Obstetric Shock Index as a Predictor of Adverse Maternal Outcomes in Patients With Postpartum Haemorrhage at Kenyatta National Hospital Between January 2021 and December 2022. An Analytical Cross-sectional Study

Abstract

Background: Postpartum hemorrhage (PPH) accounts for a significant portion of avoidable obstetric deaths and is responsible for a third of maternal mortalities worldwide. The obstetric shock index (OSI) has been deemed to be a key tool for predicting and determining the risk of adverse maternal outcomes in mothers who develop PPH. Its application in the Kenyan context has not been assessed limiting its consideration as a measure for urgent intervention to prevent adverse outcomes. Objective: To establish the accuracy of obstetric shock index versus mean arterial pressure (MAP) in predicting adverse maternal outcomes such as blood transfusion > 4 units, emergency hysterectomy, severe end-organ dysfunction, ICU admission, and maternal mortality among women diagnosed with postpartum hemorrhage at KNH between January 2021 and December 2022. Methodology: A retrospective cross-sectional study was conducted at Kenyatta National Hospital. Three hundred and forty files of women with PPH were selected. OSI (HR/SBP) and MAP (DP + 1/3(SP – DP) were calculated from the vital signs recorded within one hour of the establishment of PPH. OSI was categorized based on a cut-off of 0.9. Normal OSI <0.9 while ≥0.9 was associated with a high risk of unfavorable maternal outcomes. MAP was categorized based on a cut-off of 60mmHg, interpreted as normal if above 60mmHg while that below 60mmHg was considered a high risk for adverse maternal outcomes. Sensitivity analysis was done to investigate and compare the positive predictive value, negative predictive value, specificity, and sensitivity of OSI and MAP in predicting maternal outcomes in PPH patients. The area under receiver operating curve (AUROC) was computed to evaluate the performance of each parameter in predicting adverse outcomes. Results: The composite adverse outcomes were similar between those who had an OSI of >0.9 versus those who had an OSI of <0.9 (49.5% vs 50.5%). Among those who had OSI >0.9, 57.4% were admitted into ICU, 55.9% died, 50.6% were transfused, and an emergency hysterectomy was performed in 46%. OSI was found to be a significant predictor of ICU admission with an accuracy of 57%. The AUROC established that OSI performed better (AUROC= 0.529) in predicting adverse outcomes compared to MAP (AUROC=0.472). Conclusion and recommendations: OSI was found to predict adverse outcomes accurately 52.9% of the time and was found to be a significant predictor of ICU admission. Shock index was superior to MAP in predicting adverse outcomes. These findings are an indication that OSI should be considered while developing early warning criteria and could assist in early identification of critical patients, modifying care decisions, and reducing adverse events, by facilitating prompt resuscitation and escalation in patients who develop postpartum hemorrhage.