Abstract
Background
Preeclampsia with severe features (PEC-SF) is a hypertensive disorder of pregnancy associated with significant maternal and perinatal morbidity. It is defined by heterogenous criteria, including severe hypertension and signs of end-organ dysfunction. Whether individual criteria confer differential risk for adverse perinatal outcomes remains unclear. This gap limits clinicians’ ability to risk stratify patients, guide management decisions, and provide individualized counseling regarding prognosis.
Objective
To estimate the association between individual diagnostic criteria in preeclampsia with severe features and adverse perinatal outcomes.
Study Design
This is a retrospective cohort study of patients with antepartum PEC-SF who delivered at two academic medical centers from 2016–2017. Exposures included the number and type of severe criteria present at diagnosis. Our primary outcome was a composite of placental abruption, preterm birth < 34 weeks, a small for gestational age infant, or fetal demise. Secondary outcomes included preterm birth < 37 weeks, delivery type, or < 7 days of latency from diagnosis to delivery among those diagnosed at < 33 weeks. A multivariate logistic regression model adjusted for maternal age, race, and pre-existing medical conditions estimated the adjusted odds ratios (aOR) and 95% confidence interval (CI) for the primary outcome.
Results
There were 704 patients with PEC-SF. Of these, 608 (86.4%) had severe hypertension, 246 (34.9%) reported neurological symptoms, and 332 (47.2%) presented with ≥ 2 severe criteria. Of 704, 371 (52.7%) had a single criterion, most often severe hypertension (81.1%). The primary outcome was associated with severe hypertension (aOR: 1.36, 95% CI: 1.22, 1.52), thrombocytopenia (aOR 1.28, 95% CI 1.13, 1.45), progressive renal insufficiency (aOR 1.16, 95% CI 1.03, 1.31), and impaired liver function (aOR 1.15, 95% CI 1.04, 1.28). Unplanned cesarean delivery was associated with severe hypertension (aOR:1.42, 95% CI: 1.02, 1.98) and impaired liver function (aOR: 1.13, 95% CI: 1.02, 1.25). The odds of the primary composite outcome increased with each additional criterion; among patients with severe hypertension, the association between the primary outcome was significant among those with two to three additional severe features (aOR 1.47, 95% CI 1.30, 1.67).
Conclusion
Our findings demonstrate that not all severe features carry equal odds of adverse pregnancy outcome and highlight the importance of considering both the type and number of criteria when risk stratifying patients and guiding clinical management.
Keywords: Preeclampsia, hypertension, perinatal outcomes, severe features, preterm birth
TWEETABLE STATEMENT
In preeclampsia with severe features, laboratory markers of end-organ dysfunction and hypertension are associated with adverse perinatal outcomes.
CONDENSATION
Severe hypertension and laboratory markers of end-organ dysfunction, but not neurological symptoms, are associated with the composite adverse perinatal outcomes in preeclampsia with severe features.
INTRODUCTION
Preeclampsia is a hypertensive disorder of pregnancy with significant short-term and long-term complications for both mother and infant. In the United States, preeclampsia occurs in 2–4% of pregnancies,1 and based on 2017–2019 Centers for Disease Control and Prevention data, up to 7% of maternal deaths are related to hypertensive disorders of pregnancy.2
The American College of Obstetricians and Gynecologists (ACOG) define preeclampsia as new-onset hypertension after 20 weeks’ gestation with either proteinuria (≥300 mg/24 hour urine, urine protein: creatinine ratio ≥0.3 mg/dL) or signs of end-organ dysfunction. Preeclampsia with severe features (PEC-SF) is defined by the presence of severe hypertension and/or evidence of end-organ dysfunction, including neurologic symptoms, renal insufficiency, hepatic dysfunction, thrombocytopenia, or pulmonary edema. We defined neurological symptoms consistent with ACOG criteria to be headache unresponsive to analgesics and persistent visual disturbances. PEC-SF also includes hemolysis, elevated liver enzymes, and low platelet count (HELLP) syndrome, and eclampsia, and can be diagnosed in the absence of proteinuria if a severe feature is present.3
Compared with preeclampsia without severe features, PEC-SF is associated with a greater risk for adverse perinatal outcomes.4 Adverse maternal outcomes include higher rates of cesarean delivery, placental abruption, increased length of stay in the hospital, intensive care unit (ICU) admission, and mortality.3–9 Adverse fetal outcomes include increased risk of preterm birth and impaired fetal growth.6
Although ACOG defines criteria for PEC-SF, the relative frequency of individual criteria and independent associations with adverse perinatal outcomes remain unclear.
Despite established associations between PEC-SF and adverse maternal and neonatal outcomes, it remains unclear whether individual severe features confer differential risk. This limits clinicians’ ability to risk stratify patients, guide management decisions, and provide individualized counseling regarding prognosis.10 Therefore, we aimed to evaluate the relationship between individual PEC-SF diagnostic criteria and adverse perinatal outcomes. We hypothesized that severe hypertension and neurologic symptoms would be associated with higher risk than those with laboratory abnormalities, and that the presence of multiple severe features would demonstrate a stronger association with adverse outcomes compared to single or two criteria.
MATERIALS AND METHODS
This was a retrospective cohort study of patients diagnosed with PEC-SF, either prior to or during delivery hospital admission at two tertiary care centers. Institutional Review Board (IRB) approval was obtained from both institutions. The study included patients with PEC-SF from 12/27/2015 to 08/25/2017 at one center and 01/03/2016 to 12/30/2017 at the other. Data were manually abstracted from patients’ electronic health records (EHR). Extracted EHR study data were collected and managed using REDCap electronic data capture tools hosted at one of the study centers. Both centers provided institutional review board approval with a waiver of Health Insurance Portability and Accountability Act (HIPAA) authorization and written consent.
Eligible patients were identified as having a diagnosis of PEC-SF, HELLP syndrome, or eclampsia at delivery hospitalization or postpartum readmission up to six weeks after delivery, using ICD-10 codes O14.1, O14.2, and O15.0, respectively. Patients were included if, on chart review, the diagnosis of PEC-SF was confirmed, they were 18 years or older and delivered a singleton infant at one of the two centers within the study date range. Patients with more than one pregnancy meeting the eligibility criteria during the study period were included for each pregnancy and abstracted separately. Exclusion criteria included patients who did not have delivery admission data available in the EHR, patients who were found not to have a diagnosis of PEC-SF, those with multiple gestations, or those who were identified as duplicates between the two medical centers.
Diagnosis of PEC-SF was based on the ACOG Practice Bulletin on gestational hypertension and preeclampsia3. Exposures of interest included the number and type of severe criteria present at the time of diagnosis. Patients were identified as having only 1 defining criteria, 1 or more, 2 or more, and so on. Patients with multiple criteria were included in each of the relevant severe categories for the analysis. Data collected from the EHR included maternal demographic data, obstetric history, other pre-existing medical conditions, timing and diagnosis of PEC-SF, medication use during pregnancy, mode of delivery, and perinatal outcomes.
Our primary composite outcome was a composite of adverse placentally-mediated perinatal events, defined as the occurrence of one or more of the following: placental abruption, preterm birth prior to 34 weeks’ gestation, infant diagnosed as small for gestational age (SGA, birthweight less than 10th percentile for gestational age), and intrauterine fetal demise (IUFD).12,13 Secondary outcomes included preterm birth prior to 37 weeks’ gestation, delivery by unscheduled/unplanned cesarean delivery, or fewer than 7 days of latency from time of diagnosis of PEC-SF to delivery among those diagnosed at < 33 weeks’ gestation.
The prevalence of maternal demographic, pre-existing medical conditions, and occurrence of each severe criteria were summarized using descriptive statistics. A multivariable logistic regression model was used to estimate the association between individual criteria for PEC-SF and primary composite outcome, individual components of the composite, and the secondary outcomes. We adjusted for maternal age, race, and presence of pre-existing medical conditions, which we selected based on established associations with preeclampsia severity and adverse perinatal outcomes.3,11 These conditions included cardiac disease, diabetes, chronic hypertension, sleep apnea, asthma and other chronic lung disease, renal disease, lupus, history of DVT/PE, alcohol use disorder, HIV/AIDS, pre-diabetes, bleeding disorders, autoimmune disease, placenta previa, placenta accreta, anemia, prior bariatric surgery, gastrointestinal disease, mental health disorders, and thyrotoxicosis. Each individual severe feature was included as an independent predictor in the model, and individuals may have had more than one severe feature. Maternal obesity (body mass index (BMI) ≥ 30 kg/m2) and tobacco use were not included in the primary adjusted model due to high rates of missingness or unreliable documentation in the EHR. To assess whether omission of these variables meaningfully affected our findings, we conducted a sensitivity analysis incorporating both covariates, imputing missing values as non-obese and no current tobacco use. This conservative approach was chosen to preserve sample size while providing an estimate of the potential influence of these factors on the primary results. Because severe hypertension was present in 86.4% of the cohort, it was difficult to determine independent association between the primary outcome and accumulating severe criteria independent of severe hypertension itself. To address this, we conducted a secondary analysis restricted to patients with severe hypertension, using those with severe hypertension as their only severe feature as the reference group. By holding severe hypertension constant, this approach allowed us to isolate the additional risk associated with other severe features and assess whether increasing end-organ involvement confers risk beyond that associated with severe hypertension alone.
The results are reported as odds ratios (OR) with 95% confidence intervals (CI). A p-value less than 0.05 was considered statistically significant. All statistical analyses were conducted using R, version 4.1.0.
RESULTS
During the study period, 931 patients were identified as having a diagnosis of PEC-SF. Of these, 704 (76%) met study inclusion criteria and are included in this analysis (Figure 1). The median age was 30 years; 284 (40.3%) self-identified as Black and 266 (37.8%) self-identified as white (Table 1). The average pre-pregnancy BMI in the cohort was 30.0 kg/m2, and 548 (77.8%) identified as non-Hispanic (Table 1). The most common pre-existing comorbid medical conditions included chronic hypertension and pulmonary disorders in 158 (22.4%) and 123 (17.5%) of the cohort, respectively. Of the 704, 371 (52.7%) had only one defining criterion: 301 (81.1%) severe hypertension, 36 (9.7%) headache/neurological symptoms, 22 (5.9%) impaired liver function, 7 (1.9%) thrombocytopenia, and 5 (1.3%) progressive renal insufficiency. Overall, 608 (86.4%) had severe hypertension, 246 (34.9%) reported headache unresolved with analgesics or neurological symptoms, and 332 (47.2%) had 2 or more severe criteria.
Figure 1:
STROBE flow chart
Table 1:
Demographic and medical characteristics of study cohort
| Total Cohort (n=704) | |
|---|---|
| N (%) or Median [25th, 75th percentile) | |
| Age (years) | 30 [25, 34] |
| Race | |
| American Indian | 7 (1.0) |
| Asian | 10 (1.4) |
| Black | 284 (40.3) |
| Multiple | 2 (0.3) |
| Other | 114 (16.2) |
| Pacific Islander | 1 (0.1) |
| Unknown | 20 (2.8) |
| White | 266 (37.8) |
| Ethnicity | |
| Hispanic | 152 (21.6) |
| Non-Hispanic | 548 (77.8) |
| Not documented | 4 (0.6) |
| Pre-Pregnancy Body Mass Index (kg/m2) | 30.0 [25.1, 38.2] |
| Tobacco use during pregnancy | 58 (8.2) |
| Selected pre-existing medical conditions | |
| Chronic hypertension | 158 (22.4) |
| Diabetes mellitus | 72 (10.2) |
| Gestational diabetes (current pregnancy) | 60 (8.5) |
| Pulmonary disorders | 123 (17.5) |
| History of renal disease | 17 (2.4) |
| Systemic lupus erythematosus | 8 (1.1) |
| Deep vein thrombosis/pulmonary embolism | 9 (1.3) |
| Gestational age at diagnosis* | 36 [32, 38] |
| Diagnosis made in postpartum period | 49 (7.0) |
| Presence of severe criteria ** | |
| Severe hypertension | 608 (86.4) |
| Thrombocytopenia | 78 (11.1) |
| Progressive renal insufficiency | 65 (9.2) |
| Impaired liver function | 135 (19.2) |
| Pulmonary edema | 15 (2.1) |
| Headache/neurological symptoms | 246 (34.9) |
| Two or more criteria | 332 (47.2) |
| Three or more criteria | 89 (12.6) |
| HELLP syndrome | 74 (10.5) |
| Eclampsia | 9 (1.3) |
Includes only those with antepartum diagnosis
Patients may have had more than one criterion therefore total exceeds 704
In the unadjusted analysis, the primary composite outcome was most frequent in patients whose qualifying criteria were thrombocytopenia or pulmonary edema, occurring in 49/78 (62.8%) and 11/15 (66.7%) of patients, respectively (Table 2). Of the individual components of the primary composite outcome, preterm delivery < 34 weeks’ gestation was most common among those whose qualifying criteria were thrombocytopenia, progressive renal insufficiency, and/or pulmonary edema, occurring in 37 (47.4%), 31 (47.7%), and 7 (46.7%) of patients with those qualifying criteria, respectively (Table 2). IUFD occurred most frequently among those with pulmonary edema, though event counts were small (Table 2).
Table 2.
Prevalence of primary and secondary composite outcomes by presence of specific criteria of preeclampsia with severe features*
| Severe hypertension N= 608 (86.4%) |
Thrombocytopenia N= 78 (11.1%) |
Progressive renal insufficiency N= 65 (9.2%) |
Impaired liver function N= 135 (19.2%) |
Pulmonary edema N= 15 (2.1%) |
Headache or neurological symptoms N= 246 (34.9%) |
|
|---|---|---|---|---|---|---|
| Primary composite outcome | 243 (40.0%) | 49 (62.8%) | 36 (55.4%) | 67 (49.6%) | 10 (66.7%) | 93 (37.8%) |
| Preterm delivery< 34 weeks | 178 (29.3%) | 37 (47.4%) | 31 (47.7%) | 53 (39.3%) | 7 (46.7%) | 78 (31.7%) |
| Small for gestational age | 79 (13.0%) | 15 (19.2%) | 5 (7.7%) | 19 (14.1%) | 1 (6.7%) | 19 (7.7%) |
| Placental abruption | 22 (3.6%) | 6 (7.7%) | 2 (3.1%) | 4 (3.0%) | 2 (13.3%) | 6 (2.4%) |
| Intrauterine fetal demise | 15 (2.5%) | 5 (6.4%) | 4 (6.2%) | 4 (3.0%) | 2 (13.3%) | 3 (1.2%) |
| Secondary outcomes | ||||||
| Preterm delivery < 37 weeks | 345 (56.7%) | 52 (66.7%) | 42 (64.6%) | 82 (60.7%) | 10 (66.7%) | 156 (63.4%) |
| Mode of delivery | ||||||
| Spontaneous Vaginal | 271 (44.6%) | 20 (25.6%) | 19 (29.2%) | 60 (44.4%) | 5 (33.3%) | 113 (45.9%) |
| Forceps-Assisted Vaginal | 9 (1.5%) | 0 (0.0%) | 1 (1.5%) | 1 (0.7%) | 0 (0.0%) | 5 (2.0%) |
| Vacuum-Assisted Vaginal | 3 (0.5%) | 2 (2.6%) | 1 (1.5%) | 1 (0.7%) | 0 (0.0%) | 0 (0.0%) |
| Planned/scheduled Cesarean | 217 (35.7%) | 42 (53.8%) | 25 (38.5%) | 54 (40.0%) | 6 (40.0%) | 82 (33.3%) |
| Unplanned/unscheduled Cesarean | 107 (17.6%) | 14 (17.9%) | 19 (15.4%) | 18 (14.1%) | 3 (20.0%) | 46 (18.7%) |
| Latency from diagnosis to delivery < 7d** | 31/168 (18.5%) | 3/36 (8.3%) | 5/31 (16.1%) | 4/47 (8.5%) | 0/7 (0.0%) | 8/70 (11.4%) |
Percent of patients with each criterion may add to > 100% because some had multiple criteria
Analysis limited to patients diagnosed at < 33 weeks’ gestation
In the adjusted analysis, the primary composite outcome was significantly associated with presence of severe hypertension (aOR: 1.36, 95% CI: 1.22, 1.52), thrombocytopenia (aOR: 1.28, 95% CI: (1.13, 1.45), progressive renal insufficiency (aOR: 1.16, 95% CI: 1.03, 1.31), and impaired liver function (aOR: 1.15, 95% CI: 1.04, 1.28) (Table 3). Among the components of the primary composite outcome, preterm birth at less than 34 weeks’ gestation was significantly associated with severe hypertension (aOR: 1.29, 95% CI: 1.17, 1.42), thrombocytopenia (aOR: 1.20, 95% CI: 1.07, 1.35), progressive renal insufficiency (aOR: 1.23, 95% CI: 1.10, 1.38), impaired liver function (aOR: 1.17, 95% CI: 1.07, 1.28), and neurologic symptoms (aOR: 1.12, 95% CI: 1.05, 1.20) (Table 3). Neurological symptoms were not significantly associated with the primary composite outcome (aOR: 1.07, 95% CI: 1.00, 1.16). Infants who were SGA were significantly associated with severe hypertension (aOR: 1.09, 95% CI: 1.01, 1.17) and thrombocytopenia (aOR: 1.10, 95% CI: 1.01, 1.20) (Table 3). Among secondary outcomes, severe hypertension (aOR: 1.21, 95% CI: 1.08, 1.36) and neurological symptoms (aOR: 1.16, 95% CI: 1.07, 1.25) were significantly associated with preterm birth at less than 37 weeks’ gestation (Table 3). Unscheduled/unplanned cesarean delivery was significantly associated with severe hypertension (aOR:1.42, 95% CI: 1.02, 1.98) and impaired liver function (aOR: 1.13, 95% CI: 1.02, 1.25) (Table 3). When looking at latency to delivery < 7 days, no individual criterion predicted rapid delivery among those diagnosed <33 weeks.
Table 3.
Odds ratio of primary and secondary composite outcomes by presence of severe criteria, adjusted for maternal age, race, and one or more pre-existing medical conditions
| Adjusted Odds Ratio (95% CI) | |
|---|---|
| Primary composite outcome | |
| Severe hypertension | 1.36 (1.22, 1.52) |
| Thrombocytopenia | 1.28 (1.13, 1.45) |
| Progressive renal insufficiency | 1.16 (1.03, 1.31) |
| Impaired liver function | 1.15 (1.04, 1.28) |
| Pulmonary edema | 1.27 (0.99, 1.62) |
| Headache/neurological symptoms | 1.07 (1.00, 1.16) |
| Individual components of primary composite outcome | |
| Preterm delivery < 34 weeks | |
| Severe hypertension | 1.29 (1.17, 1.42) |
| Thrombocytopenia | 1.20 (1.07, 1.35) |
| Progressive renal insufficiency | 1.23 (1.10, 1.38) |
| Impaired liver function | 1.17 (1.07, 1.28) |
| Pulmonary edema | 1.13 (0.90, 1.42) |
| Headache/neurological symptoms | 1.12 (1.05, 1.20) |
| Small for gestational age | |
| Severe hypertension | 1.09 (1.01, 1.17) |
| Thrombocytopenia | 1.10 (1.01, 1.20) |
| Progressive renal insufficiency | 0.93 (0.85, 1.02) |
| Impaired liver function | 1.01 (0.94, 1.08) |
| Pulmonary edema | 0.92 (0.78, 1.10) |
| Headache/neurological symptoms | 0.95 (0.90, 1.00) |
| Placental abruption | *NP |
| Intrauterine fetal demise | *NP |
| Secondary outcomes | |
| Preterm delivery < 37w | |
| Severe hypertension | 1.21 (1.08, 1.36) |
| Thrombocytopenia | 1.13 (0.99, 1.29) |
| Progressive renal insufficiency | 1.11 (0.98, 1.27) |
| Impaired liver function | 1.10 (0.98, 1.22) |
| Pulmonary edema | 1.02 (0.79, 1.32) |
| Headache/neurological symptoms | 1.16 (1.07, 1.25) |
| Unscheduled/unplanned cesarean delivery | |
| Severe hypertension | 1.42 (1.02, 1.98) |
| Thrombocytopenia | 1.07 (0.98, 1.17) |
| Progressive renal insufficiency | 1.05 (0.94, 1.16) |
| Impaired liver function | 1.13 (1.02, 1.25) |
| Pulmonary edema | 0.96 (0.88, 1.04) |
| Headache/neurological symptoms | 0.96 (0.78, 1.18) |
| Latency < 7days * | |
| Severe hypertension | 1.03 (0.82, 1.31) |
| Thrombocytopenia | 0.94 (0.80, 1.11) |
| Progressive renal insufficiency | 1.01 (0.86, 1.18) |
| Impaired liver function | 0.91 (0.78, 1.06) |
| Pulmonary edema | 0.87 (0.65, 1.17) |
| Headache/neurological symptoms | 0.90 (0.80, 1.01) |
NP=not performed as outcome was too rare to preserve logistic regression model integrity
Analysis limited to patients diagnosed at < 33 weeks’ gestation
In the adjusted analysis evaluating the association between the number of qualifying criteria and the primary and secondary outcomes, greater than 1 qualifying criterion was significantly associated with the primary composite outcome, preterm birth at less than 37 weeks’ gestation, and preterm birth at less than 34 weeks’ gestation (Table 4). In the sub-analysis restricted to the 608 patients with severe hypertension, using those with severe hypertension as the reference group, (Table 5) we found the odds of the primary composite outcome increased with each additional criterion; however, unlike the main analysis, the odds of the primary composite outcome for the sub-analysis only reached significance for patients with severe hypertension and 2–3 additional severe feature criteria (aOR 1.47, 95% CI 1.30–1.67). A similar significant graded association for preterm birth at less than 37 weeks and preterm birth at less than 34 weeks was seen in the sub-analysis.
Table 4.
Odds ratio of primary and secondary composite outcomes by number of severe criteria, adjusted for maternal age, race, and one or more pre-existing medical conditions
| Number of criteria | Primary composite outcome | Preterm birth < 34 weeks | Small for gestational age | Preterm birth < 37 weeks | Unplanned/unscheduled cesarean delivery | < 7 days latency |
|---|---|---|---|---|---|---|
| 1 | Reference group | |||||
| 2 | 1.11 (1.03, 1.20) | 1.17 (1.09, 1.25) | 0.95 (0.90, 1.00) | 1.16 (1.08, 1.26) | 1.05 (0.99, 1.12) | 0.97 (0.85, 1.11) |
| 3–4 | 1.50 (1.34, 1.69) | 1.51 (1.36, 1.68) | 1.01 (0.93, 1.10) | 1.31 (1.16, 1.48) | 0.99 (0.91, 1.09) | 0.84 (0.72, 0.98) |
| ≥ 5 | 1.64 (1.04, 2.59) | 1.56 (1.07, 2.27) | 1.07 (0.80, 1.43) | 1.37 (0.89, 2.11) | 1.59 (1.14, 2.21) | 0.79 (0.51, 1.23) |
Table 5.
Odds ratio of primary and secondary composite outcomes by number of additional severe criteria among patients with severe hypertension, adjusted for maternal age, race, and one or more pre-existing medical conditions
| Number of criteria | Primary composite outcome | Preterm birth < 34 weeks | Small for gestational age | Preterm birth < 37 weeks | Unplanned/unscheduled cesarean delivery | < 7 days latency |
|---|---|---|---|---|---|---|
| Severe HTN only | Reference group | |||||
| Severe HTN + 1 | 1.08 (1.00, 1.18) | 1.16 (1.08, 1.25) | 0.93 (0.88, 0.99) | 1.18 (1.08, 1.28) | 1.06 (0.99, 1.13) | 0.98 (0.85, 1.12) |
| Severe HTN + 2–3 | 1.47 (1.30, 1.67) | 1.49 (1.33, 1.67) | 1.00 (0.92, 1.10) | 1.30 (1.14, 1.47) | 0.98 (0.89, 1.09) | 0.83 (0.71, 0.98) |
| Severe HTN + 4–5 | 1.58 (0.99, 2.52) | 1.77 (1.15, 2.72) | 0.87 (0.62, 1.20) | 1.62 (1.01, 2.61) | 1.44 (0.99, 2.09) | 0.79 (0.51, 1.24) |
Eighty-nine (12.6%) were missing BMI and 23 (3.27%) were missing tobacco use. In sensitivity analysis, we imputed missing data for maternal BMI ≥ 30 kg/m2 and current tobacco use as nonobese (< 30 kg/m2) or negative tobacco use and found minimal differences from the primary analysis (Supplementary Table 1).
DISCUSSION
PRINCIPAL FINDINGS
In our diverse study cohort of patients with preeclampsia with severe features, we examined the association between individual severe features and adverse perinatal outcomes. We hypothesized that severe hypertension and neurological symptoms would show the strongest association with adverse outcomes. Contrary to this, neurological symptoms, despite being among the most common qualifying criteria, were not significantly associated with the primary composite outcome after adjusting for maternal age, race, and one or more pre-existing medical conditions. Instead, severe hypertension, thrombocytopenia, progressive renal insufficiency, and impaired liver function were each significantly associated with the primary composite outcome. This suggests that laboratory and hemodynamic abnormalities may better reflect the degree of end-organ involvement driving adverse perinatal outcomes.
Restricting our sub-analysis to patients with severe hypertension allowed us to determine whether adverse outcomes were driven by hypertension alone or by the presence of additional severe features. Because nearly all patients in our cohort had severe hypertension, the dose–response relationship observed in the full cohort may have primarily reflected the effect of hypertension rather than the independent contribution of other criteria. However, among patients with severe hypertension, the odds of preterm birth increased as additional severe features accumulated. This finding suggests that adverse perinatal outcomes are not driven by severe hypertension alone, but rather by the cumulative burden of end-organ dysfunction.
RESULTS
Patients with preeclampsia with severe features are at an increased risk for adverse perinatal outcomes, including preterm birth, low birth weight, cesarean delivery, placental abruption, and stillbirth.3–8 Our findings add to this literature by identifying which severe features drive these risks and suggest that end-organ involvement reflected in laboratory values rather than symptomatic criteria (headache and neurological findings) may be more central to placentally mediated adverse outcomes. This is consistent with a large retrospective cohort study from Ethiopia involving over 1,000 patients with preeclampsia with severe features that found that despite being the most common criteria, headache and visual changes were not significant predictors of adverse maternal outcomes.14 They found that elevated creatinine levels were independently associated with a composite of severe adverse maternal outcomes including ICU admission, pulmonary edema, dialysis, and death. Unlike our analysis, they did not examine association between criteria and adverse fetal/neonatal outcomes.14
Our findings are consistent with a 2011 study comparing perinatal outcomes of severe preeclampsia, eclampsia, and HELLP syndrome–characterized by thrombocytopenia and hepatic dysfunction–that reported stronger associations between HELLP syndrome and adverse outcomes including lower birth weights, perinatal mortality, and earlier gestational age at delivery.15 However, that study used outdated criteria to define PEC-SF compared to our analysis using current ACOG definition for PEC-SF.
CLINICAL IMPLICATIONS
These results have important clinical implications. Since severe-range blood pressures are nearly ubiquitous in PEC-SF, risk stratification depends more on the presence and number of additional affected organ systems than on hypertension itself. Furthermore, the unadjusted prevalence of the primary composite outcome was highest among patients with thrombocytopenia or pulmonary edema, indicating that although these features are less common, they may identify particularly high-risk patients. Finally, beyond placentally mediated perinatal outcomes, we found that severe hypertension and impaired liver function were associated with an increased risk of unplanned cesarean delivery.
The association between thrombocytopenia and small for gestational age neonates, while statistically significant, was modest in magnitude. This finding should be interpreted with caution, and it is not possible to draw conclusions regarding a distinct biological pathway linking thrombocytopenia to fetal growth restriction based on this result alone. In contrast, neurological symptoms, although among the most common qualifying criteria, did not reach significance for the primary composite in adjusted analysis. This suggests that after adjustment for confounding factors, objective markers of systemic end-organ involvement may better capture placentally mediated adverse perinatal outcomes than subjective symptom-based criteria. Severe hypertension remained significantly associated across multiple outcomes, consistent with its central role in preeclampsia pathophysiology and its well-established link to placental dysfunction.16
RESEARCH IMPLICATIONS
These findings should be used to inform future studies aimed at exploring phenotype-specific pathways of disease progression in PEC-SF and raises the question of whether laboratory defined end-organ dysfunction in addition to hypertension warrants heightened clinical attention for perinatal risk stratification beyond what is currently reflected in standard management.3 Additionally, because severe hypertension was nearly universal in our cohort, we were unable to characterize perinatal risk among patients who meet preeclampsia with severe feature criteria, but do not have severe hypertension. A larger powered study could help determine whether the patterns we observed here would extend to patients with PEC-SF but not severe hypertension. Future research incorporating placental pathology may also help clarify the pathophysiologic mechanisms linking individual severe features to adverse outcomes. Prospective studies are needed to evaluate how the evolution of individual criteria over time relates to perinatal outcomes.
STRENGTHS AND LIMITATIONS
This study’s strengths include its large sample size examining placental mediated outcomes of PEC-SF, as these are clinically significant outcomes to providers and patients. Our primary and secondary outcomes reflect meaningful outcomes, increasing translational value for clinicians. Use of EHR data reflects a large and real-world population thus capturing a wide range of clinical presentations, also increasing clinical value. Our outcomes were ascertained through detailed manual chart review rather than administrative or diagnostic-code data, improving the accuracy of both exposure and outcome classification.
This study has several limitations. First, its retrospective design introduces the possibility of bias due to incomplete or inaccurate documentation and limits our ability to adjust for unmeasured confounders. Important clinical factors such as antihypertensive therapy, magnesium sulfate use, and provider decision-making were not included in the analysis and may have influenced outcomes.
Although the cohort is large, some outcomes such as intrauterine fetal demise and placental abruption, were rare, limiting statistical power to evaluate their association with specific severe features. While these outcomes appeared more common among patients with pulmonary edema, thrombocytopenia, or progressive renal insufficiency, larger studies are needed to better characterize these relationships. We were also unable to assess the timing of onset of severe features, which may be important in understanding disease progression and risk.
Finally, the analysis included multiple exposures and outcomes, raising the possibility of type I error. We did not adjust for multiple comparisons to avoid obscuring potentially meaningful clinical associations; instead, effect sizes and confidence intervals are presented to aid interpretation. Additionally, because severe hypertension was nearly universal in this cohort, its ability to independently discriminate risk was limited.
CONCLUSIONS
In conclusion, our findings demonstrate that not all severe features are comparable in their association with adverse perinatal outcomes, with pulmonary edema and thrombocytopenia, progressive renal insufficiency, and impaired liver function increasing risk of adverse perinatal outcomes. This highlights the importance of considering both the type and the burden of criteria when risk stratifying patients and guiding clinical management.
Supplementary Material
AJOG AT A GLANCE.
- Why was this study conducted?
- Preeclampsia with severe features is defined by several different diagnostic criteria, but whether individual criteria–or the number of criteria present–differ in their association with adverse perinatal outcomes is not well understood.
- What are the key findings?
- Severe hypertension and laboratory markers of end-organ dysfunction (thrombocytopenia, progressive renal insufficiency, and impaired liver function), but not neurological symptoms, were associated with the composite adverse perinatal outcome.
- The risk of adverse outcomes increased as the number of severe criteria accumulated.
- What does this study add to what is already known?
- Not all severe features carry equal risk; both the type and the cumulative burden of criteria are relevant to risk stratification in preeclampsia with severe features.
Acknowledgements:
This work was made possible by a grant from Duke Clinical & Translational Science Institute (CTSI), Duke-UNC-Chapel Hill CTSA Consortium Collaborative Translational Research Pilot Grants Program (Meng, Quist-Nelson, Federspiel), UNC Division of Maternal Fetal Medicine Research Funds
Disclosures:
JF receive grant support from Eugene Kennedy Shriver National Institute of Child Health and Human Development, Canadian Institute of Health Research, and consulting fees from Invent-VTE, Noma.AI, Matria Health Technology, Sanofi, Grifols. MLM receives consulting fees from Becton Dickinson.
REFERENCES
- 1.Ananth CV, Keyes KM, Wapner RJ. Pre-eclampsia rates in the United States, 1980–2010: age-period-cohort analysis. BMJ. 2013;347:f6564. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Centers for Disease Control and Prevention. Archive App. Accessed October 4, 2025. https://archive.cdc.gov/#/details?archive_url=https://archive.cdc.gov/www_cdc_gov/maternal-mortality/php/data-research/mmrc-2017-2019.html
- 3.American College of Obstetricians and Gynecologists. Gestational hypertension and preeclampsia: ACOG Practice Bulletin No. 222. Obstet Gynecol. 2020;135(6):e237–e260. [DOI] [PubMed] [Google Scholar]
- 4.Kongwattanakul K, Saksiriwuttho P, Chaiyarach S, Thepsuthammarat K. Incidence, characteristics, maternal complications, and perinatal outcomes associated with preeclampsia with severe features and HELLP syndrome. Int J Womens Health. 2018;10:371–377. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Li C, Gou W, Liu J, Liu D, Li X, Chen Q. Length of stay in the ward following a preeclamptic pregnancy. J Hum Hypertens. 2021. [DOI] [PubMed] [Google Scholar]
- 6.Ødegård RA. Preeclampsia and fetal growth. Obstet Gynecol. 2000;96(6):950–955. [PubMed] [Google Scholar]
- 7.Pasokpuckdee K, Boriboonhirunsarn D. Incidence of preeclampsia and cesarean section rate according to the Robson classification. Cureus. 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Oud L. Epidemiology of pregnancy-associated ICU utilization in Texas: 2001–2010. J Clin Med Res. 2017;9(2):143–153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.MacKay AP, Berg CJ, Atrash HK. Pregnancy-related mortality from preeclampsia and eclampsia. Obstet Gynecol. 2001;97(4):533–538. [DOI] [PubMed] [Google Scholar]
- 10.Reddy M, Fenn S, Rolnik DL, Mol BW, da Silva Costa F, Wallace EM, Palmer KR. The impact of the definition of preeclampsia on disease diagnosis and outcomes: a retrospective cohort study. Am J Obstet Gynecol. 2021;224(2):217.e1–217.e11. [DOI] [PubMed] [Google Scholar]
- 11.Leonard SA, Kennedy CJ, Carmichael SL, Lyell DJ, Main EK. An expanded obstetric comorbidity scoring system for predicting severe maternal morbidity. Obstet Gynecol. 2020;136(3):440–449. doi: 10.1097/AOG.0000000000004022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Borella F, Marozio Luca, Bertschy G, et al. Placenta-mediated pregnancy complications in women with a history of late fetal loss and placental infarction without thrombophilia: risk of recurrence and efficacy of pharmacological prophylactic interventions. A 10-year retrospective study. Journal of maternal-fetal and neonatal medicine/Journal of maternal-fetal & neonatal medicine 2023;36(1). doi: 10.1080/14767058.2023.2183748 [DOI] [PubMed] [Google Scholar]
- 13.Skeith L, Blondon M, Ní Áinle F. Understanding and Preventing Placenta-Mediated Pregnancy Complications. Hamostaseologie 2020;40(3):356–363. doi: 10.1055/a-1184-8388 [DOI] [PubMed] [Google Scholar]
- 14.Abraha HE, Teka H. Association of severity features and laboratory biomarkers with adverse maternal outcomes in pre-eclamptic women: a retrospective cohort study in an Ethiopian tertiary hospital. BMJ Open. 2025;15(4):e089935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Yıldırım G, Güngördük K, Aslan H, Gül A, Bayraktar M, Ceylan Y. Comparison of perinatal and maternal outcomes of severe preeclampsia, eclampsia, and HELLP syndrome. J Turk Ger Gynecol Assoc. 2011;12(2):90–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Mancia G, Hall JE. Introduction to a compendium on the pathophysiology and treatment of hypertension. Circ Res. 2019;124(7):967–968. doi: 10.1161/CIRCRESAHA.119.314953 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.