Study setting
The survey was conducted within the MMP catchment area in Chikwawa district, southern Malawi from December 2015 to May 2016. Chikwawa district is located within the East African Great Rift valley and experiences frequent flooding of its main river, Shire, during the rainy season. The rainy season, occurring from December to May leads to peaking of malaria transmission. The main project focuses on communities surrounding the Majete Wildlife Reserve with a projected total population of approximately 150,000 (African Parks, 2014). MMP aims to reduce malaria burden in this area through a behavioural change communication approach implemented by trained community volunteers.
The surveys were conducted in health-care facilities, known as health centres, which are either public (funded by government), or non-public (funded by non-governmental institutions, individuals or religious organizations). The health centres provide basic primary health care including diagnosis and treatment of uncomplicated malaria. Patients presenting at a health centre, requiring hospital admission or surgical procedures, are referred to the district hospital for secondary health care. The health centres are staffed by at least one medical assistant or clinical officer, a nurse-midwife technician, and community health workers known as Health Surveillance Assistant (HSA). Medical assistants and clinical officers are a cadre of clinicians with an 18-month certificate and 3-year diploma training in clinical medicine, respectively; nurse-midwife technicians possess a diploma in nursing and midwifery following 3-year training. HSAs have basic disease surveillance training and on the job training in programmes such as immunization, community case management including treatment of malaria, HIV counselling and testing, etc. Some HSAs are based at the facility while others are in the community providing assorted basic primary care such as immunization of children, non-invasive family planning methods, and integrated management of childhood illnesses (IMCI) in village clinics. The Malawi Ministry of Health recommend one HSA for every 1000 population.
For uncomplicated malaria diagnosis, the national treatment guidelines [
26] recommend confirmation of cases using a malaria Rapid Diagnostic Test (recommended histidine-rich protein-2 (HRP 2) products:
SD Bioline malaria Ag Pf or
Paracheck malaria Ag Pf) or microscopy. The first-line treatment for confirmed uncomplicated malaria in children above 5 kg body weight and pregnant women in the second and third trimester is artemether–lumefantrine (AL); for children less than 2 months or less than 5 kg body weight and pregnant women in the first trimester, a combination of clindamycin and quinine is recommended. Sulfadoxine–pyrimethamine (SP) and artesunate–amodiaquine (ASAQ) are reserved for intermittent preventive treatment in pregnancy and uncomplicated malaria treatment failure, respectively. Testing with RDT and treatment with AL are free in both public and non-public facilities as government provides test kits and drugs. Non-public facilities may charge a small consultation fee or require individuals to pay for non-malarial drugs, such as painkillers. During the surveys, village clinics were providing uncomplicated malaria treatment to children below 5 years old based only on fever or history of fever without a confirmatory test (although plans were underway to make confirmatory testing mandatory in village clinics).
Data collection and analysis
Enumeration of households in the main project’s catchment area was conducted in the area prior to data collection from August to November 2014. Geolocations of all health centres in the Majete Wildlife Reserve perimeter were collected using Global Positioning System (GPS) devices on Samsung Galaxy Tab 3 running Android 4.1 Jellybean Operating System, on open data kit (ODK) platform.
Data were collected by two qualified medical doctors using a structured questionnaires administered to the health facility personnel in-charges, guardians of eligible children, and health management information systems (HMIS) officer based at DHO. The health facility in-charges were interviewed on availability and cadre of health personnel, medical supplies and equipment, and services provided. The focus for medical supplies was availability of AL and RDT kits. AL is packaged in blister packs containing six fixed-doses for one treatment course per body-weight category as follows: one tablet for 5–14.9 kg; two tablets for 15–24.9 kg; three tablets for 25–34.9 kg; and four tablets for the above 35 kg category. Availability of these blisters and RDT kits, was confirmed with records on stock-cards and drug store. Availability of equipment included weighing scales, thermometers, and glucose and haemoglobin monitors. The study team observed and recorded patient flow at each health facility. The HMIS officer, based at the DHO, was interviewed on the structure and reporting of health information and its functionality. The officer is responsible for receiving all reports from all health facilities in the district and ensuring it is sent to a web-based, District Health Information System (DHIS2).
For the exit interviews, all children aged 5 years old and below, presenting at the health facility on any of the 3 working days the study team spent at the facility, were eligible for participation. Children and their guardians were consecutively referred to the study team by the health service provider after receiving all treatment. Children requiring surgical attention, or severely ill, or attending special clinics were excluded. The health workers were only informed that the participants would be interviewed about MMP project and reassessed for another study; the main purpose of the survey was not disclosed to the health workers.
Socio-demographic information including child’s age, presenting symptoms and their duration, and past medical treatment were obtained from the guardian during the interview. The reported time taken travelling from home to the health facility was also recorded for a few guardians. Weight and temperature were re-measured using a calibrated analogue weighing scale and digital electronic thermometer, respectively. Details of care, provided during that visit, including malaria RDT results, and drugs and doses prescribed, were captured from the participant health passport book—a record of a person’s health care-related information. The children were not retested for malaria. All data were entered into an electronic case report form on a Samsung Galaxy™ tablet through the ODK platform.
A checklist following Malawi treatment guidelines for management of malaria [
26] was used to assess prescriber’s adherence to guidelines. Definition of the following terms were based on the guidelines:
1.
Uncomplicated malaria: was a fever or history of fever within the preceding 24 h and a positive RDT result. Children with any sign or symptom of severe malaria, or history of malaria treatment within the preceding 14 days were excluded.
2.
Adherence to guidelines included: testing all febrile children with malaria RDT; prescribing AL to RDT positive patients and not prescribing AL to RDT negative patients.
3.
Correct AL dose: was the dose of AL prescribed based on the appropriate weight category of the child.
SPSS version 22 [
27] was used to analyse quantitative data. QGIS version 2.8.2 (QGIS Development Team, 2015—Open Source Geospatial Foundation) was used to map straight line distance between furthest households to a health facility.
After data collection and analysis, the findings were disseminated to the health facility personnel, i.e. nurses and clinicians, and the District Health Management Team (DHMT). DHMT was comprised of the District Health Officer, District Nursing Officer (in charge of district nursing services), District Medical Officer (in charge of district clinical services), and malaria coordinator (in charge of all malaria related activities in the district). A formal discussion, henceforth “dissemination discussion”, on the survey findings, including factors affecting health care practices, was conducted; feedback from the discussion is also reported in this paper.