, nurse-midwives complained that CHWs’ limited theoretical understanding of the principles made teaching clinical procedures to Level 2 paraprofessional CHWs challenging. A nurse-midwife in Nigeria explained, "And you know that this medical line, sometimes it will be very hard if you don’t know the theoretical aspects of it, for you to be good in practical, it goes together. When you [CHW] know the theory, when they teach you the practical, you will be able to understand." Similarly, the different stakeholders acknowledged the inadequacy of Level 1 paraprofessional CHWs and explained how this affected their health education role. A midwife in Kenya narrated how limited understanding shapes outcomes of health education sessions: "Most of them find it very challenging to go there and explain a certain method [contraceptives] because they don’t have most of the information…they are asked some very difficult questions like side effects so that they can assure them [service recipients] maybe it is a normal side effect. Maybe some mothers have experienced it, and that is why they are changing to another method."

Reasons advanced for inadequate CHW knowledge fell under four categories:

1. a. Limited training opportunities for CHWs in remote regions: CHWs providing services in remote areas complained of limited training opportunities. CHWs in rural and remote areas explained that CHWs in the suburban areas were attending training sessions at their expense because those in the suburban areas were more visible and geographically closer to the district/state programme office, where training invitations were prepared and mailed. This challenge was more popular in the African study countries and explained by a CHW in Malawi: "Sometimes the people from the facility [suburban areas] would go to a workshop on behalf of the other people without their knowledge pretending to be from these areas [rural areas] making us miss information. In the end, the HSAs and the community will suffer."
2. b. Non-availability of equipment for knowledge application: According to CHWs, the non-availability of MNH-relevant equipment and tools prevented them from translating their theoretical to experiential knowledge. A CHW in Nigeria talked about lacking requisite equipment and commodities for applying knowledge gained from a training session, thereby leading to progressive depreciation of newly acquired knowledge: "In that training, we were able to learn so many things, but coming down to our facility, those things [equipment] are not here for us to put them to practice the way we saw them there [at the training]. So, the knowledge begins to die gradually. But in fact, that programme [training] was very, very interesting."
3. c. Long stretch theoretical training without community-level practice between modules: There were insinuations that the approach to training Level 1 paraprofessional CHWs in Africa may be contributing to poor knowledge retention. According to the training manual [29], the entire training curriculum was delivered within a few days to weeks rather than spacing training days. This approach deprived CHWs of the opportunity to practise new concepts and gain experiential knowledge before learning new ones. A programme staff in Kenya explained, "…whatever you are to train them for can be done in a day, a day each during the six weeks rather than you do it in three days or five days at a stretch… when you go for one week you are giving these CHVs so many things some of them will not practise them." In contrast, Level 1 paraprofessional CHWs in South-East Asia (ASHA) undergo training that is spaced-out and assessed a few months after pre-service training for gain in experiential knowledge and to determine if they have been practising the knowledge gained from the previous training. This assessment would qualify them for additional training, as explained by a programme staff in India: "We assess the knowledge level of ASHAs from the induction training. We identify ASHAs who have been practising the things because they recall them, and those who do not practise can’t recall them and may not continue for further training. For skills, we ask them, show me how you measure temperature, how do you wrap the baby…"
4. d. Proliferation of roles limiting proficiency in assigned roles: Another threat to knowledge and skill retention was the additional roles assigned formally or informally to CHWs by governmental and non-governmental organisations, thereby preventing them from attaining proficiency in previously assigned roles. A programme staff in Bangladesh explained, "Whatever new thing that is coming to our community, this will be given first to the family welfare assistants [CHW]. She is given IMCH [Integrated Maternal and Child Health] training; she is given nutrition training. They have to know everything because they are the door worker… And […] they are only involved with our family planning services. One person cannot do the jobs of many people. It is a challenge for them."

“Not having enough”.

Across study countries, CHWs were challenged by insufficient infrastructural support, which negatively affected service delivery and coverage, but the popularity of this challenge varied with CHW category and primary workstation:

1. a. Lack of accommodation for Level 2 paraprofessional CHWs resulted in shorter working hours: Level 2 paraprofessional CHWs who may not be from the community served were often challenged by a lack of habitable accommodation within the community. As explained by various stakeholders, there were remote communities with low literacy rates with hardly any community member with the required academic qualification for the CHW training school. To ensure that health facilities in these areas were operational, CHWs from other communities were often posted to provide services and were expected to make their accommodation arrangement. This challenge was mainly in the African study countries where being a resident or in marriage to a community member was not a requirement for CHWs, unlike South-East Asian study countries (Table 2). A programme staff in Nigeria explained that CHWs may not be from the community served, "There are some communities that don’t have [individuals with requisite educational qualification for CHW training school]. They need a health facility, and the government has made provision for a health facility there. So, if they don’t have, you have no choice but to post somebody [CHW] from elsewhere". The CHWs reported that they often found it challenging to reside within the community. Therefore, they resorted to shortening their working hours to create adequate travel time to their residence outside the community and limit their availability for emergencies. Another programme staff in Nigeria submitted that "Most of the facilities are not offering 24-hour services because they don’t have the manpower, as the staff don’t have accommodation to stay in the community, but delivery does not tell you when it will come; it can come at any time." Still, CHWs were often discouraged from staying overnight at the health facility because of security challenges such as hoodlums’ attacks. A community health committee member in Nigeria explained, "This building behind was built to accommodate the health workers to stay, to be able to work at night, but the abattoir hoodlums did not allow them to. All these doors were vandalised."
2. b. Inadequate transportation support with consequent lower service coverage and additional work pressure: CHWs complained of the bad roads leading to hard-to-reach communities. They narrated how they would walk long distances to deliver services thereby increasing their non-productive time and lowering service coverage. Worse affected were the Level 1 paraprofessional CHWs who are expected to deliver services through home visits. A Level 1 paraprofessional CHW in Bangladesh lamented, "We face problems carrying logistics. We do not have any vehicle, and in many communities, there is no road to reach them, so we need to carry the logistics and medicine on our own". Furthermore, poor transportation support caused additional work pressure for CHWs who were expected to accompany women in labour to the hospital but were forced to conduct unsupervised deliveries (despite not having any midwifery training) when the women progressed to an advanced stage of labour while walking to the health facility. In Malawi, a community health committee member narrated, "The problem that we have here is transport. This sometimes makes women deliver in the bush on their way to the hospital."
3. c. Poor availability of supplies limits service coverage: Across study countries, there were complaints about the non-availability of work tools such as gumboots and umbrellas, which limited their capacity for service coverage during the rainy season, especially among Level 1 paraprofessional CHWs whose role largely depends on conducting home visits. A programme staff in Kenya explained, "We have no control over the weather, but what if they were provided with gumboots, raincoats, and an umbrella, how nice would this have been."

“Not feeling wanted”.

Various stakeholder groups underscored the importance of social acceptance in CHW service coverage. Still, they clarified that this was not guaranteed due to socio-cultural beliefs, the influence of demographic factors and visit time conflicts:

1. a. Local socio-cultural beliefs influenced health-seeking behaviour and, consequently, health outcomes: Across study countries, service recipients prioritised or preferred spiritual to clinical solutions for medical conditions. This sometimes resulted in non-compliance with health education sessions delivered by CHWs. A CHW in India explained, "Among some nomadic castes, like Nandiwale, they don’t come forward to take injection [anti-tetanus] unless they perform God’s ritual. They come at a later stage, such as in the 7th or 8th month of pregnancy…they used to deliver at home." Similarly, there were occasions when non-compliance and poor health-seeking behaviour resulted in poor health outcomes. A CHW in Kenya narrated, "There is a woman who delivered at home, and the baby died when we reached there, and we wanted to take the mother to the hospital [but] they refused and wanted to pray for her. So, they started praying and eventually the mother died."
2. b. Gender difference between CHWs and their service recipients affected CHW social acceptance: Across the study countries, participants reported preference for MNH care provided by female CHWs, reflecting the socio-cultural norms guiding male-female interactions. In Kenya, a programme staff explained, "Even men were not very comfortable with male CHVs [CHWs] coming to talk to their women, sometimes in the houses… but prefer if it were women…" Nonetheless, male CHWs in African study countries provided MNH services to female service recipients, including management of labour and delivery, unlike in South-East Asian study countries. Consequently, there was poor social acceptance and utilisation of the MNH services provided by male CHWs as explained by a programme staff in Nigeria. “Because in the northern part of Nigeria, any labour room that is managed by a male, especially in the communities, you won’t see anybody [pregnant women], they will continue seeing their TBAs."
3. c. Intra-community differences in language/dialect limited communication: Across study countries, there was a diversity in languages and dialects within communities, which affected communication. For example, in India, a lady health worker explained the difficulty CHWs experience while interacting with individuals who may not speak a popular national language apart from their local dialect. She stated: "Although ASHA is from the same community, but every 12 km, the language differs, maybe regarding tone or meanings. Some words differ in meanings, though they are commonly used. So, ASHA has to take care of all these things, especially when interacting with people from the low educational background or illiterate community."

Furthermore, the age difference between CHWs and their service recipients also contributed to the communication barrier faced by CHWs, as the vocabulary and choice of words tended to vary across age groups. A community health committee member in Kenya explained, "We have an age barrier; for example, if they [CHWs] go to a young girl’s house, sometimes, they want information from her. There are some words she uses that they don’t understand, there is a small barrier."

1. d. Undue familiarity and short training duration resulted in CHW services being undervalued: In contrast to Level 2 paraprofessional CHWs, Level 1 paraprofessional CHWs with shorter training and predominantly community members were often perceived as lacking the requisite proficiency to perform their assigned roles. This was partly explained by the perceived inadequacy of their training and undue familiarity from being co-community members. A programme staff in India explained that since ASHAs "belong to that village, community members are of the perception that she won’t know anything."
2. e. Time conflict with the daily routine of service recipients: Across the study countries, CHWs complained that the timing of community clinics and home visits often conflicted with recipients’ routine schedules. Particularly, female Level 1 paraprofessional CHWs complained that their domestic responsibilities prevent them from conducting early-morning home visits. For instance, ASHAs in India explained that they were expected to do their domestic chores before leaving their homes in the morning, which often meant they did not leave home early enough before service recipients leave for their farms or income-generating activities: "…if you visit the community before 11 o’clock it is only then that you can meet people. If you go after that, you find all the houses locked because everybody goes to the farm. How can we go to visit them before 9 o’clock because we have to look after our household chores also?".