Table 7 Results for the parameter estimations for the Navrongo data for sigmoid models \(\mathcal {S}\) and \(\mathcal {S}_{RE}\), compared to the results of the original analysis [16] and a previous reanalysis [39]. In brackets, the 95%CI are given. The contamination range only exists for \(\mathcal {S}\) and \(\mathcal {S}_{RE}\). The last column indicates the increase in mortality after 100 m away from the nearest discordant household. For \(\mathcal {S}_{RE}\) this is nonlinear, hence the increase after 100 m and 200 m is reported
From: Analysis of contamination in cluster randomized trials of malaria interventions
Effectiveness | Contamination range | Increase mortality shift 100m; 200m | |
|---|---|---|---|
\(\mathcal {S}_{RE}\) | 16.6%, [2.2, 30.7]% | 0.198 km, [0.092, 1.088] km | 5.6%, [0.2,15.5]%; 2.4%,[0.1,1.4]% |
\(\mathcal {S}\) | 19.0%, [7.7, 28.1]% | 0.170 km, [0.051, 0.495] km | – |
Original [16] | 17.0%, [0.0, 31.0]% | - | 6.7%, [1.8,11.4]% |
Reanalysis [39] | 18.0%, [5.0, 30.0]% | - | 1.7%, [0.6,2.6]% |