Cooray, Shamil D. MBBS; Edmonds, Sally M. MBChB; Tong, Stephen MBBS, PhD; Samarasekera, Sumudu P. MBBS; Whitehead, Clare L. MBChB
Eclampsia accounts for 12% of maternal deaths per year worldwide, the majority in developing countries.1 The global commitment to fulfilling the United Nations' fifth Millennium Development Goal of reducing maternal mortality by 75% between 1990 and 2015 requires a concerted effort to better understand eclampsia.2 In countries such as Tanzania, there are as many as 67 cases of eclampsia reported per 10,000 births.3
Currently, there are no reliable tests to predict which patients will develop eclampsia.4 Clinical signs and symptoms preceding the development of eclampsia are nonspecific but may be a surrogate marker of end-organ dysfunction and disease severity.5 Therefore, symptoms may be useful to provide some warning that a patient is at imminent risk for eclampsia.
To date, the body of literature investigating clinical symptoms associated with eclampsia includes retrospective studies of data acquired indirectly through chart review.6–10 As such, these retrospective studies might have been prone to reporting bias, skewing the reported incidences of prodromal symptoms.
A contemporaneous structured interview of patients with eclampsia would overcome this bias. However, it would be difficult to undertake a prospective study in the developed world given the incidence of eclampsia is low. Because eclampsia is 10-times to 30-times more common in developing countries than in high-income countries, it may be more feasible to prospectively study signs and symptoms preceding eclampsia in developing countries.1 Therefore, we prospectively recorded symptoms preceding eclampsia via a targeted observational study in rural Tanzania to determine whether there were symptoms predictive of eclampsia.
MATERIALS AND METHODS
We performed a prospective study of all consecutive patients admitted with an eclamptic seizure between May 1, 2007 and April 30, 2008, to Teule Hospital, Muheza, Tanga region, Tanzania. Institutional approval at Teule Hospital was obtained. Eclampsia was defined as a witnessed tonic-clonic seizure in the third trimester of pregnancy or the puerperium in the presence of hypertension or proteinuria or both. Seizures attributable to the alternative diagnoses were excluded: central nervous system infections including cerebral malaria (after investigation by blood film and lumbar puncture); a history of seizures or epilepsy; medications, or illicit drug use; or intracranial hemorrhage.
Teule Hospital is a 330-bed rural district designated hospital serving a population of approximately 280,000 in the Muheza district and is the referral center for women from neighboring regions. Maternity care is provided by midwives, trained maternal and child health care aides, and a specialist trained in the United Kingdom (S.E.); at local dispensaries, health centers, and Teule Hospital. Routine antenatal care for uncomplicated pregnancies consists of four scheduled visits. The majority of women deliver at home or with a traditional birth attendant at local primary care facilities. The hospital itself delivers 3,000 women annually.
A customized case record was collected for each case of eclampsia by a single investigator (S.E.). Demographic characteristics, clinical presentation, and obstetric and medical history were collected at the time of presentation by interview, physical examination, and review of antenatal records. All women were specifically asked about the presence or absence of severe headache, visual disturbance, epigastric pain, nausea, or vomiting before seizure. All information was obtained directly from the women and when possible was further substantiated by questioning family members and the birth attendant as appropriate. The gestational age was based on symphyseal–fundal height, last menstrual period (if known), and neonatal condition. Hypertension was defined as systolic blood pressure (BP) higher than 140 mm Hg, diastolic BP higher than 90 mm Hg, and severe hypertension was defined as a systolic BP higher than 160 mm Hg or diastolic BP higher than 110 mm Hg. Hypertension was treated with oral methyldopa or intravenous hydralazine. Diazepam or magnesium sulfate was used for seizure therapy, depending on local availability and patient response.
Statistical analyses were evaluated using Graph-pad Prism Software. Descriptive statistics and Fisher exact test were applied as appropriate.
There were 3,267 deliveries at Teule Hospital during the study period. Forty-six women presented with an eclamptic seizure (1.4%). Forty-two (91%) women had been receiving ongoing prenatal care. Mean maternal age was 22 years (range 16–42). Thirty (65%) women were primigravid. Of the 16 multigravidas, eight had new partners and paternity was unknown in two cases (Table 1).
Twenty-four women (52%) had antepartum seizures, with a mean gestational age at time of seizure of 35 weeks (range 28–42 weeks). Seventeen (71%) of those had their seizure outside of hospital. Only four women were already hospitalized with diagnosed preeclampsia when the seizure occurred. Intrapartum seizures occurred in seven (15%) women, with a mean gestational age of 37 weeks (range 36–40). Seizures occurred postpartum in 15 (33%) women. Of these, nine occurred within 48 hours of delivery, and the remaining six (13% overall) had a late onset seizure, occurring more than 48 hours after delivery. The interval between delivery and late postpartum seizure ranged from 3 to 13 days.
At the first antenatal appointment, all women were normotensive (mean systolic BP 111 mm Hg and diastolic BP 71 mm Hg). During hospitalization for eclampsia, 30 women (65%) had a systolic/diastolic BP higher than 160/110 mm Hg (severe hypertension), with the remaining 16 women (35%) having a systolic/diastolic blood pressure higher than 140/90 mm Hg (moderate hypertension).
Twenty-nine women were treated with magnesium sulfate (63%). The remainder was treated with diazepam, primarily because of limited access to supplies of magnesium sulfate.
Headache and visual disturbance were the most common prodromal symptoms, reported in 37 (80%) and 21 (45%) women, respectively. None of the women had a history of significant headaches. Epigastric pain occurred in nine (20%) women before seizure and only one woman reported severe nausea or vomiting. The number of women reporting a prodromal headache or visual disturbance was not altered by the degree of hypertension: 80% of women with severe hypertension and 81% of women with moderate hypertension reported headache; likewise, 43% of women with severe hypertension described visual disturbances, as did 50% of women with moderate hypertension (Table 2). The presence of headache and visual disturbance was also not altered by the timing of seizure: both headache and visual disturbance occurred at a similar frequency whether eclampsia occurred antepartum intrapartum, or postpartum, 80% and 87%, respectively, for headache and 45% and 47% for visual disturbance (Table 3). Eight (17%) women were completely asymptomatic before seizure.
Of the cohort, there were three maternal deaths (7%). They all occurred after eclamptic seizures initially occurring at home, followed by recurrent seizures on admission to hospital despite treatment with anticonvulsants. There were 10 (22%) perinatal deaths,: eight stillbirths and two neonatal deaths (within 48 hours of birth). Eight perinatal deaths were preterm, ranging from 28 to 36 weeks of gestation. Postmortem examination to determine cause of death was not performed for either maternal or perinatal deaths because of resource limitations.
Clinical signs and symptoms in preeclampsia may be better than laboratory investigations to predict adverse maternal outcomes including eclampsia.11 Because delivery is the only cure for preeclampsia, clinicians must balance the benefit to the fetus of increasing gestation against the maternal risk of adverse events including eclampsia. It would be advantageous to be able to predict eclampsia to optimize the timing of delivery. Furthermore, given that an intervention (magnesium sulfate) exists that halves the risk of seizure,12 prospective characterization of warning symptoms preceding imminent eclampsia is worthwhile.
Although there have been a number of studies examining the prodrome of eclampsia, they have been retrospective and subject to reporting bias.6–10 Establishing symptoms that precede eclampsia could be more accurately ascertained by direct questioning at the time of seizure. However, given that eclampsia is uncommon in developed countries, a prospective study of eclampsia would only be feasible in low-income countries, where eclampsia is more prevalent.
We undertook such a study, specifically asking all women who had experienced an eclamptic seizure whether they experienced symptoms listed on a proforma sheet. Recording of symptoms was inclusive of all women presenting across an entire year and, importantly, was systematic and contemporaneous (data were collected around the time of the seizure). Consequently, the incidence of different symptoms we report here may be more accurate than those previously reported in retrospective studies.
Previous studies have shown rates of premonitory symptoms vary from as low as 41% to as high as 91%.6–8 However, these have been performed by case note review. We have prospectively confirmed eclamptic seizures are preceded by a prodrome in 83% of cases.
Furthermore, seizures were most commonly preceded by neurological symptoms. Headache and visual disturbance were the most common symptoms, occurring in 80% and 45% of women before eclampsia, respectively. This is not surprising given that headache and visual disturbance probably reflect the development of cerebral edema and vasospasm of cerebral and retinal vessels.13 Neurological symptoms are often the most common premonitory symptom of eclampsia, but rates in previous retrospective studies have varied from 50% to 90% of women with eclampsia.6–8 Neurological symptoms may be useful at predicting eclampsia; a recent review of women with established preeclampsia or hemolysis, elevated liver enzymes, low platelets syndrome, or both, gave an odds ratio of 3.6 (95% confidence interval 1.2–10.4) for headache to predict eclampsia.11 In a further study, the sensitivity for headache to predict eclampsia was as high as 0.98 (95% confidence interval 0.87–1), although the specificity was lower.5
Epigastric pain, nausea, and vomiting are believed to reflect hepatocellular involvement from periportal or parenchymal necrosis, liver capsule stretching, and hemorrhage.15 Symptoms may reflect the biochemical markers of liver dysfunction, although we were unable to correlate with transaminases in a significant proportion of our cohort because of resource limitations. In this study, 20% of women reported epigastric pain and 2% reported severe nausea or vomiting before the onset of eclampsia. Unsurprisingly, reporting of hepatocellular symptoms tends to be lower than neurological symptoms in eclampsia but is nevertheless important in predicting adverse outcomes, potentially reflecting the development of hemolysis, elevated liver enzymes, low platelets syndrome.6–11 The recently published fullPIERS model gave epigastric pain an odds ratio of 2.92 for predicting adverse outcomes in preeclampsia, although not specifically eclampsia.16
In contrast, hypertension and proteinuria have been shown to be poor predictors of eclampsia, despite forming the criteria for diagnosis and being included in all international classifications of severity.11,16 In our study, all patients had documented hypertension on arrival at hospital or during their admission; however, only 52% had proteinuria. This is similar to previous studies, which also have described eclampsia in the absence of either hypertension or proteinuria.3,6–8 This emphasizes that isolated testing for hypertension or proteinuria alone may fail to identify imminent eclampsia. In contrast, 83% of women with eclampsia had symptoms and hypertension or proteinuria. These women may have been identified before seizure by including screening for symptoms.
In the developed world, many patients will be admitted with preeclampsia and monitored expectantly. The trigger to deliver and initiate seizure prophylaxis varies between clinicians. In studies in the United States and Europe, despite admission with a diagnosis of severe preeclampsia, magnesium sulfate prophylaxis rates vary from as low as 10% to 60%.14,16 Given the high frequency of prodromal neurological symptoms in women with eclampsia, seizure prophylaxis should be encouraged in patients with prodromal symptoms. However, the absence of symptoms should not be falsely reassuring because seizures may still occur. In low-resource settings, access to magnesium sulfate firstly must be made readily available. Unfortunately, in our institution it is in limited supply and despite the clear benefit of its use in both prophylaxis and treatment, only half of all patients were treated with magnesium sulfate.
Interestingly, 17% of women with severe hypertension were completely asymptomatic before seizure. These patients may not have had the typical progression from mild gestational hypertensive disease to severe preeclampsia but rather a rapidly fulminating course with early neurological sequelae. If so, it suggests there will always be a proportion of women for whom it may not be possible to administer timely magnesium sulfate prophylaxis because of the lack of warning symptoms.
Thirteen percent of women with eclampsia presented significantly postpartum, more than 48 hours after delivery, and all had seizures that occurred at home. Notably, all women with late postpartum eclampsia were symptomatic before seizure, suggesting there may be an opportunity to prevent some postpartum seizures from occurring if women seek medical care. Thus, the new onset of symptoms postpartum merits keen medical attention and consideration of magnesium sulfate.
Despite the clear advantage of this study in collecting contemporaneous data via structured interview at the time of seizure, there are limitations. The study was performed in a low-resource setting, and patients were included only if they experienced eclampsia. We are therefore unable to comment on the prevalence of symptoms in the general obstetric or gestational hypertensive population. However, a much larger prospective study could focus on improving the specificity of this low-cost pragmatic approach to identify those patients who are potentially at risk for eclampsia. This may lead to the development and implementation of predictive tools for both the developing and developed world; however, such a study would be challenging to conduct.
In conclusion, neurological symptoms commonly precede eclampsia. Education in both the developed and developing world should emphasize, to patients and clinicians, the importance of reporting prodromal symptoms. In the developing world, this may prompt patients to seek medical attention and encourage health care providers to triage women to referral centers appropriately. Earlier presentation and timely management may reduce maternal deaths from eclampsia in low-resource settings. In the developed world, reporting symptoms may prompt the clinician to consider commencing magnesium sulfate and initiate delivery, improving rates of prophylaxis and reducing morbidity. Finally, postpartum eclampsia should not be ignored. All patients should be encouraged to seek medical attention with the new onset of premonitory symptoms, even after delivery.
1. Duley L. The global impact of pre-eclampsia and eclampsia. Semin Perinatol 2009;33:130–7.
3. Urassa DP, Carlstedt A, Nystrom L, Massawe SN, Lindmark G. Eclampsia in Dar es Salaam, Tanzania–incidence, outcome, and the role of antenatal care. Acta Obstet Gynecol Scand 2006;85:571–8.
4. Thangaratinam S, Langenveld J, Mol BW, Khan KS. Prediction and primary prevention of pre-eclampsia. Best Pract Res Clin Obstet Gynaecol 2011;25:419–33.
5. Thangaratinam S, Gallos ID, Meah N, Usman S, Ismail KM, Khan KS. How accurate are maternal symptoms in predicting impending complications in women with preeclampsia? A systematic review and meta-analysis. Acta Obstet Gynecol Scand 2011;90:564–73.
6. Douglas KA, Redman CW. Eclampsia in the United Kingdom. BMJ 1994;309:1395–400.
7. Knight M. Eclampsia in the United Kingdom 2005. BJOG 2007;114:1072–8.
8. Mahmoudi N, Graves SW, Solomon CG, Repke JT, Seely EW. Eclampsia: a 13-year experience at a United States tertiary care center. J Womens Health Gend Based Med 1999;8:495–500.
9. Katz VL, Farmer R, Kuller JA. Preeclampsia into eclampsia: toward a new paradigm. Am J Obstet Gynecol 2000;182:1389–96.
10. Sibai BM. Eclampsia. VI. Maternal-perinatal outcome in 254 consecutive cases. Am J Obstet Gynecol 1990;163:1049–54.
11. Cavkaytar S, Ugurlu EN, Karaer A, Tapisiz OL, Danisman N. Are clinical symptoms more predictive than laboratory parameters for adverse maternal outcome in HELLP syndrome? Acta Obstet Gynecol Scand 2007;86:648–51.
12. Altman D, Carroli G, Duley L, Farrell B, Moodley J, Neilson J, et al.. Do women with pre-eclampsia, and their babies, benefit from magnesium sulphate? The Magpie Trial: a randomised placebo-controlled trial. Lancet 2002;359:1877–90.
13. Matsuda H, Sakaguchi K, Shibasaki T, Takahashi H, Kawakami Y, Furuya K, et al.. Cerebral edema on MRI in severe preeclamptic women developing eclampsia. J Perinat Med 2005;33:199–205.
14. Zwart JJ, Richters A, Ory F, de Vries JI, Bloemenkamp KW, van Roosmalen J. Eclampsia in the Netherlands. Obstet Gynecol 2008;112:820–7.
15. Rolfes DB, Ishak KG. Liver disease in toxemia of pregnancy. Am J Gastroenterol 1986;81:1138–44.
16. von Dadelszen P, Payne B, Li J, Ansermino JM, Broughton Pipkin F, et al.. Prediction of adverse maternal outcomes in pre-eclampsia: development and validation of the fullPIERS model. Lancet 2011;377:219–27.