Univariate and multivariate logistic regression analysis for treatment failure comparing each group. 95% confidence intervals in parenthesis.

1Adjusted odds ratios (AOR) adjust for refugee village, sex and age.
2Restricted analysis excludes all failures during treatment period (months 0–1).
3Restricted analysis excludes all failures occurring before 33 days for placebo group, 44 days for 14 day PQ group and 91 days for 8 week group. Time under observation is censured at 244 days post restriction period to give equal follow-up period.
4p<0.001
5p<0.01