Study setting: The study evaluation was undertaken by the Navrongo Health Research Centre (NHRC), a field station of the Research and Development Division of the Ghana Health Service (12). The NHRC runs the Navrongo Health and Demographic Surveillance System (NHDSS) a platform designed to provide an efficient platform for evaluating health and social interventions (12). The implementation of the intervention and control was done in two administrative districts; Lawra district, Upper West Region and West Mamprusi in Northern Region both located in the northern part of Ghana (12–14). Figure 1 shows the location of the NHRC and the two administrative districts in the respective regions. The study setting is characterized by dryness due to its proximity to the Sahel and Sahara. The vegetation is predominantly Savannah grasslands with clusters of highly drought-resistant trees. There are two main seasons in the area, the dry and wet seasons. The peak of the dry season stretches from November to March and the wet season from April to October with a peak in July to September which coincides with malaria transmission. In northern Ghana cases of clinical malaria are seasonal caused mainly by Plasmodium falciparum and transmitted by anopheles’ mosquitoes with An. gambiae s.l. and An. funestus being the predominant vector species. (15–17). As part of efforts to reduce malaria transmission in Ghana, several interventions have been put in place, including indoor residual spraying and free distribution of long-lasting insecticide-treated bed nets. Intensive education on compliance and use of the available malaria interventions have also been taking place (11). Although these efforts have contributed to the reduction in malaria deaths, high case fatalities, and decreased testing rates have still been attributed to Northern Ghana (11).
Study areas: The study districts have similar weather and vegetation types as described above making them well suited for this study. The intervention area was the Lawra district (13) which lies in the North-Western corner of the Upper West Region of Ghana between latitudes 20 25"W and 2°45"W and longitudes 10°20" N and 11°00"N. The total population of the area was about 100,929 people. The main ethnic group in Lawra is the Dagaaba.
The district had five health centers, 10 Community-based Health Planning and Services (CHPS) compounds (18), three pharmacy shops, one clinic and one hospital which provides health services to the people in the area. The control area was at the West Mamprusi District (14), one of twenty districts in the Northern Region of Ghana and lies within longitudes 0°35’W and 1°45’W and latitudes 9°55’N and 10°35’N. The had a total population of about 180,877. There was one polyclinic, one health center, three private clinics, 6 CHPs compounds, and 15 pharmacy shops in this area. The people are mainly of Mamprusi ethnicity in the West Mamprusi district.
Study design: The study used a quasi-experimental design with before and after cross-sectional surveys. The implementation was during the main malaria transmission season from July 2015 to the end of the study in December 2015. The duration of participation for each child was six months. Enrolment was conducted over 2 weeks period and the post-implementation evaluation surveys lasted for 2 weeks. The study population included children aged 3 months to 59 months who were eligible for enrolment in the selected households and communities. Two cross-sectional surveys, one at baseline and the other at end line after 6 months were used for evaluation. The baseline survey took place in both study areas before drug administration. The enrolled children were followed up between the two surveys for episodes of severe and mild malaria. Other parameters documented included weight, height, malaria parasitemia, hemoglobin level and temperature.
Sample size estimation and Sampling strategy: The study was designed to have at least 90% power to detect a reduction in the incidence of severe malaria of 50% assuming the incidence rate of 0.04 per child per transmission season (i.e., 8 per 1000 child-months, based on data from 2009 from district health records showing there were 290 positive cases with fever or history of fever and a positive RDT among children under five years in the district from June- December, with an estimated total population of children of 7,000. District health records from 2010 showed a higher rate of about 0.1 per child per month, or 0.5per child per 5-month transmission season. The sample size calculation was based on the impact of SMC on severe malaria as a primary outcome. Effectiveness assumption is based on studies that showed a reduction in severe malaria by 65 to 86% (6–9). To demonstrate a 75% reduction in the proportion of severe malaria cases in the intervention districts a minimum number of 14 clusters per arm were required.
A cluster was taken as the community of residence as outlined by the Ghana Health Service District Health Directorate. Each community was a well-delineated area with defined boundaries. Using an average proportion of 0.36 of children under 5 years developing severe malaria and 50 children per cluster with an inter-cluster variation of 0.26, approximately 14 clusters were needed per arm to demonstrate a 75% reduction in the proportion of under-five children developing severe malaria at a power of 90 percent and a two-sided significance level of 5%.
All districts in the Upper West region were listed and one district (Lawra) was selected at random for evaluation of SMC. The control district, West Mamprusi, was selected at random from the Northern Region which had the same interventions against malaria as the Upper West Region apart from SMC. All communities in the selected districts were subsequently listed and fourteen out of 141 and 152 communities from Lawra District and West Mamprusi Districts respectively were randomly selected using the random sample selection command in STATA 12 statistical software to form` the study clusters. Each selected community was defined as the cluster, the unit of analysis. All the households in each selected community were labeled and listed. All households with eligible children were listed and 40 households randomly selected. In each selected household all eligible children were selected for participation and out of these the children whose guardians gave parental consent and were available for the 4-month duration of the study were selected for inclusion. Those children whose guardians could not give consent or were not likely to stay in the community for the duration of the study were excluded. A minimum of 50 children per village and 700 children per arm were recruited into the study. Where the community was too small to provide a minimum of 50 participants, the nearest community was added to it to make up a bigger community.
Study drug administration: Study drugs were supplied by National Malaria Control Programme. Formulations of Sulphadoxine -Pyrimethamine 500/25mg tablet (SP) and Amodiaquine (AQ) 153mg tablet were used. Infants 3–11 months old were dosed with half of a 153mg tablet of Amodiaquine given once daily for three consecutive days and a single dose of half of 500/25mg tablet of SP. Children 12–59 months were given a full tablet of 153mg Amodiaquine base given once daily for three consecutive days and a single dose of a full tablet of 500/25mg tablet of SP. The administration was done by direct observation therapy and given by community health volunteers. Children received medication for four consecutive months at monthly intervals between July and November.
To determine the proportion of children who experienced an adverse event, study participants in the intervention district were visited at home daily for three days during the administration of SMC and two days after the last dose in each round of drug administration and a side effects questionnaire was completed to document and quantify adverse events that might have occurred since receiving the trial medication. Grading of the severity of adverse events was done by trained field workers. An adverse reaction was graded as mild (grade 1) if it was easily tolerated and moderate to severe (grade 2) if it interfered with normal activity or required treatment.
Malaria surveillance: A passive surveillance system to monitor malaria episodes was set up at the district hospitals and health centers in the study areas. Fieldworkers were stationed at health facilities as well as in the communities in both the intervention and control districts to pick up any episodes of mild or severe malaria by checking for malaria parasitemia before treatment was given. Participants were visited every fortnight to maintain their interest in the study and encourage them to seek health care promptly if any sickness developed. Field workers visited the children once a week during the period of drug administration to enquire about their health and completed a morbidity form if a child had any illness. If a child had a history of fever or vomiting within the past 48 hours the parents were advised to take their child to the nearest health facility for examination and treatment. Any mortality that occurred outside the health facility was to be investigated and a verbal autopsy questionnaire administered to help ascertain the cause of death at the Navrongo Health Research Centre (12).
Field workers were recruited and trained to do home visitations and in addition, they were trained to prepare thick and thin films, which were collected by the trial team for microscopic examination to crosscheck the RDT results. They were also trained to record and report any adverse events. In each community, trained community health workers (CHW) provided malaria case management services. Persons with fever were encouraged to attend the health post or health centre to be tested with an RDT and were treated with artemether-lumefantrine if tested positive. Consultations were recorded in a health facility register to document the test results and treatment given. In the intervention area, if a child was unwell on the day of the SMC visit, caregivers were asked to bring the child to the health post or hospital for testing with malaria RDT; those who tested positive were treated with artemether-lumefantrine and those who tested negative received SMC and were then referred to the nearest hospital.
Participants who presented at the health facilities with fever were similarly tested with an RDT and had thin and thick blood films taken by the nurse, which were read later. Treatment of study participants seen at the health facilities for other conditions was carried out per national guidelines. Registers for recording SMC administration with a list of all children enrolled in each village were made available for each field worker. In SMC villages, the dose of SP and AQ administered to each child was recorded in a register. All consultations for illness were also recorded in the registers and other details including date, symptoms, RDT results and treatment in case report forms. Besides, in each health facility, consultation records were reviewed by the field workers to identify all consultations of children from study villages.
Laboratory methods: Capillary blood was used for diagnosis for this study. To obtain this, the participant's heel or thumb was wiped using a swab moisturized in 70% alcohol. A sterile lancet was used to make a prick on the participant’s thumb or heel and squeezed gently to obtain a large drop of blood. To ensure good staining and standardization of reporting, the blood sample was transferred from the participant to a sterile microscope slide using a bulb pipette. One drop was placed in the middle of the slide and another about 15mm to the right of the slide for both the thick and thin films. These films were prepared using a card showing illustrations of a thick and thin-film as a guide. The blood was then allowed to air dry by placing slides in a horizontal position. Thin films were fixed by placing slide horizontally on a bench with a small drop of methyl alcohol such that the thick film did not get fixed in the process. The slides were stained using a 10% Giemsa solution for ten minutes. The stains from the slides were washed using running water and allowed to dry. One of the two slides from each participant was examined using x100 oil immersion microscopic lens and the other archived for confirmatory testing. A sample was considered negative only after 200 high power fields had been read. Parasite counts were converted to parasites per microliter (µl), assuming a white blood cell count of 8000 leukocytes per µl of blood. In instances where there were discrepancies in the findings in a slide between the two initial technicians (positive or negative or a 50% or more difference in parasite density), a senior microscopist read the slide and his reading was deemed to be the correct reading. Hemoglobin was measured using Hemocue ® (Hb 801 system)
Data management and Statistical analysis: Village registers that listed each child’s record in the census were printed for health workers to record SMC administration and consultations for illness. All consultations for illness were recorded and the register checked for completeness during weekly supervision visits. Completed forms by trained personnel on the study were checked by field supervisors and data managers for consistency and accuracy before logging it out for data entry. All data collected were entered twice into a database using EpiData software. Again, automatic checks for consistency and range errors were done, and queries resolved before the dataset was locked for analysis. Other data management procedures such as taking the consent of the participants before participation in the research was also done. Effects of SMC on the prevalence of parasitemia, gametocyte carriage, mean hemoglobin concentration, and proportions of anemia (Hb < 110g/l) and severe anemia (Hb < 60g/l) were estimated from the survey at the end of the transmission season. The results were presented in tables and figures. Analyses were performed using Stata version 14 (Stata Corp, College Station, Texas). Descriptive statistics are presented for continuous variables.
Ethics approval and Community engagement: This study was conducted in full conformity with the current revision of the Declaration of Helsinki and with local regulatory requirements, to offer maximum protection to the subject. Members of the study team held community engagement meetings with community, administrative, and religious leaders to explain the aims and activities of the study. This was also done to seek community consent and to promote community ownership of the program. Staff then visited each household to explain the study in the local language, provided an information sheet, and sought signed consent from parents. Children whose parents consented were enrolled, and mothers/caregivers were issued with a unique identification card bearing the details and study number for each eligible child in their care. The cards were used to identify children if malaria was diagnosed and, for those in clusters randomized to receive SMC, to document SMC courses of treatment received. A data and safety monitoring board were convened to oversee the project and review data on tolerability and safety.