Expectant management of early onset, severe pre-eclampsia: perinatal outcome
Correspondence: Dr D. R. Hall, Department of Obstetrics and Gynaecology, Tygerberg Hospital and University of Stellenbosch, P. O. Box 19081, Tygerberg, 7505 South Africa.
Objective To evaluate the perinatal outcome of expectant management of early onset, severe pre-eclampsia.
Design Prospective case series extending over a five-year period.
Setting Tertiary referral centre.
Population All women (n= 340) presenting with early onset, severe pre-eclampsia, where both mother and the fetus were otherwise stable.
Methods Frequent clinical and biochemical monitoring of maternal status with careful blood pressure control. Fetal surveillance included six-hourly heart rate monitoring, weekly Doppler and ultrasound evaluation of the fetus every two weeks. All examinations were carried out in a high care obstetric ward.
Main outcome measures Prolongation of gestation, perinatal mortality rate, neonatal survival and major complications.
Results A mean of 11 days were gained by expectant management. The perinatal mortality rate was 24/1000 (≥ 1000 g/7 days) with a neonatal survival rate of 94%. Multivariate analysis showed only gestational age at delivery to be significantly associated with neonatal outcome. Chief contributors to neonatal mortality and morbidity were pulmonary complications and sepsis. Three pregnancies (0.8%) were terminated prior to viability and only two (0.5%) intrauterine deaths occurred, both due to placental abruption. Most women (81.5%) were delivered by caesarean section with fetal distress the most common reason for delivery. Neonatal intensive care was necessary in 40.7% of cases, with these babies staying a median of six days in intensive care.
Conclusion Expectant management of early onset, severe pre-eclampsia and careful neonatal care led to high perinatal and neonatal survival rates. It also allowed the judicious use of neonatal intensive care facilities. Neonatal sepsis remains a cause for concern.
Many authors have emphasised the lack of uniformity in management approaches to severe pre-eclampsia1–3. The clinical course of severe pre-eclampsia may be associated with progressive deterioration in both maternal and fetal conditions. Thus, because delivery is the only way of arresting the disease, there is broad agreement on delivery in the presence of multi-organ dysfunction, fetal distress or once a gestation of 34 weeks has been reached. However, hypertensive disease is an important cause for delivery of very low birthweight babies4 and delivery at early gestations is associated with high perinatal mortality and morbidity resulting from prematurity5–7. In addition, recent research suggests that fetal lung maturity8, as well as fetal neurologic and physical development9, are not accelerated in pregnancies complicated by pre-eclampsia.
In 1989 Derham et al.10 demonstrated that the most important factor for survival of a healthy baby was gestational age, which was strongly correlated to birth-weight. In a case control study Oláh et al.11 showed fewer neonatal complications when women with early, severe pre-eclampsia were managed expectantly. Two randomised controlled trials have also shown that expectant management of women with early onset, severe pre-eclampsia improves the neonatal outcome of their babies12,13.
Visser et al.14 and Sibai15 have recently cautioned that the number of reported women whose early onset, severe pre-eclampsia was managed expectantly is limited. This study was planned to investigate the current perinatal outcome of the expectant management approach. The maternal outcome has been documented separately. The ultimate goal remains the safety of the mother and the delivery of a live infant who will not require intensive and prolonged neonatal care.
Over a five-year period (April 1992 to March 1997) all women with singleton pregnancies presenting with early onset (≥ 24 and < 34 weeks of gestation), severe pre-eclampsia, where both the mother and the fetus were otherwise stable, were studied. Exclusion criteria included major maternal complications, such as eclampsia, pulmonary oedema, haemolysis, elevated liver enzymes and low platelet syndrome (HELLP), placental abruption and severe renal impairment, or fetal distress.
The definitions of hypertension and proteinuria used were those put forward by Davey and MacGillivray16 and accepted by the International Society for the Study of Hypertension in Pregnancy. Severe hypertension was accepted as a diastolic blood pressure measurement of ≥ 120 mmHg on one occasion or a diastolic blood pressure measurement of ≥ 110 mmHg on two occasions, four hours or more apart. Blood pressure was carefully controlled to levels below 160/110 mmHg, using a stepwise approach with up to three oral anti-hypertensive agents, namely methyldopa, prazosin and nifedipine. Magnesium sulphate prophylaxis was not considered unless the woman developed eclampsia or was to be delivered.
The population served by the study hospital is comprised chiefly of persons of mixed racial origin from a low socioeconomic background. The ratio of neonatal intensive care beds to births in the hospital is 0.5/100017. Recruited women were admitted to a high care ward for intensive, noninvasive monitoring of the maternal and fetal status. Gestation was determined by means of last menstrual period, obstetric ultrasound or both. Not all women underwent an early (< 24 weeks) ultrasound scan. Maternal monitoring included four-hourly blood pressure measurement, clinical evaluation twice daily and daily urine tests for protein and glucose. A full blood count, renal function tests, liver function tests, and 24-hour urine collection for creatinine clearance and protein quantification, were all performed twice weekly.
Fetal viability was set at a gestation of ≥ 28 weeks with a minimum estimated mass (by ultrasound) of 800 g. The fetus was monitored by six-hourly nonstress tests (from viability), weekly Doppler velocimetry of the umbilical artery (beginning at 24 weeks) and ultrasound evaluation of growth and amniotic fluid every second week. Betamethazone (12 mg) was given intramuscularly to women at the end of the 27th week of gestation, repeated after 24 hours and given weekly thereafter until 33 weeks of gestation or delivery. Failure to control blood pressure or the development of major maternal or fetal complications, were indications for delivery. Women reaching a minimum gestation of 34 weeks without complications were delivered electively. The following fetal complications were regarded as indications for delivery: fetal distress on cardiotocograph (baseline variability < 5 over 60 minutes, repeated late decelerations or both); absent end diastolic flow velocity (Doppler ultrasound) before 26 weeks; reversed end diastolic flow velocity; and other problems such as intrauterine death.
Major neonatal complications included death, hyaline membrane disease grades III/IV, confirmed septicaemia, pneumonia, necrotising enterocolitis, bronchopulmonary dysplasia, convulsions, pulmonary haemorrhage, intraventricular haemorrhage grades III/IV and hypoxic ischaemic encephalopathy. These complications were identified retrospectively from detailed casenote contents after death or discharge.
Birth centiles based on the local population18 were used in the analysis. Data are expressed as mean with standard deviation and median with range. Differences in means were analysed using the two-tailed, Student's t test but for data not normally distributed the Mann-Whitney U test was used. The χ2 test was applied to qualitative variables with Yate's correction when the expected number was from 5 to 10, or with Fisher's method when it was < 5. A P value of less than 0.05 was regarded as significant. The study protocol was approved by the local ethics committee.
During the trial period 33,832 women were delivered in the study hospital, while 2390 (7.1%) women with any form (mild or severe, early or late onset) of pre-eclampsia were managed. Three hundred and forty women were entered into the study. The mean and median numbers of days gained (antepartum) by expectant management were 11 and 9, respectively, with a standard deviation of 7 and a range of 1–47 days. There were no maternal deaths and only three women (0.8%) were admitted to the adult intensive care unit. Fetal distress was the most common delivery indication (44.4%), with maternal reasons accounting for 25.9% of the deliveries. Fifty-four (15.9%) women reached the point of elective delivery, while combined maternal/fetal reasons and spontaneous labour accounted for the remaining delivery indications. Induction of labour was attempted in 103 (30.7%) of the women with a viable pregnancy. Overall 63 (18.5%) women were delivered vaginally and 277 (81.5%) by caesarean section.
Three hundred and thirty five (98.5%) babies were born alive, with 19 (5.6%) of these having an Apgar score < 6 at five minutes. The mean and median birth-weights were both 1400 g, with a standard deviation of 371 g and a range of 512–2585 g. The sex ratio was skewed towards the female gender (male babies, 124 (36.5%); female babies, 216 (63.5%)).
Three pregnancies (0.8%) were terminated prior to viability. In these cases a combination of complications (absent end diastolic velocity in the umbilical artery; severe growth retardation; haemolysis, elevated liver enzymes and low platelet syndrome; severe renal impairment and loss of blood pressure control) developed during the expectant management. Only two (0.5%) babies died in utero, the cause of death being sudden placental abruption in both cases. One intrauterine death occurred at 28 weeks and the other one day before planned induction at 34 weeks. Sixty-nine women (20.2%) in the study experienced placental abruption.
One hundred and thirty-four babies (40.7%) were admitted to the neonatal intensive care unit. These babies stayed a median number of six days with a range of 1–74 days. Of the babies transferred to the unit, 116 (86.5%) were ventilated. The median number of days of ventilation was four with a range of 1–68 days. The periods of time for which babies were admitted to intensive care are as follows: 1–7 days (59%); 8–14 days (25%); 15–21 days (7%); 22–28 days (1%); and 29–74 days (8%). The perinatal mortality rate was 24 per thousand births (≥ 1000 g/7 days) or 44.1 per thousand births (≥ 500 g/7 days) in the two categories. The contribution of intensive care at specific delivery gestations, neonatal survival to 28 days of age and overall perinatal survival are shown in Table 1. Included in this table under exclusions are the three terminations of nonviable pregnancies, two intrauterine deaths and four babies who were born alive but did not qualify for neonatal intensive care for the following reasons: two had a gestational age < 28 weeks; one died an hour after delivery before transfer to the neonatal intensive care unit; and one had a birth mass too low for the unit.
Table 1. Perinatal outcome and contribution of intensive care at specific periods of gestation (viability/neonatal intensive care (NNIC) set at 28 weeks with minimum mass 800 g). Values are given as n or n (%), unless otherwise indicated.
There were 10 early and 10 late neonatal deaths, while five deaths occurred during the post-neonatal period before discharge. The details of these losses are shown in Table 2. The major neonatal complications are documented in Table 3.
Table 2. Details of neonatal deaths. Values are given as n, unless otherwise indicated. NNIC = neonatal intensive care.
|Early neonatal deaths (days 1–7)|| || || |
| 28||764||1||Severe prematurity, no NNIC|
| 27||840||1||Severe prematurity, no NNIC|
| 28||1100||1||Severe pulmonary complications|
| 27||910||2||Severe prematurity, no NNIC|
| 28||1000||3||Intra-ventricular haemorrhage|
| 34||1380||3||Severe pulmonary complications|
| 28||860||4||Severe pulmonary complications|
| 31||1100||6||Congenital cardiac abnormality|
|Late neonatal deaths (days 8–28)|| || || |
| 27||1100||8||Severe pulmonary complications|
| 28||560||10||Growth retardation|
| 31||1322||14||Necrotising enterocolitis|
| 32||1110||14||Necrotising enterocolitis|
| 34||1680||17||Necrotising enterocolitis|
| 32||914||23||Necrotising enterocolitis|
|Infant deaths (day 29–discharge)|| || || |
| 29||760||39||Necrotising enterocolitis|
| 28||940||52||Necrotising enterocolitis|
Table 3. Major neonatal complications.
|With major complication||113||34|
|Hyaline membrane disease Grade III/IV||50||14.9|
|Intra-ventricular haemorrhage Grade III/IV||2||0.5|
|Hypoxic ischaemic encephalopathy||0||0|
|Neonatal death ≤ day 7||10||2.9|
|Neonatal death day 8–28||10||2.9|
According to the Ballard evaluations, 54% of babies were described as small for gestational age. Thirty-six percent of babies, however, did not undergo Ballard evaluations after birth. When the birthweights of babies who had undergone early ultrasound scans were plotted on the local population based centile chart, 36% were noted to be small for gestational age. The median number of days spent in the hospital until death or discharge by babies, excluding the three terminations and two intrauterine deaths, was 14 (range 1–130) days.
In this study a multivariate analysis was performed in order to identify the most important factors influencing neonatal outcome. Neonatal outcomes were described as good or poor, poor being early or late neonatal deaths, as well as infant deaths. Intrauterine deaths and terminations of pregnancy were excluded. Eighteen factors were initially analysed. Thereafter, factors still considered to be important in determining neonatal outcome were entered into the model. These were gestational age, pre-eclampsia or superimposed pre-eclampsia, number of oral anti-hypertensive agents, elective or emergent delivery, abdominal or vaginal delivery, number of doses of betamethazone, cardiotocographic pattern indicating fetal distress prior to delivery, amniotic fluid volume and small for gestational age babies. Gestational age at delivery remained the only factor significantly associated with neonatal outcome after adjustment for all others. The odds ratio was 0.617 and the confidence interval 0.488–0.779. The number of doses of betamethazone, elective or emergent delivery and delivery route, were some of the variables that did not significantly correlate with neonatal outcome.
High perinatal survival rates were achieved in this group of women with early onset, severe pre-eclampsia. Overall perinatal survival (Table 1) was 72% at 28 weeks (delivery gestation), 92% at 29 weeks and higher survival at greater gestations. The perinatal mortality rate was 24/1000 (≥ 1000 g/7 days) with a neonatal survival rate of 95% from 29 weeks of gestation at delivery. This compares favourably with recently reported survival rates from a developed country, where few infants of > 28 weeks of gestational age without lethal anomalies died19. To evaluate the good perinatal outcome, the antenatal and neonatal periods of care need to be examined. The major reported cause of sudden intrauterine death in severe pre-eclampsia is placental abruption20. As there is still no proven method of preventing this devastating condition, it is important to identify it early in order to save the fetus21. In this study 20% of cases were complicated by placental abruption. Frequent fetal heart rate monitoring was used to identify the fetal distress caused by placental abruption and only two intrauterine deaths occurred. These two deaths occurred despite six-hourly testing, following the recommendation of Chari et al.22 of daily antenatal testing in women with severe pre-eclampsia.
The neonatal outcome has been linked to neonatal intensive care facilities and gestational age at birth. Neonatal intensive care remains a key requirement in the management of early onset, severe pre-eclampsia, but this service is expensive and limited, particularly in developing countries. Astute utilisation of neonatal intensive care is therefore essential. Odendaal et al.12 in a randomised controlled trial found expectant management to be associated with a smaller number of neonates requiring ventilation (P < 0.05). Sibai et al.13 in a similar, but larger, trial found that expectant management was associated with a lower incidence of admission to a neonatal intensive care unit (P= 0.002) and a lower mean number of days spent there (P < 0.0001). In the index study only 40% of the babies were admitted to the neonatal intensive care unit, staying a median number of six days. Table 1 shows that a decreasing percentage of babies who qualified for neonatal intensive care (as defined by gestation) actually needed it, even at early gestations. Thus, at 30 weeks of gestation, when all the babies qualified for neonatal intensive care, only 55% needed this facility. This illustrates efficient utilisation of this scarce resource.
Only gestational age at delivery was significantly associated with neonatal outcome in the multivariate analysis. In another multivariate analysis performed by Witlin et al.23 neonatal outcome in women with severe pre-eclampsia or eclampsia between 24 and 34 weeks of gestation was directly correlated with increasing birthweight (P < 0.001), while increasing gestational age correlated with a reduction in respiratory distress syndrome (P= 0.001). Interestingly, neonatal outcome did not correlate with the use of steroids.
In 1987 Odendaal et al.24 published a case series in which 129 women with early onset, severe pre-eclampsia were managed expectantly. A mean number of 11 days were gained with a perinatal mortality rate of 97/1000 being reported. After a subsequent randomised controlled trial12 which showed that expectant management of such patients improved the neonatal outcome, this same group then reported further improvements in perinatal mortality rates (62/1000 for birthweights (1000 g) with this policy. In the latter study there were six maternal deaths among 1001 women over 10 years20. This perinatal mortality rate is comparable to those reported by Visser et al.25 who reported 107 per 1000 births for patients with admission gestations ranging from 25–35 weeks and Oláh et al.11 who stated 71 per 1000 births in their expectant management arm between 24 and 32 weeks of gestation. The perinatal mortality rate of 24/1000 (≥ 1000 g/7 days) or 44.1 (≥ 500 g/7 days) in the index study was brought about by further refinements to the expectant management approach and no maternal deaths occurred.
Of more practical importance (than the perinatal mortality rate) is the prospective outcome for the fetus at each completed week of gestation after viability. In the index study, viability was defined as 28 weeks of gestation with a minimum fetal mass of 800 g or 27 weeks of gestation or a lower mass under exceptional circumstances. Several authors have demonstrated an increase in perinatal survival as gestation increases5,10,14. In 1987 Odendaal et al.24 showed increased perinatal survival as birthweight increased and later another South African group, Moodley et al.26, demonstrated improved survival with increasing gestational age. These data are of importance in the management of women with severe pre-eclampsia, remote from term and for whom the same illness has a high chance of re-occurring in a subsequent pregnancy. It is thus imperative that the obstetrician managing these difficult cases is aware of the benefits to be gained by prolonging the intrauterine period, and the chances of fetal survival in the lower gestational age range.
With regard to major neonatal complications most studies have concentrated on reporting the intrauterine and neonatal deaths involved. Odendaal et al.12 and Sibai et al.13 reported lower numbers of neonatal complications in randomised controlled trials. In the index study severe hyaline membrane disease and sepsis were the main culprits. Both of these factors further emphasise the importance of expectant management for this category of patient. Schiff et al.8 found that fetuses of women with pre-eclampsia do not exhibit accelerated pulmonary maturation. Septicaemia is a particular problem for preterm infants in over-populated wards and delivery of more mature babies would thus help to address this problem. Other noninfective complications were uncommon. It must be mentioned that an outbreak of extended spectrum, beta-lactamase producing, resistant Klebsiella pneumoniae occurred in the hospital nurseries during the study period27.
Finally, expectant management can only be undertaken by an experienced team offering continuous monitoring and care. It is best that such women are moved at an early stage to specialised regional centres28. The expectant approach needs to be used with discretion and in the knowledge that in some women it will achieve little gain in time. All such women are hospitalised because complications cannot be accurately predicted and may develop rapidly.
The perinatal outcome during this study can be ascribed to careful management of the pregnant woman29, frequent assessment of maternal and fetal wellbeing, as well as refinements in neonatal intensive care. Gestational age at delivery proved to be a key factor for the neonate. Prolonged expectant management of women with severe pre-eclampsia is not practised in most centres. With improvements in neonatal care, certain clinicians regard the delivery of women with severe pre-eclampsia beyond 30 weeks to be in the best interests of not only the mother but also the fetus30. Although tempting, it would seem unwise to shift our limits of viability and elective delivery while the current approach remains safe for the mother, neonatal intensive care facilities are limited and sepsis related problems prevalent for the baby. This situation may, however, differ in other institutions.
Our study showed that expectant management of early onset, severe pre-eclampsia and careful neonatal care led to high perinatal survival rates. It also allowed the judicious use of neonatal intensive care facilities. Neonatal sepsis remains a cause for concern.
The authors would like to thank Sister M. H. Carstens for collection of the data and the Medical Superintendent of Tygerberg Hospital for permission to publish. The study was supported by the South African Medical Research Council.