Table 3

Effect of the intervention on cognitive outcomes at 15-month follow-up (classes 4 and 5 only)*

Trial endpointsSummary statisticsCrude change from baselineEffect estimate: Intervention vs Control schools†
Difference (95% CI); P value
Control schoolsIntervention schoolsControl schoolsIntervention schoolsBasic model‡Fully adjusted model‡
Cognitive outcomes in February 2012:
Code transmission:
 Single-digit task, z-score1.39 (1.00)1.34 (1.00)+1.32+1.410.08 (−0.05 to 0.22)0.2230.08 (−0.06 to 0.21)0.260
 Double-digit task, z-score0.37 (1.00)0.53 (1.07)+0.34+0.570.23 (0.10 to 0.36)<0.0010.23 (0.10 to 0.36)0.001
Visual search task, z-score§0.74 (0.97)0.81 (1.02)+0.79+0.76−0.09 (−0.30 to 0.13)n/s−0.07 (−0.28 to 0.15)n/s
Numeracy task, z-score§0.66 (0.95)0.66 (0.93)+0.63+0.690.00 (−0.15 to 0.15)n/s0.01 (−0.15 to 0.15)n/s
Vocabulary task, z-score§0.67 (1.41)0.70 (1.35)+0.68+0.690.02 (−0.26 to 0.31)n/s0.00 (−0.28 to 0.30)n/s
Writing task, z-score§−0.20 (1.25)−0.28 (1.43)−0.24−0.24−0.07 (−0.47 to 0.32)n/s−0.07 (0.74 to −0.47)n/s
  • Data are mean (SD) unless specified otherwise.

  • *Includes data from children in classes 4 and 5 recruited into the cognitive cohort in 80 schools, age range 6–16 years.

  • †Children in intervention schools received the educational intervention and an insecticide-treated net in April–May 2011, followed by intermittent parasite clearance in December 2011. Cognitive outcomes are compared with schoolchildren in control communities that had received nets as part of a government-run universal coverage campaign in May 2011.

  • ‡All statistical analyses account for clustering within schools, repeated measures within individuals over time and practice effects. Fully adjusted analyses control for age, sex and school class.

  • §Analyses additionally control for different versions of the tasks administered to classes 4 and 5. 95% CI are from bias-corrected bootstrap analyses using 2000 bootstrap samples, and n/s indicates that the results are not statistically significant at p=0.05.