10.1 statistical package® (The R Foundation for Statistical

Computing, Vienna, Austria) to obtain general prevalence and 95% confidence intervals estimated. After parasitological examination, regardless of infection status, all persons were treated with a standard dose of praziquantel (40 mg/kg) (ShinPoong Pharma., Seoul, Republic of Korea) and a single 400 mg tablet of albendazole (GSK, London, UK), or a half tablet for children aged under two years. On the basis of a positive blood film, or Paracheck© test, non-pregnant women and children were offered Lonart (20 mg/120 mg artemether/lumefrantrine medication; Cipla, Mumbai, India) while pregnant women were offered quinine sulphate tablets (Zest Pharma, Madhya Pradesh, India), as supervised by the project nurse find more and monitored the following day. A total of 15 GPS-data loggers (I-GotU GT-120, Mobile Action, UK) were available for this study. After completing a brief baseline acceptance survey questionnaire, mothers selected at random were requested

to carry this small unit (dimension 44.5 x 28.5 x 13 mm, weight 20 g) back to their homestead, returning it to the field medical team the same day. The unit was powered by a rechargeable 230 mAh Lithium-ion battery which, if set for GPS-data logging at 3-minute intervals, lasts for up to three days before needing recharging. The units were ‘locked’ electronically to avoid any external tampering. Upon receipt of the unit, data were offloaded onto a personal computer onsite as GPX files which were then used directly in GoogleEarth 5 (Google Inc., CA, USA) and ArcView 9.3 (ESRI, CA, USA) GIS. Using learn more the log it was possible to ascertain, more easily, the position of the homestead. Whilst identity records were kept anonymous, the infection status of each mother and child was used to annotate the maps to reveal any micro-patterning. To investigate the positional accuracy of the I-GotU device, the lead author accompanied 15 mothers back to their household whilst carrying a Garmin Oregon 550t handheld unit (Garmin, KS, USA). These track logs were later downloaded and directly compared against those obtained from

the I-GotU. To identify clustering of infection, from a spatial scan statistic (Satscan v9.1) was performed.23 Based on an expectation of Poisson distribution of cases of infection amongst all possible subject locations surveyed, the spatial scan statistic considered whether the number of cases in an area was excessively high or low. The scan consisted of placing circles of varying radius distances centred at each subject’s household location, and computing ratios of observed to expected cases. Both clusters of high and low prevalence were searched for in the scan. The scan statistic was performed separately for schistosomiasis, hookworm, and malaria prevalence. Additionally, a scan was performed for multiple parasite infection, i.e., persons with more than one type of parasite infection.