Home > Clinical Guides > Hypertension in Pregnancy: The... > Medical management of severe...

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Women's and Children's Health (UK). Hypertension in Pregnancy: The Management of Hypertensive Disorders During Pregnancy. London: RCOG Press; 2010 Aug. (NICE Clinical Guidelines, No. 107.)

10Medical management of severe hypertension or severe pre-eclampsia in a critical care setting

10.1. Introduction

Severe pre-eclampsia continues to cause maternal and perinatal morbidity. The UK Confidential Enquiries into Maternal Death have consistently reported substandard care in the management of these women. Protocols and guidelines have been developed in most units and more recently supported by guidance in this area from the Royal College of Obstetricians and Gynaecologists (RCOG). This section reviews the evidence for the acute management of severe pre-eclampsia that is conducted within a critical care setting, or what is more usually known as high-dependency care. In most circumstances, this occurs following a decision to end the pregnancy.

A single literature search was conducted for the various interventions: antihypertensive drugs, anticonvulsant drugs, steroids for HELLP syndrome (to prolong pregnancy) and for fetal lung maturation, fluid therapy and operative birth (caesarean section). The population studied was women with severe pre-eclampsia, eclampsia, severe hypertension or HELLP syndrome. The search identified 3379 references, of which 152 were retrieved.

10.2. Anticonvulsants

Clinical effectiveness

Six high-quality publications were identified.164–169 [EL = 1++] Four of these were Cochrane systematic reviews164–167 and the remaining two were separate publications that reported follow-up data from a single large double-blind RCT,168;169 which was included in one of the Cochrane systematic reviews.167 Of the Cochrane systematic reviews, one examined magnesium sulphate and other anticonvulsants for the prevention of eclampsia in women with pre-eclampsia,167 and the other three compared magnesium sulphate with other anticonvulsants for the treatment of eclampsia.164–166

Prevention of eclampsia

Magnesium sulphate versus placebo or no treatment

A Cochrane systematic review167 [EL = 1++] investigated the differential effects of magnesium sulphate (intramuscular or intravenous) when compared with placebo or no treatment for the care of women with pre-eclampsia. A subgroup analysis by severity of pre-eclampsia was also conducted: severe pre-eclampsia was defined as two or more signs or symptoms of imminent eclampsia, or blood pressure of 170/110 mmHg or higher and 3+ proteinuria, or, if on antihypertensive treatment, 150/110 mmHg or higher and 2+ proteinuria, or if the individual study authors described them as having severe pre-eclampsia. Those who did not meet any of the above criteria were classified as not having severe pre-eclampsia, which for the purpose of this guideline is reported as mild or moderate pre-eclampsia.

Six RCTs were included in the review (n = 11–444 women). One multicentre RCT (the Magpie trial) involved 10–141 women. Other smaller trials were conducted in the USA, South Africa and Taiwan. The quality of the studies included in this review ranged from excellent to poor. In the largest study, concealment of allocation was secure and completeness of follow-up was 99%. In one trial, the procedure used for trial entry did not give secure concealment of allocation and 17% of women were lost to follow-up. Apart from the Magpie trial, few studies attempted to blind administration of the allocated treatment.

Women with severe pre-eclampsia

In women with severe pre-eclampsia, magnesium sulphate was statistically significantly better than none/placebo in preventing eclampsia (three RCTs, n = 3555: RR 0.37; 95% CI 0.22 to 0.64). No statistically significant differences were found between the two groups in terms of maternal death, serious maternal morbidity, pulmonary oedema, placental abruption or kidney dialysis. The stillbirth and neonatal death rates were not statistically significantly different between the two groups.

Women with mild or moderate pre-eclampsia

Results for the mild or moderate pre-eclampsia subgroup showed that magnesium sulphate was statistically significantly better than none/placebo in preventing eclampsia (four RCTs, n = 3889: RR 0.44: CI 0.28 to 0.69). Other outcomes, however, were not statistically significantly different between the two groups (maternal death, serious maternal morbidity, stillbirth and neonatal death).

Follow-up for women (outcomes at 2 years)

A large RCT (the Magpie trial)168 [EL = 1++] investigated the prognosis and possible unexpected adverse events related to the use of magnesium sulphate in the cohort of women with pre-eclampsia in the original trial.170 In the Magpie trial, 7927 women with pre-eclampsia before birth or 24 hours postpartum (diastolic blood pressure 90 mmHg or higher, systolic blood pressure 140 mmHg or higher, proteinuria 1+ or more) were randomised to receive either magnesium sulphate (intravenous or intramuscular) or identical placebo regimens. Of the 4782 women contacted for the follow-up study, 3375 women participated (reasons for exclusions were the feasibility of following up in some centres, women discharged without a surviving child, and women who opted out of centres that contacted fewer than 20% of families). Women were randomised either via a central telephone service or consecutively numbered sealed treatment packs stratified by centre. A computer-generated allocation sequence was used. The baseline characteristics of the women in the two groups at trial entry were comparable.

The primary outcome reported was death or serious morbidity related to pre-eclampsia. No statistically significant difference in the primary outcome was found between the two groups (58 of 1650 versus 72 of 1725: RR 0.84; 95% CI 0.60 to 1.18). This difference remained non-statistically significant when ‘death’ and ‘serious morbidity’ outcomes were analysed separately. Subgroup analyses were conducted for the primary outcome to see whether the results were affected by the severity of pre-eclampsia (severe versus mild moderate), the randomisation (before delivery versus after delivery) or the respective country’s perinatal mortality index (high, middle or low). Results were consistent across all subgroups.

The only outcomes for which the difference between the magnesium sulphate and placebo groups achieved statistical significance was gynaecological problems, for which the risk was higher in the magnesium group (RR 1.59; 95% CI 1.17 to 2.16).

Follow-up for children (outcomes at 18 months)

In another publication169 [EL = 1++] from the Magpie trial, the authors investigated whether giving magnesium sulphate to women with pre-eclampsia had effects on the child’s chance of developing major neurosensory disability (18 months follow-up). This follow-up study contacted 4483 children, of whom 3283 ultimately participated (reasons for exclusion were those not eligible for follow-up, or those born at centres where follow-up was not thought possible).

The primary outcome reported was death or non-congenital neurosensory disability. No statistically significant difference in the primary outcome was found between babies born to mothers treated with magnesium sulphate and those born to mothers treated with placebo (245 of 1635 versus 233 of 1648: RR 1.10; 95% CI 0.93 to 1.29). The difference remained non-statistically significant when ‘death’ and ‘neurosensory disability’ outcomes were analysed separately (death: 226 of 1635 versus 206 of 1648: RR 1.06; 95% CI 0.90 to 1.25; neurosensory disability: ten of 1409 versus 27 of 1442: RR 0.72; 95% CI 0.40 to 1.29).

Subgroup analyses were conducted for the primary outcome to see whether the results were affected by the severity of pre-eclampsia at trial entry (severe, moderate, mild), gestation at birth (up to 33 weeks, more than 33 weeks) or the country’s perinatal mortality index (high, middle, low). ). Results were consistent across all subgroups.

No statistically significant differences were found between the two groups in terms of having isolated speech delay or other significant disability.

Cost effectiveness

A literature search identified 100 studies and four were ordered. Only one study171 met the inclusion criteria. The study was a multinational trial-based economic evaluation of the Magpie trial. Outcome and hospital resource use data were available for the trial period from the 33 participating countries. The study was an international study coordinated from the UK. The GDG believes that the study represented practice that was relevant to the UK. Country-specific unit costs were collected as part of the study and converted into USD at 2001 prices using national consumer price indices. The conversion of the reported CPI in USD at 2001 to prices in GBP 2009 was done using a CPI conversion calculator.172 Cost effectiveness was estimated for three categories of country grouped by gross national income (GNI) into high-, middle- and low-GNI countries using a regression model. Uncertainty was explored using probabilistic sensitivity analysis. Results of the high-income countries that are relevant to the UK were abstracted.

Using magnesium sulphate to prevent eclampsia in women with pre-eclampsia costs, on average, $86 (approximately £60) and results in reductions in hospital resource use, due to the lower risk of eclampsia, worth an average of $20 (approximately £14) per woman. Because overall the reduction in healthcare expenditure per pregnancy is less than the cost of the magnesium sulphate treatment, the net health service cost is higher for the intervention group than for the control group. Thus the incremental healthcare cost to prevent a case of eclampsia is $21,202 (approximately £14,752).

The cost-effectiveness acceptability curves show the probability of prophylactic magnesium sulphate being cost effective as a function of the decision-maker’s willingness to pay to prevent a case of eclampsia against the alternative of not providing prophylactic anticonvulsant. Eighty percent certainty about the cost effectiveness of the intervention was not reached, even if decision-makers would be willing to pay more than $50,000 (approximately £34,800) per case of eclampsia prevented. A subgroup analysis by severity of pre-eclampsia showed that it would approximately halve the cost per case of eclampsia prevented since the absolute benefit from treatment is huge. The estimated ICER would fall to $11,149; (approximately £7,760) (95% CI £500 to £59,200).

The authors concluded that magnesium sulphate for pre-eclampsia is cost effective in the prevention of eclampsia in high-GNI countries. Cost effectiveness substantially improves if it is used only for severe pre-eclampsia. This was a well-conducted economic analysis with results that were well presented. Although NICE’s preferred measure of outcome is a QALY, the study did not consider this; however, the GDG believes this approach would be unlikely to change the conclusions of the analysis since eclampsia is a good proxy for both the quality and the quantity of life that would generate the QALYs.

Evidence statement

A Cochrane review [EL = 1++] showed that in women with either severe or mild/moderate pre-eclampsia, magnesium sulphate was statistically significantly better than no treatment/placebo in preventing eclampsia. However, there were no statistically significant differences in other outcomes, including maternal death and serious maternal morbidity.

A well-conducted economic analysis found that magnesium sulphate was cost effective in preventing eclampsia when compared with placebo in women with pre-eclampsia. The cost effectiveness improved with severity of pre-eclampsia.

A large RCT [EL = 1++] investigated the long-term effects of magnesium sulphate used in pre-eclampsia in the mothers (at 2 years follow-up) and their babies (at 18 months follow-up) in comparison with placebo. The trial found no statistically significant differences between the mothers or the babies of the two groups in the primary outcomes studied (mothers: death or serious morbidity potentially related to pre-eclampsia; babies: death or non-congenital neurosensory disability). Subgroup analysis by severity of pre-eclampsia was consistent across all subgroups. The only outcome for which the difference between the two groups of mothers achieved statistical significance was ‘gynaecological problems’, for which the risk was higher in the magnesium sulphate group. No statistically significant differences were found in the babies for any of the other studied outcomes (isolated speech delay or significant disability).

Clinical effectiveness

Treatment of eclampsia

Three Cochrane systematic reviews studied the use of magnesium sulphate in women with eclampsia compared with diazepam,164, phenytoin165 and lytic cocktail166 (lytic cocktail is no longer used in UK clinical practice). For a better overview of the available evidence, results for the primary outcomes of these reviews are presented in Tables 10.1a (maternal outcomes) and 10.1b (fetal outcomes).

Table 10.1a. Maternal outcomes reported in systematic reviews of treatment for women with eclampsia – magnesium sulphate compared with diazepam, phenytoin and lytic cocktail (reported as RRs with 95% CIs).

Table 10.1a

Maternal outcomes reported in systematic reviews of treatment for women with eclampsia – magnesium sulphate compared with diazepam, phenytoin and lytic cocktail (reported as RRs with 95% CIs).

Table 10.1b. Fetal outcomes reported in systematic reviews of treatment for women with eclampsia – magnesium sulphate compared with diazepam, phenytoin and lytic cocktail (reported as RRs with 95% CIs).

Table 10.1b

Fetal outcomes reported in systematic reviews of treatment for women with eclampsia – magnesium sulphate compared with diazepam, phenytoin and lytic cocktail (reported as RRs with 95% CIs).

Magnesium sulphate versus diazepam

A Cochrane systematic review investigated the effects of magnesium sulphate (intramuscular or intravenous) compared with diazepam.164 [EL = 1++] Participants were women with eclampsia at trial entry before or after delivery, who had singleton or multiple pregnancies, and who may have had an anticonvulsant before trial entry.

Seven RCTs were included in the review (n = 1441 women). Most trials included women with both antepartum and postpartum eclampsia. Overall, about half the women in this review had also had an anticonvulsant before trial entry. The treatment regimens all included a loading dose and maintenance therapy. Three trials were of good quality; adequacy of concealment of allocation was unclear in four other trials. The largest contribution to the Cochrane systematic review was from a good-quality RCT (the Collaborative Eclampsia Trial),173 which contributed 910 of the 1441 women in the review (63%). One study was available only as an unpublished report; another study was available as an abstract and an unpublished report. None of the trials could include blinding after randomisation because of the type of intervention.

Magnesium sulphate showed better results than diazepam in women with eclampsia. Both ‘maternal death’ and ‘recurrence of convulsions’ outcomes were statistically significantly less likely in the magnesium sulphate group compared with the diazepam group (maternal death: six RCTs, n = 1336; RR 0.59; 95% CI 0.37 to 0.94; recurrence of convulsions: seven RCTs, n = 1441; RR 0.44; 95% CI 0.34 to 0.57).

Babies of women treated with magnesium sulphate were statistically significantly less likely to stay in neonatal care (variously reported in the primary studies as NICU or special care baby unit (SCBU)) for longer than 7 days (three RCTs, n = 631; RR 0.66; 95% CI 0.46 to 0.95) and to be intubated at place of birth (two RCTs, n = 591; RR 0.67; 95% CI 0.45 to 1.00) when compared with babies born to mothers treated with diazepam. Besides, magnesium sulphate babies were statistically significantly less likely to score less than 7 in Apgar scale measured at both 1 minute (two RCTs, n = 597; RR 0.75; 95% CI 0.65 to 0.87) and 5 minutes after delivery (two RCTs, n = 597; RR 0.72; 95% CI 0.55 to 0.94).

Magnesium sulphate versus phenytoin

A Cochrane systematic review investigated the effects of magnesium sulphate (intramuscular or intravenous) compared with phenytoin.165 [EL = 1++] Participants were women with eclampsia at trial entry either before or after delivery, who had singleton or multiple pregnancies, and who may have had an anticonvulsant before trial entry.

Six RCTs were included in the review (n = 897) which mainly comprised women with antepartum eclampsia (only 17% were postpartum). About 80% of the women had received an anticonvulsant before trial entry. Five trials were small, and one was large (the Collaborative Eclampsia Trial).173 The Collaborative Eclampsia Trial contributed 777 of the 897 women in the Cochrane systematic review (87%). The methodological quality of the Collaborative Eclampsia Trial was good but concealment of allocation in the small trials was not adequate or not reported clearly. None of the trials could include blinding after randomisation because of the type of intervention.

The recurrence of convulsions was statistically significantly less likely in the magnesium sulphate group compared with the phenytoin group (five RCTs, n = 895; RR 0.31; 95% CI 0.20 to 0.47). Women in the magnesium sulphate group were statistically significantly less likely to be admitted to intensive care units (one RCT, n = 775; RR 0.67; 95% CI 0.50 to 0.89). They were also statistically significantly less likely to be given supportive mechanical ventilation (one RCT, n = 775; RR 0.66; 95% CI 0.49 to 0.90).

Babies born to women treated with magnesium sulphate were statistically significantly less likely to be admitted to NICU (one RCT, n = 518; RR 0.73; 95% CI 0.58 to 0.91) and were statistically significantly less likely to either die or to be admitted to NICU for more than 7 days (composite outcome of one RCT, n = 518; RR 0.53; 95% CI 0.33 to 0.86). Furthermore, fewer babies born to women treated with magnesium sulphate compared with babies born to women treated with phenytoin scored less than 7 in Apgar at 1 minute (one RCT, n = 518; RR 0.78; 95% CI 0.66 to 0.93). However, the Apgar score less than 7 at 5 minutes did not show a statistically significant difference.

Magnesium sulphate versus lytic cocktail

A Cochrane systematic review investigated the differential effects of magnesium sulphate (intramuscular or intravenous) compared with any combination of drugs known as ‘lytic cocktail’ regardless of their constituents or how they were administered.166 [EL = 1++] Participants were women who had eclampsia at trial entry, which could have been before or after delivery, who had singleton or multiple pregnancies, and who may have had an anticonvulsant before trial entry.

Two RCTs were included in the review (n = 199 women). For one study, the randomisation procedure was described, although it is unclear whether there was any central record of the envelopes or whether the envelopes were to be used in a particular sequence. One woman with uncertain diagnosis was excluded from the analysis. The other study was only available as an abstract, and there was no information about concealment of allocation or how outcome was assessed. Some additional information about the interventions and outcomes for this study was obtained by recording data from the poster presentation. The lytic cocktail in both trials was a combination of pethidine, promethazine and chlorpromazine.

The recurrence of convulsions was statistically significantly less likely in the magnesium sulphate group compared with the phenytoin group (two RCTs, n = 198; RR 0.09; 95% CI 0.03 to 0.24). Women in the magnesium sulphate group had statistically significantly fewer cases of coma at more than 24 hours (one RCT, n = 108; RR 0.04; 95% CI 0.00 to 0.74) and of respiratory depression (two RCTs, n = 198; RR 0.12; 95% CI 0.02 to 0.91). Fetal or infant deaths were statistically significantly lower in the magnesium sulphate group (two RCTs, n = 177; RR 0.45; 95% CI 0.26 to 0.79).

Evidence statement

A Cochrane review [EL = 1++] showed that in women with eclampsia, magnesium sulphate had statistically significantly better results than diazepam in preventing maternal death and recurrence of convulsions. Babies of women treated with magnesium sulphate were statistically significantly less likely to stay in neonatal care (variously reported in the primary studies as NICU or SCBU) for more than 7 days, to be intubated at place of birth or have an Apgar score less than 7 at both 1 minute and 5 minutes from delivery.

A Cochrane review [EL = 1++] showed that in women with eclampsia, magnesium sulphate has statistically significantly better results than phenytoin in preventing recurrence of convulsions. They were also statistically significantly less likely to be admitted to ICU or to be given supportive mechanical ventilation. No statistically significant results were found between the two groups in preventing maternal death. Babies born to women treated with magnesium sulphate were statistically significantly less likely to be admitted to neonatal care (variously reported in the primary studies as NICU or SCBU), to stay there for more than 7 days or to die there after > 7 days.

A Cochrane review [EL = 1++] showed that in women with eclampsia, magnesium sulphate has statistically significantly better results than a cocktail of lytic agents in preventing recurrence of convulsions, having a coma after more than 24 hours or having respiratory depression. Fetal or infant deaths were statistically significantly lower in the magnesium sulphate group.

GDG interpretation of the evidence

The evidence supported the use of magnesium sulphate in severe pre-eclampsia to prevent progression to eclampsia, as the number needed to treat to prevent one eclamptic fit was 50, whereas in women who have pre-eclampsia with mild or moderate hypertension, 100 women would need to be treated to avoid an eclamptic fit. There was no difference for the mother or fetus in other outcome measures. Regarding recurrence, there was clear evidence from RCTs and systematic reviews that magnesium sulphate treatment in eclampsia reduces the incidence of further eclamptic fits. There was also clear evidence from systematic reviews that magnesium sulphate is more effective than phenytoin, diazepam and lytic cocktail in preventing further eclamptic fits (lytic cocktail is no longer relevant to UK clinical practice). The GDG’s view is that treatment with magnesium sulphate is likely to be cost effective: it is cheaper and easier to administer than phenytoin, and it requires less follow-up nursing care than diazepam, which has sedative effects.173 The GDG’s view is that the regimen for administration of magnesium sulphate should be the intravenous regimen used in the Collaborative Eclampsia Trial,173 because this trial contributed much of the evidence for the effectiveness of magnesium sulphate and was of better methodological quality than the other included studies. The intravenous regimen used in the Collaborative Eclampsia Trial173 was:

  • a loading dose of 4 g given intravenously over 5 minutes, followed by an infusion of 1 g/hour maintained for 24 hours
  • recurrent seizures should be treated with a further dose of 2–4 g given over 5 minutes.

Most trials that compared the effectiveness of magnesium sulphate with phenytoin or diazepam also involved monitoring of respiration rate, urine output and tendon reflexes, but not serum, in women undergoing treatment.164;165

Recommendations

If a woman in a critical care setting who has severe hypertension or severe pre-eclampsia has or previously had an eclamptic fit, give intravenous magnesium sulphate.*

Consider giving intravenous magnesium sulphate* to women with severe pre-eclampsia who are in a critical care setting if birth is planned within 24 hours.

If considering magnesium sulphate* treatment, use the following as features of severe pre-eclampsia:

Use the Collaborative Eclampsia Trial§ regimen for administration of magnesium sulphate:*

  • loading dose of 4 g should be given intravenously over 5 minutes, followed by an infusion of 1 g/hour maintained for 24 hours
  • recurrent seizures should be treated with a further dose of 2–4 g given over 5 minutes.

Do not use diazepam, phenytoin or lytic cocktail as an alternative to magnesium sulphate* in women with eclampsia.

*

In this guideline, drug names are marked with an asterisk if they do not have UK marketing authorisation for the indication in question at the time of publication (August 2010). Informed consent should be obtained and documented.

§

The Eclampsia Trial Collaborative Group (1995) Which anticonvulsant for women with eclampsia? Evidence from the Collaborative Eclampsia Trial. Lancet 345:1455–63.

10.3. Antihypertensives

Clinical effectiveness

The population considered here included women with severe hypertension. No separate analyses were done for women with severe pre-eclampsia, severe chronic hypertension or chronic hypertension with superimposed pre-eclampsia. Eight studies were identified that compared various antihypertensive agents.174–181

One of these studies was a Cochrane systematic review174 [EL = 1++] of all randomised trials (quasi-randomised designs were excluded) that looked at any comparison of one antihypertensive agent with another regardless of dose, route of administration or duration of therapy. Comparisons of alternative regimens of the same agent and of alternative agents within the same class of drug were not included. Participants were women with severe hypertension (diastolic blood pressure of 105 mmHg or higher and/or systolic blood pressure of 160 mmHg or higher) during pregnancy requiring immediate treatment. Postpartum women were excluded.

The overall number of RCTs included was 24 (n = 2949 women). All trials were small, apart from one (n = 1750) that compared nimodipine with magnesium sulphate.

The antihypertensive drugs evaluated in these trials were hydralazine, calcium-channel blockers (nifedipine, nimodipine, nicardipine and isradipine), labetalol, methyldopa, diazoxide, epoprostenol, ketanserin, urapidil, magnesium sulphate, prazosin and isosorbide. Most drugs were given either intravenously or intramuscularly, except nifedipine, nimodipine, isosorbide and prazosin, which were given orally. Dosage varied considerably between studies, in both amount and duration.

Most of the included trials were small. Only three studies recruited more than 100 women. Several trials were conducted in countries where English is not widely used. Only five trials (n = 314 women) had adequate concealment of allocation. Most of the others did not give adequate information about how or whether the allocation to treatment group was concealed. For most trials, the identity of the allocated drug could only be blinded after trial entry with use of a double placebo. This was stated to have been conducted in one study (50 women). In another two, the comparison was stated to have been blinded.

The review identified 12 different comparisons:

Six other trials were identified that were not included in the Cochrane review – four176;179181 were EL = 1+ and two177;178 were EL = 1−. These trials studied five comparisons:

There is another well-conducted meta-analysis of RCTs175 [EL = 1++] that compared hydralazine with other antihypertensive drugs in pregnant women with moderate to severe hypertension (moderate: diastolic blood pressure of 100–109 mmHg; severe: diastolic blood pressure of 110 mmHg or higher). Twenty-one RCTs were included (n = 1085 women). The randomisation method was adequate in 11 trials while it was unknown or inadequate in the other trials. Blinding was applied in four trials. The other 17 were either not blinded (11 trials) or blinding was not reported (six trials). Five of these studies had women with moderate hypertension (one trial, n = 30: labetalol versus hydralazine; two trials, n = 59: urapidil versus hydralazine; two trials, n = 100: ketanserin versus hydralazine).

The meta-analysis identified five comparisons (labetalol, calcium-channel blockers, ketanserin, urapidil or epoprostenol versus hydralazine). There is an overlap in the included trials with the above-mentioned Cochrane review. However, the adverse effects and persistent high blood pressure outcomes were reported in more detail in this meta-analysis.

Overall, there were 15 different comparisons between a variety of antihypertensive drugs. Table 10.2 provides an overview of all the available evidence. Results for the primary outcomes of all included studies are presented in Tables 10.3 to 10.10. These tables present comparisons based on evidence available from two or more difference sources (the Cochrane systematic review, the meta-analysis or additional individual trials).

Table 10.2. Source and level of evidence for comparisons between the various antihypertensive agents.

Table 10.2

Source and level of evidence for comparisons between the various antihypertensive agents.

Table 10.3. Evidence from the Cochrane review, meta-analysis and individual trials for labetalol versus hydralazine (reported as RRs with 95% CIs).

Table 10.3

Evidence from the Cochrane review, meta-analysis and individual trials for labetalol versus hydralazine (reported as RRs with 95% CIs).

Table 10.4. Evidence from the Cochrane review, meta-analysis and individual trials for calcium-channel blockers versus hydralazine (reported as RRs with 95% CIs).

Table 10.4

Evidence from the Cochrane review, meta-analysis and individual trials for calcium-channel blockers versus hydralazine (reported as RRs with 95% CIs).

Table 10.5. Evidence from the Cochrane review, meta-analysis and individual trials for ketanserin versus hydralazine (reported as RRs with 95% CIs).

Table 10.5

Evidence from the Cochrane review, meta-analysis and individual trials for ketanserin versus hydralazine (reported as RRs with 95% CIs).

Table 10.6. Evidence from the Cochrane review, meta-analysis and individual trials for urapidil versus hydralazine (reported as RRs with 95% CIs).

Table 10.6

Evidence from the Cochrane review, meta-analysis and individual trials for urapidil versus hydralazine (reported as RRs with 95% CIs).

Table 10.7. Evidence from the Cochrane review, meta-analysis and individual trials for epoprostenol versus hydralazine (reported as RRs with 95% CIs).

Table 10.7

Evidence from the Cochrane review, meta-analysis and individual trials for epoprostenol versus hydralazine (reported as RRs with 95% CIs).

Table 10.8. Evidence from the Cochrane review, meta-analysis and individual trials for labetalol versus calcium-channel blockers (reported as RRs with 95% CIs).

Table 10.8

Evidence from the Cochrane review, meta-analysis and individual trials for labetalol versus calcium-channel blockers (reported as RRs with 95% CIs).

Table 10.9. Evidence from the Cochrane review for comparisons between various antihypertensives (reported as RRs with 95% CIs).

Table 10.9

Evidence from the Cochrane review for comparisons between various antihypertensives (reported as RRs with 95% CIs).

Table 10.10. Evidence from individual RCTs for comparisons between various antihypertensives (reported as RRs with 95% CIs).

Table 10.10

Evidence from individual RCTs for comparisons between various antihypertensives (reported as RRs with 95% CIs).

Table 10.10 presents comparisons based on evidence available in one source only (i.e. individual RCTs).

Labetalol versus hydralazine

The Cochrane review174 [EL = 1++] included three RCTs (n = 69) that compared labetalol with hydralazine. No statistically significant differences were found between the two drugs.

The meta-analysis175 [EL = 1++] included five RCTs (n = 156) that compared labetalol with hydralazine. Women treated with labetalol were statistically significantly more likely to have persistent high blood pressure in comparison with those treated with hydralazine (four RCTs, n = 126: RR 3.4; 95% CI 1.0 to 12.5). However, they were less likely to have hypotension (four RCTs, n = 122: RR 0.2; 95% CI 0.0 to 0.9) or to suffer from side effects (five RCTs, n = 156: RR 0.3; 95% CI 0.2 to 0.6).

A non-blinded randomised trial from Panama176 [EL = 1+] that compared labetalol with hydralazine included 200 women (100 in each arm) with severe hypertension (blood pressure of 160/110 mmHg or higher), at 24 weeks of gestation or later with no concurrent antihypertensive therapy. Labetalol was given intravenously: 20 mg bolus, followed by 40 mg if not effective within 20 minutes, followed by 80 mg every 20 minutes up to a maximum dose of 300 mg (five doses). Hydralazine was given intravenously: 5 mg slow bolus and repeated every 20 minutes up to a maximum of five doses. The study showed no statistically significant differences between the two drugs either in the effectiveness of hypertension control or in the appearance of adverse effects.

Calcium-channel blockers versus hydralazine

The Cochrane review174 [EL = 1++] included six RCTs (n = 313) that compared calcium-channel blockers with hydralazine. Women treated with calcium-channel blockers were statistically significantly less likely to have persistent high blood pressure than those treated with hydralazine (five RCTs, n = 263: RR 0.33; 95% CI 0.15 to 0.70). No other statistically significant differences were found.

The meta-analysis175 [EL = 1++] included nine RCTs (n = 619) that compared calcium-channel blockers with hydralazine. Babies born to women treated with calcium-channel blockers were statistically significantly less likely to have fetal heart rate decelerations than those born to women treated with hydralazine (six RCTs, n = 360: RR 0.2; 95% CI 0.1 to 0.6). No other statistically significant differences were found.

Nifedipine versus hydralazine

A non-blinded quasi-randomised trial177 [EL = 1−] from Ghana compared nifedipine with hydralazine. Women were numbered as they attended, with odd-numbered women joining the nifedipine group and even-numbered women joining the hydralazine group. The study included 79 women with severe pre-eclampsia (blood pressure of 160/110 mmHg or higher and proteinuria 1+ or more) who were at 28 weeks of gestation or later. Nifedipine was given sublingually (10 mg capsule) to 49 women. This was repeated every 30 minutes if blood pressure remained above 160/110 mmHg. After that, 10 mg tablets were given orally every 6–8 hours until delivery. Hydralazine was given intravenously (5 mg bolus) and was repeated at intervals determined by blood pressure measurements. When diastolic pressure stabilised at around 90–100 mmHg, 20–80 mg hydralazine tablets in divided doses were administered until delivery. The study showed that women on nifedipine were statistically significantly less likely to develop persistent high blood pressure than women treated with hydralazine (RR 0.28; 95% CI 0.11 to 0.71). No other statistically significant results were found.

Isradipine versus hydralazine

A small non-blinded quasi-randomised trial178 [EL = 1−] from Jamaica included 39 women with severe pre-eclampsia (blood pressure of 160/110 mmHg or higher, proteinuria 1+ or more) who were at 28 weeks of gestation or later. Isradipine was infused at 0.15 g/kg per minute* over 6 hours to a total maximum dose of 2.8 mg for 20 women. When diastolic pressure was controlled below 100 mmHg, slow-release tablets were started (5 mg, twice a day). Hydralazine was infused at 2 mg/kg/hour to a maximum dose of 20 mg, followed by oral alpha-methyldopa 500 mg three times a day for 19 women. The study only reported one outcome, caesarean section, which showed no statistically significant difference between the two groups.

Ketanserin versus hydralazine

The Cochrane review174 [EL = 1++] included four RCTs (n = 200) that compared ketanserin with hydralazine. Women treated with ketanserin were statistically significantly more likely to have persistent high blood pressure than those treated with hydralazine (three RCTs, n = 180: RR 4.79; 95% CI 1.95 to 11.73). However, they were statistically significantly less likely to suffer adverse effects from the drug (three RCTs, n = 120: RR 0.32; 95% CI 0.19 to 0.53) or to develop HELLP syndrome (one RCT, n = 44: RR 0.20; 95% CI 0.05 to 0.81). No other statistically significant differences were found.

The meta-analysis175 [EL = 1++] included four RCTs (n = 190) that compared ketanserin with hydralazine. Women treated with ketanserin were statistically significantly less likely to suffer from adverse effects than those treated with hydralazine (two RCTs, n = 64: RR 0.4; 95% CI 0.2 to 0.7). No other statistically significant differences were found.

Urapidil versus hydralazine

The Cochrane review174 [EL = 1++] included two RCTs (n = 59) that compared urapidil with hydralazine. No statistically significant differences were found.

The meta-analysis175 [EL = 1++] included two RCTs (n = 59) that compared urapidil with hydralazine. No statistically significant differences were found.

Epoprostenol versus hydralazine

The Cochrane review174 [EL = 1++] included one RCT (n = 47) that compared epoprostenol with hydralazine. No statistically significant differences were found.

The meta-analysis175 [EL = 1++] included one RCT (n = 47) that compared epoprostenol with hydralazine. No statistically significant differences were found.

Labetalol versus calcium-channel blockers

The Cochrane review174 [EL = 1++] included one RCT (n = 60) that compared labetalol with nicardipine. No statistically significant differences were found.

A double-blind RCT179 [EL = 1+] (n = 50) from the USA compared labetalol with nifedpine (n = 25 in each group). Women at 24 weeks of gestation of later with severe pre-eclampsia or chronic hypertension with superimposed pre-eclampsia, either intrapartum (n = 29) or within 24 hours postpartum (n = 21), were included. Severe hypertension was defined as sustained systolic blood pressure of 170 mmHg or higher or diastolic blood pressure of 105 mmHg or higher on repeat measurements 15 minutes apart. Women were randomly assigned to receive either nifedipine or labetalol. Nifedipine 10 mg was give orally with repeated doses of 20 mg every 20 minutes up to a maximum of five doses. Labetalol was given intravenously (20 mg) followed by escalating doses of 40 mg then 80 mg up to a maximum of five doses. The study showed no statistically significant differences in side effects, Apgar score less than 7 at 5 minutes or umbilical artery pH less than 7.0 between the two groups.

Labetalol versus methyldopa

The Cochrane review174 [EL = 1++] included one RCT (n = 74) that compared labetalol with methyldopa. No statistically significant differences were found.

Labetalol versus diazoxide

The Cochrane review174 [EL = 1++] included one RCT (n = 90) that compared labetalol with diazoxide. Women treated with labetalol were statistically significantly less likely to have maternal hypotension than those treated with diazoxide (one RCT, n = 90: RR 0.06; 95% CI 0.00 to 0.99). No other statistically significant differences were found.

Nitrates versus magnesium sulphate

The Cochrane review174 [EL = 1++] included one RCT (n = 36) that compared nitrates with magnesium sulphate. No statistically significant differences were found.

Nifedipine versus chlorpromazine

The Cochrane review174 [EL = 1++] included one RCT (n = 60) that compared nifedipine with chlorpromazine. No statistically significant differences were found.

Nifedipine versus prazosin

The Cochrane review174 [EL = 1++] included one RCT (n = 130) that compared nifedipine with prazosin. No statistically significant differences were found.

Nimodipine versus magnesium sulphate

The Cochrane review174 [EL = 1++] included two RCTs (n = 1683) that compared nimodipine with magnesium sulphate. Women treated with nimodipine were statistically significantly less likely to develop persistent high blood pressure than those treated with magnesium sulphate (one RCT, n = 1650: RR 0.84; 95% CI 0.76 to 0.93). For specific side effects, women treated with nimodipine were statistically significantly less likely to report ‘flushing’ than those treated with magnesium sulphate (one RCT, n = 1650: RR 0.22; 95% CI 0.12 to 0.40). No other statistically significant differences were found.

Diazoxide versus hydralazine

An RCT180 [EL = 1+] from Australia compared diazoxide with hydralazine (n = 97, 50 versus 47). Women requiring intravenous antihypertensive treatment (97 antenatal period, 27 postnatal period) were randomised to receive either diazoxide (15 mg boluses every 3 minutes until pressure was controlled or 300 mg was given) or hydralazine (5 mg boluses every 20 minutes for up to three doses). Four women in each group were prescribed two oral medications before and after the administration of intravenous medications. The authors reported 24 drug administration protocol violations. The study showed no statistically significant differences between the two groups.

Nitroglycerine versus nifedipine

A double-blind RCT181 [EL = 1+] from Mexico compared nitroglycerine with nifedipine (n = 32, 16 each arm). Women at 24 weeks of gestation or later with uncomplicated severe pre-eclampsia and with no history of chronic hypertension, use of antihypertensive therapy or life-threatening fetal heart-rate changes were eligible to enter the trial. Thirty-two eligible women were randomly allocated to receive either nitroglycerine infusion (5 micrograms/minute) with increases in dose of 5 micrograms/minute every 5 minutes or nifedipine capsules (10 mg) every 30 minutes. Both groups received a loading dose of magnesium sulphate 4 g/250 ml dextrose 5% in water (D5W) intravenously, followed by an intravenous infusion of 1 g/hour for up to 8 hours postpartum. The study showed no statistically significant differences in side effects, caesarean section, post-delivery bleeding above 1000 ml or Apgar score less than 7 at 1 minute and 5 minutes between the two groups.

Evidence statement

A Cochrane systematic review and a published meta-analysis considered the effectiveness of antihypertensives for treatment of severe hypertension. [EL = 1++] Both were based on a large number of studies, although the emphasis of the analyses differed between the two; the Cochrane systematic review compared pairs of antihypertensive agents, whereas the meta-analysis focused specifically on comparisons between hydralazine and other antihypertensive agents.

Labetalol versus hydralazine

The Cochrane review [EL = 1++] showed no statistically significant differences between the two drugs in the primary and secondary outcomes set by the GDG.

The meta-analysis [EL = 1++] showed that women treated with labetalol were statistically significantly more likely to develop persistent high blood pressure than those treated with hydralazine. However, they were less likely to have maternal hypotension and suffer from side effects.

The individual RCT [EL = 1+] showed no differences between the two drugs in primary and secondary outcomes.

Calcium-channel blockers versus hydralazine

Both the Cochrane review [EL = 1++] and an individual extra RCT [EL = 1−] showed that women treated with calcium-channel blockers were statistically significantly less likely to develop persistent high blood pressure than those treated with hydralazine.

The meta-analysis [EL = 1++] showed that babies of women treated with calcium-channel blockers were statistically significantly less likely to have fetal heart decelerations than those treated with hydralazine. No other statistically significant results were found.

Ketanserin versus hydralazine

The Cochrane review [EL = 1++] showed that women treated with ketanserin were statistically significantly more likely to develop persistent high blood pressure but were less likely to have side effects or develop HELLP syndrome than those treated with hydralazine.

The meta-analysis [EL = 1++] showed that women treated with ketanserin were statistically significantly less likely to have side effects. No other results were statistically significantly different between the two groups.

Urapidil versus hydralazine

Both the Cochrane review [EL = 1++] and the meta-analysis [EL = 1++] showed no statistically significant differences between the two groups in the primary and secondary outcomes.

Epoprostenol versus hydralazine

Both the Cochrane review [EL = 1++] and the meta-analysis [EL = 1++] showed no statistically significant differences between the two groups in the primary and secondary outcomes.

Labetalol versus calcium-channel blockers

Both the Cochrane review and an extra individual RCT [EL = 1+] showed no statistically significant differences between the two groups in the primary and secondary outcomes.

Labetalol versus diazoxide

The Cochrane review showed that women treated with labetalol were statistically significantly less likely to develop hypotension than those treated with methyldopa. No other statistically significant differences were found.

Labetalol versus methyldopa

The Cochrane review showed no statistically significant differences between the two groups in the primary and secondary outcomes.

Nitrates versus magnesium sulphate

The Cochrane review showed no statistically significant differences between the two groups in the primary and secondary outcomes.

Nifedipine versus chlorpromazine

The Cochrane review showed no statistically significant differences between the two groups in the primary and secondary outcomes.

Nifedipine versus prazosin

The Cochrane review showed no statistically significant differences between the two groups in the primary and secondary outcomes.

Nimodipine versus magnesium sulphate

The Cochrane review showed that women treated with nimodipine were statistically significantly less likely to develop persistent high blood pressure than those treated with magnesium sulphate. They were also less likely to suffer from ‘flushing’ as a side effect. No other statistically significant differences were found.

Diazoxide versus hydralazine

Individual RCT [EL = 1+] showed no statistically significant difference in primary and secondary outcomes between the two groups.

Nitroglycerine versus nifedipine

Individual RCT [EL = 1+] showed no statistically significant difference in primary and secondary outcomes between the two groups.

GDG interpretation of the evidence

There are no placebo controlled trials of antihypertensive treatment in women with severe pre-eclampsia in a critical care setting to inform the GDG but the consensus was that lowering blood pressure in women with severe hypertension is necessary. There did not appear to be any evidence that one particular antihypertensive agent was preferable in lowering blood pressure or in adverse outcomes for the mother or the fetus.

The GDG have recommended the commonly used antihypertensive regimens. There is no clear advantage in the route of delivery of antihypertensive therapy in the trials but the GDG agreed that route of administration could be oral or intravenous for labetalol, oral for nifedipine and intravenous for hydralazine.

Labetalol is the only drug licensed for the treatment of hypertension in pregnancy.

The side effect profile for these drugs was similar with no drug showing a clear advantage in minimising side effects. However, there is some advantage of labetalol over hydralazine for all maternal side effects, but the overall numbers in the studies was small.

Preloading or co-administration using no more than 500 ml of intravenous crystalloid fluid reduces the risk of sudden severe hypotension seen with intravenous hydralazine and may be considered prior to birth. Although there are few data on pulmonary oedema in the trials the main indication for the prevention of sudden hypotension is protection of the fetal circulation. There is less justification for fluid loading following birth.

Overall the cost of treatment was considered by the GDG. Although there is little difference between the costs of different antihypertensives, oral administration is likely to be cheaper than intravenous administration. The GDG noted that the mode of administration would depend on the condition of the woman, but where feasible oral administration should be preferred to intravenous administration because it is likely to be cost effective.

The evidence is not available to support a specific target blood pressure, nor the time to achieve that blood pressure. The GDG consensus was to avoid a rapid and precipitate fall in the maternal blood pressure and to closely observe the woman for side effects and response to treatment. The GDG considered a fall in blood pressure to 150/80–100 mmHg appropriate with maintenance of the blood pressure at this level to avoid placental underperfusion.

Recommendations

Treat women with severe hypertension who are in critical care during pregnancy or after birth immediately with one of the following:

In women with severe hypertension who are in critical care, monitor their response to treatment:

Consider using up to 500 ml crystalloid fluid before or at the same time as the first dose of intravenous hydralazine in the antenatal period.

In women with severe hypertension who are in critical care, aim to keep systolic blood pressure below 150 mmHg and diastolic blood pressure between 80 and 100 mmHg.

This guideline assumes that prescribers will use a drug’s summary of product characteristics (SPC) to inform decisions made with individual patients. Drugs for which particular attention should be paid to the contraindications and special warnings during pregnancy and lactation are marked with † and detailed in Section 1.6.

Research recommendation

What is the most clinically effective antihypertensive agent for severe pre-eclampsia in a critical care setting?

Why this is important

The choice of antihypertensive treatment in severe hypertension in the critical care setting has evolved historically rather than scientifically and there are few useful comparisons. Dosage and route of administration vary, as does use of different routes or doses from those shown to be effective in trials.

Effective and safe control of severe hypertension is the most important aspect of critical care management, as the main cause of maternal death is the consequence of poorly controlled hypertension. Randomised controlled trials should evaluate antihypertensive treatments (labetalol, nifedipine and hydralazine) for women with severe hypertension in pregnancy in the critical care setting. Comparisons should be made between the different antihypertensives, with assessment against outcomes such as persistence of severe hypertension after completion of therapy or by the need for additional treatment, maternal side effects and the effect on the fetus and baby.

10.4. Corticosteroids for fetal lung maturation

Clinical effectiveness

A Cochrane systematic review investigated the effect of antenatal corticosteroids for accelerating fetal lung maturation in women at risk of preterm birth.182 [EL = 1++] A subgroup analysis of the review presented data for women with hypertensive syndromes in pregnancy. The review assessed all RCTs comparing antenatal corticosteroid administration (betamethasone, dexamethasone or hydrocortisone) with placebo or no treatment given to women before anticipated preterm birth. Quasi-randomised trials were excluded. Trials that tested the effect of corticosteroid along with other co-interventions were also excluded.

Five RCTs were included in the ‘women with hypertension syndromes in pregnancy’ subgroup analysis. One trial (n = 220) included only women with severe pre-eclampsia. The other trials included all women with preterm birth but with results for those with hypertension in pregnancy syndromes reported separately. Methods of randomisation were properly described in two of these trials but not stated in the other three.

Babies from pregnancies complicated by hypertension syndromes treated with corticosteroids had a statistically significantly reduced risk of neonatal death (two RCTs, n = 278 babies; RR 0.50; 95% CI 0.29 to 0.87), respiratory distress syndrome (five RCTs, n = 382 babies; RR 0.50; 95% CI 0.35 to 0.72) and cerebroventricular haemorrhage (two RCTs, n = 278 babies; RR 0.38; 95% CI 0.17 to 0.87). They were also statistically significantly less likely to need mechanical ventilation (one RCT, n = 200 babies: RR 0.62; 95% CI 0.41 to 0.91) or to have systemic infection in the first 48 hours of life (one RCT, n = 200 babies: RR 0.46; 95% CI 0.26 to 0.84). In pregnancies complicated by hypertension syndromes, no statistically significant differences between groups treated with antenatal corticosteroids and controls were reported for combined fetal and neonatal death, fetal death, birthweight, chorioamnionitis or puerperal sepsis. The Cochrane review did not report any direct comparisons between different types of corticosteroids (betamethasone, dexamethasone and hydrocortisone).

A large non-randomised retrospective study has suggested that babies exposed to betamethasone antenatally have less neonatal cystic periventricular leucomalacia than those exposed to antenatal dexamethasone.183 [EL = 2−] Another historical cohort study reported a statistically significant reduction in the number of neonatal deaths with the use of dexamethasone compared with betamethasone (OR 1.66; 95% CI 1.07 to 2.57; P < 0.05).184 [EL = 2−]

Evidence statement

A Cochrane review [EL = 1++] showed that antenatal corticosteroids in women with hypertensive syndromes statistically significantly reduced the risk of neonatal death, respiratory distress syndrome and cerebroventricular haemorrhage. Babies of women treated with corticosteroids were also less likely to need mechanical ventilation or have infections in the first 48 hours of life.

Two retrospective studies [EL = 2−] showed that betamethasone was associated with fewer neonatal adverse effects (neonatal deaths or cystic periventricular leucomalacia) than dexamethasone.

GDG interpretation of the evidence

There is good evidence to suggest that the use of steroids antenatally in pregnancies complicated by hypertensive disorders will enhance fetal lung maturity and reduce the incidence of the complications of preterm birth, especially respiratory distress syndrome, when the pregnancy is at less than 34 weeks. The evidence is less clear when the pregnancy is between 34 and 37 weeks, but the GDG considers that there is likely to be benefit in this group of women. The preferred steroid is two doses of betamethasone 12 mg administered intramuscularly 24 hours apart, with betamethasone being preferred over dexamethasone because it is associated with fewer neonatal adverse effects (neonatal death and cystic periventricular leucomalacia); the two drugs are similarly priced and so the recommendation to use betamethasone is likely to be cost effective.

In formulating the recommendations, the GDG noted the results of the Antenatal Steroid for Term Elective Caesarean Section (ASTECS) study, which showed that babies born after 37 weeks by elective caesarean section also benefit from antenatal corticosteroid administration.185

Recommendation

If birth is considered likely within 7 days in women with pre-eclampsia:

  • give two doses of betamethasone * 12 mg intramuscularly 24 hours apart in women between 24 and 34 weeks
  • consider giving two doses of betamethasone* 12 mg intramuscularly 24 hours apart in women between 35 and 36 weeks.
*

In this guideline, drug names are marked with an asterisk if they do not have UK marketing authorisation for the indication in questionat the time of publication (August 2010). Informed consent should be obtained and documented.

10.5. Corticosteroids to manage HELLP syndrome

Clinical effectiveness

Corticosteroids have been used in women (antepartum and postpartum) diagnosed with HELLP syndrome. One Cochrane systematic review186 [EL = 1++] studied two comparisons: dexamethasone plus standard treatment versus standard treatment alone, and dexamethasone versus betamethasone. One additional RCT187 [EL = 1+] compared dexamethasone with placebo while another RCT188 [EL = 1+] compared dexamethasone with betamethasone.

Dexamethasone plus standard treatment versus standard treatment alone

A Cochrane review investigated the effects of corticosteroids in women with HELLP syndrome (diagnosed clinically and by biochemical parameters) during pregnancy or shortly after delivery.186 [EL = 1+] All RCTs and trials that used pseudo-randomised methods, such as alternate allocation, were included. Five studies were included, three of which employed adequate randomisation and allocation concealment methods. However, blinding was not described in any. There was significant loss to follow-up in one study. Only 25 out of the original 40 participants randomised were accounted for in the results section. Intention to treat analysis was not performed in this study. The other studies had no loss to follow-up.

No statistically significant differences were found in maternal death or neonatal deaths. No cases of maternal morbidity were reported in either group (liver haematoma or rupture, pulmonary oedema, kidney failure or placental abruption). There were no statistically significant differences in the likelihood of having perinatal intraventricular haemorrhage, respiratory distress syndrome or retrolental fibroplasias. No intracerebral haemorrhagic events or necrotising enterocolitis were recorded.

In secondary outcomes, no statistically significant differences were found in postpartum sepsis, caesarean section or increase in platelet count over 48 hours. However, there were statistically significant differences in the mean number of hospital stay days post-randomisation (one RCT, n = 30: WMD −4.50 days; 95% CI −7.13 to −1.87 days) and time interval from randomisation to delivery (one RCT, n = 25: WMD 26.00 hours; 95% CI 17.17 to 34.83 hours), both of which were in favour of women allocated to dexamethasone treatment.

A Colombian double-blind RCT compared the efficacy of dexamethasone with placebo for the treatment of women (pregnant or puerperal) who developed hypertension during pregnancy and met the criteria for HELLP syndrome classes 1 and 2.187 [EL = 1+] One hundred and thirty-two women were randomised to receive either dexamethasone (n = 66) or placebo (n = 66). The baseline characteristics of women in the two groups were comparable. Randomisation was done by the use of stratified and random permuted blocks of four, and concealment of allocation was ensured by using opaque envelopes. Dexamethasone 10 mg was given intravenously every 12 hours until delivery and three further times after delivery. Women in the placebo group were given sterile water at a similar schedule.

There was no statistically significant difference in maternal mortality between the two groups (three of 66 versus one of 66: RR 3.0; 95% CI 0.32 to 28.1). There were also no statistically significant differences between the two groups in the maternal complications of acute kidney failure, oliguria, pulmonary oedema, eclampsia, infections or the need for platelets or plasma transfusion. The mean duration of hospitalisation of women was not statistically significantly different between the two groups. No statistically significant differences were found in the time to recovery of platelet counts (hazard ratio 1.2; 95% CI 0.8 to 1.8), LDH (hazard ratio 0.9; 95% CI 0.5 to 1.50) or AST (hazard ratio 0.6; 95% CI 0.4 to 1.1).

The results related to both pregnant and puerperal groups. Stratified analysis showed no differences in the occurrence of complications, recovery of laboratory parameters, transfusion need or duration of hospitalisation.

Dexamethasone versus betamethasone

There was only one study from the Cochrane review described above186 [EL = 1+] that compared dexamethasone with betamethasone (n = 40). No maternal death occurred. Perinatal mortality was not statistically significantly different between the two groups (RR 0.95; 95% CI 0.15 to 6.08). There were no cases of liver haematoma or rupture, pulmonary oedema or placental abruption in either group. There was a statistically significant difference in maternal oliguria (RR 0.06; 95% CI 0.00 to 0.93) in favour of women randomised to dexamethasone. No statistically significant differences were found in neonates’ need for ventilatory support or having respiratory distress syndrome. No cases of intracerebral haemorrhage or necrotising enterocolitis were recorded.

There were statistically significant differences in favour of women allocated to dexamethasone in the adjusted time-average change from baseline in the following secondary outcomes: the mean arterial pressure decrease (WMD −7.50 mmHg; 95% CI −8.37 to −6.63 mmHg), the mean increase in urinary output (WMD 24.80 ml/day; 95% CI 19.58 to 30.02 ml/day), the mean increase in platelet count (WMD 8.10 × 109/litre; 95% CI 6.23 to 9.97 × 109/litre), the mean decrease in LDH activity (WMD −4.20 U/litre; 95% CI −88.22 to −20.18 U/litre) and the mean decrease in AST activity (U/L) (WMD −30.30 U/litre; 95% CI −36.06 to −24.54 U/litre).

The number of women needing acute antihypertensive therapy in the dexamethasone group differed statistically significantly compared with those allocated to betamethasone (RR 0.29; 95% CI 0.12 to 0.73).

There were no statistically significant differences between the two groups with regard to the number of neonates with a Apgar score less than 7 at 5 minutes, neonatal sepsis, neonatal hyperbilirubinaemia or mean time to discharge.

An RCT in the USA compared the efficacy of dexamethasone with betamethasone for the treatment of women with HELLP syndrome first manifesting itself in the postpartum period.188 [EL = 1+] Women who developed HELLP syndrome or any other manifestation of pre-eclampsia in the antepartum period were excluded. Thirty-six women were randomised to receive either dexamethasone 10 mg intravenously every 12 hours (n = 18) or betamethasone 12 mg intramuscularly every 24 hours (n = 18). The baseline characteristics of women in the two groups were comparable except for LDH level, which was statistically significantly higher in the dexamethasone group (1831.7 ± 1140.6 U/litre versus 1193.6 ± 496.4 U/litre; P < 0.05). Randomisation was by sequentially numbered sealed opaque envelopes constructed from a random number table.

The time to discharge from the obstetric recovery room was not statistically significant between groups. Reduction in mean arterial blood pressure was more pronounced in the dexamethasone group compared with the betamethasone group (−15.3 ± 1.4 mmHg versus −7.5 ± 1.4 mmHg; P < 0.01). Women in the dexamethasone group required statistically significantly less antihypertensive treatment than the betamethasone group (one of 18 versus nine of 18: RR 0.11; 95% CI 0.02 to 0.79) and also had a decreased need for readmission to the obstetric recovery room (none of 18 versus four of 18: RR 0.11; 95% CI 0.006 to 1.924).

Evidence statement

In women with HELLP syndrome during pregnancy or shortly after delivery, a Cochrane review [EL = 1++] showed that the use of corticosteroids was no different from placebo in terms of maternal or neonatal complications. However, women who were allocated to corticosteroids stayed in hospital for statistically significantly shorter periods and had statistically significantly shorter time intervals between randomisation and delivery. An RCT [EL = 1+] also showed no difference in maternal or neonatal complications between women treated with corticosteroids and placebo. Hospital duration and time to recovery for platelets, LDH and AST were also similar in both groups. The results were found in both pregnant and puerperal groups.

When comparing dexamethasone with betamethasone use in women with HELLP syndrome (antenatally or postnatally), a Cochrane review [EL = 1+] showed no statistically significant difference in the two groups in terms of maternal or neonatal complications. However, those treated with dexamethasone had statistically significantly higher time-average change in arterial pressure decrease, urinary output increase, platelet count increase, and LDH and AST decrease. They were also statistically significantly less likely to need acute antihypertensive therapy. An RCT [EL = 1+] in women with postpartum HELLP syndrome showed that those treated with dexamethasone were more likely to have reduction in arterial blood pressure than those treated with betamethasone. They were also less likely to require antihypertensive treatment or to need readmission to the obstetric recovery room.

GDG interpretation of the evidence

There is high-quality evidence that corticosteroids used in the management of HELLP syndrome do not improve any clinically important outcomes either antenatally or postnatally. Two studies into the use of corticosteroids in HELLP syndrome had different conclusions with respect to antenatal and postnatal stays, which may be an important clinical outcome.

Recommendation

Do not use dexamethasone or betamethasone for the treatment of HELLP syndrome.

Research recommendation

Does the use of dexamethasone in HELLP syndrome have clinical utility?

Why this is important

HELLP syndrome is a variant of severe pre-eclampsia where hypertension is less marked but where there is severe involvement of both the liver and the coagulation system. In addition to the usual complications of severe pre-eclampsia there is a risk of liver failure and bleeding.

Studies carried out to determine if steroid injections improve laboratory results have been relatively small and have not clearly shown clinically important benefits. Randomised controlled trials should be carried out in women with HELLP syndrome to assess the clinical utility of dexamethasone compared with placebo control based on outcomes associated with HELLP syndrome (delay to birth; time to hospital discharge following birth; severe maternal complications; serious neonatal complications and long-term outcomes).

10.6. Fluid balance and volume expansion

Clinical effectiveness

An RCT conducted in the Netherlands investigated the use of a volume expansion protocol in women with severe hypertensive disorders of pregnancy (severe pre-eclampsia, HELLP syndrome, and concomitant IUGR) who presented with a viable singleton pregnancy at a gestational age between 24 and 34 weeks.189 [EL = 1+] Exclusion criteria included severe fetal distress or lethal fetal congenital abnormalities, language difficulties, or if plasma volume expansion had already been given.

Women were randomly allocated by use of computer within two bands of gestational age (between 24+0 and 29+6 weeks, and between 30+0 and 33+6 weeks) into either the volume expansion group (n = 111) or the no volume expansion group (n = 105). The software concealed the group allocation until the woman’s details had been entered. Reasons for leaving the study were reported. Baseline characteristics of women in two groups were comparable.

The volume expansion group received 250 ml of 6% hydroxy-ethylstarch (HES) over 4 hours twice a day. Antihypertensives (intravenous ketanserine) were used to achieve diastolic blood pressure of 85–95 mmHg. Additional medication (oral labetalol, methyldopa and nifedipine and occasionally intravenous dihydralazine) was used when necessary. Restricted amounts of sodium chloride 0.9% were infused with medications in between the infusions of HES. Fluid treatment was discontinued if clinical signs of pulmonary oedema were observed.

In the no volume expansion group, antihypertensives (methyldopa) were used to achieve diastolic blood pressure of 95–105 mmHg. Additional medication (oral labetalol, nifedipine and intravenous ketanserine and occasional intravenous dihydralazine) was used when necessary. Restricted amounts of sodium chloride 0.9% were infused with intravenous medication.

Magnesium sulphate was used for preventing and treating eclampsia. One course of intramuscular betamethasone (two doses of 11.4 mg with a 24 hour interval in between) was given when delivery was considered imminent before 32 weeks of gestation.

There was a trend towards a longer pregnancy in the control group (by 10.5 days; 95% CI 0.2 to 440 days) compared with the treatment group (7.4 days; 95% CI 0.1 to 35 days; P = 0.054). There was no difference in fetal or postnatal death. Liveborn neonates for women in the volume expansion group were statistically significantly more likely to need ventilation or respiratory support (78 of 98 versus 60 of 98: RR 1.3; 95% CI 1.08 to 1.57). There was no statistically significant difference in major maternal morbidity but there were statistically significantly more caesarean sections in the treatment group (96 of 98 versus 88 of 98: RR 1.10; 95% CI 1.02 to 1.17). Neither neurological scores nor composite neonatal morbidity differed statistically significantly (neonatal morbidities: respiratory distress syndrome, chronic lung disease, intraventricular haemorrhage, progressive ventricular dilation, necrotising enterocolitis, sepsis/meningitis or patent ductus arteriosus). However, episodes of neonatal morbidity were statistically significantly higher in the treatment group (93 of 98 versus 80 of 98: RR 1.26; 95% CI 1.05 to 1.30).

Babies (n = 172) born to women in the RCT discussed above were followed up for a year (n = 82 treatment, n = 90 control).190 [EL = 1+] The follow-up study assessed the mental and psychomotor development of the babies using the Touwen Scale and the Bayley Scales of Infant Development II that includes two standardised development indices: the Mental Development Index (MDI) and the Psychomotor Development Index (PDI). Adverse neurodevelopmental infant outcome was defined as an MDI/PDI score < 70 and/or an abnormal Touwen score. The mean score was not different between the randomisation groups on any of these scales. There was no difference in the number of cases shown as moderately or severely delayed by the Bayley test and nor was there a difference in the cases shown as suspect or abnormal in the Touwen test.

A Dutch case control study compared the results of nulliparous women with severe pre-eclampsia who were treated with a volume expansion protocol with those receiving no volume expansion treatment.191 [EL = 2+] Women with known pre-existing hypertensive, cardiac or kidney disease were excluded. Cases (n = 57) and controls (n = 57) were recruited from two medical centres in the Netherlands and matched retrospectively according to gestational age at admission (maximum 1 week difference). Characteristics at admission for the two groups were comparable.

The volume expansion group was admitted to ICU for central haemodynamic monitoring. If the pulmonary capillary wedge pressure (PCWP) was less than 10 mmHg and/or the cardiac index was less than 3.5 litres/minute per m2, women received intravenous pasteurised plasma (250 ml/hour) to maintain the PCWP at 10–12 mmHg and a cardiac index of 3.5–4.6 litres/minute per m2. If the cardiac index remained below 3.5 and the diastolic blood pressure above 100 mmHg, women received intravenous dihydralazine (1 mg/hour), followed by hourly increments of 1 mg. Methyldopa was used when the desired reduction was not obtained. After stabilisation, women were transferred to the ward where plasma volume expansion and antihypertensive treatments were continued: bed rest, continuous monitoring, and diazepam where eclampsia was thought to be imminent or convulsions occurred; diet was unrestricted. Women in the control group had bed rest, no intravenous fluids, and a diet with less than 400 mg sodium per 24 hours. Women with symptoms of headache, upper abdominal pain or visual disturbances received phenobarbital 30 mg orally three times a day. Antihypertensive medication was given when diastolic blood pressure reached and remained above 115 mmHg (intravenous dihydralazine). Intravenous magnesium sulphate was administered as anticonvulsant treatment.

No statistically significant differences were found in prolongation of pregnancy between the two groups. SGA infants (less than 2.3 percentile) were statistically significantly less frequent in the volume expansion group than in the control group (five of 57 versus 19 of 57: OR 0.19; 95% CI 0.07 to 0.56). However, babies born to women in the volume expansion group were statistically significantly more likely to need artificial ventilation (27 of 57 versus eight of 57: OR 5.51; 95% CI 2.22 to 13.70) and to have patent ductus arteriosus (nine of 57 versus two of 57: OR 5.16; 95% CI 1.06 to 25.04). Other neonatal complications were not statistically significantly different between the two groups. For maternal complications, no statistically significant differences were found for HELLP syndrome, placental abruption, pulmonary oedema, postpartum cardiomyopathy or postpartum renal insufficiency.

Evidence statement

In women with severe hypertension during pregnancy, an RCT [EL = 1+] that compared women who received volume expansion treatment with those who received no volume expansion treatment showed no statistically significant difference in major maternal morbidity, but there were more caesarean sections in the treatment group. On a 1-year follow-up of the babies, no statistically significant differences were found in mental or psychomotor development of babies from the two groups. The use of volume expansion treatment was not statistically significantly different from the no volume expansion protocol in terms of fetal or postnatal death. Neither neurological scores nor composite neonatal morbidity differed statistically significantly between liveborn neonates for women from the two groups. However, episodes of neonatal morbidity were statistically significantly higher in the treatment group. Babies born to women in the treatment group were also statistically significantly more likely to need ventilation or respiratory support.

A case control study [EL = 2+] showed no statistically significant difference in prolongation of pregnancy between the two groups. For maternal complications, no statistically significant differences were found between the two groups. SGA infants were statistically significantly less frequent in the volume expansion group than in the control group. However, babies born to women in the volume expansion group were statistically significantly more likely to need artificial ventilation and to have patent ductus arteriosus. Other neonatal complications were not statistically significantly different between the two groups.

GDG interpretation of the evidence

The two studies reviewed both suggested that neonatal morbidity may be higher when maternal fluid expansion is used. In one study there was a reduction in the incidence of SGA babies. There were no obvious maternal advantages.

The Confidential Enquiry into Maternal Deaths in the UK reported six deaths in 1994–96 due to adult respiratory distress syndrome (ARDS) that appeared to be related to poor fluid management in women with eclampsia or pre-eclampsia.192 Recommendations made on the basis of these reported deaths advised that senior medical involvement and care was essential when intravenous fluids were being considered. This advice is thought to have resulted in the fact that by 2003–05 no deaths due solely to fluid mismanagement and ARDS were reported.192

The GDG’s view is that volume expansion (fluid loading) should be used only if hydralazine (a vasodilator) is the antenatal antihypertensive. Fluid loading in women taking hydralazine will help to reduce severe hypotension.

Recommendations

Do not use volume expansion in women with severe pre-eclampsia unless hydralazine is the antenatal antihypertensive.

In women with severe pre-eclampsia, limit maintenance fluids to 80 ml/hour unless there are other ongoing fluid losses (for example, haemorrhage).

10.7. Caesarean section versus induction of labour

Clinical effectiveness

Caesarean section without labour versus labour induction

A Nigerian RCT compared caesarean section with labour induction in primigravida with singleton cephalic presentation and antenatal or imminent eclampsia and a closed cervical os.193 [EL = 1−] Fifty women were randomised to have caesarean section (n = 25) or labour induction (n = 25).

Labour was induced using misoprostol (50 mg) and women were re-evaluated after 4 hours. If the woman went into labour, another 50 mg of misoprostol was inserted and the second stage of labour was shortened by the use of outlet forceps. If labour did not start, induction was considered to have failed and emergency caesarean section was offered. All women were sedated with intravenous diazepam and slow boluses of intravenous hydralazine if diastolic blood pressure was above 110 mmHg.

Misoprostol failure was recorded in four of 25 women (16%) and they were subsequently delivered by caesarean section. The mean duration of admission was statistically significantly longer in the caesarean section group (10.1 days versus 6.08 days; P = 0.05; no SD reported). There were no more maternal complications in the caesarean section group (eight of 25 versus two of 25: RR 4.0; 95% CI 0.94 to 17.00). Apgar scores at 1 minute and 5 minutes, babies’ admission to NICU, perinatal mortality and maternal mortality did not differ statistically significantly between the groups.

A retrospective cohort study in the USA looked at outcomes of infants born after labour induction compared with those delivered by caesarean section without labour.194 [EL = 2+] The study included 278 liveborn very low birthweight (750–1500 g) infants (n = 145 labour induction, n = 133 caesarean section without labour) delivered for women who had severe pre-eclampsia. Women received intramuscular magnesium sulphate for seizure prophylaxis and intravenous hydralazine for severe hypertension. No glucocorticoids were given for fetal lung maturation. Baseline characteristics for the women in the two groups were statistically significantly different in terms of age and nulliparity.

Both birthweight and gestational age were statistically significantly lower in the caesarean section group (birthweight: 1131 ± 232 g versus 1235 ± 185 g; P = 0.001, gestational age: 29.9 ± 2.3 weeks versus 30.8 ± 2.6 weeks; P = 0.004). After adjustment for birthweight and gestational age, logistic regression analysis showed the OR for Apgar score less than or equal to 3 at 5 minutes to be statistically significantly different (induction group: OR 6.1; 95% CI 1.1 to 32.2). The ORs for umbilical artery blood pH less than or equal to 7.0, respiratory distress syndrome, sepsis, intraventricular haemorrhage, seizures and neonatal deaths were not statistically significant.

Vaginal birth versus caesarean section after labour induction

An chart review study in the USA investigated outcomes of 306 women who underwent elective caesarean section (n = 161), caesarean section after labour induction (n = 75) and vaginal delivery after labour induction (n = 70).195 [EL = 3] Participants were women who had severe pre-eclampsia and with single liveborn babies (24–34 weeks of gestation). Maternal age, parity and gestational age at delivery were comparable between the groups.

No statistically significant differences were found after induction between caesarean section and vaginal delivery in Apgar score less than 7 at 5 minutes or endometritis. Total hospital stay was also no different between the two groups but, after excluding three women who had an unusually prolonged hospital stay (longer than 400 hours) for unrelated medical conditions (SLE nephritis in two women and sickle cell disease in the third), total hospital stay became statistically significantly higher in the caesarean section group (130.0 ± 41.1 hours versus 109.7 ± 44.3 hours; P = 0.005).

Evidence statement

When comparing caesarean section without labour with labour induction, an RCT [EL = 1−] showed no statistically significant difference in reported maternal or neonatal complications. However, women allocated to caesarean section stayed for statistically significantly longer periods in the hospital. A retrospective cohort study [EL = 2+] showed odds for Apgar score less than or equal to 3 at 5 minutes to be statistically significantly lower in the caesarean section group. However, the odds for neonatal complications including umbilical artery blood pH less than or equal to 7.0, respiratory distress syndrome, sepsis, intraventricular haemorrhage, seizures and neonatal deaths were not statistically significant.

When comparing vaginal birth after labour induction with caesarean section after labour induction, a chart review study [EL = 3] showed no difference between the two groups in reported outcomes (Apgar score less than 7 at 5 minutes and endometritis). Hospital stay, however, was statistically significantly longer in those who underwent caesarean section.

GDG interpretation of the evidence

Poor-quality small studies seemed to indicate little advantage to caesarean section and in one study women undergoing caesarean section had longer postnatal stays. However, it was felt that flaws in the studies available meant that there were no reliable data to inform the GDG and it was felt that mode of delivery would be best decided on both clinical circumstance and the woman’s preference.

Recommendation

Choose mode of birth for women with severe hypertension, severe pre-eclampsia or eclampsia according to the clinical circumstances and the woman’s preference.

10.8. Indications for referral to critical care levels

There are no studies into specific indications for care of women with severe hypertensive disorders during pregnancy in specific critical care settings.

The GDG has adapted existing definitions and guidance for critical care produced by the Intensive Care Society to reflect the range of disease severity in pre-eclampsia and gestational hypertension.

Recommendation

Offer women with severe hypertension or severe pre-eclampsia referral to the appropriate critical care setting using the following criteria:

Level 3 care
Level 2 careStep-down from level 3 or severe pre-eclampsia with any of the following complications:
Level 1 care

Adapted from Intensive Care Society, Standards and Guidelines 2002.

Footnotes

*

The authors reported that the dosage was 0.15 g/kg per minute over 6 hours, but this appears to be a typographical error and the results should therefore be treated with caution.

Copyright © 2011, Royal College of Obstetricians and Gynaecologists.

No part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK [www.cla.co.uk]. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

Cover of Hypertension in Pregnancy
Hypertension in Pregnancy: The Management of Hypertensive Disorders During Pregnancy.
NICE Clinical Guidelines, No. 107.
National Collaborating Centre for Women's and Children's Health (UK).
London: RCOG Press; 2010 Aug.

Download

NICE (National Institute for Health and Care Excellence)

PubMed Health Blog...

read all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...