OS038. Pre-conceptional preparation of women with chronic hypertensionof I–II stage II of the degree of risk of CHF 0–I and recurrent bacterial vaginosis

Abstract

Introduction: Pre-eclampsia remains the second leading cause of maternal death globally, with most of these deaths occurring in LMICs. The burden of disease could be reduced with improved clinical management through the development of simple, low-cost tools that allow care providers to accurately predict adverse events within a timeframe that can inform and guide care. Objectives: To develop and validate a clinical outcome prediction model for use in women with a hypertensive disorder of pregnancy (HDP) using only symptoms and clinical signs, to predict outcomes within 48 h of assessment. Methods: This study used a prospective cohort design to collect data on women with an HDP at 7 International study sites. Women were included in the cohort if they had hypertension (systolic BP 140 mmHg or diastolic BP 90 mmHg) with or without proteinuria (dipstick >1+ or >0.3 g/24 h) or suspected HELLP syndrome. Candidate predictor variables were selected that were a priori known to be measurable, reliable, and available in LMIC settings. The model was developed using logistic regression analysis of all selected predictor variables against the dependent variable of death or severe maternal morbidity. Performance of the final model was assessed based on discrimination ability, stratification capacity and calibration. Results: From July 1, 2008 to January 31, 2012, 1540 women were recruited to the cohort. 295 (19.2%) developed one or more components of the adverse maternal outcome at any time and 174 (11.3%) within 48 h of admission. The miniPIERS model includes the predictor variables: maternal weight on admission; the symptoms of chest pain/dyspnoea, headache, visual disturbances, epigastric pain/RUQ pain, nausea/vomiting; systolic BP; and dipstick proteinuria. The model performs accurately with an area under the receiver operating characteristic curve (AUC ROC) of 0.78 (95% CI 0.75,0.80) (Fig. 1). Calibration and stratification capacity are good. When a predicted probability of 30% is used as a threshold for a positive test the positive likelihood ratio of the test is 7.18 (95% CI 5.32,9.69) showing the model can be used as a rule in test for adverse maternal outcome. Conclusion: The miniPIERS model accurately discriminates between women with a HDP who will and will not go on to develop an adverse event within 48 h of assessment. Next steps include final model refinement and internal and external validation of the model. Funding: The Bill & Melinda Gates Foundation, UNDP/ UNFPA/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproductive HealthandResearch (RHR),USAID, FIGO, Pre-eclampsia Foundation and the Rockefeller Foundation, Canadian Institutes of Health Research and the Child and Family Research Institute.