Procedures and patients
The study was performed in the area of Bandim on the outskirts of Bissau, Guinea-Bissau. Parents from Bandim attending the Bandim Health Centre with children weighting more than 7.5 kg and having fever or other symptoms compatible with malaria and stating that the children had not taken any antimalarial drug during the previous week were informed of the study. Included children had a thick film examined for malaria. Children with convulsions, severe vomiting, severe anaemia, a severe concurrent infection or who for other reasons were considered in need of hospital care were not eligible.
Children with mono-infection with P. falciparum and 20 or more parasites per 200 leukocytes (= 800/μl assuming a leukocyte count of 8000/μl) were enrolled from June 2004 to July 2006. Children were allocated to one of six 5 kg-interval weight groups or to a group of children with a body-weight above 37.5 kg. For each weight group, numbered boxes had been prepared at random with either paracetamol tablets or undistinguishable placebo tablets corresponding to approximately 50 mg paracetamol per kg bodyweight per day for three days. The children were given study numbers consecutively within each weight-group. The first dose of paracetamol/placebo from the tablet-box corresponding to the study number of the child was given by an experienced nurse at the health centre. For logistic reasons (home-visits could not be performed at intervals matching the intervals between medications) the mothers were then given the tablet-box containing tablets until the end of day 2 and were carefully instructed on how to give the tablets. A health worker visited the house on day 3 and asked to see the tablet-box and counted the number of remaining tablets. The randomisation numbers were kept separately at the Department of Paediatrics in Kolding, Denmark. An interim analysis was made by one of the researchers not involved in the recruitment of the patients when a total of 40 and 140 children had completed the follow-up to ensure that none of the groups had an unacceptable high rate of recurrent parasitaemia using the results of previous in-vivo chloroquine studies as guidelines [16–18].
Following recommendations of the National Malaria Programme at the time of the study, 25 mg/kg bodyweight of chloroquine was split into daily doses of 10, 10 and 5 mg/kg and given days 0, 1 and 2 for treatment of malaria. An experienced nurse ascertained that the tablets were swallowed and recorded the time the dose was given. If the child did not turn up as planned, a project health worker went to the house to ensure the medication. If the child vomited within 30 minutes after receiving the tablets, the dose was repeated. If the child vomited again during the next 30 minutes, a dose of 10 mg Quinimax® dihydrochloride (Sanofi Winthrop, Gentilly, France) per kg bodyweight was given intramuscularly and the child was referred to hospital.
Tablets with 160 mg chloroquine-phosphate (= 100 mg chloroquine base) were obtained from Recip SB, 12054 Årsta, Sweden. Paracetamol tablets of 500 mg and undistinguishable placebo tablets were obtained from GlaxoSmithKline, Dungarvan, Irland.
The children or their mothers were asked for any symptoms twice daily on day 0, day 1 and day 2 and in the evening on day 3. At the same time, the axillary temperature was measured using an electronic thermometer.
Follow-up
Following treatment, the children were visited once weekly until day 35 for thick and thin films. The mothers were interviewed about any medication given since the last visit, and the health worker assessed the condition of the child. During the study period, the parents were asked to bring the child to the health centre in case of any illness to ensure early detection of clinical malaria. All treatments were free during the study-period. Recrudescent infections were treated according to the national recommendations which were then sulphadoxine/pyrimethamine or, for severe cases, quinine.
Microscopy
Before inclusion, a thick and a thin film were examined. Thick films were also examined on day 3, 7, 14, 21, 28 and 35 after initiation of the treatment and whenever a child sought medical attention at the Bandim Health Centre due to fever or other symptoms suggestive of malaria. To ensure the quality of the microscopy, 10% of the thick films were re-examined by an experienced laboratory technician. The parasites were counted per 200 leukocytes. If abundant, the number of leukocytes was counted per 500 parasites.
Polymerase chain reaction
Each time a thick smear was made, approximately 100 μL of blood was also put onto filter paper, dried and then stored in a separate sealed plastic bag. Approximately 50 μl of blood was cut from the filter papers and DNA was extracted using an ABI Prism 6100 Nucleic Acid Prep Station (Applied Biosystems, Fresno, CA) according to the manufacturer's recommendations. Extracted DNA was frozen at -20°C until amplification by PCR. Merozoite surface protein 1 (pfmsp1) and merozoite surface protein 2 (pfmsp2) were amplified by PCR as described previously [19]. PCR products were separated on agarose gels (Amresco, Solon, OH), stained with ethidium bromide and visualised under UV trans-illumination (BioRad GelDoc System, BioRad, Hercules, CA). Pfmsp1 and pfmsp2 were amplified from DNA collected on the day of inclusion and the day of failure. Recrudescent infections were defined as reappearance of at least one band from both pfmsp1 and pfmsp2 during the follow-up or the re-appearance of a band in the successfully amplified gene if either pfmsp1 or 2 failed. Re-infections were defined as those where no band re-occurred. Whenever a child received antimalarial medication elsewhere, the blood sample drawn on the last visit was analysed.
Outcome measures
Early treatment failure (ETF) was defined as 1) development of severe malaria or appearance of danger signs on day 1 to 3; 2) positive malaria film and temperature ≥ 37.5°C on day 3; or 3) parasitaemia on day 3 ≥ 25% of the parasitaemia on day 0. Late clinical failure (LCF) was defined as either; 1) fever and a positive malaria film; 2) a history of fever and clinical symptoms of malaria as well as a positive malaria film. Late parasitological failure (LPF) was defined as reappearing parasiteamia on day 7 or later for children without ETF or LCF. Adequate clinical and parasitological response (ACPR) was defined as absence of parasitaemia without the child meeting any of the criteria of ETF, LCF or LPF.
The primary outcome was the cumulative PCR-uncorrected ACPR rate until day 35 and PCR-corrected failure rates on day 28 and 35. Secondary outcomes were the ETF rates, the symptoms as reported by the mothers, the temperature measured during the treatment with paracetamol/placebo; and possible clinical adverse effects of the medication given.
Statistical analysis
An intention-to-treat analysis was performed to answer the question as to whether the treatments were valuable for the children [20, 21]. Therefore, all children admitted to hospital on the day of inclusion were considered as ETF, and all children treated by a third person with an anti-malarial during follow-up were considered as LCF. Children who could not be found during medication or the first follow-up visit due to unknown addresses were considered to have violated the entry criteria of living in Bandim and were therefore excluded.
A per protocol analysis was done to evaluate the efficacy of the treatments. Children admitted to hospital on the day of inclusion were then excluded as they were considered to be so sick that they had been included in violation of the inclusion criteria. Children treated outside the study during follow up without having re-parasitaemia confirmed by a positive malaria film and/or a positive PCR-analysis were considered withdrawal of consent and excluded on the day of retreatment.
The data is in the form of grouped failure-times, the outcome of interest being time until parasitaemia during follow-up. Mantel-Haenszel weighted odds ratios are given. Loss to follow-up between two analysis times has been treated by censoring at the beginning of the time interval. Other differences between and within the two groups have been tested using the Chi-square test and Mann-Whitney U test.
Ethics
The protocol stated that any study group should be terminated if parasitaemia reappeared in 50% or more of at least 40 children. Ethical clearance was obtained from Direcção de Higiene e Epidemiologia, Ministério da Saúde Publica in Guinea-Bissau (030/DHE/2004) and the Central Ethical Committee in Denmark (2004-7041-11). Children and/or their parents were informed of the study orally as the literacy rate is low. Written information in Creole was given on request. The information was standardised and in accordance with the principles of the Helsinki declaration. This trial was registered at ClinicalTrials.gov [22] with the study ID: PSB-2004-paracetamol (NCT00137566).