Global Research Review: Issue 7

Published on July 10, 2013
On behalf of the Emergency Medicine Literature Review Group

This cluster-randomized field trial evaluated the effectiveness of green bananas for in-home management of acute and prolonged childhood diarrhea. The authors identified 72 clusters in the rural Mirsarai sub-district of Bangladesh, each containing approximately 3,000 persons. Eighteen clusters were randomly chosen and assigned to standard diarrhea care based on WHO guidelines (12 clusters) and standard care plus green bananas (6 clusters). To evaluate children with prolonged diarrhea (defined as symptoms longer than seven days), the 12 standard care clusters were further randomized to standard care (6 clusters) and standard care plus green bananas (6 clusters), commencing only after the seventh day of symptoms. Nonhospitalized children aged 6-36 months with active diarrhea were followed for 14 days by a network of local women and trained fieldworkers. Over 20 months, 2968 children were enrolled, of whom 198 were further evaluated for prolonged diarrhea.The cumulative probability of symptomatic cure was higher in the green banana group for both acute diarrhea (hazard ratio = 0.63, P < 0.001) and prolonged diarrhea (hazard ratio = 0.38, P < 0.001). Recovery rates of children receiving green bananas were significantly higher for acute diarrhea at day 3 (79.9% vs. 53.3%, P <0.001) and at day 7 (96.6% vs. 89.1%, P < 0.001). Children with prolonged diarrhea also benefited from green bananas, with recovery rates at day 10 of 79.8% vs. 51.9% (P < 0.001), and at day 14 of 93.6% vs. 67.2% (P < 0.001).

Green banana is an inexpensive traditional remedy for childhood diarrhea, and was shown in this study to be effective in decreasing the duration of illness for children treated for diarrhea at home. The cluster randomization, large sample size, low rate of attrition (0.1% dropout rate), well-defined study protocol, and use of an intention-to-treat analysis led to very strong internal validity of the conclusions. The lack of blinding, and reliance on self-reported data from mothers were identified as potential sources of bias and addressed by the authors. Data on maternal education and breastfeeding habits, which are other known confounders in diarrheal illness, were notably missing from this study. Nevertheless, the study findings provide health care professionals in resource-limited settings with a potentially useful adjunct to the management of childhood diarrhea, which is a common source of infant and child morbidity and mortality in the developing world. Further research is necessary to determine whether other similar starch-based remedies might also be efficacious when green banana is not available. Generalizability of the findings to other regions, and the relative cost-effectiveness of the treatment compared to other therapies for diarrhea, must also be studied before widespread implementation of green banana can be recommended at the policy level.

The authors of this study attempt to quantify the injury burden from insurgent conflict to the population within the Baghdad Governorate of central Iraq. The study was conducted from October 2009 to November 2009. By describing the injury burden, including the indirect injury from the breakdown of infrastructure, the authors hope to help guide appropriate areas of injury prevention and treatment. Previous attempts to quantify injuries in the country have focused mainly on direct mortality. In this study, a two-stage, cluster sample was used and households were randomly selected for inclusion. The Iraqi Ministry of Health staff then administered a cross-sectional survey to the heads of the households. The survey asked participants to recall all injuries incurred in the last three months, and injuries resulting in death in the last year. The authors define injury as any type of physical harm to an individual that created the loss of at least one day of normal activity, or that caused the individual to seek health care treatment. The Iraqi Ministry of Health staff administered the survey to a total of 1,172 households, obtaining data regarding 7,396 individuals. For the three month recall period, there were 103 reported injuries, three of which were injury-related death. There were seven cases of injury-related deaths within the 12 month recall period. As such, there was a 1.7% (95% CI = 0.7% to 3.5%) proportionate mortality for injuries. Of these injuries, only 8.9% were recorded as intentional. Injury incidence was much greater in men, displaced persons, and those with less education.

By describing the injury burden through the application of a cross-sectional household survey, this study emphasizes that a significant proportion of injury is related to conflict and infrastructure breakdown. Furthermore, it is evident that intentional injury, which is often publicized and discussed, makes up only a small proportion of the total injury burden. In an area of conflict, study limitations include survey responders’ fear to participate, and the dynamic state of security during any specific time period. It should be noted that extremely insecure zones were deliberately avoided, providing some selection bias. Finally, this was a retrospective study based on recall and selfreporting. As such, overall injury burden, intentional injury, and mortality-related injury were all most likely underestimated. It is also important to acknowledge that a state of conflict often serves to prevent other forms of otherwise more common injuries such as motor vehicle-related injuries. This study effectively emphasizes the injury burden in an area of conflict, including the effect of social and infrastructure breakdown.


This article originally appeared in issue 7 of Emergency Physicians International

e.g. Global, Research, or India
e.g. Features, Opinion, or Research
Browse by Category
    Most Popular
      Download Latest Issue
      Issue 19