Hypertensive disorders of pregnancy are a major cause of pregnancy-related morbidity and mortality and are the second leading cause of maternal mortality, accounting for 17% of maternal deaths in California.1 Though stroke in pregnancy is an uncommon cause of maternal death, nearly one half of the strokes that occur in pregnancy are associated with preeclampsia or eclampsia.2 Traditional teaching has been that stroke risk increases with eclampsia (seizure) and with severe diastolic hypertension,3,4 with little emphasis on elevations in systolic blood pressure. Major society guidelines as recently as the early 2000s included systolic blood pressure in the diagnostic criteria for severe preeclampsia, but only recommended antihypertensive treatment for severe elevations of diastolic blood pressure.3 However, systolic blood pressure is associated with stroke risk in the general adult population, and more recent data have suggested that severe elevations in systolic blood pressure are also significant contributors to stroke in pregnancy, as well as other morbidities including heart failure and cardiomyopathy.5,6 These observations have led to changes in the recommendations of major practice organizations—a paradigm shift—with regard to management of acute hypertension in the obstetric patient, now addressing both severe systolic and diastolic hypertension.7–10 This present study was undertaken to describe the clinical characteristics of women who died from preeclampsia-related strokes in California from 2002 to 2007, with particular attention to blood pressure elevation, and to identify patient- and health care provider–related contributions and chances to alter the outcome among these deaths.
This is a retrospective review of all pregnancy-related preeclampsia stroke deaths identified by the California Pregnancy-Associated Mortality Review from 2002 to 2007 (Fig. 1). Pregnancy-related deaths were defined as “deaths directly related to the physiological changes in pregnancy or causes aggravated by the pregnancy or its management” (Box 1). In prior work, pregnancy-related deaths attributable to preeclampsia were identified by the California Pregnancy-Associated Mortality Review for the years from 2002 to 2007, using methodology described previously.1,9,11–13 In brief, when death certificates, autopsy, and toxicology reports suggested a pregnancy-related death, trained abstractors used standardized forms to abstract medical records and generate case summaries. Case summaries included detailed information regarding demographic characteristics, medical and obstetric history, and a chronologic timeline of clinical events leading up to each woman's death. The timeline included the patient's reported symptoms, physical exam findings, abnormal vital signs, laboratory values, medication administrations, and imaging and pathology reports whenever available.
Classification of Preeclampsia Pregnancy-Related Deaths
California birth cohort
Women with a live birth or reported fetal death within 1 calendar year of pregnancy, as identified in the California vital statistics birth cohort file, during the study time period.
Death during pregnancy or in the year after birth from causes directly related to the physiologic changes in pregnancy or causes aggravated by the pregnancy or its management.
Preeclampsia pregnancy-related death
Death in which a hypertensive disorder of pregnancy, caused or aggravated by pregnancy, is the initiating or critical pathologic event leading to death.
Case summaries were reviewed by the California Pregnancy-Associated Mortality Review committee, an interdisciplinary group including representatives from obstetrics, maternal–fetal medicine, anesthesiology, neonatology, midwifery, labor and delivery nursing, emergency medicine, and cardiology. For each case, committee members arrived at consensus determinations after review that were captured via data collection tools during the review meeting.12 These determinations included cause of death and identified risk factors, contributing factors, including health care system-, health care provider–, and patient-related factors, the degree to which the death might have been prevented (ie, chance to alter outcome), and opportunities for quality improvement. The California Pregnancy-Associated Mortality Review operationalized preventability as the chance to alter outcome, with response options: strong, good, some, none, and unable to determine. Cases with a good or strong chance to alter outcome help set priorities for prevention and quality improvement efforts in California and thus these categories were combined in subsequent analyses.9
For the present study, preeclampsia deaths were further examined and classified into nine different clinical pathways: stroke (hemorrhagic or ischemic); hepatic failure; cardiac failure; multi-organ failure, pulmonary edema; eclampsia; respiratory failure; or hemorrhage or disseminated intravascular coagulation. Two physician experts from the California Pregnancy-Associated Mortality Review committee confirmed these clinical pathway categorizations.
For the present analysis, case summaries of the pregnancy-related preeclampsia stroke deaths from 2002 to 2007 prepared as part of the California Pregnancy-Associated Mortality Review process were independently reviewed by a senior fellow in maternal–fetal medicine (A.E.J.). The review began by identifying the timing of stroke for each case, as determined by the onset of neurologic symptoms or physical examination findings consistent with an acute intracranial process. Clinical factors occurring before the time of stroke, including signs and symptoms, laboratory abnormalities, documented blood pressure values, and treatments (antihypertensives and magnesium sulfate), were abstracted for analysis, when available. In cases of uncertainty regarding the timing of stroke or other variables, a senior perinatologist was consulted (M.L.D.), and the data were coded by consensus.
Laboratory results before the time of stroke were noted, as available, with attention to abnormalities in platelet count, liver function tests, serum creatinine, serum ammonia, and serum lactate dehydrogenase (LDH). The presence or absence of hemolysis, elevated liver enzymes, and low platelet count (HELLP) syndrome was defined by platelet count below 100,000/microliters plus liver transaminases twice the normal limit before the occurrence of stroke. Evidence of hemolysis was not considered requisite to this diagnosis for the purpose of this analysis, as laboratory assessment for hemolysis (serum LDH or trend in blood count) was not available in most cases.
Symptoms abstracted based on documentation in the medical record included headache, visual changes, abdominal pain, and nausea–vomiting. For the purposes of this review, women were considered to have eclampsia when a tonic-clonic seizure occurred with a documented period of relatively normalized behavior before the onset of stroke-like symptoms. In cases of acute and sustained decompensation coincident with seizure-like activity, eclampsia was not considered to precede the stroke, as seizure-like activity in these cases could be secondary to stroke itself. Similarly, any seizure-like activity after the onset of stroke was not considered to represent eclampsia in this analysis.
Administration of antihypertensive agents before the time of stroke was recorded, along with the initial medication and dose administered, and whether any additional doses were administered before the onset of stroke. “Standard” antihypertensive treatments were considered to be those currently recommended by the American College of Obstetricians and Gynecologists (ACOG) as first line treatments: labetalol 20 mg intravenous (IV), hydralazine 5–10 mg IV, or nifedipine 10 mg orally; at the time of these mortalities, labetalol 200 mg orally was the preferred agent in patients lacking IV access, and therefore this treatment is also considered standard for the purpose of this analysis. Blood pressure values before stroke were recorded; the temporal relation between individual blood pressure values and timing of administration of antihypertensives was not available. Administration of magnesium sulfate before the time of stroke was also recorded.
Data analysis was conducted using IBM SPSS Statistics 20.0. For demographic and perinatal characteristics, comparisons were made between all-cause preeclampsia-related deaths and the cohort of California women with a live birth or reported fetal death within a calendar year of pregnancy (as identified in the California vital statistics Birth Cohort file) during the study time period. To assess the relative risk of any given demographic characteristic and its contribution to preeclampsia death, given the occurrence in the birth population, effect size was calculated by dividing the proportion of the total exposed (deaths in each group/total deaths) by the proportion of the unexposed (live births among each group/total live births). Effect size was not calculated for groups with fewer than five deaths. Statistical testing was not performed on the subset of preeclampsia-related stroke deaths given the small sample size; instead, these data are presented descriptively. All cases with missing data were excluded from analysis; thus, the total sample size may vary but the percentages add up to 100.
California Pregnancy-Associated Mortality Review protocols were approved by the State of California Committee for the Protection of Human Subjects; this project was considered exempt by the Stanford University Institutional Review Board, given that the analysis is of de-identified data on deceased persons.
From 2002 to 2007, there were 333 pregnancy-related maternal deaths in California. Of these, 54 were associated with preeclampsia or eclampsia and in 61% (33/54), stroke was identified as the cause of death. Imaging or autopsy or both results were available in 31. Hemorrhagic stroke occurred in 29 cases and ischemic stroke occurred in two cases. Among the remaining 21 preeclampsia pregnancy-related mortalities, 15 were attributed to organ failure, three to pulmonary edema, two to eclampsia, and one to hemorrhage or disseminated intravascular coagulation (Fig. 1).
Table 1 shows the demographic characteristics among preeclampsia stroke deaths and preeclampsia all-cause deaths compared with California Birth Cohort data. Within the preeclampsia deaths, women aged 35 or older were more than twice as likely than those aged 20–34 to die (effect size 2.2 [95% CI 1.5–3.1] vs 0.8 [95% CI 0.6–1.0]). Women who died from preeclampsia or eclampsia were more likely than those in the California Birth Cohort to be aged 35 years or older (37%/17%; effect size 2.2 [95% CI 1.5–3.1]). The women who died from preeclampsia were more likely than women among the California Birth Cohort to have Medi-Cal as their payer source (59%/47%; effect size 1.3 [95% CI 1.0–1.6]).
Table 2 shows perinatal characteristics, and comparisons are reported for all-cause preeclampsia-related deaths and the California Birth Cohort. Women who died of preeclampsia were more likely than women from the California Birth Cohort to be overweight (but not obese) (46%/25%; effect size 1.9; [95% CI 1.4–2.5]). Cesarean delivery was more common in preeclampsia-related deaths compared with the California Birth Cohort (77%/30%; effect size 2.6; [95% CI 2.2–3.0]). Women who died from preeclampsia were almost 12 times as likely as the California Birth Cohort to have a very preterm delivery (before 34 weeks of gestation) (39%/3%; effect size 11.9; [95% CI 8.5–16.6]).
The majority of women who died from stroke were multiparous (19/33 cases, 58%), delivered at before 37 weeks of gestation (20/33 cases, 61%), and delivered by cesarean (24/33 cases, 73%). Delivery before 34 weeks of gestation occurred in 36% (12/33) cases. Intrauterine fetal demise was present in 9% (3/33) of cases. Preexisting chronic hypertension was documented in five cases.
Stroke occurred antenatally in 33% (11/33), intrapartum in 18% (6/33) and postpartum in 46% (15/33); there was insufficient information to characterize the timing of stroke in one case. Among postpartum strokes, 12 occurred within 72 hours of delivery and three occurred more than 72 hours after delivery. Antenatal stroke occurred out-of-hospital in five out of six cases. Postpartum stroke occurred out-of-hospital in three of 12 cases. Eclampsia occurred in 36% of cases (10/28); clinical data were insufficient to characterize five cases. For patients in whom laboratory data was available, HELLP was present in 38% (9/24) of cases.
Information regarding blood pressure elevation before the event was available in 26 cases. Systolic blood pressure in the time preceding the stroke was more than 160 mm Hg in 96% (25/26) of cases. Diastolic blood pressure was more than 105 mm Hg in 73% (19/26) of cases and 110 or higher in 65% (17/26) of cases. Maximum systolic blood pressure varied widely (median 196, range 134–238 mm Hg). Maximum diastolic blood pressure also varied widely (median 113, range 79–148 mm Hg). In the single case in which the maximal blood pressure recorded before stroke was below 160/105 mm Hg, normal blood pressure with headache had been noted in clinic that day; the woman was sent to the hospital via private transportation and stroke occurred en route. Initial blood pressure on admission did demonstrate severe systolic and diastolic hypertension 248/130 mm Hg (data not shown).
Information on the type and timing of initial treatment was available for 31 cases (Table 3). Less than half (15/31 cases, 48%) received any antihypertensive treatment before the occurrence of stroke. Of the 15 women receiving antihypertensive treatment, 53% (9/15) received treatment other than the currently recommended first-line therapies by ACOG. For those receiving a nonstandard therapy, the majority received a dose lower than the current recommendation (7/9 cases, 78%). Among women treated with an antihypertensive medication, 87% (17/31) were administered at least one additional dose of antihypertensive before the occurrence of stroke. Magnesium sulfate prophylaxis was administered in 61% of cases.
The presence of symptoms (headache, visual changes, abdominal pain, and nausea–vomiting) was assessed, and data were available for 30 cases (Table 4). At least one symptom preceded stroke in all 30 cases. In 63% of cases (19/30), symptoms were present for more than 24 hours before the occurrence of stroke. The most common symptom was headache (26/30 cases, 87%). Many women had more than one symptom.
Laboratory data (platelet count, liver function tests, serum creatinine, international normalized ratio, fibrinogen, ammonia, and LDH) were available for 24 cases before the occurrence of stroke (Table 4). One or more abnormal values were noted in 79% of cases 79% (19/24). Liver transaminases twice the normal limit occurred most commonly, in 71% (17/24) of cases, followed by platelet counts below 100,000/microliters in 42% (10/24) of cases.
Chance to alter outcome and contributing factors were considered for the 33 preeclampsia-related stroke deaths. In one case, the available data were insufficient to characterize. A good-to-strong chance to alter outcome was determined in 21 (66%) stroke cases, similar to the chance to alter outcome observed among all preeclampsia related deaths (62%). Health care provider– and patient-related contributing factors are depicted in Figures 2 and 3. For preeclampsia-related stroke deaths, a delayed response to clinical warning signs was identified in 30 of 33 cases (91%). Ineffective treatment was identified in 25 of 33 cases (76%). Similar proportions of these factors were observed when all preeclampsia deaths were considered (93% and 70%, respectively). For all preeclampsia deaths, delay in seeking care (41%), underlying medical condition (35%), and lack of knowledge (35%) were the most frequently observed patient-related factors.
In this in-depth, statewide review of all cases of preeclampsia-associated stroke mortality cases from 2002 to 2007, stroke was the underlying cause of death in the majority of deaths associated with preeclampsia or eclampsia. Severe systolic hypertension (more than 160 mm Hg) preceded stroke in all but one case—and that woman arrived at the hospital with a stroke and severe hypertension. Additionally, all women demonstrated at least one symptom (headache, visual changes, abdominal pain, and nausea or vomiting) before the occurrence of stroke. Despite the near-universal occurrence of concerning signs and symptoms, few women in our series received antihypertensive therapy before the occurrence of stroke.
Although stroke is an established cause of morbidity among women with preeclampsia,6,14 the factors that contribute to mortality within this subset of acutely ill women are uncertain. In this series, severe systolic hypertension was present in 96% of the cases. Though symptoms were present in all cases before stroke, many symptoms such as nausea and vomiting are common in laboring women. The common occurrence of symptoms has been demonstrated in other studies of postpartum preeclampsia and eclampsia.15–17 Other clinical correlates, such as HELLP (present in 38% of cases), eclampsia (observed in 36% of cases), and individual laboratory abnormalities, were not consistently observed in this review, and this is in keeping with prior publications.6 In 2005, a retrospective review of preeclampsia-related stroke by Martin et al6 identified severe systolic hypertension in a majority of cases, with severe diastolic hypertension occurring less frequently. Subsequent publications have also demonstrated increasing rates of severe hypertension and severe maternal morbidity.18 In response to growing evidence that systolic hypertension contributes to morbidity and mortality in preeclampsia and eclampsia, ACOG and other organizations have updated guidelines and released treatment algorithms to address severe systolic blood hypertension.7,8,10,19 The findings of our current study reaffirm this paradigm shift to emphasize the significance of systolic blood pressure elevation in preeclampsia-associated morbidity.
Limitations of this study include the retrospective nature and the heterogeneity of medical evaluation in each case. A small number of patient records was missing, and two women lacked autopsy or imaging studies to confirm the diagnosis. However, trained abstractors created detailed summaries from original medical records and a large, experienced multidisciplinary panel determined cause of death by consensus. It is possible that women with severe elevations in diastolic blood pressure were successfully treated, and thus are not represented in this sample of mortalities. Notably, nearly all cases in this series were complicated by hemorrhagic stroke. Higher rates of ischemic stroke have been reported with pregnancy and in preeclampsia,3,6,20 but hemorrhagic stroke results in higher mortality,21 likely increasing its representation in this cohort.21 Similar findings were reported by the United Kingdom's Confidential Enquiry into Maternal Deaths (2016) who also identified hemorrhagic stroke as the leading cause of mortality among women dying from hypertensive disorders of pregnancy.22
It is important to note the mortalities in our review occurred before the release of current ACOG guidelines for the management of preeclampsia (and treatment of severe systolic hypertension). We observed low rates of appropriate antihypertension treatment in this cohort, which has been described in other case series from the same era.6 Recent work examining a cohort of severely hypertensive women during a preeclampsia quality collaborative showed high rates of appropriate antihypertensive treatment (greater than 80%) with very low rates of stroke,18 and the United Kingdom has demonstrated reductions in preeclampsia- and eclampsia-related morbidity after implementation of contemporary standardized guidelines,23 suggesting that newer guidelines have been effective in increasing early treatment and reducing morbidity. It is possible that had these cases occurred after the implementation of the current guidelines for treatment of severe hypertension, some of these mortalities may have been avoided.
From 2002 to 2007, the preeclampsia pregnancy-related death rate in California was 1.6 deaths per 100,000 live births. In comparison, the rate in the United Kingdom from 2009 to 2013 was 0.25 per 100,000, and after the implementation of national guidelines, it fell to 0.08 per 100,000 for 2012–2014.22 Comprehensive, standardized guidelines targeting antenatal, intrapartum, and postpartum diagnosis and management of preeclampsia has been put forth in California.8 One-quarter of women in this series experienced stroke outside of the hospital. Algorithms directed at treatment of acute hypertension only benefit patients within reach of medical practitioners. “Lack of knowledge” was frequently identified as a patient-related factor in this series, and changes to discharge requirements, follow-up parameters, and an emphasis on patient education may affect mortalities. Room for improvement of care persists, and it remains to be seen whether the paradigm shift to emphasize systolic hypertension paired with implementation of guideline-based toolkits will result in decreases in mortality due to preeclampsia-related stroke.
1. Main EK, McCain CL, Morton CH, Holtby S, Lawton ES. Pregnancy-related mortality in California: causes, characteristics, and improvement opportunities. Obstet Gynecol 2015;125:938–47.
2. Sharshar T, Lamy C, Mas JL. Incidence and causes of strokes associated with pregnancy and puerperium: a study in public hospitals of Ile de France. Stroke in Pregnancy Study Group. Stroke 1995;26:930–6.
3. Diagnosis and management of preeclampsia and eclampsia. ACOG Practice Bulletin No. 33. American College of Obstetricians and Gynecologists. Int J Gynaecol Obstet 2002;77:67–75.
4. Report of the National High Blood Pressure Education Program Working Group on high blood pressure in pregnancy. Am J Obstet Gynecol 2000;183:S1–22.
5. Lindenstrom E, Boysen G, Nyboe J. Influence of systolic and diastolic blood pressure on stroke risk: a prospective observational study. Am J Epidemiol 1995;142:1279–90.
6. Martin JN Jr, Thigpen BD, Moore RC, Rose CH, Cushman J, May W. Stroke and severe preeclampsia and eclampsia: a paradigm shift focusing on systolic blood pressure. Obstet Gynecol 2005;105:246–54.
7. Emergent therapy for acute-onset, severe hypertension during pregnancy and the postpartum period. ACOG Committee Opinion No. 767. American College of Obstetricians and Gynecologists. Obstet Gynecol 2019;133:e174–80.
8. Druzin ML, Shields LE, Peterson NL, Cape V. Preeclampsia toolkit: improving health care response to preeclampsia (California Maternal Quality Care Collaborative toolkit to transform maternity care). Stanford (CA): California Maternal Quality Care Collaborative; 2013.
9. Mitchell C, Lawton E, Morton C, McCain C, Holtby S, Main E. California pregnancy-associated mortality review: mixed methods approach for improved case identification, cause of death analyses and translation of findings. Matern Child Health J 2014;18:518–26.
10. Gestational hypertension and preeclampsia. ACOG Practice Bulletin No. 202. American College of Obstetricians and Gynecologists. Obstet Gyncol 2019;133:e1–25.
11. Hameed AB, Lawton ES, McCain CL, Morton CH, Mitchell C, Main EK, et al. Pregnancy-related cardiovascular deaths in California: beyond peripartum cardiomyopathy. Am J Obstet Gynecol 2015;213:379 e1–10.
12. The California pregnancy-associated mortality review: report from 2002 to 2007 maternal death reviews. Sacramento (CA): California Department of Public Health, Maternal Child and Adolescent Health Division; 2017.
13. The California pregnancy-associated mortality review. Report from 2002 and 2003 maternal death reviews. Sacramento (CA): California Department of Public Health, Maternal Child and Adolescent Health Division; 2011.
14. Cantwell R, Clutton-Brock T, Cooper G, Dawson A, Drife J, Garrod D, et al. Saving mothers' lives: reviewing maternal deaths to make motherhood safer: 2006–2008. The eighth report of the Confidential Enquiries into Maternal Deaths in the United Kingdom. BJOG 2011;118(suppl 1):1–203.
15. Al-Safi Z, Imudia AN, Filetti LC, Hobson DT, Bahado-Singh RO, Awonuga AO. Delayed postpartum preeclampsia and eclampsia: demographics, clinical course, and complications. Obstet Gynecol 2011;118:1102–7.
16. Chames MC, Livingston JC, Ivester TS, Barton JR, Sibai BM. Late postpartum eclampsia: a preventable disease? Am J Obstet Gynecol 2002;186:1174–7.
17. Matthys LA, Coppage KH, Lambers DS, Barton JR, Sibai BM. Delayed postpartum preeclampsia: an experience of 151 cases. Am J Obstet Gynecol 2004;190:1464–6.
18. Kilpatrick SJ, Abreo A, Greene N, Melsop K, Peterson N, Shields LE, et al. Severe maternal morbidity in a large cohort of women with acute severe intrapartum hypertension. Am J Obstet Gynecol 2016;215:91.e1–7.
19. American College of Obstetricians and Gynecologists. Hypertension in pregnancy. Washington, DC: American College of Obstetricians and Gynecologists; 2013.
20. Pathan M, Kittner SJ. Pregnancy and stroke. Curr Neurol Neurosci Rep 2003;3:27–31.
21. Andersen KK, Olsen TS, Dehlendorff C, Kammersgaard LP. Hemorrhagic and ischemic strokes compared: stroke severity, mortality, and risk factors. Stroke 2009;40:2068–72.
22. Knight M, Nair M, Tuffnell D, Kenyon S, Shakespeare J, Brocklehurst P, et al. Saving lives, improving mothers' care—surveillance of maternal deaths in the UK 2012-2014 and lessons learned to inform maternity care from the UK and Ireland confidential enquiries into maternal deaths and morbidity 2009–2014. Oxford (UK): National Perinatal Epidemiology Unit, University of Oxford; 2016.
23. Tuffnell DJ, Jankowicz D, Lindow SW, Lyons G, Mason GC, Russel IF, et al. Outcomes of severe pre-eclampsia/eclampsia in Yorkshire 1999/2003. BJOG 2005;112:875–80.