Health systems supports for community case management of childhood illness: lessons from an assessment of early implementation in Malawi

Health Surveillance Assistants are a cadre of CBHW in Malawi that was originally established in the 1960s to support smallpox vaccinations [28]. HSAs are formal non-clinician health workers who are salaried by the government, are required to have 10 to 12 years of education, and undergo a 10-week basic training [29]. The government recruited over 5,000 new HSAs between 2002 and 2010 through grants from the Global Fund to Fight AIDS, Tuberculosis, and Malaria. HSAs now serve a catchment area of approximately 1,000 people and their main functions are to provide health education and sanitation within the communities where they are posted [29]. Although some HSAs serve urban areas, the majority are posted to rural communities and are supervised by Assistant Environmental Health officers (AEHOs)--non-clinicians under the MOH’s Preventive Health Section and based at health facilities–and Senior HSAs–HSAs posted in communities who have been promoted to a supervisory role. HSAs are expected to receive a supervision visit from an AEHO or Senior HSA in their community each month; additional supervision occurs during quarterly review meetings.

HSAs began to provide CCM services in Malawi in September 2008 in ten districts receiving technical support and funding from WHO and UNICEF for early CCM implementation. Selected HSAs received six days of training to treat uncomplicated cases of malaria, pneumonia, and diarrhea following a job aid, known as the sick child recording form, with an algorithm adapted from the Integrated Management of Childhood Illness (IMCI) guidelines [30]. The six-day trainings were led centrally at the district level by trained clinicians from district hospitals. Implementation plans specified that HSAs should receive a wooden drug box, with a lock and initial drug supplies, on the last day of training. Six of the ten early districts with the strongest levels of CCM implementation (based on the number of HSAs trained and equipped to operate village health clinics) were selected for inclusion in this early evaluation of the program. The six districts represent all three of Malawi’s regions–two from the southern region, three in the central region, and one in the northern region (see Figure 1). Facility-based health services in these districts are provided by the Malawi MOH, which provides services free of charge, and the Christian Health Association of Malawi (CHAM), a non-profit organization that charges user fees for most services and drugs provided at its clinics. Additional details about the HSA program and components of the CCM package are available elsewhere [29].

This study was conducted in partnership with the Ministry of Health as a part of an independent evaluation of the Catalytic Initiative, led by the Johns Hopkins University Institute for International Programs and the Malawi National Statistics Office. The design followed an explanatory-sequential approach [31] by collecting quantitative data on the types and levels of health systems supports, followed by a qualitative study to gain greater understanding of the support systems, barriers, and facilitating factors. Quantitative data on the quality of care and the types and levels of health system supports were collected from a random sample of HSAs operating village health clinics in the six selected districts through a cross-sectional survey, with results on quality of care available elsewhere [32]. Data were collected by three-person survey teams, composed of CCM trainers from the Ministry of Health, trained and supervised by Johns Hopkins researchers and MOH managers. Briefly, the teams visited sampled HSAs in the rural posts where they conduct village health clinics to observe sick child consultations, record the availability of drugs, and to ask the HSA to recall the supervision visits he or she received and any drug stock outs in the past 3 months using a standard questionnaire. In order to minimize bias in data collection, survey teams were sent to different districts than those in which they worked. A sample size of 132 HSAs was calculated to allow observation of 369 sick child consultations and estimate the proportion of sick children correctly managed with +/− 6%, assuming 95% confidence and a design effect of 1.5 [32]. Data collection took place during a four-week period in October and November 2009, approximately one year after CCM implementation began. The indicators for drugs and equipment were selected based on the functional requirements of the village health clinic as indicated by the national CCM treatment algorithm [27] and were calculated for the overall sample and each district using Stata 10 [33]. Differences between districts were tested using Fischer’s exact test for proportions and the Kruskall-Wallis rank-sum test for interval indicators.

At the time of the study there had been no systematic documentation of how the CCM program was being implemented at district level to support interpretation of quantitative results. We therefore conducted follow up interviews and focus groups to obtain information on program implementation strategies and challenges, focusing on the areas of CCM training, supervision, job aids, and drug supply. These qualitative data were collected over a four-week period in November and December 2009 by two independent Malawian researchers and a Johns Hopkins researcher, with permission and introductions from Ministry of Health officials. A pilot of the qualitative protocol and interview guides, conducted in a district excluded from the study, identified the relevant personnel in each district to include the District Health Officer, IMCI Coordinator, Administrator, Environmental Health Officer, Assistant Environmental Health Officers, Health Center Clinicians, and Senior HSAs. Although the study team interviewed the IMCI Coordinators--the manager with primary responsibility for the CCM program at district level--for all six districts included in the survey, study timing did not permit visits to all districts. Four of the six districts--Lilongwe, Kasungu, Phalombe, and Ntcheu--representing high- and low-performing districts in terms of supervision and drug supply based on the preliminary results from the quantitative survey, were selected for in-depth interviews with other health personnel. In-depth investigation in these four districts was expected to reveal variations in implementation strategies and challenges between districts.

During visits to each of the four districts, the study team attempted to interview each of the individuals occupying these positions. Focus groups with HSAs operating village health clinics were also conducted in the four selected districts. Interview and focus group methods are described in detail elsewhere [34]. Three members of the research team coded all transcripts using a standard coding index. Analysis procedures for exploring themes in the data included summarizing health system strategies (e.g., drug supply delivery method) for each district, developing implementation timelines, and placing coded data in charts by district and respondent cadre. Reported constraints for delivering health system supports were analyzed for underlying factors using root cause analysis techniques [35]. Data from the quantitative and qualitative components of the study were triangulated, and associations between qualitative data on health systems strategies and quantitative outcomes were assessed for using Fischer’s exact test. Preliminary results were reviewed with stakeholders at national and district level, including representatives of DHMTs in all participating districts. This study was approved by the Institutional Review Boards at the Johns Hopkins Bloomberg School of Public Health and the Malawi National Health Sciences Research Committee.

Following national workshops to adapt the CCM algorithm and train trainers in September 2008, the ten districts selected for early implementation started training with HSAs posted to the most “hard-to-reach” areas. Hard-to-reach areas are defined by each DHMT (usually the Environmental Health Officer), based on the general criterion offered by the MOH of approximately 7 km or farther from a health facility. DHMTs were responsible for determining the pace of training sessions and rollout of CCM in their district, with support and feedback from WHO and UNICEF offices. In qualitative assessments, managers from several districts indicated that they were purposefully introducing the program slowly in order to identify and correct problems with health system supports. As said by one district IMCI Coordinator: “It isn’t possible to just train everybody because we also have to consider the quality of the services they will be delivering . . . We suggest . . . to first sort out supervision and other logistics.”

Despite awareness of the need for supervision and other health system supports among some managers, these inputs and processes were still under development throughout the first year of the CCM program. Following the 6-day CCM training, HSAs were to begin operating their village health clinics (VHCs) when they received their wooden drug box and initial supply of drugs. The initial drug boxes were delivered directly to the HSAs’ post or the nearest health center from the district level. In most districts, there were substantial delays between the time that the HSAs completed training and when they received their drug box (Figure 2). Four districts (Lilongwe, Kasungu, Phalombe, and Mzimba) each had 40 or more HSAs trained by November 2008 who did not receive drug boxes until March 2009. After the delivery of the initial drug supply with the drug box, the MOH’s strategy was to integrate CCM drug resupplies into the routine drug supply system. HSAs disbursed the same drugs, in the same packaging, as health facilities, and were to collect resupplies of drugs from their nearest health facility, which in turn requisitioned resupplies from national and regional medical stores.

During the CCM program’s first year, the national IMCI office had not yet provided guidance on supervision protocols to districts, such as standardized checklists or guidelines for activities to be conducted during supervision. District IMCI coordinators reported a common understanding in the qualitative interviews that the expectations from the national level were that HSAs should receive a follow-up visit in their communities within 6 weeks of training, followed by monthly CCM-specific supervision visits. However, similar to provision of drug supplies, implementation of supervision visits was delayed in most districts. None of the six districts began CCM supervision visits earlier than 4 months after the first HSAs were trained, and some HSAs did not receive any supervision visit for 8 months following training.

Three of the six drugs included in the CCM algorithm in Malawi--cotrimoxazole, Coartem (Lumefantrine-artemether, an antimalarial combination therapy),^a and oral rehydration solution (ORS)--are critical for treatment of the priority childhood illnesses of malaria, pneumonia, and diarrhea, and are therefore the focus of this analysis^b. Among the 131 HSAs enrolled in the survey, 68.7% [CI: 60.0, 76.5] had all three critical drugs available at their respective CCM clinics at the time of the survey team’s visit (Table 3). The percentage of HSAs with all three drugs on the day of the survey team visit ranged from 100% in Nsanje [CI: 84.5, 100.0] to 52.4% in Kasungu [CI: 29.8, 74.3]. When asked to recall stock outs in the previous 3 months, 46.6% of HSAs reported having a stock out of any one of the three critical drugs [CI: 37.8, 55.5]. HSAs reported that Coartem, the only drug which is provided to HSAs in age-specific dosing, was frequently out of stock in the dosage for children aged 36 to 60 months (2x6), with 45.8% [CI: 37.1, 54.7] reporting a stock out in the past 3 months. Over 40% of HSAs also reported a stock out of ORS [42.7%; CI: 34.1, 51.7] but only 10.7% [CI: 6.0, 17.3] of HSAs reported a Cotrimoxazole stock out in 3 months (Additional file 1: Table S2).

Among the 111 HSAs who received a resupply of drugs in the prior 3 months, the majority restocked at the health center [81%; CI: 72.5, 87.9]. The location of drug resupply in the three months previous to the survey also varies significantly by district. Lilongwe is the only district where 100% of resupplied HSAs received their resupply from a health center, while 81% of HSAs restocked at the district hospital in Ntcheu district and 39% of HSAs received drug stocks from a visiting supervisor in Phalombe (Additional file 1: Table S2).

Informants reported important challenges in implementing the official drug supply strategy for the CCM program, which stipulates that HSAs should visit their nearest health center to request drug stocks (Table 4). Health center clinicians in several districts reported that they resisted supplying HSAs with drugs at first because they had not been informed about the CCM program and drug supply policies. Additionally, informants at all levels noted that the health centers themselves experience periodic drug shortages and stock outs. When drug stocks at the health center are inadequate, clinicians are unable or unwilling to provide scarce supplies to HSAs. According to a health center in-charge, “The only problem that there is about drug supply [is] we don’t have drugs here at the health center.” HSAs in some districts indicated that mistrust had developed between themselves and health center clinicians over the sharing of drug supplies. One HSA stated: “In my opinion I feel that it is just the selfishness of those medical assistants. They say ‘we have no drugs’ but they are operating their services so it is very difficult to say that this problem is from central medical stores to health facilities.”

Distance was a barrier to drug supply for many HSAs who reported difficulties traveling to the health center to collect drugs and in some cases spending their own money for transport. Transport difficulties exacerbate the challenge with drug stocks at health centers, as expressed by one HSA: “When we go to the government facility, most of the times drugs are out of stock and this is an embarrassment to us because we travel a long distance.” The challenge of distance was even greater for HSAs reporting to health centers managed by CHAM, who were expected to travel to the nearest government health center to collect drugs which often was further away than the CHAM facility serving their catchment area.

In several districts, managers took various initiatives to try to reduce the constraints to drug supply. In some cases, these practices served to bolster the official drug supply policy, and in other cases the practices were outside the official policy. In support of the official policy, districts undertook activities that increased the involvement in or awareness of the CCM program among health center staff. Three districts held one-off review or orientation meetings to inform the health center clinicians about the CCM program, while another district included the clinicians during the initial community sensitization activities conducted by the district-level managers. Practices that also facilitated drug supply, but were outside of the official drug supply policy, included: 1) allowing HSAs to collect drugs directly from the district hospital; 2) delivering drugs to the HSAs in their communities; and 3) establishing agreements with facilities operated by CHAM. These efforts, although positive, did not show a significant association with drug supply or drug stock-outs in this study.

Table 5 presents results related to the frequency and type of supervision received by the sampled HSAs. Overall, levels of routine CCM supervision reported by HSAs were low. Within the period of 3 months prior to the survey, less than 60% of the HSAs reported receiving any type of supervision visit [58.6%, CI: 49.6, 67.2], and only about 40% reported a visit specific to CCM [38.3%; CI 29.8, 47.3]. Moreover, only 15.6% of HSAs received a CCM supervision visit that included the supervisor observing the HSA while managing a sick child.

Among the 49 HSAs reporting supervision specific to CCM in the previous 3 months, IMCI coordinators were the most frequently-cited cadre conducting the supervision visit. Some HSAs also reported that Assistant Environmental Health Officers, IMCI trainers, and pharmacy technicians provided CCM supervision (data not shown). The most commonly reported supervisors for the 75 HSAs receiving any supervision visit in the previous 3 months were Environmental Health Officers (30%) and Senior HSAs (21%). HSAs who received a CCM supervision visit in their communities reported that, during the supervisory visit, supervisors checked records (83.7% percent of visits), corrected the HSA’s work (71.4%), answered the HSA’s questions (67.3%), and provided positive feedback (63.3% of visits). Observation of the HSA performing case management was the least frequently conducted activity by supervisors (36.7% of supervision visits).

Although the MOH initially advocated monthly supervision of each VHC, the survey results demonstrated that supervision visits were in fact far less frequent. Every IMCI Coordinator interviewed for this study acknowledged that their district was unable to meet this standard of supervision, and the majority considered monthly supervision to be an unattainable target. Common supervision challenges cited by manager informants included supervisors’ busy schedules, lack of transportation, and the coordination required to undertake supervision in teams (teams often included Pharmacy Technicians, Environmental Health Officers, and Malaria Coordinators). Districts teams typically visited as many HSAs as possible during a one- to two-week period set aside for supervision and rarely spent enough time at each site to observe the HSA managing a sick child. In order to improve supervision frequency, two districts provided CCM training to routine HSA supervisors, such as Assistant Environmental Health Officers and Senior HSAs. However, supervision frequency was not significantly higher in these districts.

Both HSAs and managers agreed that supervision is valuable for the program, and that the current levels of supervision are inadequate. However, HSAs desired different outcomes from supervision than those expected by IMCI coordinators, Environmental Health Officers, and other managers. Program managers generally considered supervision as an opportunity to correct any mistakes that HSAs may be making: “for them to do a quality job throughout, they need to be supervised frequently . . . because if we leave them without supervision, they may miss some concepts.” The majority of HSAs welcomed opportunities for receiving feedback to “know whether [we] are doing fine or not”, but also expressed a strong desire for supervisors to help them solve the problems encountered while operating CCM clinics, particularly in terms of supplies and resource needs. The lack of problem solving during supervision visits resulted in frustration among HSAs, as expressed by one focus group participant:

“When they came to supervise me I told them what I was lacking in my clinic, but until now there is nothing that has changed. So I feel like I have not been supervised because when you are supervised you expect to see some change.” (HSA, Focus Group #4)

A key finding that emerged from the qualitative data on health systems supports for CCM in Malawi is the importance of the job aid developed for the program, known as the Sick Child Recording Form (SCRF). This form was adapted in Malawi based on a similar form developed by WHO as a standard part of IMCI training and implementation [36]. During the CCM training, HSAs are taught to conduct assessment following steps on the form, and to make treatment decisions using the decision rules included on the form. Informants in the qualitative study had overwhelmingly positive opinions about the SCRF. Additionally, in over 90% of sick child consultations observed during the quality of care survey study, HSAs made reference to a hard copy of the SCRF while managing sick children. HSAs reported that they liked using the SCRF for their CCM work, as the following comments by HSAs illustrate:

“I like using the guide line chart because it acts as my sign post. Whenever I am confused, I consult it to know where I am lost and then I am in a better position to do what I am supposed to.”

“I should say that the Sick Child Recording Form is the mwini filimu (the major actor) in this program.”

Managers also considered the SCRF to be important for the quality of CCM services. One IMCI coordinator said, “I think HSAs are doing a good job, and basically it is because they are using the Sick Child Recording Form.”