ABSTRACT
BackgroundNeonatal deaths still account for a large proportion of child deaths. Understanding the extent and characteristics of neonatal deaths in Sub-Saharan Africa is a significant challenge with poor quality, inaccurate record-keeping in these settings. The process of audit is considered the foundation of quality improvement. With mHealth gaining prominence in many hospital departments, the NeoTree data collection and quality improvement system is a digital innovation deployed at Kamuzu Central Hospital (KCH), Neonatal Unit, Malawi, from April 2019, facilitating automated, real-time, prospective audit via a data dashboard.ObjectivesTo describe the pattern of admissions and outcomes in a Malawian neonatal unit over a one-year period via a prototype data dashboard, using a bedside electronic data collection and quality improvement system: NeoTree.MethodsAn electronic audit of admissions to KCH, Neonatal Unit (1st May 2019 to 31st April 2020) was carried out. Data were collected prospectively by nursing staff at the point of care, using electronic forms in the NeoTree app on tablet devices. Admission and outcome forms contained embedded reminders, education and training regarding newborn care according to national and international guidelines. Data were exported from the tablets to a cloud database. The data were then reviewed, visualised, and retrieved via a Microsoft Power BI dashboard. Data cleaning and descriptive statistics were executed using R.ResultsA total of 2732 neonates were admitted and 2413 (88.3%) had an outcome recorded electronically using the NeoTree app. Of 2413 whose outcome was known, 1899 (78.7%) were discharged alive, 12 (0.5%) were referred to another hospital and 10 (0.4%) left the hospital before being discharged. One fifth (n=492) infants died, giving an overall case fatality rate of 204/1000 admissions. Of 492 deaths, the commonest causes of death were prematurity with respiratory distress (n=252, 51%), neonatal sepsis (n=116, 23%), and neonatal encephalopathy (n=80, 16%). Almost half (45%) of deaths occurred within the first 72 hours of admission. The most common perceived modifiable factors around deaths were inadequate monitoring of vital signs and management of sepsis. Monthly trends were tracked in relation to a change in admission criteria and the COVID-19 pandemic. 202 (8.1%) neonates were HIV exposed and 143 (70.8%) of these received nevirapine as prophylaxis leaving 59 (29.2%) vulnerable to vertical infection. Data collected on the NeoTree app were on average 96% complete across all data fields and coverage of admissions, discharges and deaths was 97%, 99% and 91% respectively when compared with the ward logbook.ConclusionsThis is the first electronic, point-of-care audit of neonatal admissions and outcomes to a neonatal unit, collected by health professionals using a mobile app, reported via a data dashboard. This system achieved high coverage and completeness of data over a one-year period suggesting that M-health quality improvement systems such as the NeoTree can provide and improve quality of data for audit, timely reporting, and effective decision making regarding neonatal care in low resource settings, even during significant external stresses such as the COVID-19 pandemic. A larger scale evaluation of impact on quality of care and case fatality rate is planned.