Acute renal failure complicating severe preeclampsia requiring

admission to an obstetric intensive care unit

Andrew J. Drakeley, MRCOG, Paul A. Le Roux, MBChB, John Anthony, FCOG(SA), and
James Penny, MD, MRCOG
Cape Town, South Africa

OBJECTIVE: To determine risk factors and outcomes for women with severe preeclampsia and renal failure.
STUDY DESIGN: Retrospective study from 1995 to 1998 of all women with renal failure who were admitted
to the obstetric intensive care unit at Groote Schuur Hospital, South Africa. A total of 89 women were identi-
fied with severe preeclampsia defined as blood pressure ≥160/110 mm Hg and ≥2+ proteinuria, renal failure
defined as a creatinine level of ≥1.13 mg/dL, and oliguria defined as <100 mL urine produced in 4 hours; 72
charts were available for analysis. A comparison was made between the 3 groups, which were defined by the
maximum recorded creatinine levels.
RESULTS: Of the 72 women, 31 women (43%) were primiparous and 41 (57%) were multiparous. Median
gestation at delivery was 32 weeks (range, 21-40 weeks). The median maximum creatinine was 3.85 mg/dL
(range, 1.13-12.50 mg/dL). Twelve women (16%) had a history of chronic renal disease or hypertension, and
36 women (50%) had HELLP syndrome and 23 (32%) abruptio placentae. All women with severe renal im-
pairment had either abruptio placentae or hemolysis, elevated liver enzymes, and low platelet count (HELLP)
syndrome. Perinatal mortality was 38% (27/72). However, in this series only 7 women (10%) required dialysis
in the short term and none required long-term dialysis or kidney transplant. There were no maternal deaths.
CONCLUSIONS: In women with severe preeclampsia and renal failure, major obstetric complications were
common and perinatal outcome was poor. However, the need for dialysis was infrequent, with only 10%
women requiring transient dialysis, and there were no cases of chronic renal failure that required dialysis or
kidney transplant. (Am J Obstet Gynecol 2002;186:253-6.)

Key words: Preeclampsia, renal failure, pulmonary artery catheterization

The incidence of acute renal failure has declined over
recent decades as a result of improved antenatal care and
the virtual elimination of post-abortal sepsis. Thirty years
ago the incidence of acute renal failure was 1 in 2000
pregnancies, and it is now 1 in 10,000 pregnancies.1,2
All obstetricians are faced with the problem of acute
renal failure in the second half of pregnancy. The causes
now prevalent include hypertensive disorders, hemor-
rhage, sepsis, intrinsic renal disease, and other rare preg-
nancy-related disorders. The management of patients
with acute renal failure caused by preeclampsia or
eclampsia differs geographically, according to the avail-
ability of resources and different perceptions of the dis-
ease.
From the Department of Obstetrics and Gynaecology, Groote Schuur Hos-
pital.
Received for publication March 16, 2001; revised June 18, 2001; ac-
cepted September 12, 2001.
Reprint requests: AJ Drakeley, MRCOG, Department of Obstetrics & Gy-
naecology, Liverpool Women’s Hospital, Crown Street, Liverpool L8 7SS,
UK. E-mail: adrakeley@yahoo.com or anthonyj@uctgsh1.uct.ac.za.
Copyright 2002, Mosby, Inc. All rights reserved.
0022-9378/2002 $35.00+0 6/1/120279
doi:10.1067/mob.2002.120279
Patients with preeclampsia have vasoconstriction, he-
moconcentration, and a reduced intravascular volume;
therefore, they are particularly vulnerable to the effects
of blood loss. Antepartum hemorrhage (eg, abruptio pla-
centae) and postpartum hemorrhage reduce plasma vol-
ume and may lead rapidly to renal ischemia. Acute renal
failure in preeclampsia is often associated with placental
abruption.3
The largest obstetric review of renal failure of which we
are aware was published in 1990.3 This study, conducted
over 12 years, included 9600 women with hypertension:
1433 women had preeclampsia and 251 had eclampsia. Of
those women described in the study, only 31 developed
acute renal failure, all in the postpartum period. From
these 31 cases there were 2 maternal deaths, and 50% of
the patients in the preeclamptic group required dialysis.
All patients had acute tubular necrosis. In the chronic hy-
pertensive group with superimposed preeclampsia, 42%
required dialysis and 3 had cortical necrosis. Most impor-
tant is that all patients in the preeclamptic group had nor-
mal renal function at long-term follow-up (average, 4
years). The investigators concluded that appropriate treat-
ment in the women with preeclampsia resulted in the ab-
sence of long-term renal impairment. It is noteworthy that

253
254 Drakeley et al
these investigators managed their patients in a dedicated
obstetric intensive care unit (ICU) with facilities for inva-
sive monitoring. Another review by Sibai et al3 of 254
eclamptic patients showed that of the 12 patients requir-
ing dialysis, 1 patient died. The remaining 11 patients had
reversible acute tubular necrosis.
Renal failure is a serious complication of pregnancy,
with significant associated morbidity and mortality. We
evaluated the pregnancy-associated and, more particu-
larly, the preeclampsia-associated incidence and sequelae
of renal failure in a tertiary-care obstetric unit.
The objectives of our study were to determine manage-
ment, complications, and outcomes for women with se-
vere preeclampsia and acute renal failure in pregnancy.
Methods
Groote Schuur Hospital is a tertiary referral unit for
the entire metropolitan area of the Western Cape Penin-
sula of South Africa that covers 28,000 deliveries per
annum. All medical problems that complicate pregnancy
are referred to the obstetric service at Groote Schuur
Hospital. The Maternity Centre at the hospital has an
obstetrician-led ICU, with facilities for ventilation and in-
vasive monitoring. At the time of this study, the majority
of hospital admissions were for complications of severe
preeclampsia. Other indications for admission were
hypovolemic shock caused by obstetric hemorrhage and
septic shock. A database of all ICU admissions is main-
tained, which documents indications for admission, com-
plications, need for invasive monitoring, and outcome.
Also included are biochemical and hematologic parame-
ters together with daily fluid balance.
The obstetric ICU database was interrogated to identify
women with renal impairment (defined as a serum crea-
tinine level of ≥1.13 mg/dL and oliguria of <100 mL of
urine in 4 hours). To ensure complete identification of
all cases, the records of the biochemistry department
were checked for all women who were admitted to the ob-
stetric ICU. Finally, the renal unit database was searched
to ensure that all cases of severe disease and dialysis for
obstetric patients were identified. Women with less severe
renal impairment (eg, creatinine levels of <2.26 mg/dL)
who were cared for at Groote Schuur Hospital may not
have been identified if they were not admitted to the ob-
stetric ICU.
Women were divided into three groups on the basis of
the highest creatinine level recorded in the first 7 days
after admission: group 1, 1.13-2.25 mg/dL (100-199
µmol/L); group 2, 2.26-6.78 mg/dL (200-600 µmol/L);
and group 3, >6.78 mg/dL (>600 µmol/L). The thresh-
olds for division were arbitrary but facilitated compar-
isons between women with different severities of disease.
The groups were compared for antenatal characteristics
and for postnatal outcomes. All women included in the
study were managed according to established depart-
February 2002
Am J Obstet Gynecol
mental protocols that remained consistent for the dura-
tion of the study.
All women with severe preeclampsia (persistent hyper-
tension at ≥160/110 mm Hg and ≥2+ proteinuria on dip-
stick) underwent a standardized management protocol.
Magnesium sulfate was prescribed to all women with se-
vere preeclampsia, and hydralazine was the hypotensive
agent of choice. Oliguria was intially managed with col-
loid fluid plasma substitute challenges. If there was no re-
sponse, then low-dose dopamine was prescribed and, if
necessary, a pulmonary artery flotation catheter was in-
serted to aid in hemodynamic management. Maternal
medical records were analyzed and data were extracted
pertaining to the indication for admission and progress
while in the ICU. Details of postnatal appointments were
also recorded, including blood pressure, proteinuria,
and renal biochemistry. For women who were transferred
to the renal unit for dialysis, renal clinic charts were ana-
lyzed for recovery of renal function to normal levels. Dis-
cussion with the renal specialists with regard to data
interpretation and presentation was undertaken before
commencement of the study.
Results
During the 3-year study period, Groote Schuur Hospi-
tal provided tertiary care to a maternity population of
94,500 women, of whom 5200 had hypertensive compli-
cations of pregnancy. A total of 588 women were admit-
ted to the obstetric ICU. Of these, 89 had a blood
pressure reading of ≥160/110 mm Hg, a reading on
urine dipstick of ≥2+ proteinuria, and a serum creatinine
level of ≥1.13 mg/dL.
Patient records were located for only 72 women; the re-
maining 17 records could not be found. Of the 72 women
identified, 38 had mildly elevated maximum creatinine
levels of between 1.13 and 2.25 mg/dL; 19 women had
moderately elevated creatinine levels of between 2.26 and
6.78 mg/dL, and 15 had severely elevated creatinine levels
of >6.78 mg/dL (Table I). In the majority of women, pre-
eclampsia was the underlying pathologic condition (67 of
72 cases). Other diagnoses included one each of diabetic
ketoacidosis, myocardial infarction, mixed mitral valve
disease, extra-uterine pregnancy and pulmonary embolus.
None of the women had a maximum creatinine level ex-
ceeding 2.26 mg/dL. In all of the remaining women, pre-
eclampsia was the sole cause of acute renal failure. Fig 1
shows the pattern of creatinine levels in women within
each of the three groups in the first 7 days postpartum
(mild, diamond; moderate, square; and severe, triangle).
The diagnoses contributing to ICU admission are shown
in Table II. In some cases, multiple disorders contributed
to the woman’s renal compromise (eg, preeclampsia plus
abruptio placentae plus eclampsia plus HELLP).
In 14 of the 15 women with maximum creatinine levels
of >6.78 mg/dL, the women had either HELLP syndrome
Volume 186, Number 2 Drakeley et al 255
Am J Obstet Gynecol

Table I. Baseline characteristics for women grouped ac-

cording to creatinine levels

Creatine level (mg/dL)

1.13-2.25 2.26-6.78 >6.78

Characteristic (n = 38) (n = 19) (n = 15)

Primigravid (%) 44 42 43
Mean blood pressure 126/76 128/78 126/82
at booking
(mm Hg)
N (%) without 5 (13) 3 (15) 1 (7)
prior antenatal
care
Median weeks 32 (22-40) 32 (21-38) 32 (23-40)
of gestation
at delivery (range)
Fig. Creatinine trend over 1 week for the 3 groups of women.

(n = 11) or placental abruption (n = 5, of whom 2 also Table II. Diagnoses contributing to admission to inten-
had HELLP syndrome). In one woman with HELLP syn- sive care unit according to creatinine levels
drome, the nadir of the platelet count was 103  109/L,
Creatinine level (mg/dL)
with maximum levels of lactate dehydrogenase and ala-
nine transaminase of 3200 U/L and 229 U/L, respec- 1.13-2.25 2.26-6.78 > 6.78
tively. One patient did not have HELLP or a placental Diagnosis (n = 38) (n = 19) (n = 15)
abruption, but had intrinsic renal disease exacerbated by HELLP syndrome 12 9 11
preeclampsia. Abruptio placentae 9 (1 with 9 (1 with 5 (2 with
After diagnosis of acute tubular necrosis (persistent olig- HELLP) HELLP) HELLP)
Pulmonary edema 3 1 1
uria and rising creatinine level), standard management of Eclampsia 8 2 2
fluid restriction and close monitoring of blood chemistry Antenatal hypertension 2 2 3
was instituted. Seven women required dialysis; in 5 cases Antenatal chronic renal 2 1 1
failure
the indication was uremia. Two women required acute
dialysis because of fluid overload and pulmonary edema Data are number of women. HELLP, Hemolysis, elevated liver
that failed to respond to conventional management. None enzymes, and low platelet count.
of the women required dialysis because of electrolyte im-
balance. In all of the cases referred for dialysis there was Table III. Treatment according to creatinine level
subsequent improvement in renal function. None of the Creatinine level (mg/dL)
women required long-term dialysis, the longest treatment
being 2 weeks. Nine women required treatment for hyper- Treatment 1.13-2.25 2.26-6.78 >6.78
kalemia (potassium level of >7 mmol/L): 4 were treated
Calcium resonium 0 2 3
with intravenous insulin and dextrose and 5 were given Insulin/dextrose 0 2 2
calcium resonium ion exchange resin. Forty-six women re- Dialysis 0 0 7
quired pulmonary artery flotation catheters, invariably to PAFC (Swan-Ganz catheter) 20 12 14
CVP line 1 1 0
aid fluid management (Table III).
Renal impairment associated with preeclampsia was as- Data are number of women. CVP, Central venous pressure;
sociated with a high incidence of perinatal mortality, PAFC, Pulmonary artery flotation catheter.
overall 38% (27/72), ranging from 26% in group 1 to
47% in group 3. There were no maternal deaths in this Table IV. Neonatal outcome and follow up for women
cohort of women by the end of the study period (maxi- with preeclampsia
mum follow-up 31/2 years). In addition, there was a high
Creatinine level (mg/dL)
incidence of hypertension (≥140/90 mm Hg) at 3-month
follow-up, although persistent proteinuria was relatively 1.13-2.25 2.26-6.78 >6.78
uncommon (Table IV). In all but one woman who had re- (n = 38) (n = 19) (n = 15)
quired renal dialysis, the creatinine level had returned to
Perinatal death, n (%) 10 (26%) 10 (53%) 7 (47%)
within normal limits by the time of discharge. In one case Hypertensive at 6 of 8 who 3 of 9 who 5 of 9 who
the last recorded creatinine level was 1.82 mg/dL. 3-month follow up attended attended attended
Because of limited and oversubscribed neonatal re- Proteinuric at 0 of 8 who 5 of 9 who 0 of 9 who
3-month follow up attended attended attended
sources, neonates with birth weights of <1000 g are not
256 Drakeley et al
ventilated. The severity of disease, late presentation, and
birth weight restriction for ventilation all contribute to
these high neonatal mortality rates.
Comment
This current study is unique in several aspects. The rel-
ative rarity of acute renal failure nowadays, compared
with data from 30 years ago is confirmed (1:1060). This
study reflects a higher rate of renal failure than does the
Sibai’s3,4 series; however, Sibai restricted his study to
women with HELLP syndrome. The difference in the in-
cidence of renal failure may be due to referral pattern,
population, or the definition employed. The current
study is a true population-based study because we run a
single and complete tiered obstetric service for an entire
metropolitan region. The incidence of HELLP syndrome
in our groups of acutely ill hypertensive women admitted
to the labor ward at Groote Schuur is 23%.5 The inci-
dence of women with renal failure who require dialysis is
low and reliably quantifiable (7:94,500 or 1:13,500).
The prevalence of preeclampsia as a cause of acute
renal failure has been confirmed. The rarity of preexist-
ing medical disorders as contributory factors leading to
acute renal failure is in contrast to data presented in
some earlier published reports and may reflect an overall
improvement in the management of other conditions, es-
pecially those associated with an infective etiology.1,2
The most severe forms of renal failure occurred in
complicated preeclampsia, invariably HELLP5,6 syn-
drome and placental abruptions. It is uncertain whether
the favorable outcome reflects an aggressive fluid and
management protocol or the natural history of the con-
dition. It is noteworthy that the risk of requiring dialysis is
highest in patients for whom the creatinine level doubles
in the first 24-48 hours post hospital admission, in the
presence of a major complication of preeclampsia.
This retrospective review cannot assess whether our
approach to management was beneficial (or harmful)
in either the short or long term. According to our un-
derstanding of the pathophysiology, the use of invasive
monitoring is consistent with goals of rational manage-
ment. A wider review in the same obstetric intensive
care unit showed this type of monitoring to be both
February 2002
Am J Obstet Gynecol
beneficial and relatively safe.6,7 The incidence of inva-
sive monitoring overall was low and if accepted to be
necessary would favor arguments that support the cre-
ation of dedicated units for the care of these women.
However, although the management protocol cannot
be evaluated in this series, it is reasonable to assert that
women with renal failure complicating severe pre-
eclampsia should be managed in a dedicated unit with
appropriate expertise to ensure that the good results
from this and series by Sibai et al (REF) are achieved.
The absence of recognized long-term renal failure was
notable. However, because of population movement, the
duration of follow-up was inadequate. Although there was
no requirement for immediate long-term dialysis, im-
proved survival may be at the expense of chronic renal
failure, renovascular hypertension, and reduced long-
term survival.
In conclusion, given the circumstances of the study,
acute renal failure complicating severe preeclampsia is a
rare condition. Intensive hemodynamic and renal inter-
vention in this group of patients was associated with a
good outcome. The management applied appears to be
effective, although comparative data are few and long-
term sequelae are unknown. Future studies should focus
on comparative interventions as well as controlled cohort
outcome studies.
REFERENCES
1. Lindheimer MD, Katz AI, Ganeval E. Acute renal failure in
pregnancy. In: Brenner BM, Lazarus JM, editors. Acute renal
failure. 2nd ed. New York: Churchill Livingstone; 1988. p. 597.
2. Madias NE, Donohoe JF, Harrington JT. Post ischaemic acute
renal failure. In: Brenner BM, Lazarus JM, editors. Acute renal
failure. 2nd ed. New York: Churchill Livingstone; 1988. p. 251.
3. Sibai BM, Villar MA, Mabie M. Acute renal failure in hyper-
tensive disorders of pregnancy. Am J Obstet Gynecol 1990;
162:777-83.
4. Sibai BM. Eclampsia: maternal-perinatal outcome in 254 consec-
utive cases. Am J Obstet Gynecol 1990;163:1049-54.
5. Weinstein L. Syndrome of hemolysis, elevated liver enzymes, and
low platelet count: a severe consequence of hypertension in
pregnancy. Am J Obstet Gynecol 1982;142:159-67.
6. Anthony J, Johanson R, Dommisse J. Critical care management
of severe preeclampsia. Fetal Matern Med Rev1994;6:219–29.
7. Gilbert WM. Towner DR, Field NT, Anthony J. The safety and
utility of pulmonary artery catheterization in severe preeclamp-
sia and eclampsia. Am J Obstet Gynecol 2000;182:1397-403.