Leveraging item-level accuracy and reaction time to address ceiling effects in the measurement of inhibitory control in preschool-aged children.

Preschool-aged children's performance on inhibitory control tasks is typically represented by the overall accuracy of their item responses (e.g., mean proportion correct). However, in settings where children vary widely in age or ability level, inhibitory control tasks are susceptible to ceiling effects, which undermine measurement precision. We have previously demonstrated a general approach for scoring inhibitory control tasks that combines item-level accuracy and reaction-time information to minimize ceiling effects. Here, we extend that approach by incorporating additional item-level reaction time data from an adjunct (simple reaction time) task. We contrast three approaches for scoring inhibitory control tasks, two of which rely exclusively on item accuracy information and a third which also considers item reaction time information. We demonstrate the impacts of these different approaches to scoring with two inhibitory control tasks that were included in a recent evaluation of the Red Light, Purple Light intervention in preprimary classrooms in Nairobi County, Kenya. We limited our study to children who met inclusion criteria at pre-test (N = 418; 51% male; mean age = 4.8 years) or post-test (N = 386; 51% male; mean age = 4.8 years). Children's performance on individual inhibitory control tasks was strongly correlated regardless of the scoring approach (rs = 0.73-0.97 across two tasks). However, the combined accuracy and reaction time scores eliminated ceiling effects that were common when only accuracy information was used. The combined accuracy and reaction time models also distinguished item-level RT into inhibitory control and processing speed components, which are distinct constructs. Results are discussed with respect to the challenges and nuances of the estimation and interpretation of inhibitory control task scores with children of varied ages and ability levels.

Testing the Efficacy of the Red-Light Purple-Light Games in Preprimary Classrooms in Kenya.

This study adapted and tested the efficacy of the Red-Light Purple-Light (RLPL) games for improving executive function (EF) skills in preprimary classrooms in Nairobi, Kenya. A cluster randomized controlled trial was used to evaluate the efficacy of the adapted RLPL intervention. Specifically, 24 centers (including 48 classrooms) were randomized to the RLPL or a wait-list control condition. Consistent with previous studies, participating classrooms delivered 16 lessons across an 8-week intervention period. A total of 479 children were recruited into the study. After exclusions based on child age and data quality, 451 and 404 children (90% retention) had completed computerized assessments of EF skills at pre- and posttest assessments, respectively. Children in the RLPL centers did not demonstrate any improvements in EF skills relative to their peers in the wait-list control condition (Cohen's ds = -0.14 to 0.03, all ps > 0.20). Exploratory tests of moderators (language of assessment, grade, school type, baseline ability) were also all null. Results are discussed with respect to measurement limitations and contextual factors that may explain the null results of RLPL on EF skills in young children in Kenya.

Measuring executive function skills in young children in Kenya: Associations with school readiness.

Most of what is known about the association between children's executive function (EF) and school readiness skills is derived from research conducted in Western countries. We tested whether these associations were evident in a middle-income country context. Participants were 1,480 children, aged 4-7 years old, who participated in an endline assessment of the Tayari program, an early childhood education (ECE) model that is being delivered by the Kenyan education system. High rates of task completion, low rates of floor effects, and high rates of assessor quality ratings supported the feasibility of large-scale direct assessments of EF with young children. Assessor ratings of children's attention-related behaviors during testing were positively associated with their performance on EF tasks (rs = 0.12-0.27). An EF composite score was not related to demographic factors or to children's exposure to the Tayari program. However, the EF composite score was uniquely associated with performance-based measures of early literacy (ß = 0.18, 95% confidence interval [CI] = 0.05, 0.31), early numeracy (ß = 0.16, 95% CI = 0.07, 26), and social-emotional competencies (ß = 0.12, 95% CI = 0.03, 0.20), even after adjustment for multiple covariates. These results are discussed with respect to the ways in which EF skills inform ongoing efforts to invest in ECE in low- and middle-income countries.

Measuring executive function skills in young children in Kenya.

Interest in measuring executive function skills in young children in low- and middle-income country contexts has been stymied by the lack of assessments that are both easy to deploy and scalable. This study reports on an initial effort to develop a tablet-based battery of executive function tasks, which were designed and extensively studied in the United States, for use in Kenya. Participants were 193 children, aged 3-6 years old, who attended early childhood development and education centers. The rates of individual task completion were high (65-100%), and 85% of children completed three or more tasks. Assessors indicated that 90% of all task administrations were of acceptable quality. An executive function composite score was approximately normally distributed, despite higher-than-expected floor and ceiling effects on inhibitory control tasks. Children's simple reaction time (ß = -0.20, p = .004), attention-related behaviors during testing (ß = 0.24, p = .0005), and age (ß = -0.24, p = .0009) were all uniquely related to performance on the executive function composite. Results are discussed as they inform efforts to develop valid and reliable measures of executive function skills among young children in developing country contexts.

Designing for Scale: Reflections on Rolling Out Reading Improvement in Kenya and Liberia.

Since 2008, the Ministries of Education in Liberia and Kenya have undertaken transitions from small-scale pilot programs to improve reading outcomes among primary learners to the large-scale implementation of reading interventions. The effects of the pilots on learning outcomes were significant, but questions remained regarding whether such large gains could be sustained at scale. In this article, the authors dissect the Liberian and Kenyan experiences with implementing large-scale reading programs, documenting the critical components and conditions of the program designs that affected the likelihood of successfully transitioning from pilot to scale. They also review the design, deployment, and effectiveness of each pilot program and the scale, design, duration, enabling conditions, and initial effectiveness results of the scaled programs in each country. The implications of these results for the design of both pilot and large-scale reading programs are discussed in light of the experiences of both the Liberian and Kenyan programs.

Evaluation of impact of long-lasting insecticide-treated bed nets and point-of-use water filters on HIV-1 disease progression in Kenya.

OBJECTIVES: Among HIV-1-infected individuals in Africa, coinfection with malaria and diarrhoeal disease may be associated with more rapid HIV-1 disease progression. We sought to determine whether the use of long-lasting insecticide-treated bed nets and simple point-of-use water filters can delay HIV-1 disease progression. DESIGN: A prospective cohort study. SETTING: Two HIV care sites in Kenya. PARTICIPANTS: HIV-1-infected adults not yet meeting criteria for antiretroviral therapy. INTERVENTIONS: One group received the standard of care, whereas the other received long-lasting insecticide-treated bed nets and water filters. Individuals were followed for up to 24 months. MAIN OUTCOME MEASURES: The primary outcome measures were time to CD4 cell count less than 350 cells/µl and a composite endpoint of time to CD4 cell count less than 350 cells/µl and nontraumatic death. Time to disease progression was compared using Cox proportional hazards regression. RESULTS: Of 589 individuals included, 361 received the intervention and 228 served as controls. Median baseline CD4 cell counts were similar (P=0.36). After controlling for baseline CD4 cell count, individuals receiving the intervention were 27% less likely to reach the endpoint of a CD4 cell count less than 350 cells/µl (hazard ratio 0.73; 95% confidence interval 0.57-0.95). CD4 cell count decline was also significantly less in the intervention group (-54 vs. -70 cells/µl per year, P=0.03). In addition, the incidence of malaria and diarrhoea were significantly lower in the intervention group. CONCLUSION: Provision of a long-lasting insecticide-treated bed net and water filter was associated with a delay in CD4 cell count decline and may be a simple, practical and cost-effective strategy to delay HIV-1 progression in many resource-limited settings.

Empiric deworming to delay HIV disease progression in adults with HIV who are ineligible for initiation of antiretroviral treatment (the HEAT study): a multi-site, randomised trial.

BACKGROUND: Co-infection with HIV and helminths is common in sub-Saharan Africa and findings from previous studies have suggested that anthelmintic treatment might delay immunosuppression in people with HIV. We aimed to assess the efficacy of empiric deworming of adults with HIV in delaying HIV disease progression. METHODS: In this non-blinded randomised trial, we enrolled adults (aged ≥18 years) with HIV who did not meet criteria for the initiation of antiretroviral treatment from three sites in Kenya. Using a computer-generated sequence, we randomly assigned (1:1) eligible participants to either empiric albendazole every 3 months plus praziquantel annually (treatment group) or to standard care (control group). Participants were followed up for 24 months. We measured CD4 cell counts every 6 months and plasma HIV RNA annually. The primary endpoints were a CD4 count of less than 350 cells per µL and a composite endpoint consisting of the first occurrence of a CD4 count of less than 350 cells per µL, first reported use of antiretroviral treatment, and non-traumatic deaths. We compared these measures by use of Cox proportional hazards regression and Kaplan-Meier survival analyses. Primary analysis was done by intention to treat. The trial was registered with ClinicalTrials.gov, number NCT0050722. FINDINGS: Between Feb 6, 2008, and June 21, 2011, we enrolled and followed-up 948 participants; 469 were allocated to the treatment group and 479 to the control group. All participants were provided with co-trimoxazole prophylaxis. Median baseline CD4 cell counts and HIV RNA concentrations did not differ between groups. We recorded no statistically significant difference between the treatment and control groups in the number of people reaching a CD4 count of fewer than 350 cells per µL (41·6 events per 100 person-years vs 46·2 events per 100 person-years; hazard ratio 0·89, 95% CI 0·75-1·06, p=0·2) or the composite endpoint (44·0 events per 100 person-years vs 49·8 events per 100 person-years; 0·88, 0·74-1·04, p=0·1). Serious adverse events, none of which thought to be treatment-related, occurred at a similar frequency in both groups. INTERPRETATION: Our findings do not suggest an effect of empiric deworming in the delaying of HIV disease progression in adults with HIV in an area where helminth infection is common. Alternative approaches are needed to delay HIV disease progression in areas where co-infections are common.

Prevalence and correlates of helminth co-infection in Kenyan HIV-1 infected adults.

BACKGROUND: Deworming HIV-1 infected individuals may delay HIV-1 disease progression. It is important to determine the prevalence and correlates of HIV-1/helminth co-infection in helminth-endemic areas. METHODS: HIV-1 infected individuals (CD4>250 cells/ul) were screened for helminth infection at ten sites in Kenya. Prevalence and correlates of helminth infection were determined. A subset of individuals with soil-transmitted helminth infection was re-evaluated 12 weeks following albendazole therapy. RESULTS: Of 1,541 HIV-1 seropositive individuals screened, 298 (19.3%) had detectable helminth infections. Among individuals with helminth infection, hookworm species were the most prevalent (56.3%), followed by Ascaris lumbricoides (17.1%), Trichuris trichiura (8.7%), Schistosoma mansoni (7.1%), and Strongyloides stercoralis (1.3%). Infection with multiple species occurred in 9.4% of infections. After CD4 count was controlled for, rural residence (RR 1.40, 95% CI: 1.08-1.81), having no education (RR 1.57, 95% CI: 1.07-2.30), and higher CD4 count (RR 1.36, 95% CI: 1.07-1.73) remained independently associated with risk of helminth infection. Twelve weeks following treatment with albendazole, 32% of helminth-infected individuals had detectable helminths on examination. Residence, education, and CD4 count were not associated with persistent helminth infection. CONCLUSIONS: Among HIV-1 seropositive adults with CD4 counts above 250 cells/mm(3) in Kenya, traditional risk factors for helminth infection, including rural residence and lack of education, were associated with co-infection, while lower CD4 counts were not.