This study was conducted a study in Lambaréné from July 2004 to June 2005 on a cohort of children participating in an ongoing trial at the Medical Research Unit of the Albert Schweitzer Hospital (HAS), Gabon. Lambaréné is a small town of approximately 30,000 inhabitants located near the equator in the Central African rainforest belt. Malaria transmission is moderate and perennial. 95% of all malaria infections are caused by P. falciparum and the entomological inoculation rate is about 50 infective bites per person per year [22, 23].
The study subjects are a subgroup of participants in a prospective Intermittent Preventive Treatment intervention (IPTi) trial [24, 25] – administering treatment doses of sulphadoxine – pyrimethamine (or placebo) to children at three, nine and 15 months and following them up monthly for 30 months.
All participants of the IPTi trial were included who had uncomplicated malaria, defined as the presence of asexual parasitaemia of P. falciparum with a rectal temperature of at least 38.5 °C or a history of fever in the last 48 hours. Written or documented oral consent of the parents or the guardians of the children were obtained at enrolment into the main study, usually after birth in the maternity ward of the Albert Schweitzer Hospital or the Public Regional Hospital in Lambaréné. Ethical clearance was obtained from the Ethics Committee of the International Foundation of the Albert Schweitzer Hospital in Lambaréné.
Two groups of children were compared; those that received artesunate plus amodiaquine combination under supervision and those who received the combination unsupervised. Assignment of the two groups was not randomized. According to the IPTi protocol, all study subjects with falciparum malaria were treated with three daily doses of artesunate plus amodiaquine combination from July 2004. The first dose was administered at the Medical Research Unit and the second and third doses at home. This treatment was classified as unsupervised. Then, from December 2004 all patients received all three doses of artesunate-amodiaquine combination under supervision at the Medical Research Unit. This treatment was classified as supervised.
All cases that had (i) a mixed infection or (ii) received a sufficient dose of another antimalarial drug a week prior or during the 28-day post-malaria follow-up period or (iii) taken the IPTi study medication a week prior or during the 28-day post-malaria follow-up period were excluded from intention to treat (ITT) analysis. If they vomited the artesunate plus amodiaquine combination twice, they were excluded from the evaluability analysis.
Treatment consisted of a daily oral dose of artesunate 4 mg/kg body weight plus amodiaquine 10 mg/kg body weight given for three days (ARSUCAM™ provided by Sanofi Synthélabo). Artesunate and amodiaquine were supplied in tablets which were crushed then mixed with sugar into syrup and given orally. A full dose of artesunate-amodiaquine combination was re-administered after 30 minutes if the child either spat the medication out or vomited within one hour.
The day of treatment was considered as day 0. Patients were invited for follow up visits on days 2 and 28. A clinical and laboratory assessment was done on these follow up visits. These included a thick blood smear and a full blood count. The Giemsa-stained thick smears which were read by at least two experienced microscopists. Parasitaemia was quantified (number/μL) by the Lambaréné method . A smear was declared negative only after ≥100 visual fields were scrutinized.
We advised the parent or guardian to administer doses as seen or practiced with the first dose, and to report back to the hospital if the child vomited or refused to take the medication. We also encouraged the parents or guardians to return back to the hospital at any time in case the child's health appeared to deteriorate.
Definition of study end points
Our primary outcome was defined as a parasitological cure on day 28. Failure was defined as persistent parasitaemia or reappearance of parasites during the follow-up period of 28 days. Our secondary outcome was the safety of the artesunate-amodiaquine combination drug measured as the frequency of adverse events. These were defined as any signs or symptoms or any abnormal laboratory value not present on day 0 or becoming worse during follow up and were judged by the clinicians in the study with respect to severity and relationship to study drug.
Full blood count
We measured the haemoglobin (Hb), white blood cell count (WBC) and neutrophils count on days 0 and 28 with an automated analyser (Cell-Dyn 3000™, Abbott Diagnostics Santa Clara, CA).
MSA-1 and MSA-2 genotyping
Filter-paper blots (Glass fibre filters: Schleicher & Schuell MicroScience, Dassel, Germany) were taken on day 0 and on the day of recurrent parasitaemia for polymerase chain reaction (PCR). For optimal differentiation between strains, and as described before [27, 28], we genotyped parasites for merozoite surface antigens MSA-1 and MSA-2 as two non-linked genotypical markers, in order to distinguish between re-infection and recrudescence in the case of reappearance of parasites during the follow-up period of 28 days.
Efficacy was assessed by evaluability (according-to-protocol, ATP) and ITT (intention-to-treat) analysis. Cure rates were calculated from the number of patients with clinical and parasitological cure by day 28 divided by total number of patients in the ITT population or per protocol population respectively. The according-to-protocol population was defined as children who completed the 3-day regimen of daily artesunate plus amodiaquine and had a day 28-follow up visit (+/- 1 week). There were 61 cases eligible for ATP analysis. The intention-to-treat population was defined as children who took at least the first dose. There were 89 cases eligible for ITT analysis. Based on a previous randomized trial of artesunate plus amodiaquine in our study area , a PCR-corrected day-28 cure rate of 90 % in the observed group was assumed.
Any data inconsistencies were reconciled, and the data were analyzed with the statistical software JMP 5.0 and Stata 8.2 (StataCorp, Texas, USA). A descriptive analysis of both observed and unobserved group was made, and continuous data (age, temperature and haemoglobin concentrations) between groups by unpaired t-test and categorical variables with a chi-square test were compared.