BackgroundThis study was part of the work to prepare for a cluster-randomized controlled trial to evaluate the effect of combining indoor residual spraying and long-lasting insecticidal nets on malaria incidence. A pilot study was done to estimate the variations of malaria incidence among villages, combined with entomological collections and an assessment of susceptibility to insecticides in malaria vectors.MethodsA cohort of 5309 residents from four kebeles (the lowest government administrative unit) in 996 households was followed from August to December 2013 in south-central Ethiopia. Blood samples were collected by a finger prick for a microscopic examination of malaria infections. A multilevel mixed effect model was applied to measure the predictors of malaria episode. Adult mosquitoes were collected using light traps set indoors close to a sleeping person, pyrethrum spray sheet catches and artificial outdoor pit shelters. Enzyme-linked immunosorbent assays were used to detect the sources of mosquito blood meals, while mosquito longevity was estimated based on parity. The World Health Organization’s tube bioassay test was used to assess the insecticide susceptibility status of malaria vectors to pyrethroids and carbamates.ResultsThe average incidence of malaria episode was 4.6 per 10,000 person weeks of observation. The age group from 5 to 14 years (IRR = 2.7; 95 % CI 1.1–6.6) and kebeles near a lake or river (IRR = 14.2, 95 % CI 3.1–64) were significantly associated with malaria episode. Only 271 (27.3 %) of the households owned insecticide-treated nets. Of 232 adult Anopheles mosquitoes collected, Anopheles arabiensis (71.1 %) was the predominant species. The average longevity of An. arabiensis was 14 days (range: 7–25 human blood index days). The overall human blood index (0.69) for An. arabiensis was higher than the bovine blood index (0.38). Statistically significant differences in Anopheline mosquitoes abundance were observed between the kebeles (P = 0.001). Anopheles arabiensis was susceptible to propoxur, but resistant to pyrethroids. However, An. pharoensis was susceptible to all pyrethroids and carbamates tested.ConclusionsThis study showed a high variation in malaria incidence and Anopheles between kebeles. The observed susceptibility of the malaria vectors to propoxur warrants using this insecticide for indoor residual spraying, and the results from this study will be used as a baseline for the trial.

BackgroundFacility-based delivery has gained traction as a key strategy for reducing maternal and perinatal mortality in developing countries. However, robust evidence of impact of place of delivery on maternal and perinatal mortality is lacking. We aimed to estimate the risk of maternal and perinatal mortality by place of delivery in sub-Saharan Africa.MethodsWe conducted a systematic review of population-based cohort studies reporting on risk of maternal or perinatal mortality at the individual level by place of delivery in sub-Saharan Africa. Newcastle-Ottawa Scale was used to assess study quality. Outcomes were summarized in pooled analyses using fixed and random effects models. We calculated attributable risk percentage reduction in mortality to estimate exposure effect. We report mortality ratios, crude odds ratios and associated 95% confidence intervals.ResultsWe found 9 population-based cohort studies: 6 reporting on perinatal and 3 on maternal mortality. The mean study quality score was 10 out of 15 points. Control for confounders varied between the studies. A total of 36,772 pregnancy episodes were included in the analyses. Overall, perinatal mortality is 21% higher for home compared to facility-based deliveries, but the difference is only significant when produced with a fixed effects model (OR 1.21, 95% CI: 1.02-1.46) and not when produced by a random effects model (OR 1.21, 95% CI: 0.79-1.84). Under best settings, up to 14 perinatal deaths might be averted per 1000 births if the women delivered at facilities instead of homes. We found significantly increased risk of maternal mortality for facility-based compared to home deliveries (OR 2.29, 95% CI: 1.58-3.31), precluding estimates of attributable risk fraction.ConclusionEvaluating the impact of facility-based delivery strategy on maternal and perinatal mortality using population-based studies is complicated by selection bias and poor control of confounders. Studies that pool data at an individual level may overcome some of these problems and provide better estimates of relative effectiveness of place of delivery in the region.Electronic supplementary materialThe online version of this article (doi:10.1186/1471-2458-14-1014) contains supplementary material, which is available to authorized users.

No abstract