Maternal near-miss and mortality associated with hypertensive disorders of pregnancy remote from term: a multicenter observational study in Ghana

BACKGROUND: Maternal death rates remain high in many low- and middle-income countries. Hypertensive disorders of pregnancy account for 18% of maternal mortality in Ghana. The maternal near-miss approach was designed to evaluate severe (acute) complications in pregnancy, which is useful to detect potential areas for clinical care improvement. OBJECTIVE: This study aimed (1) to determine the incidence of severe maternal complications, maternal near-miss cases, and mortality cases associated with hypertensive disorders of pregnancy remote from term and (2) to assess the health system’s performance indicators for the management of hypertensive disorders of pregnancy remote from term in middle-income country referral hospitals. STUDY DESIGN: This study was nested in the ongoing Severe Preeclampsia adverse Outcome Triage study, a multicenter observational cohort study, and included women recruited from December 1, 2017, to May 31, 2020, from 5 referral hospitals in Ghana. Women aged >16 years, admitted to the hospital with hypertensive disorders of pregnancy, with gestational age between 26 and 34 weeks were eligible. Near miss was defined according to the World Health Organization and sub-Saharan African near-miss criteria. Descriptive statistics of pregnancy and maternal and perinatal outcomes up to 6 weeks after delivery of women with severe maternal outcomes were presented for maternal deaths and maternal near-miss casigurees and compared with that of women without severe maternal outcomes. The health system’s maternal and perinatal performance indicators were calculated. RESULTS: Overall, 447 women with hypertensive disorders of pregnancy were included in the analyses with a mean maternal age of 32 (±5.8) years and mean gestational age at recruitment of 30.5 (±2.4) weeks. Of these patients, 46 (10%) had gestational hypertension, 338 (76%) had preeclampsia, and 63 (14%) had eclampsia. There were 148 near-miss cases (33.1%) and 12 maternal deaths (2.7%). Severe maternal outcomes constituted complications from severe preeclampsia (80/160 [50%]) and eclampsia (63/160 [39.4%]). Concerning organ dysfunction, hematologic and respiratory dysfunctions constituted 59/160 [38.6%] and 23/160 [14.8%] respectively. Nearly all women had a cesarean delivery (347/447 [84%] and 140/160 [93%] in the severe maternal outcome group) and delivered prematurely (83%, with 178/379 [93%] at <32 weeks of gestation). Stillbirth and neonatal deaths occurred in 63 of 455 women (14%) and 81 of 392 women (19%), respectively, constituting a stillbirth ratio of 161 per 1000 live births and neonatal mortality rate of 207 per 1000 live births as there were 392 live births in this cohort. Overall, the intensive care unit admission rate was 12.7% (n=52/409); moreover, 45 of 52 women (86.5%) admitted to the intensive care unit had severe maternal outcomes. The maternal death ratio was 3100 per 100,000 live births, the maternal near-miss–to–mortality ratio was 12.3, and the mortality index was 8%. CONCLUSION: Maternal near miss and maternal and perinatal mortalities were common in women with hypertensive disorders of pregnancy remote from term in referral hospitals in Ghana. Providing appropriate patient-centered and multidisciplinary quality care for these women is crucial in improving pregnancy outcomes. Context-tailored interventions should be considered in the clinical management of complications associated with hypertensive disorders of pregnancy in resource-limited settings. Further research on interventions to improve timely referral and reduce in-hospital delays in care provision is recommended to facilitate emergency care services for women with hypertensive emergencies. This analysis was nested within the ongoing Severe Preeclampsia adverse Outcome Triage (SPOT) study, a multicenter observational prospective cohort study in Ghana, which aims to validate previously developed risk prediction models for the management of women with preeclampsia and other HDP.18,19 Of note, 4 major referral hospitals in the Greater Accra Region (Greater Accra Regional Hospital [Ridge Hospital], Korle-Bu Teaching Hospital, La General Hospital, and Tema General Hospital) and 1 hospital in the Eastern Region of Ghana (Koforidua Regional Hospital) were selected on the basis of their large patient volume and infrastructure to conduct this study. The total number of deliveries in these facilities exceeds 30,000 annually, and all hospitals have neonatal intensive care units (NICUs). Moreover, HDP are a leading cause of maternal morbidities in these facilities and account for 18% of all maternal mortalities in the country.7 Women aged ≥16 years with a diagnosis of preeclampsia or another HDP (definitions are provided in Supplementary A) at a gestational age between 26 and 34 weeks admitted to any 1 of the participating facilities were eligible for participation in the SPOT study. The exclusion criteria were spontaneous active labor at admission and occurrence of any of the adverse maternal outcomes before meeting the inclusion criteria or collecting the independent variables. In this analysis, all women who were recruited between December 1, 2017, and May 31, 2020, were included. Women who died during admission or within the follow-up period of 6 weeks because of pregnancy-related complications were classified as maternal deaths (MDs). In all surviving women, those who met the WHO or sub-Saharan African (SSA) MNM criteria were considered MNM. Supplementary B provides an overview of the criteria. This approach was chosen because of the appropriateness of SSA in this context.20 The fulfillment of at least 1 criterion was enough to consider a woman as MNM.9,17 The SSA criterion on severe complications of abortion was not applicable as only pregnant women with a gestational age >26 weeks were considered eligible. Several other criteria were not included because of (1) limited access to laboratory tests (ie, pH or lactate), (2) nonrecording of observations that were not commonly documented in medical files (ie, acute cyanosis, gasping, or jaundice), and (3) other data that were not included in the case report forms of the SPOT study (ie, respiratory rates, urine production, loss of consciousness, cardiopulmonary resuscitation, or severe malaria). Definitions of clinical conditions and diseases that were included as maternal outcomes (eg, severe postpartum hemorrhage and severe preeclampsia) followed WHO MNM guideline definitions (Supplementary A).9 Intensive care unit (ICU) was defined as a ward where mechanical ventilation and administration of continuous vasoactive drugs were possible. This included an extended stay at the postoperative recovery room >6 hours, considering the limited availability of actual ICU departments.21 Body mass index was calculated on the basis of height in meters and weight in kilogram at first booking in antenatal care (ANC). All MNM cases and MDs conjointly were categorized as “severe maternal outcomes” (SMOs). Women who did not experience MD or near miss were considered as the comparison group. Data on near-miss cases, MDs, SMO cases, stillbirths, and neonatal mortality cases were presented as ratio per 1000 live births. The MNM mortality ratio (=MNM cases/MDs), mortality index (=MDs/SMO cases × 100%), and ICU admission rate (which is equal to the number of women admitted to the ICU/all included women) in total and among SMO cases were calculated to assess complexity and performance of care. All ratios are listed in Supplementary A.9 Trained research assistants prospectively collected data from medical records supplemented by face-to-face interview of the women to complete the information that were not initially obtained from the medical records, using standardized data collection forms designed for the SPOT study. Information regarding sociodemographic characteristics (eg, ethnicity, religion, marital status, and the highest level of education), medical history, obstetrical history (especially previous pregnancy complications), and information regarding current pregnancy and ANC services provided were recorded within 24 hours after admission. Symptoms and clinical signs of organ dysfunction were documented at the time of admission and daily during hospitalization. When delivery occurred during admission, circumstances of delivery, required interventions and maternal and neonatal outcomes were recorded. In case of discharge before delivery or data collection completion, information regarding pregnancy outcomes was collected at follow-up. Data of any readmissions before the end of pregnancy were added to the study file. Late maternal complications and neonatal outcomes were obtained at follow-up, during a routine visit, 6 weeks after delivery. All available data at the time of analysis were considered for this study. Baseline characteristics and maternal and pregnancy outcomes for all women with HDP were presented using descriptive statistics for women without SMOs, women with SMOs, near-miss cases, and MDs. Categorical variables were presented as frequency (percentage), whereas continuous variables were presented as mean (standard deviation) and transformed into categorical groups when necessary. P values were calculated using the chi-square, Fisher exact, or unpaired 2-samples Wilcoxon test. Outcomes among the 5 study sites were compared using stratified analyses. Missing values and inconsistent data were cross-checked, source documents consulted, and missing data were excluded in the analyses. All analyses were executed using R statistics (version 4.0.2; R Foundation for Statistical Computing, Vienna, Austria). The SPOT study protocol was approved by the Ghana Health Service Ethical Review Committee (protocol ID GHSERC-GHSERC015/09/17) and the Ethical and Protocol Review Committee of the College of Health Sciences, University of Ghana (protocol ID GHSERC-CHS-EtM.4-P1.2/2017-2018). All participants gave their written informed consent.