Eclampsia, a comparison within the International Network of Obstetric Survey Systems

Authors


Abstract

Objective

To compare incidences, characteristics, management and outcome of eclampsia in the Netherlands and the UK.

Design

A comparative analysis of two population-based prospective cohort studies.

Setting

All hospitals with consultant-led maternity units.

Population

Women with eclampsia in the Netherlands (226) and the UK (264).

Methods

Comparison of individual level data from national studies in the Netherlands and the UK (LEMMoN 2004–06; UKOSS 2005/06).

Main outcome measures

Incidence, maternal complications and differences in management strategies.

Results

Incidences of eclampsia differed significantly between both countries: the Netherlands 5.4/10 000 deliveries versus UK 2.7/10 000 (relative risk [RR] 1.94, 95% confidence intervals [95% CI] 1.6–2.4). The proportion of women with a preceding diagnosis of pre-eclampsia was comparable between both countries (the Netherlands 42%; UK 43%), as was the proportion who received magnesium sulphate prophylaxis. Women in the Netherlands had a significantly higher maximum diastolic blood pressure (111 mmHg versus 95 mmHg, < 0.001); significantly fewer received anti-hypertensive medication (16% versus 71%; RR 0.2, 95% CI 0.1–0.3) and were treated less often with magnesium sulphate after their first fit (95% versus 99%; RR 0.96, 95% CI 0.92–0.99). Maternal death occurred in three cases in the Netherlands compared with zero in the UK.

Conclusions

The incidence of eclampsia in the Netherlands was twice as high compared with the UK when using uniform definitions. Women with eclampsia in the Netherlands were not managed according to guidelines, particularly with respect to blood pressure management. Changes in management practice may reduce both incidence and poor outcomes.

Introduction

The incidence of maternal mortality and morbidity shows conflicting trends in the last decade in high-income countries.[1-5] For example, the maternal mortality ratio in the Netherlands has risen in the past decade, whereas in the same period this ratio declined in the UK. Many theories have been proposed for this contrasting increase, including rising maternal age, obesity and chronic hypertension.[1, 2] Examining the rates more closely it is clear that this difference is almost entirely the result of hypertensive disease of pregnancy. While in the Netherlands the level of substandard care in all cases of direct maternal mortality decreased from 78% to 63%, this decline was not present in women with hypertensive disease of pregnancy.[1, 2, 6, 7] A difference in quality of care might therefore explain some of the between-country differences in maternal mortality and morbidity due to hypertensive disease.

Both the Netherlands (LEMMoN 2004–06) and the UK (UKOSS 2005–06) have undertaken nationwide registration studies of eclampsia during comparable periods.[8, 9] A comparison of the incidences calculated by both studies shows an incidence of eclampsia in the Netherlands of 6.2 per 10 000 deliveries compared with 2.7 per 10 000 deliveries in the UK. There may be several reasons for this difference, including the use of different definitions of the condition. We could locate no detailed comparisons between the two countries in terms of characteristics of women with eclampsia, their management and outcome. Hence, it is not known whether there is a true difference in incidence and if so, whether this is the result of different populations or management strategies.

The aim of this study was therefore to compare the incidence of eclampsia between the Netherlands and the UK and to describe characteristics, management and outcomes of women with eclampsia.[10]

Methods

Details of women with eclampsia in the Netherlands were retrieved from a nationwide cohort study of Severe Maternal Morbidity in the Netherlands (LEMMoN) between 1 August 2004 and 1 August 2006; all Dutch obstetrics units participated. Requests for notification of cases of eclampsia, along with other types of severe maternal morbidity, were sent to the local coordinators from the LEMMoN study on a monthly basis. Cases were communicated to the National Surveillance Centre for Obstetrics and Gynaecology in a web-based form by mentioning date of birth and initials of the women. Absence of a case was also registered. Reminders were sent to nonresponders every month until they had returned the monthly notification cards. After notification, a case report form was sent to the local coordinator and returned to the research team, accompanied by anonymised photocopies of all relevant parts of the hospital case notes and correspondence.

Cases of eclampsia in the UK were retrieved from the study into eclampsia using the UK Obstetric Survey System (UKOSS) between February 2005 and February 2006. In this study, nominated clinicians in each hospital with a consultant-led maternity unit were sent a card each month, which, together with a series of other conditions, included a simple tick-box to indicate a case of eclampsia and a box indicating ‘nothing to report’. They were asked to return the card each month whether or not there had been any women who had an eclamptic episode. When a card was received indicating a case of eclampsia, the reporting clinician was sent a data collection form requesting further details to confirm the case definition and describe risk factors, management and outcomes. All data were reported anonymously. Cases that did not meet the case definition were excluded. In both the Netherlands and the UK, any case of eclampsia occurring outside a hospital will be referred to the hospital and so should have been notified to either study.

Crucial for a meaningful comparison is the comparability between the definitions used in both studies. The definition used in the Netherlands was more liberal compared with the UKOSS definition (Figure 1). Therefore, in this comparison we included only those Dutch cases that met the inclusion criteria of the UKOSS study.

Figure 1.

Definitions used for eclampsia.

The individual level data from both studies were used to compile an aggregated data set in IBM SPSS Statistics, release 19.0.0 (SPSS Inc., Chicago, IL, USA). In circumstances where a variable from either country was not directly comparable with the other, both variables were recoded into one unifying variable. The main outcome measures were incidence, case fatality rate, putative risk factors and management differences. The use of magnesium sulphate was compared for the prevention of eclampsia (magnesium sulphate prophylaxis) before the first fit and treatment of eclampsia (magnesium sulphate treatment). Proportions are presented as percentages, skewed distributions are presented as medians with interquartile ranges. For categorical data the differences are presented as relative risks (RR) with 95% confidence intervals (95% CI). We calculated P values with the chi-square test, unless the cell count was <5, in which case we used Fisher's exact test. For continuous data with a non-normal distribution we used the Mann–Whitney U test. Ethical approval was obtained for the original study from the Ethics Committee in both countries. Further approval was not required for this secondary analysis of anonymous data.

Results

All hospitals with a maternity unit in the Netherlands (98) and the UK (229) participated. The mean monthly card return rate was 97% in the Netherlands and 91% in the UK. There was a total of 358 874 maternities in the Netherlands (2004–06) compared with 775 186 in the UK (2005–06) during the studied timeframes. In the Netherlands, there were 222 cases of eclampsia compared with 214 in the UK. Applying the UKOSS criteria led to the exclusion of 30 cases from the Netherlands: nine because of insufficient data and 21 because only one (instead of at least two) of the four features from the UKOSS definition were present. In total there were 406 cases available for analysis (Figure 2). This resulted in an incidence of 5.4 per 10 000 deliveries (95% CI 4.6–6.2) in the Netherlands versus 2.7 per 10 000 (95% CI 2.4–3.2) in the UK (RR 1.94, 95% CI 1.6–2.4).

Figure 2.

Included cases of eclampsia.

Characteristics of the women are shown in Table 1. The women that experienced eclampsia in the Netherlands (NL) were significantly older (NL 30 years versus UK 25 years; < 0.001), consisted of significantly fewer women with a body mass index >30 kg/m2 (NL 9% versus UK 17.0%; RR 0.5, 95% CI 0.3–0.9) and consisted of more women with multiple pregnancies (NL 10% versus UK 4%; RR 2.4, 95% CI 1.1–5.1). Also, women in the Netherlands were less likely to smoke (NL 8% versus UK 18%; RR 0.5, 95% CI 0.3–0.8). (Table 1) No contemporaneous control women were selected as part of these studies; specifically, we do not have data on the proportion or characteristics of women with pre-eclampsia in the two countries.

Table 1. Comparison of characteristics of women with eclampsia in the Netherlands and UK
 The Netherlands (n = 192)United Kingdom (n = 214)RR95% CIP value
  1. ALT, alanine aminotransferase; AST, aspartate aminotransferase; CS, caesarean section; CVA, cerebrovascular accident; NA, not applicable.

  2. Data are number of women n (% of available) or median (25th and 75th quartiles).

  3. a

    NL number based on diastolic and/or systolic blood pressure; UK number based only on diastolic blood pressure.

  4. b

    Incidence expressed as incidence per 100 000 deliveries.

Maternal characteristics
Woman's age (years)30 (25–35)25 (20–31)< 0.001
Non-Caucasian59/192 (31%)57/211 (27%)1.10.81.60.44
Nulliparous135/191 (71%)157/213 (74%)1.00.91.20.51
Multiple pregnancy19/191 (10%)9/214 (4%)2.41.15.10.030
Smoking16/192 (8%)38/214 (18%)0.50.30.80.005
Body mass index (≥ 30)15/162 (9%)32/184 (17.0%)0.50.30.90.029
History of epilepsy2/192 (1%)5/214 (2%)0.40.12.30.45
Hypertension before pregnancy12/192 (6%)8/214 (4%)1.70.74.00.26
Signs
Diagnosed pre-eclampsia before eclampsia80/192 (42%)91/214 (43%)1.00.81.20.92
Premonitory signs
Headaches115/178 (65%)119/191 (62%)1.00.91.20.67
Visual disturbance49/165 (30%)48/212 (23%)1.30.91.90.13
Diastolic blood pressure
Median of highest measured (mmHg)111 (105–120)95 (80–104)< 0.001
≥110 mmHg130/187 (70%)89/174 (51%)1.41.11.6< 0.001
Meeting criteria for severe hypertension134 (70%)89 (51%)aNANANANA
Proteinuria at admission (g/l/24 hour)2.0 (0.59–5.00) (n = 94)1.2 (0.67–1.99) (n = 17)0.10
Highest proteinuria (g/l/24 hour) (available)5.0 (0.60–5.38) (n = 107)1.2 (0.40–2.50) (n = 16)0.27
Lowest platelet count106 (58–183)176 (98–226)<0.001
Highest AST84 (40–372)46 (33–151)0.013
Highest ALT58 (19–234)33 (18–132)0.012
Eclampsia
First fit antepartum77/192 (40%)97/214 (45%)0.90.71.10.32
First fit intrapartum56/192 (29%)41/214 (19%)1.51.12.20.020
First fit postpartum59/192 (31%)76/214 (36%)0.90.71.10.31
Recurrent fits46/192 (24.0%)54/210 (26%)0.90.71.30.73
Management
Magnesium sulphate prophylaxis (entire study population)19/192 (10%)12/214 (6%)1.80.93.50.13
Pre-eclampsia before eclampsia9/80 (11%)9/91 (10%)1.10.52.70.81
Magnesium sulphate treatment (entire study population)181/192 (95%)211/214 (99%)0.960.920.990.026
Anti-hypertensive treatment
Entire population31/192 (16%)151/214 (71%)0.20.10.3< 0.001
Of women with diastolic >110 mmHg31/130 (24%)56/89 (63%)0.40.30.5< 0.001
Gestational age at delivery
Antepartum37+5 (30–36)38+0 (30–37)0.63
Pre-eclampsia before eclampsia37+3 (30–35)37+1 (30–37)0.72
Pregnancy ended iatrogenically
Entire population154/192 (80%)162/214 (76%)1.10.951.180.28
Antepartum (induction of labour/CS)72/77 (94%)96/97 (99%)0.90.91.00.07
Delivery by CS
Prelabour60/76 (79%)84/97 (87%)0.90.81.00.22
After onset of labour29/116 (25%)20/117 (17%)1.50.92.40.15

Interval fit to delivery

Intrapartum (minutes) (hours:minutes)

60

(0:10–2:25)

48

(0:21–0:83)

0.17

Interval fit to delivery

Antepartum (minutes) (hours:minutes)

420

(2:00–42:00)

160

(0:57–8:08)

0.008

Interval fit to delivery

Ante partum; only CS delivery (hours:minutes)

240

(1:01–15:04)

127

(0:53–6:03)

0.006

Interval fit to delivery in minutes

Ante partum; only vaginal delivery (hours:minutes)

1800

(15:10–56:12)

779

(6:48–18:39)

0.003
Outcome
Maternal deathb0.840.000.11
CVA in eclampsiab1.950.651.60.54.80.56
Perinatal mortality9/192 (5%)5/214 (2%)2.00.75.90.20

The proportion of women with eclampsia who had pre-eclampsia diagnosed before eclampsia was comparable between countries—42% in the Netherlands compared with 43% in the UK. Also, premonitory symptoms were present in comparable proportions in both countries. Women in the Netherlands had a higher highest diastolic blood pressure before eclampsia compared with the UK (NL 111 mmHg versus UK 95 mmHg; < 0.001), a lower platelet count nadir (NL 106 × 109/l versus UK 176 × 109/l; < 0.001), a higher highest aspartate aminotransferase (NL 84 IU/l versus UK 46 IU/l; P = 0.013) and a higher highest alanine aminotransferase ALT (NL 58 IU/l versus UK 33 IU/l; = 0.012). In the Netherlands, 130 women had a diastolic blood pressure >110 mmHg (70%) compared with 90 (51%) in the UK (RR 1.4, 95% CI 1.1–1.6). In the Netherlands, 134 (70%) of the women met the criteria for severe hypertensive disease as defined by The Dutch Society of Obstetrics and Gynaecology by use of either systolic or diastolic blood pressure or both (systolic blood pressure ≥160 mmHg and/or diastolic blood pressure ≥110 mmHg). Since there were no systolic blood pressures available for the UK women, severe hypertension as defined by the Royal College of Obstetricians and Gynaecologists was based only on diastolic blood pressures. Therefore, no comparison could be performed (Table 1).

In both countries, most women had their first eclamptic episode before birth. In the Netherlands, however, significantly more women had their first fit during labour (NL 29% versus UK 19%; RR 1.5, 95% CI 1.1–2.2). The number of women with recurrent fits was comparable between both nations (Table 1).

There was no difference in the use of magnesium sulphate prophylaxis between both countries. Magnesium sulphate was used to treat eclampsia in 95% of the women in the Netherlands compared with 99% in the UK (RR 0.96, 95% CI 0.92–0.99). Only 16% of the women in the Netherlands received anti-hypertensive medication versus 71% in the UK (RR 0.2, 95% CI 0.1–0.3). In the subgroup of women with a diastolic blood pressure of >110 mmHg, anti-hypertensives were also administered significantly less frequently in the Netherlands (NL 24% versus UK 63%; RR 0.4, 95% CI 0.3–0.5; Table 1).

Gestational age at delivery was comparable between both countries for both the entire group of women that developed eclampsia as well as for the subgroup with pre-eclampsia diagnosed before eclampsia. Similarly, labour was induced or pregnancy was ended by planned caesarean section in comparable proportions of women in the Netherlands versus the UK.

We analysed the interval between fit and delivery to approximate the time needed to stabilise the woman and accomplish delivery. The time between eclampsia and delivery for the intrapartum group was not significantly different with a median of 60 minutes in the Netherlands versus 48 minutes in the UK. In the women with antepartum eclampsia there was a significant difference in interval with 420 minutes in the Netherlands versus 160 minutes in the UK (= 0.008). When subdividing the antepartum cases among women who ultimately delivered by caesarean section or vaginal delivery, both remained significantly different with a longer interval in the Netherlands (caesarean section NL 240 minutes versus UK 127 minutes, P = 0.006; vaginal delivery NL 1800 minutes versus UK 779 minutes, = 0.003; Table 1).

During the study period, three maternal deaths occurred in women with eclampsia in the Netherlands compared with none in the UK. This resulted in an incidence of 0.84 per 100 000 for the Netherlands and 0.00 per 100 000 for the UK; this difference was not statistically significant. One woman died at home after repeated refusal of admission, two others died in hospital after term delivery.

In the Netherlands, 7 (1.95 per 100 000) women developed a cerebrovascular accident compared with 5 (0.65 per 100 000) women in the UK, not significantly different. There were no differences in perinatal death between both countries.

Discussion

Main findings

The incidence of eclampsia differed considerably during the studied timeframe, and was higher in the Netherlands compared with the UK. From an international perspective, this incidence was also higher than in the countries surrounding the Netherlands including all the Nordic countries.[11, 12]

Strengths and weaknesses

This study to our knowledge is the first international comparison of eclampsia based on individual level data from International Network of Obstetric Survey Systems (INOSS). This form of analysis makes inter-country comparison of management strategies possible for rare disorders of pregnancy, such as eclampsia.

Due to the retrospective nature of this analysis, not all variables from either country had comparable definitions or could be recoded into unifying definitions. This led sometimes to small numbers in some of the analysed variables or to noncomparability such as eclampsia outside hospital or more details concerning ethnic background. Furthermore, neither study included data on women with pre-eclampsia who ultimately did not develop eclampsia, therefore it is not known whether the differences found occur among all women with hypertensive disease of pregnancy.

Interpretation

The characteristics of the affected women within both countries are not entirely similar. The number of women with eclampsia who received anti-hypertensive treatment differed between the countries, with only 16% in the Netherlands compared with 71% in the UK. Systolic blood pressure measurements were not available for the women from the UK. In the Netherlands, 134 of the 192 women (69%) had severe hypertension based on both systolic and/or diastolic blood pressure as defined by Dutch guidelines, suggesting that use of anti-hypertensives was too conservative. The aim of anti-hypertensive treatment is to avert further maternal complications such as potentially fatal cerebral haemorrhage. In the Netherlands, seven women had a cerebrovascular accident of whom three ultimately died. In comparison, five women developed a cerebrovascular accident in the UK, without any fatalities. There has been much controversy concerning the threshold for the use of anti-hypertensive treatment.[13-16] The Dutch Society of Obstetrics and Gynaecology guidelines advise treating severe hypertension defined as a systolic blood pressure >160 mmHg and/or a diastolic blood pressure >110 mmHg.[17] This is comparable to the UK guideline during the studied timeframe. The guidelines differ in the recommendation for treatment in cases of non-severe pre-eclampsia, in that the Dutch guideline states there is insufficient evidence to start treatment. Which is in line with the updated recommendations of the American Congress of Obstetricians and Gynecologists taskforce, stating that anti-hypertensive medication should not be administered in mild gestational hypertension or pre-eclampsia.[18] Nevertheless, in the Netherlands more than two-thirds of the women had high blood pressure warranting the use of anti-hypertensives according to nationwide guidelines, but only 24% of these women received the appropriate treatment. This suggests a high non-compliance with guidelines resulting in substandard care. The development of guidelines on blood pressure management in nonsevere pre-eclampsia is hampered by a lack of reliable data on the benefits of lowering blood pressure in this subgroup coupled with possible negative effects such as reduced placental perfusion and resulting fetal compromise.[19-23]

The use of magnesium sulphate prophylaxis was not significantly different. Data on use of magnesium sulphate prophylaxis in all pregnant women are not available for both countries. Therefore, it is not clear whether prophylaxis failed more frequently in the Netherlands. There was a significant difference in magnesium sulphate treatment, which was still not used in 5% of cases in the Netherlands compared with only 1% in the UK. Guidelines in both countries are comparable on guidance for magnesium sulphate treatment and prophylaxis, such as loading dose and continuous infusion.[17, 24] Women with eclampsia in the Netherlands were as likely as women in the UK to have premonitory symptoms that may be suggestive of impending encephalopathy, and were equally likely to have received magnesium sulphate prophylaxis. This does not therefore suggest that clinicians in the Netherlands were less likely to respond to these premonitory symptoms. However, we do not have data on women with pre-eclampsia without eclampsia, which would be key to investigating whether there is a difference between the two countries in the way symptoms of concern in women with pre-eclampsia are acted upon.

Interestingly, the number of women with eclampsia who had labour induced or pregnancy ended by elective caesarean section was comparable in both countries. Surprisingly, there was a substantial difference in time to achieve a stable clinical condition and delivery between both countries. This is illustrated by higher blood pressures and worse blood results, such as a lower platelet count and higher liver enzymes. The longer interval between convulsions and delivery in the Netherlands suggests that even when women had developed eclampsia there was still less urgency to expedite delivery. This restrained attitude towards delivery may have been part of the routine practice and if so, may also have been applied to women with pregnancy-induced hypertension or pre-eclampsia, possibly leading to the higher incidence of eclampsia.

There was no significant difference in maternal mortality ratio attributable to eclampsia; there were three cases in the Netherlands compared with zero in the UK. Similarly, perinatal mortality was almost twice as frequent, though the difference was not statistically significant. The rates of postpartum haemorrhage could not be compared because of different definitions. Other forms of maternal morbidity could not be compared in this analysis but would be expected to be high given that previous studies estimate that 70% of women with eclampsia develop severe morbidities.[8, 9, 25]

In the management of severe pre-eclampsia and eclampsia, stabilising the woman before proceeding with delivery is an important element of care. The difference in interval between convulsions and delivery suggests that in the Netherlands, clinicians take much longer to confirm that the woman is stable. Dutch guidelines do not include parameters to determine whether a hypertensive woman was stable nor do they address a timeframe in which to achieve this. Therefore, a longer interval appears to be accepted to ensure stability, but could ultimately lead to more complications. The data from this study suggest that stabilisation, including a rigorous treatment of elevated blood pressure, should start immediately and that the interval between eclampsia and delivery should be an integral part of clinical decision making.

Conclusion

These data confirm that the incidence of eclampsia during the studied timeframe was twice as high in the Netherlands compared with the UK. Different management in women with eclampsia, in particular the lower use of anti-hypertensives, suggests a less active practice in the Netherlands. It is not clear whether this reflects management of all women with hypertensive disease of pregnancy; less active management in women with pre-eclampsia could lead to a higher incidence of eclampsia. Therefore, the incidence and subsequent morbidity in the Netherlands might be reduced by changes in management, focused on more aggressive diagnosis and treatment of pre-eclampsia and more frequent use of both magnesium sulphate and anti-hypertensive treatment. This would be instigated by a stricter adherence to existing clinical guidelines. The study suggests that there may be a need for more practical guidelines and parameters to discern stable from non-stable women and the need to start stabilisation immediately after the onset of severe pre-eclampsia.

Since the LEMMoN study and other publications into maternal morbidity and mortality in the Netherlands, stricter adherence to clinical guidelines and training in obstetric emergencies has been propagated by obstetricians. To date there have been no new studies into the incidence of eclampsia. A nationwide registration in the Netherlands is therefore warranted to ascertain whether the recent changes in management practice have decreased the incidence of eclampsia. Furthermore, international comparative analysis of serious maternal illness is imperative for further advancements in the quality of obstetric care and could be propagated through INOSS.[26]

Disclosure of interests

The authors have no potential conflicts of interest to disclose.

Contribution to authorship

TS, MK, KB and JvR designed the study and revised the manuscript. JJZ and JvR provided the data from the severe maternal morbidity study (LEMMoN). MK provided the data from the UKOSS eclampsia study (UKOSS Eclampsia 2005). TS coordinated data collection, coded the data, carried out the analysis and wrote the first draft of the paper. MK, JvR, KB, JK and PB corrected and revised the manuscript. All authors gave final approval of the manuscript.

Details of ethics approval

Ethics approval is not applicable.

Funding

No funding was received.

Acknowledgements

The authors thank the members of both the UKOSS and the LEMMoN study group and the reporting clinicians in both countries who notified cases to both studies and made data available for this analysis. We also thank the members of the INOSS for their advice on the study.

Ancillary