148 - Validation of the Pediatric Resuscitation and Trauma Outcome (PRESTO) Model in Injury Patients in Tanzania

Poster Number: 148
Publication Number: 148.111

Elizabeth M. Keating, University of Utah, Salt Lake City, UT, United States; Joao V. Souza, State University of Maringa, Maringa, Parana, Brazil; Dalton B. Costa, Universidade Federal de Ciências da Saúde de Porto Alegre - UFCSPA, Porto Alegre, Rio Grande do Sul, Brazil; Arthi Kozhumam, Duke Global Health Institute, Durham, NC, United States; Modesta P. Mitao, Kilimanjaro Christian Medical University College, Kilimanjaro, Tanzania, Kilimanjaro, Tanzania; Cecilia S. Anthony, Kilimanjaro Christian Medical University College, Kilimanjaro, Kilimanjaro, Tanzania; Blandina T. Mmbaga, Kilimanjaro Christian Medical University College, Kilimanjaro, Kilimanjaro, Tanzania; Catherine A. Staton, Duke University School of Medicine, Durham, NC, United States; Joao Vissoci, Duke University School of Medicine, Hillsborough, NC, United States

Background: Sub-Saharan Africa has the world's highest rate of unintentional injury deaths among children. The Pediatric Resuscitation and Trauma Outcome (PRESTO) model predicts short-term mortality using initial patient assessment variables available in low-resource settings: patient age, hypotension via systolic blood pressure (SBP), heart rate (HR), oxygen saturation, need for supplemental oxygen (SO), and neurologic status (AVPU).

Objective: We sought to validate and assess the prognostic performance of the PRESTO model for pediatric injury patients at a tertiary-level referral hospital in Northern Tanzania.

Design/Methods: This is a cross-sectional study of 398 pediatric patients from a prospective trauma registry in Northern Tanzania from November 2020 and November 2021. Data was collected by trained research assistants and input into REDCap©. We performed exploratory data analysis of sociodemographic variables and developed a logistic regression (LR) model to predict mortality using R(version 4.1)©. The LR model was evaluated using the area under the Receiver Operating Characteristic Curve (AUC), with specificity and sensitivity used to select the decision-rule threshold (t) value.

Results: 398 participants were enrolled with a median age of 7 years old (IQR 3.41-11.18, unimodal mode of 7). Most participants were male (n=258, 65%) and the most common mechanism of injury was road traffic injury (n=141, 35%). In-hospital mortality was 7.1% (n=30). Most participants (n=326, 86%) were classified as alert on the AVPU scale, had normal SBP (n=351, 98%), and had a median HR of 107 (IQR 88.5-124). 36 (9.2%) patients needed SO via nasal cannula, nasopharyngeal airway, or intubation. The LR model executed following the original PRESTO model revealed that AVPU status, HR, and SO were statistically significant (p < 0.05) to predict in-hospital mortality. The model fit to our study population revealed an AUC of 0.89, sensitivity of 0.80, and specificity of 0.80, using a decision-threshold value of t=0.055.Conclusion(s): This is the first validation of a model to predict in-hospital mortality for pediatric injury patients in Tanzania. Despite the low number of participants, our preliminary results show good predictive potential. Further research should be done to expand our injury population in order to more solidly validate this model for use in our population and other similar low-resource settings.