Piers Model for the Prediction of Adverse Maternal and Perinatal Outcomes in Preclampsia at Kenyatta National Hospital, Nairobi, Kenya

Abstract

Introduction: Preeclampsia is a subset of hypertensive disorders in pregnancy (HDP) and contributes to the top 3 causes of maternal morbidity and mortality worldwide. Because of the enormous burden of adverse maternal outcomes among patients with preeclampsia, there is a need to correctly identify women at high risk of developing adverse outcomes in time to avoid their occurrence and aid decision making around the management of preeclampsia. The PIERS model (Preeclampsia Integrated Estimate of Risk) was developed to predict adverse maternal outcomes using easy to assess predictors collected within the first 48 hours of hospital admission among patients with preeclampsia. In validation studies, the fullPIERS model had a sensitivity of 85.1% while the miniPIERS model had a sensitivity of 73.8% for adverse maternal outcomes (Uber et al, 2009). The performance of PIERS model has not been evaluated in our Kenyan setting. Objective: To determine the performance of the PIERS model in predicting the risk of adverse maternal and perinatal outcomes among patients with preeclampsia at Kenyatta National Hospital, Nairobi, Kenya. Methodology: This was a descriptive prospective cohort study. Patients admitted with preeclampsia were recruited from the labour ward and antenatal wards. These patients were interviewed using a questionnaire to determine the presence of the symptom based predictor variables and their files were analysed to get the laboratory predictor values. Enrolled patients were recruited and followed up to document development of any adverse maternal and perinatal outcomes. We estimated the performance of the mini and fullPIERS model using receiver operator curves, area under the curve. x Results: Of 197 women recruited within 48 hours of admission, 12.2 % experienced an adverse maternal and 49.7% experienced adverse perinatal outcomes. The mean maternal age was 29.1 years while the mean gestational age was 34 weeks 6 days. 97 patients (49.2%) had preeclampsia with severe features. The fullPIERS model predicted adverse maternal outcomes with AUC ROC 0.647, 95% CI 0.539-0.755 while the miniPIERS model predicted adverse maternal outcomes with AUC ROC 0.654, 95% CI 0.553-0.754 within 48 hours of inclusion. The fullPIERS model predicted adverse perinatal outcomes with AUC ROC 0.62, 95% CI 0.54- 0.71 while the miniPIERS model predicted adverse neonatal outcomes with AUC ROC 0.59, 95% CI 0.5-0.69 within 48 hours after inclusion. Conclusion: These results confirm the usability of the fullPIERS model for prediction of adverse maternal and perinatal outcomes, and the usability of the miniPIERS for the prediction of adverse maternal outcomes in women admitted with preeclampsia within the first 48 hours of admission. Additional research should target stratification of patients into those presenting with early onset preeclampsia (less than 34 weeks gestation) and those presenting with late onset preeclampsia (34 weeks of gestation and above). In addition, further studies involving multicenter sites in smaller, peripheral hospitals should be conducted to assess the performanceof this model in non-teaching/referral hospitals. KEY WORDS: Preeclampsia, predictors of adverse maternal outcome, PIERS