Omphalitis Community Based Algorithm Validation Study
Each year 4 million children die before 4 weeks of age. Infections are responsible for approximately 36% of these deaths that occur annually in low- and middle-income countries and an estimated half of newborn deaths in countries with high neonatal mortality rates. Omphalitis (umbilical cord infection) is an important cause of morbidity and mortality in developing country settings. Hospital-based studies have reported incidence estimates ranging from 2 to 77 per 1,000 hospital-born infants. Contamination of the umbilical cord stump via unsterile cord cutting, traditional cord applications (i.e. ash, soil, cow dung) or unsanitary handling may lead to serious infection that could be fatal in the newborn period.
Task-shifting to community level health workers trained and equipped with community-based algorithms has become a central WHO strategy to address common pediatric illnesses in the setting of low human resources. There is evidence that community case management works well for malaria, pneumonia and diarrhea. Historically, an algorithmic approach has been utilized less often for newborn illnesses, as signs and symptoms of neonatal illness are often vague and non-specific. Omphalitis may present as redness or presence of pus discharge, making it one of the few neonatal conditions that may have objective findings, making it amenable to algorithmic diagnosis. To date, there has been no validated community-based algorithms developed and tested in sub-Saharan Africa where the manifestations of omphalitis presentation may vary and diagnosis could be potentially more challenging in infants with darker skin color. The Zambia Chlorhexidine Application Trial (ZamCAT), an ongoing cord care trial in Southern Province, Zambia recently provided an opportunity to test a diagnostic algorithm administered by the trial’s community-based lay health workers.
Newborns aged 1-10 days presenting to the health facility for routine or sick visits were enrolled in the Omphalitis Community Based Algorithm Valdiation Study (OCAVS) and underwent two independent, parallel evaluations; first, by a community level worker and second, by a Zambian medical doctor (gold standard). A third independent assessment of a photo of the cord was performed remotely by a board-certified pediatrician. Using the on-site clinician as the gold standard, the community-based algorithm and the photo assessment have been tested for concordance and the sensitivity and specificity of the algorithm was generated. Likewise, the remote pictorial assessment was compared to the gold standard to determine reliability of diagnosis from photographs alone. The study is now in the data analysis phase with results expected in 2014.
Given the current attention to cord care at the global and national policy level, validated community-based algorithms are needed to allow primary health workers to identify cord infections and reduce associated morbidity. Health systems in resource poor countries are increasingly relying on community case management of pediatric and neonatal illnesses. Now is an ideal time to refine existing diagnostic algorithms so that new cord care training packages can include both appropriate cord care messages as well as a validated method for identifying omphalitis.
|Principal Investigator||Julie Herlihy|
|Boston University Co-Investigators||Davidson Hamer, Katherine Semrau, Kojo Yeboah-Antwi, Arthur Mazimba, Caroline Grogan|
|Dates of Research||2012-2013|