NeoSED: Early Detection of Neonatal Sepsis

Every year, there are 4 million neonatal deaths around the world, predominantly in low and middle income countries. One of the leading causes of these deaths is sepsis, which is a systemic inflammatory response to infection. Half of home delivered neonates succumb to sepsis and severe infections, indicating the magnitude of the issue. Despite efforts to move deliveries from homes to medical facilities, more than 70% of women in the two poorest wealth quintiles still deliver at home in developing countries within Sub Saharan Africa (SSA), South Asia, and Southeast Asia. In addition, neonates are discharged from the hospital as early as six hours postbirth, when their immune systems are still weak. Families in these regions depend heavily on community health workers (CHWs) to perform routine screenings for their neonates, but such screenings are for physical symptoms of general diseases and are unable to capture early signs of neonatal sepsis. Failure to rapidly detect serious neonatal infections during home care immediately after birth delays treatment, and leads to increased morbidity and mortality. Studies have shown that each hour in delay of treatment results in an approximate twofold increase in mortality rate.

The aim of our project is to achieve early risk assessment of neonatal sepsis. NeoSED is a homebased sepsis risk detection system that analyzes heart rate characteristics to output a highly sensitive sepsis risk score. NeoSED fits into the current CHW system and technological infrastructure of developing countries it features automatic data processing which allows for greater compliance, and utilizes text messages to seamlessly transmit data regarding the neonate’s status to the CHW. The NeoSED system consists of a worn device called the SepSock that houses the processing unit, heart rate sensor and user interface components. The SepSock is to be worn around a neonate’s foot semicontinuously for the first week after birth and will collect heart rate data. That data and patient history information will then be processed in situ by an algorithm, to produce a SepScore which indicates the probability that the neonate will develop sepsis. Parents can periodically send the SepScore to CHWs by plugging in their basic mobile phones to the worn device; if the score exceeds a certain threshold, an alarm message will be sent to the family members, thereby prompting them to seek critical earlier treatment.

We are confident that NeoSED has the potential to drastically reduce delays in treatment for septic neonates. On a large scale, this system may prevent hundreds of thousands of neonatal deaths due to sepsis and stands to have a great global impact.