Introduction
Background and rationale
Research questions and hypotheses
Primary research question
Secondary research questions
Description of research objectives
Primary objective
Secondary objectives
Study methods
Trial design
Randomization
Covariate/endpoint | Restriction criteria | Data source | Analytic method |
---|---|---|---|
Malaria disease incidence | Difference in mean baseline clinical case incidence between trial arms (size of difference to be assessed when data are available) | Baseline cohort | Difference in baseline disease incidence of cluster summaries between study arms |
Bednet use | Difference in mean proportion of persons slept under any net night before survey between trial arms ≤ 5 percentage points | Baseline survey | Difference in means of cluster summaries of proportion of persons of all ages slept under any net night before survey between arms |
Population | Total population size of larger trial arm no more than 10% larger than smaller arm | Enumeration datasets | Sum (pop size of clusters arm large)/sum (pop size of clusters arm small) less than 1.10 |
Urbanizationa | Number of urban clusters in each arm nearly balance | Census data using national classification (alternatively remotely sensed classification could be used (GRUMP/WorldPop)) | N in arm A ± 1 of N in Arm B |
Housing densitya | Difference in mean housing density between trial arms ≤ 0.3 SD of overall cluster level housing density | Enumeration + cluster boundaries GIS files Or Remote sensed data (GRUMP/WorldPop) plus Cluster boundaries GIS | SD (cluster estimates of housing densities) × 0.3 ≥|mean (cluster estimates housing density Arm a) – mean (cluster estimates of housing density Arm b)| |
HF location | Number of clusters with a primary care facility nearly balanced across arms | Study team documentation | N in arm A ± 1 of N in Arm B |
Altitude | Differences in mean altitude of cluster centroids between trial arms ≤ 0.3 SD of overall cluster level mean altitude | Digital Elevation Model (ASTER) combined with (GIS) shape files for cluster boundaries | SD (cluster estimates of altitude) × 0.3 ≥|mean (cluster estimates of altitude Arm a) – mean (cluster estimates of altitude Arm b)| |
Entomological data collection | Number of clusters with entomological data collection planned is exactly equal across study arms | Study team self-report | N in arm A = N in arm B |
Sample size
Mali | Kenya | Zambia | |
---|---|---|---|
Clusters per arm (overall) | 38 (76) | 35 (70) | 35 (70) |
Trial duration in calendar years (seasonality of FU) (total FU per participant time in months) | 2 years (8-month seasons) (16 months FU) | 2 years (12-month seasons) (24 months FU) | 2 years (6-month seasons) (12 months FU) |
α (type 1 error probability for 2-year trial) | 0.05 (Haybittle-Peto with one interim analysis) | 0.05 (Haybittle-Peto with two interim analyses at approx. 50% and 75%) | 0.05 (Haybittle-Peto with one interim analysis) |
Power | 88% | 80% | 80% |
Baseline incidence of clinical malaria in the target age group | 0.40 events per person-year (based on 0.6 incident events during an 8-month malaria season) (5y– < 15y) | 0.845 events per person-year during a 12-m malaria transmission season (1y– < 15y) | 0.50 events per 6-month malaria season (Jan–Jun) (1y– < 15y) |
Reduction in baseline incidence (incidence rate ratio = 0.70) | 30% | 30% | 30% |
Coefficient of variation | 0.40 | 0.40 | 0.40 |
Assumed loss of person time, including true LTFU plus loss due to exclusion of person time following each treatment with AL | 20% | 20% | 34% |
Total person-years required (number enrolled per cluster before loss-to-follow-up) | 3850 person-years (obtained by enrolling 38 individuals per cluster to account for loss-to-follow-up, followed for a total of 16 months) | 1260 person-years (obtained by enrolling 13.5 individuals per cluster to account for loss-to-follow-up, followed for a total of 24 months) | 1610 person-years (obtained by enrolling 35 individuals per cluster to account for loss-to-follow-up, followed for a total of 12 months) |
Mali | Kenya | Zambia | |
---|---|---|---|
Cluster per arm | 38 | 35 | 35 |
α (2-tailed) | 0.05 | 0.05 | 0.05 |
Power | 90% | 80% | 90% |
Baseline parasite prevalence measured by RDT among participants aged 6 months and older | 50% | 29.0%a | 50.0% |
Reduction in baseline prevalence | 30% | 30% | 30% |
ICC = intracluster correlation coefficient (coefficient of variation) | 0.16 (cv = 0.4) | 0.05 | 0.10 |
Non-response | 20% | 20% | 20% |
Sample size per cluster (accounting for non-response) | 25 (32) | 24 (30) | 16 (20) |
Total sample size per survey round/year (accounting for non-response) | 1900 (2432) | 1680 (2100)b | 1120 (1400) |
Kenya | Zambia | Mali | |
---|---|---|---|
Clusters per arm | 8 (16 total) | 10 (20 total) | 15 (30 total) |
α | 0.05 (two-tailed) | 0.05 (two-tailed) | 0.05 (two-tailed) |
β | 0.20 (> 80% power) | 0.20 (> 80% power) | 0.20 (> 80% power) |
Number of months of sampling per cluster | 12 | 8 | 24 |
Number of sampling sites (households) per cluster per month | 10 | 10 | 10 |
Case incidence cohort
Cross-sectional household survey
Passive case detection
Entomological endpoints
Framework
Statistical interim analyses and guidance
Overwhelming benefit rule
Stopping for harm
Timing of final analysis
Timing of outcome assessments
Primary and secondary efficacy outcomes
Primary outcome
Secondary outcomes
Statistical principles
Confidence intervals and p-values
Adherence and protocol deviations
Analysis populations
Multiplicity
Trial population
Screening data
Eligibility
-
Household resident
-
At least 12 months of age and less than 15 years of age at the time of enrollment (≥ 5 to 15 in Mali, to exclude those covered by Seasonal Malaria Chemoprevention).
-
Resident whose home is located within a buffer zone
-
Pregnant at the time of cohort enrollment.
-
Pregnant at any time during the cohort study.
Recruitment
Cohort study (For each cohort) | Cross-sectional study (each round) |
---|---|
Number of study clusters (by arm) | Number of Study clusters (by arm) |
Number of sampled houses (by arm) | Number of Sampled houses (by arm) |
Number of consented participants (HHs with participants) (by arm) | Number of consenting houses (by arm) |
Number of participants (HHs) randomized to each study arm | Number of completed interviews (by arm) |
Number of monthly follow-up visits conducted (by arm) | Number of tested individuals (by arm) |
Number of missing HH monthly visits (by arm) | Number of Incomplete HH surveys (by arm) |
Number of participants (HH) lost completely to follow-up (by arm) | Number of identified eligible participants not tested (by arm) |
Number of participants (HH) completing (by arm) |
Withdrawal/follow-up
Baseline patient characteristics
Characteristic | Cohort summary measure | Cross-sectional summary measure |
---|---|---|
Cluster level | ||
Number of clusters | N | N |
Cluster size | Mean N HH (total HH) | Mean N HH (total HH) |
Cluster size | Mean N residents (total N) | Mean N residents (total N) |
Cluster size (sampling areas) | Mean N residents (total N) | Mean N residents (total N) |
Cluster size (buffer zones) | Mean N residents (total N) | Mean N residents (total N) |
Baseline incidence | Mean Incidence rate of clinical malaria in baseline cohort per person month (variance) | Mean incidence rate of clinical malaria in baseline cohort per person month (variance) |
Baseline prevalence | Proportion positive by RDT for P. falciparum at baseline | |
Household level | ||
HH size | Mean N residents (SD) | Mean N residents (SD) |
LLIN ownership | Proportion HH with ≥ 1 LLIN (first interview) | Proportion HH with ≥ 1 LLIN |
LLIN ownership | Proportion HH with ≥ 1 LLIN per 2 residents (first interview) | Proportion HH with ≥ 1 LLIN per 2 residents |
Individual characteristics | ||
Age | Mean age (SD) | Proportion under five |
Sex | Proportion female | Proportion female |
HH size | Mean hh size of participant’s HH (SD) | Mean hh size of included hh (SD) |
Net use | Proportion slept under the net night before the survey | Proportion (tested population) slept under the net night before the survey |
Analysis
Outcome definitions
-
If a participant is symptomatic and positive by RDT, they are treated and the subsequent 2 weeks of follow-up time are censored.
-
If in the next month the participant is also symptomatic and again positive by RDT, they will be treated and PCR or microscopy will be used to determine if they are considered a case of persistent antigenemia or a true new clinical case
-
If PCR or microscopy in month two is positive, they are considered to have contributed the person-time between the previous visit and this visit less than 2 weeks and they are considered to contribute a second case to the numerator; two more weeks of follow-up will be censored following the second positive. In Mali, a person who is a malaria case on the day they re-enter the study does not contribute to the number of cases as no follow-up is associated with the case, i.e., they contribute neither to the numerator nor the denominator until they have contributed follow-up.
-
If PCR or microscopy is negative, then contributed follow-up time between the previous visit and the second visit with the second positive RDT is included (minus 2 weeks) and only one case is included in the numerator; however, two more weeks of follow-up time are censored after the second RDT positive (due to the required treatment).
Secondary outcomes
-
1. RDT infection prevalence
-
2. Passive incidence
-
3. Parity
-
4. Mosquito abundance
-
5. Sporozoite rate
-
6. Human landing/biting rate
-
7. EIR
Analysis methods
Primary outcome
Covariate adjusted analysis of the primary and secondary outcomes
Variable | Categorization (if applicable) | Analysis | Analysis population |
---|---|---|---|
Baseline prevalence | Calculated at cluster level | Clinical incidence, prevalence | ITT, per-protocol |
Baseline incidence | Calculated at cluster level | Clinical incidence, prevalence | ITT, per-protocol |
Rainfall (anomaly) | Summarized monthly at cluster level (lagged one month preceding) as anomaly | Clinical incidence, prevalence | ITT, per-protocol |
Season | Clinical incidence, prevalence | ITT, per-protocol | |
Year | One vs. Two | Clinical incidence, prevalence | ITT, per-protocol |
Age | Under 60 months vs. greater than 60 months | Clinical incidence, prevalence | ITT, per-protocol |
Subgroup analysis of the primary outcome
Subgroup name | Categorization | Rationale |
---|---|---|
Housing type | Closed eaves vs. Non-closed eaves | House structure may act as effect modifier by eliminating indoor biting risk independent of ATSB deployment |
Gender | Male vs. Female | Behavioral and occupational difference may act as effect modifier; to demonstrate equity of the intervention effect |
One month lagged rainfall (total m per m2 previous month) | High vs. low (≥ mean for study site (country) vs. < mean for study site (country)) | High levels of absolute rainfall may reduce impact of ATSB by increasing environmental carrying capacity for mosquito population |
Season | High vs low (four continuous months of the year with the highest clinical malaria incidence at local health facilities during the trial) vs. eight months with lower incidence | (Kenya only) |
Age | ≤ 60 months of age vs > 60 months of age | Behavioral differences by age may act as effect modifier |
Baseline prevalence | High vs. low (≥ median cluster prevalence vs. < median cluster prevalence) | Local endemicity may act as an effect modifier |