We recruited consecutive children 6–60 months of age with a history of fever from the General Outpatient Department of the Jos University Teaching Hospital in Jos, Nigeria. Jos is an urban center in central Nigeria and had a population of approximately one million persons. The General Outpatient Department is a walk-in clinic for patients seeking primary and secondary care. In 2004, a total of 35,966 patients were seen, and 16% had a diagnosis of malaria.25
The study design was an experimental, randomized, controlled trial. It was conducted during December 2004–June 2005: five months during the dry season and two months in the rainy season. Children 6–60 months of age were considered eligible for enrollment if they had a recent history of fever, malaria parasitemia with trophozoites in a peripheral blood film, anemia (hematocrit ≤ 33%), and the ability to take medication orally. Children were excluded if they had severe malaria demonstrated by inability to drink or breast feed, prostration, repeated vomiting, shock, dyspnea, or convulsions during the current illness. Children with any associated illness requiring hospitalization were also excluded.
Written informed consent was obtained from a literate parent or guardian of all enrolled children. Those persons who were unable to sign their name used their thumb print on the consent form. Ethical approval for the study was obtained from the research and ethical committee of the Jos University Teaching Hospital.
We obtained information regarding date of birth, if known, or their approximate age in months of each participant. We also obtained information regarding the mother's education, characteristics of the current illness, use of medications, number of febrile illnesses in the preceding three months, use of insecticide-treated bed nets, and the frequency of dietary intake of meat, fish, and eggs.
Each child's weight was measured on a scale that was calibrated daily with a 5-kg standard. Standing height was measured with a wall-mounted stadiometer. Length was measured in those who were unable to stand. Physical examination was performed to determine axillary temperature, clinical pallor, jaundice, respiratory rate, and pulse. The heart was auscultated for gallop rhythm and murmurs. The abdomen was examined for splenomegaly, and the liver span was measured by percussion for hepatomegaly.
Children with a history of recent fever had blood collected from a fingerprick sample. Blood was drawn into a heparinized capillary tube; additional drops were used to prepare thick and thin Giemsa-stained blood films for malaria. The capillary tube was centrifuged for five minutes, and a hematocrit reader was used to determine each child's hematocrit. Malaria parasite density was determined by counting the number of parasites corresponding to 200 leukocytes, assuming 8,000 leukocytes per microliter of blood. A hematologist evaluated the blood film for erythrocyte morphology.
Children with a positive malaria smear and a hematocrit < 33% were enrolled and randomized by using a lottery method in blocks of 20 to receive iron plus folate or folate alone in addition to antimalarial treatment. The active treatment group was given iron (ferric ammonium citrate, 2 mg/kg/day, as syrup) plus folate (5 mg/day as a tablet), and the control group was given only folate (5 mg/day). Both groups were instructed to take their assigned medications for four weeks. All enrolled children were treated with chloroquine (10 mg/kg for the first two days and 5 mg/kg on the third day) and sulfadoxine-pyrimethamine (25 mg/kg of sulfadoxine and 1.25 mg/kg of pyrimethamine) as a supervised single dose, which had been demonstrated to be effective for the treatment of malaria in this population.26
Parents were instructed to return with their children in four weeks or sooner if there was any problem. All parents were counseled on the use of insecticide-treated bed nets and other preventive measures against malaria. At the return visit, the quantity of remaining syrup and pill counts were recorded to assess compliance. We did not assess if medication was shared with others in the household. The child was weighed and a hematocrit was obtained. Children who remained anemic at the end of the study underwent further evaluation and continued treatment with iron and folate as appropriate.
A sample size of 35 children in each group was estimated to provide 95% confidence and 80% power to detect a 2% mean difference in hematocrit, which was regarded as clinically important. To allow attrition, we targeted recruitment to 40 children in each group, for a total of 80 children. Data were entered and analyzed in Epi Info 3.3 (Centers for Disease Control and Prevention, Atlanta, GA). The primary outcome variable was the change in hematocrit at the four-week visit. Student's t-test was used to compare mean values of normally distributed continuous variables at baseline between the two groups. In the case of ordinal or continuous variables with a non-normal distribution, the Mann-Whitney test was used to compare the two groups. Proportions were compared by using the chi-square test. The effect of variables on the change in hematocrit was assessed by a multiple linear regression analysis with the final hematocrit as the dependent variable and including the baseline hematocrit as one of the independent variables. P values < 0.05 were considered significant.