Evaluation of a Canadian-Adapted Modified Early Obstetric Warning System (Can-MEOWS): A Quality Improvement Initiative

OBJECTIVES: This study aimed to evaluate the performance of a Canadian-adapted Modified Early Obstetric Warning System (Can-MEOWS) in predicting severe maternal morbidity (SMM), and to examine the association between individual Can-MEOWS parameters and SMM. METHODS: weeks gestation to 6 weeks postpartum at a tertiary care centre between 2022 and 2024 with documented Can-MEOWS scores. Relative risks (RRs) and 95% confidence intervals (CIs) for the composite SMM outcome were estimated using log-binomial regression. Can-MEOWS performance was examined at maximum total score thresholds using sensitivity, specificity, positive predictive value, and negative predictive value (NPV). RESULTS: Among 13 756 encounters, 13.8% experienced an SMM event, most commonly severe hemorrhage (61.6%), sepsis (24.5%), and severe hypertensive disorders of pregnancy (8.1%). Risk of SMM increased with higher total Can-MEOWS scores (RR 1.73; 95% CI 1.40-2.13 for scores 1-4; RR 4.15; 95% CI 3.33-5.17 for scores 5-6; and RR 6.70; 95% CI 5.36-8.38 for scores ≥7) compared with a Can-MEOWS score of 0. Tachypnea, fever, hypertension, and tachycardia were significantly associated with SMM. Proteinuria demonstrated a dose-dependent association with SMM, including patients with normal blood pressure. A total score ≥1 yielded the highest sensitivity (95.4%) and NPV (93.4%), whereas a score ≥7 demonstrated the highest specificity (97.6%). CONCLUSIONS: The Can-MEOWS is a reliable tool for early identification of SMM, with high sensitivity and NPV supporting its role as a screening tool. Threshold refinements will decrease false positives and alarm fatigue. These findings support broader implementation of obstetric early warning systems in Canadian centres to improve maternal safety.