Tracking facility-based perinatal deaths in Tanzania: Results from an indicator validation assessment

• Published in: PloS one 2018-07, Vol.13 (7), p.e0201238-e0201238
• Main Authors: Plotkin, Marya, Bishanga, Dunstan, Kidanto, Hussein, Jennings, Mary Carol, Ricca, Jim, Mwanamsangu, Amasha, Tibaijuka, Gaudiosa, Lemwayi, Ruth, Ngereza, Benny, Drake, Mary, Zougrana, Jeremie, Khadka, Neena, Litch, James A, Rawlins, Barbara
• Format: Article
• Language: English
• Subjects: Adult; Babies; Biology and Life Sciences; Children & youth; Childrens health; Community development; Death; Discharge; Fatalities; Female; Fetuses; Health care facilities; Health Facilities; Heart rate; Hospital facilities; Humans; Indicators; Infant, Newborn; Information management; Information systems; Internet; Library cataloging; Male; Management information systems; Medicine and Health Sciences; Mortality; Neonates; Newborn babies; Obstetrics; Perinatal Care; Perinatal Death; Quality; Stillbirth; Surveillance; Tanzania - epidemiology; Training
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0201238

Globally, an estimated 2.7 million babies die in the neonatal period annually, and of these, about 0.7 million die from intrapartum-related events. In Tanzania 51,000 newborn deaths and 43,000 stillbirths occur every year. Approximately two-thirds of these deaths could be potentially prevented with improvements in intrapartum and neonatal care. Routine measurement of fetal intrapartum deaths and newborn deaths that occur in health facilities can help to evaluate efforts to improve the quality of intrapartum care to save lives. However, few examples exist of indicators on perinatal mortality in the facility setting that are readily available through health management information systems (HMIS). From November 2016 to April 2017, health providers at 10 government health facilities in Kagera region, Tanzania, underwent refresher training on perinatal death classification and training on the use of handheld Doppler devices to assess fetal heart rate upon admission to maternity services. Doppler devices were provided to maternity services at the study facilities. We assessed the validity of an indicator to measure facility-based pre-discharge perinatal mortality by comparing perinatal outcomes extracted from the HMIS maternity registers to a gold standard perinatal death audit. Sensitivity and specificity of the HMIS neonatal outcomes to predict gold standard audit outcomes were both over 98% based on analysis of 128 HMIS-gold standard audit pairs. After this validation, we calculated facility perinatal mortality indicator from HMIS data using fresh stillbirths and pre-discharge newborn death as the numerator and women admitted in labor with positive fetal heart tones as the denominator. Further emphasizing the validity of the indicator, FPM values aligned with expected mortality by facility level, with lowest rates in health centers (range 0.3%- 0.5%), compared to district hospitals (1.5%- 2.9%) and the regional hospital (4.2%). This facility perinatal mortality indicator provides an important health outcome measure that facilities can use to monitor levels of perinatal deaths occurring in the facility and evaluate impact of quality of care improvement activities.