Evaluating shock index for prediction of adverse maternal outcomes related to postpartum haemorrhage and maternal sepsis in Sierra Leone: a prospective observational cohort study.

Background: Postpartum haemorrhage (PPH) and sepsis account for more than half of global maternal deaths. Shock index (SI), the ratio of heart rate to systolic blood pressure, has shown superior prediction of adverse outcomes compared to other vital signs, but most studies are retrospective, with limited evidence in sepsis. We aimed to prospectively evaluate SI for adverse outcome prediction in PPH and sepsis. Methods: This prospective cohort study was undertaken between March 2022 and July 2023 in three hospitals in Sierra Leone, in 495 and 855 women with PPH or sepsis respectively. 'First' and 'worst' SI measured in each woman was compared to other vital signs, including modified shock index (MSI), for prediction of outcomes (maternal death; major procedure; lowest haemoglobin <70 g/L; blood transfusion ≥4 international units). Predictive statistics confirmed previously determined SI thresholds (<0.9; 0.9-1.69; ≥1.7), and evaluated SI for rule-out and rule-in of adverse outcomes. Findings: SI most consistently predicted adverse outcomes compared to other vital signs, including the MSI. In cases of PPH and sepsis, the risk of adverse outcomes increased with higher SI thresholds. For maternal death in PPH, the odds ratios were 5.96 (95% confidence interval [CI]: 1.82-19.53) and 10.02 (95% CI: 3.94-25.50) for the first and worst recorded SI values respectively, and in sepsis, odds ratios were 1.91 (95% CI 0.24-15.26) and 15.4 (95% CI 7.6-31.1) for first and worst recorded SI values, when comparing SI ≥ 1.7 to SI 0.9-1.69. Sensitivity ranged from 88.4% to 95.2%, and negative predictive value ranged from 73.8% to 98.9% across outcomes. The area under the receiver operating characteristic (ROC) curve indicated poorer SI prediction of maternal death in the presence of hypertension (0.86 vs 0.59, p = 0.0020). Interpretation: SI could effectively rule-in and rule-out adverse outcomes in PPH and sepsis, enabling targeted management, and appropriate resource allocation by health care workers, and informing updates to national definitions and guidelines by policy makers. Future work should further explore the impact of hypertension on SI performance. Funding: This study was funded by the UK Medical Research Council (MR/T038942/1) and the National Institute for Health and Care Research (NIHR133232).