DOI: 10.1111/j.1471-0528.2006.00908.

x
www.blackwellpublishing.com/bjog
Systematic review

Accuracy of serum uric acid in predicting

complications of pre-eclampsia:
a systematic review
S Thangaratinam,a KMK Ismail,a S Sharp,b A Coomarasamy,c KS Khanc for TIPPS
(Tests in Prediction of Pre-eclampsia Severity) review group
a Academic Unit of Obstetrics and Gynaecology, Maternity Block, University Hospital of North Staffordshire, Stoke-on-Trent, UK b NeLH

Specialist Library for ENT and Audiology, Radcliffe Infirmary, Oxford, UK c Education Resource Centre, Birmingham Women’s Hospital,
Birmingham, UK
Correspondence: Dr S Thangaratinam, Academic Unit of Obstetrics and Gynaecology, Maternity Block, University Hospital of North Staffordshire,
Stoke-on-Trent, ST4 6QG, UK. Email shakila@doctors.org.uk

Accepted 1 January 2006.

Background Pre-eclampsia is one of the largest causes of maternal
and fetal mortality and morbidity. Hyperuricemia is often
associated with pre-eclampsia.
Objective To determine the accuracy with which serum uric acid
predicts maternal and fetal complications in women with
pre-eclampsia.
Study design Systematic quantitative review of test accuracy
studies.
Search strategy We conducted electronic searches in MEDLINE
(1951–2004), EMBASE (1980–2004), the Cochrane Library
(2004:4) and the MEDION database to identify relevant articles.
A hand-search of selected specialist journals and reference lists of
articles obtained was then carried out. There were no language
restrictions for any of these searches.
Selection criteria Two reviewers independently selected the articles
in which the accuracy of serum uric acid was evaluated to predict
maternal and fetal complications of pre-eclampsia.
Data collection and analysis Data were extracted on study
characteristics, quality and accuracy to construct 2 · 2 tables with
maternal and fetal complications as reference standard. Summary
likelihood ratios for positive (LR+) and negative LR(–) test results
are generated for various threshold levels of uric acid.
Main results There were 18 primary articles that met the selection
criteria, including a total of 3913 women and forty-one 2 · 2
tables. In women with pre-eclampsia, a positive test result of uric
acid greater than or equal to a 350-mmol/l threshold predicted
eclampsia with a pooled likelihood ratio (LR) of 2.1 (95% CI 1.4–
3.5), while a negative test result had a pooled LR of 0.38 (95% CI
0.18–0.81). For severe hypertension as the outcome measure, the
LRs were 1.7 (95% CI 1.3–2.2) and 0.49 (95% CI 0.38–0.64) for
positive and negative test results, respectively, and for caesarean
section the LRs were 2.4 (95% CI 1.3–4.7) and 0.39 (95% CI 0.20–
0.76). For stillbirths and neonatal deaths the respective LRs were
1.5 (95% CI 0.91–2.6) and 0.51 (95% CI 0.20–1.3). For the
prediction of small-for-gestational-age fetus, the pooled LRs were
1.3 (95% CI 1.1–1.7) and 0.60 (95% CI 0.43–0.83) for positive and
negative results, respectively.
Author’s conclusion Serum uric acid is a poor predictor of
maternal and fetal complications in women with pre-eclampsia.
Keywords Complications, diagnostic tests, pre-eclampsia,
systematic review, uric acid.

Please cite this paper as: Thangaratinam S, Ismail K, Sharp S, Coomarasamy A, Khan K. Accuracy of serum uric acid in predicting complications of pre-eclampsia:
a systematic review. BJOG 2006; 113:369–378.

of the characteristic findings in pre-eclampsia. In clinical

Introduction
Pre-eclampsia affects approximately 2–8% of all pregnancies
and is associated with several complications.1 It remains one
of the largest single causes of maternal and fetal mortality and
morbidity.1,2 Clinical prediction of disease complications may
facilitate instigation of timely management to avert mortality
and morbidity in the mother and baby. Hyperuricemia is one
practice, uric acid determination is considered to be a part
of the workup in women with pre-eclampsia to monitor dis-
ease severity and aid management of these women. The asso-
ciation between raised serum uric acid and pre-eclamptic
pregnancy was first reported in 1917.3 Reduced uric acid
clearance secondary to reduced glomerular filtration rate,
increased reabsorption and decreased secretion may be the

ª RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology 369

 Thangaratinam et al.
reasons for elevated serum uric acid levels in women with
pre-eclampsia.4,5 The pathophysiologic mechanisms of
pre-eclampsia comprising increased trophoblastic tissue
shedding, endothelial dysfunction, and reduced blood flow
in the fetomaternal unit have also been hypothesised as the
underlying cause of hyperuricemia in this condition.6
Several studies have reported a positive correlation between
elevated maternal serum uric acid levels and adverse maternal
and fetal outcomes.7–10 However, these primary diagnostic
studies have not generally been conducted with large enough
sample size to provide precise accuracy estimates and they
vary widely in their definition of pre-eclampsia, maternal
and fetal outcomes and optimal cutoff levels of uric acid in
predicting maternal and fetal complications. There are no
systematic reviews exploring the accuracy of uric acid to pre-
dict complications of pre-eclampsia. We therefore conducted
a comprehensive systematic review to obtain precise estimates
of maternal serum uric acid levels to predict maternal and
fetal complications in women with pre-eclampsia.
Methods
The review was carried out with a prospective protocol using
widely recommended methods.11–14
Identification of studies
We searched MEDLINE (1951–2004), EMBASE (1974–2004),
Cochrane Library (2004:4) and MEDION (a database of diag-
nostic test reviews set up by Dutch and Belgian researchers)
for relevant citations. The reference lists of all known primary
and review articles were examined to identify cited articles not
captured by electronic searches. Details of the search strategy
are available from the authors. Language restrictions were not
applied. A comprehensive database of relevant articles was
constructed.
Study selection and data extraction procedures
Studies that evaluated the accuracy of maternal serum uric
acid in women with pre-eclampsia for the prediction of
maternal or fetal complications were selected in a two-stage
process. First, the electronic searches were scrutinised and full
articles of all citations that were likely to meet the predefined
selection criteria were obtained by two independent reviewers
(S.T. and K.M.K.I.). Second, final inclusion or exclusion deci-
sions were made by the reviewers (S.T. and K.M.K.I.) after
examination of these articles. Studies that met the predefined
and explicit criteria regarding population, tests, outcomes
and study design (Table 1) were selected for inclusion in
the review. When disagreements occurred, they were resolved
by consensus. In cases of duplicate publication, the most
recent and complete versions were selected. There were no
language restrictions.
370
Information was extracted from each selected article on
study characteristics, quality and accuracy results. Accuracy
data were used to construct 2 · 2 tables of uric acid result (test
positive if uric acid levels were above a threshold as defined in
the primary study, and test negative if these were below the
threshold) and maternal (eclampsia, severe hypertension,
HELLP [haemolysis, elevated liver enzymes and low platelet
count] syndrome, caesarean section) and fetal (stillbirths and
neonatal deaths, intrauterine death, small-for-gestational-age
fetus) outcomes.
Methodological quality assessment
All articles meeting the selection criteria were assessed for
their methodological quality. Quality was defined as the con-
fidence that the study design, conduct and analysis minimised
bias in the estimation of test accuracy. Based on existing
check-lists,11,12,14,30,31 quality assessment involved scrutinising
study design and relevant features of the population, test and
outcomes of the study. A study was considered to be of good
quality if it used a prospective design, consecutive enrolment,
full verification of the test result with reference standard and
had adequate test description.11,14,30,31 We excluded studies
with case–control design as these are known to result in sta-
tistical bias.30
Data synthesis
Likelihood ratios (LRs) for positive and negative test results
were calculated for each study, separately for each test
threshold. Summary LRs32 were then computed for positive
and negative test results for each individual test threshold
and for each outcome of interest. This information is clini-
cally more relevant than traditional summaries of accuracy
such as sensitivity and specificity, as LRs allow the estimation
of post-test probabilities of various complications for women
at different risk levels.33 The LR indicates by how much
a given test result raises or lowers the probability of having
the disease. The higher the LR of an abnormal test, the
greater is the value of the test. Conversely, the lower the
LR of a normal test, the greater the value of the test. An
LR >10 for an abnormal test or <0.1 for a normal test is
regarded as ‘very useful’ test accuracy, LR of 5–10 or 0.1–0.2
is regarded as ‘moderately useful’ and LR of 2–5 or 0.5–0.2 is
regarded as ‘somewhat useful’. An LR of 1–2 or 0.5–1 is only
regarded as ‘little useful’ and LR of 1 as ‘useless’. Although,
this categorisation is useful for interpretation of LRs, it
should be noted that the value of a test may vary depending
on the pre-test probability of the condition and the conse-
quences of treatment.
Heterogeneity of diagnostic odds ratio was assessed graphi-
cally using forest plot34 (not shown) and statistically using
chi-square test35 to aid in decisions on how to proceed with
quantitative synthesis.36 As, for some tests and outcomes, there
was either graphical or statistically significant heterogeneity, we
ª RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology

Table 1. Study characteristics of the trials included in the systematic review of accuracy of serum uric acid in predicting complications in
women with pre-eclampsia
Study
Study (Year) Quality
Yassaee (2003)15 Cohort, not blind,
enrolment method not known,
direction of data collection
not known, test not described
Williams and Cross-sectional, not blind,
Galerneau (2002)16 prospective, test described,
enrolment method not known
D’Anna et al. (2000)17 Cross-sectional, not blind,
retrospective, enrolment
not known, test described
Martin et al. (1999)18 Cross-sectional, not blind,
retrospective, not consecutive
enrolment, test not described
Odendaal and Cross-sectional, not blind,
Pienaar (1997)19 prospective, enrolment
method not known,
test described
Shah and Reed (1996)20 Cross-sectional, not blind,
retrospective, consecutive
enrolment, test description
not adequate
Magann et al. (1993)21 Cross-sectional, not blind,
retrospective data collection,
not consecutive enrolment,
test described
Voto et al. (1988)22 Cross-sectional, not blind,
enrolment not known,
direction of data collection
not known, test not described
Sagen et al. (1984)23 Cross-sectional, not blind,
prospective, test described,
not consecutive enrolment
ª RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology
Uric acid to predict complications of pre-eclampsia
Population
Number of Inclusion criteria
patients
103 Severe pre-eclampsia
194 Blood pressure (BP) 140/90
after 20 weeks and
proteinuria 11 or
300 mg/24 hour
94 National Working Group on
Hypertension in Pregnancy
criteria for pre-eclampsia
568 Severe pre-eclampsia
229 Severe pre-eclampsia
271 Pre-eclampsia (de novo
development of hypertensive
disorder in second half
of pregnancy). It includes
proteinuric pre-eclampsia
.300 mg/dl in 24 hour or
11 6 hour apart and
nonproteinuric pre-eclampsia
with increased BP in second-half
of pregnancy and elevated
uric acid or other systemic
involvement markers
454 Severe pre-eclampsia
125 BP 140/90 in third trimester,
mild pre-eclampsia
(BP 140–159/90–99),
severe pre-eclampsia
(BP 160/100)
72 Severe pre-eclampsia,
BP 160/110 and
proteinuria 5 g/24 hours
Test (uric acid), cutoff Outcome
350 mmol/l Eclampsia, maternal death,
caesarean section, intrauterine
death, intrauterine growth
restriction
450 and 540 mmol/l HELLP syndrome (SGOT .
40 iu/l, LDH .600 iu/l,
haemolysis on blood film,
platelets 150 3 109/l),
small for gestational age
(birthweight ,10th centile),
severe hypertension (systolic
BP 160 and/or diastolic
BP 110 on two occasions)
340 mmol/l Intrauterine growth restriction
380 and 460 mmol/l Significant maternal morbidity,
renal, hepatic and/or
gastrointestinal
520 mmol/l Small for gestational age
(Tygerberg hospital growth
curves), intrauterine death,
neonatal death, perinatal
mortality (within 7days),
caesarean section,
preterm delivery
350 mmol/l Adverse perinatal outcome
that includes: 1. perinatal
death; 2. perinatal morbidity
of prematurity due to
hypertensive disease,
moderate or severe hyaline
membrane disease, patent
ductus arteriosus,
intraventricular haemorrhage;
3. perinatal morbidity due
to uteroplacental vasculopathy,
intrauterine growth
restriction, abruption, fetal
distress needing
caesarean section
330, 460 and 600 mmol/l Class I HELLP syndrome
(Platelets ,50 000,
elevated LDH, haemolysis
and hepatic dysfunction)
350 mmol/l Severe hypertension
BP 160/100, intrauterine
growth restriction
Urate increment 50 mmol/l Perinatal distress (perinatal
in last 3 days prior death, Apgar, 7 at 1 or
to delivery, 350 mmol/l 5 mins, later development of
neonatal asphyxia, respiratory
distress syndrome,
hypoglycaemia or fits),
small for gestational
age (,10th centile),
perinatal death
(continued)
371
 Thangaratinam et al.
Table 1. (Continued )
Study
Study (Year) Quality
Liedholm et al. (1984)7 Cross-sectional, not blind,
retrospective, test described,
consecutive enrolment
Varma (1982)24 Cross-sectional, not blind,
prospective, not consecutive
enrolment, test described
Mathews et al. (1980)25 Cross-sectional, not blind,
prospective, not consecutive
enrolment, test described
Dequiedt et al. (1979)26 Cross-sectional, not blind,
prospective data collection,
enrolment not known,
test described
Fadel et al. (1969)27 Cross-sectional, not blind,
prospective, test described,
enrolment not known
Connon and Cross-sectional, not blind,
Wadsworth (1968)28 retrospective data collection,
not consecutive enrolment,
test described
Lancet and Cross-sectional, not blind,
Fisher (1956)9 enrolment not known,
direction of data collection
not known, test not described
Seitchik (1953)29 Cross-sectional, not blind,
retrospective data collection,
enrolment not known,
test described
LDH, Lactal dehydration; SGOT, Serum glutamic oxaloacetic transaminase
used random effects model meta-analysis.35 All statistical anal-
yses were performed using Stata 7.0 statistical package.
Results
Literature identification and study quality
Figure 1 summarises the process of literature identification
and selection. There were 18 primary articles7,9,15–29,37 that
met the selection criteria, consisting of 41 accuracy studies
(16 studies evaluating accuracy of serum uric acid in severe
pre-eclampsia, and 25 studies evaluating accuracy in various
grades of pre-eclampsia), including a total of 3913 women
(Figure 1). Each study’s salient features according to the pop-
ulation subgroups, test characteristics and reference standards
can be obtained from the authors. The definition of pre-
eclampsia differed widely between the studies. The test thresh-
olds varied from 240 to 600 mmol/l in individual studies. The
most common was 350 mmol/l, which was the threshold in 21
372
Population Test (uric acid), cutoff Outcome
Number of Inclusion criteria
patients
26 BP 140/90 after 20 weeks, 350 mmol/l Caesarean section,
proteinuria 11 on two use of any one
or more occasions antihypertensive,
use of two antihpertensives
(hydralazine added on
to beta blocker)
200 BP 140/90 after 24 weeks 60 mmol/l on 2 Intrauterine growth restriction
on two or more occasions consecutive samples (birthweight ,10th centile),
24 hours apart. or maximum fetal distress in labour,
330 mmol/l neonatal death, stillbirth
40 Diastolic BP 90, .trace protein 240 mmol/l before Perinatal death
34 weeks, 350 mmol/l
after 34 weeks
43 BP 140/80 with proteinuria 300 mmol/l Intrauterine growth restriction,
in third trimester and blood stillbirth, caesarean section
pressure returned to
normal postnatally
62 Pre-eclamptic women 240 and 350 mmol/l Eclampsia
BP .140/90 and/or
proteinuria in latter
half of pregnancy
124 BP 140/90 with proteinuria 350 mmol/l Severe pre-eclampsia
469 Not available 350 mmol/l Severe pre-eclampsia,
eclampsia
14 BP 140/90 mmHg 350 mmol/l Severe pre-eclampsia
studies. The commonest maternal and fetal outcomes assessed
were severity of hypertension (in eight studies) and small for
gestational age (in ten studies), respectively. The methodo-
logical quality of the included studies is given in Figure 2.
Uric acid to predict maternal outcomes
The pooled accuracy estimates for uric acid in predicting
maternal outcomes using different test thresholds and strati-
fied by different population subgroups are presented in
Table 2 as LR (LR+ and LR–). Eclampsia was evaluated as
an outcome in three studies using a cutoff level of 350 mmol/l.
The pooled estimates of LR+ and LR– of uric acid levels for
this threshold level were 2.1 (95% CI 1.4–3.5) and 0.38 (95%
CI 0.18–0.81), respectively (Figure 3a).
Six studies estimated the accuracy of uric acid levels more than
or equal to 350 mmol/l to predict severe hypertension. The pooled
LR for positive test in this group was 1.7 (95% CI 1.3–2.2) and for
negative test was 0.49 (95% CI 0.38–0.64) (Figure 3b).
ª RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology
 Uric acid to predict complications of pre-eclampsia

Total citations identified from electronic searches = 436

Citations excluded after screening titles and/or abstracts: n = 322

Articles of accuracy of serum uric acid in predicting maternal and fetal complications
in women with pre-eclampsia retrieved for detailed evaluation: n = 125

From electronic searches n = 114

From reference lists n = 11

Articles excluded with reasons

Inappropriate population n = 39
Data not extractable n = 28
Not a test accuracy study to predict complications of pre-eclampsia n = 24
Lack of original data i.e. reviews or letters n=9
Articles that cannot be retrieved or translated n=5
Duplicate publication n=2
Total excluded n = 107

Primary articles included in systematic review n = 18

Figure 1. Study selection process for systematic review of uric acid to predict maternal and fetal complications.

Consecutive enrolment 6 35

Prospective design 23 18

Blinding 41

Appropriate patient spectrum 17 24

Adequate test description 26 15

Adequate reference standard 24 17

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

, feature present; , feature absent, unclear or unreported

Figure 2. Quality of the included studies in the systematic review of accuracy of uric acid in predicting maternal and fetal complications.

ª RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology 373

 Thangaratinam et al.

Table 2. LRs for predicting maternal outcomes in women with pre-eclampsia using serum uric acid

Subgroups Maternal outcomes

and test
thresholds Eclampsia Severe hypertension Caesarean section HELLP syndrome
(mmol/l)
n LR1 LR2 n LR1 LR2 n LR1 LR2 n LR1 LR2
(95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI)

Overall accuracy across all studies*

350 3 2.1 0.38 6 1.7 0.49 2 2.4 0.39
(1.4–3.5) (0.18–0.81) (1.3–2.2) (0.38–0.64) (1.3–4.7) (0.20–0.76)
Subgroup in which study participants had varying degrees of pre-eclampsia
300 1 2.6 0.29
(0.82–8.5) (0.06–1.4)
350 2 3.0 0.43 6 1.7 0.49 1 2.2 0.19
(0.89–9.8) (0.19–1.0) (1.3–2.2) (0.38–0.64) (0.59–7.9) (0.02–1.7)
450 1 1.8 0.87 1 1.6 0.90
(0.86–3.7) (0.56–1.4) (0.73–3.3) (0.56–1.4)
540 1 1.8 0.93 1 1.9 0.92
(0.71–4.3) (0.60–1.4) (0.85–4.2) (0.81–1.0)
Subgroup in which study participants had severe pre-eclampsia
330
350 1 2.0 0.14 1 2.6 0.42
(0.85–4.8) (0.02–1.1) (1.2–5.4) (0.21–0.84)
520 1 1.3 0.97
(0.53–3.3) (0.65–1.4)

n 5 number of studies.
*Pooling was performed using a random effects model.

Caesarean section was studied as an outcome in four stud-
ies. The pooled LRs for a threshold level of 350 mmol/l were
2.4 (95% CI 1.3–4.7) and 0.39 (95% CI 0.20–0.76) for positive
and negative test results, respectively (Figure 3c).
Two studies evaluated the prediction of HELLP syndrome
using 450 mmol/l and 540-mmol/l threshold levels. The LRs
for predicting HELLP syndrome with a threshold of 450
mmol/l were 1.6 (95% CI 0.73–3.3) and 0.90 (95% CI 0.56–
1.4), and with a threshold of 540 mmol/l were 1.9 (95% CI
0.85–4.2) and 0.92 (95% CI 0.81–1.0).
Uric acid to predict fetal outcomes
Ten studies evaluated the accuracy of uric acid to predict
small-for-gestational-age fetus for various threshold levels
(Table 3). The pooled LRs for positive and negative tests for
threshold level of 350 mmol/l (n = 5) to predict small-for-
gestational-age fetus are 1.3 (95% CI 1.1–1.7) and 0.60 (95%
CI 0.43–0.83), respectively (Figure 4a).
Stillbirths and neonatal deaths were evaluated as an out-
come measure in seven studies. The pooled estimates of LR+
and LR– for uric acid levels more than or equal to 350 mmol/l
were 1.5 (95% CI 0.91–2.6) and 0.51 (95% CI 0.20–1.3),
respectively (Figure 4b). In the subgroup of women with
severe pre-eclampsia, the LR+ of uric acid for the above cutoff
and outcome was 1.65 (95% CI 0.82–3.3) and the correspond-
ing LR– was 0.21 (95% CI 0.03–1.7).
Four studies evaluated the accuracy of uric acid to predict
intrauterine death for various threshold levels of uric acid:
300 mmol/l (n = 1), 330 mmol/l (n = 1), 350 mmol/l (n = 2)
and 520 mmol/l (n = 1). Meta-analyses were not performed
since the studies could not be pooled due to the varied thresh-
old levels. The positive LRs for uric acid levels more than or
equal to 300 and 330 mmol/l were 2.7 (95% CI 0.71–9.8) and
2.8 (95% CI 0.42–18.3), respectively. The negative LRs for the
above thresholds were 0.13 (95% CI 0.01–2.4) and 0.28 (95%
CI 0.01–5.9).
All the above results were statistically homogenous.
Discussion
This review presents the best available evidence so far in
addressing the question of significance of uric acid levels as
a predictor of maternal and fetal complications in pre-
eclampsia. Although uric acid as a marker may be of value
in detecting pre-eclampsia,38 it has been identified as a poor
predictor of any complications of pre-eclampsia. The provi-
sion of LRs stratified by the severity of pre-eclampsia and
test thresholds will enable clinicians to understand the poor

374 ª RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology


Figure 3. LRs for predicting maternal outcomes using 350-mmol/l threshold level of serum uric acid. (a) Prediction of eclampsia, (b) prediction
of severe hypertension, (c) prediction of caesarean section.
clinical value of this test in predicting complications in
women with pre-eclampsia.
The validity of our review findings depends on the meth-
odology of the systematic review and the quality of the
individual studies included.12,30 An extensive literature
search was performed in relevant databases without any
language restrictions to minimise the possibility of missing
ª RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology
Uric acid to predict complications of pre-eclampsia
any studies. Methodological deficiencies like verification
bias, differential use of reference standards and case–control
design did not apply to the studies in the review, ensuring
inclusion of acceptable quality studies. We used summary
LRs to report the accuracy of tests in preference to summary
receiver operating characteristics curve39 that has limited
value in clinical interpretation of diagnostic information.
375
 Thangaratinam et al.

Table 3. LRs for predicting fetal outcomes in women with pre-eclampsia using serum uric acid

Subgroups and Fetal outcomes

test thresholds
(mmol/l)
Small for gestational age Stillbirths and neonatal deaths Intrauterine death

n LR1 (95% CI) LR2 (95%CI) n LR1 (95% CI) LR2 (95% CI) n LR1 (95% CI) LR2 (95% CI)

Overall accuracy across all studies*

350 5 1.3 (1.1–1.7) 0.60 (0.43–0.83) 4 1.5 (0.91–2.6) 0.51 (0.20–1.3)
Subgroup in which study participants had varying degrees of pre-eclampsia
300 1 2.2 (0.63–7.6) 0.38 (0.07–2.0) 1 2.7 (0.68–10.4) 0.14 (0.01–2.6) 1 2.7 (0.71–9.8) 0.13 (0.01–2.4)
330 1 2.3 (1.2–4.3) 0.62 (0.35–1.1) 1 2.8 (0.42–18.3) 0.28 (0.01–5.9) 1 2.8 (0.42–18.3) 0.28 (0.01–5.9)
350 3 1.4 (1.1–1.8) 0.65 (0.50–0.86) 2 1.4 (0.92–2.0) 0.63 (0.30–1.3)
450 1 0.89 (0.38–2.1) 1.0 (0.63–1.7)
540 1 0.71 (0.23–2.2) 1.0 (0.65–1.7)
Subgroup in which study participants had severe pre-eclampsia
350 2 1.0 (0.08–11.9) 0.68 (0.24–1.9) 2 1.65 (0.82–3.3) 0.21 (0.03–1.7) 1 2.1 (0.89–5.1) 0.07 (0.01–1.3)
520 1 0.65 (0.28–1.5) 1.1 (0.72–1.5) 1 1.6 (0.68–3.9) 0.94 (0.61–1.5) 1 1.5 (0.40–5.3) 0.93 (0.46–1.9)

n 5 number of studies.
*Pooling was performed using a random effects model; LR1 Likelihood ratio for positive test; LR2 Likelihood ratio for negative text.

A significant limitation of this review is the heterogeneity
noticed between individual studies with regard to popula-
tion, definition of pre-eclampsia, test thresholds, frequency
of testing, interval between the test and outcome and refer-
ence standards. This led us to analyse data within subgroups
defined by severity of pre-eclampsia and threshold levels,
resulting in the inclusion of a small number of studies in
the subgroup meta-analyses.
Caution is needed in interpreting quantitative estimation
of uric acid levels in relation to outcome. Apart from the
variations in the methods for estimating uric acid levels, levels
of uric acid could be raised due to the use of antihyperten-
sives.19 The outcomes could also be influenced by the differ-
ent therapeutic interventions such as use of antenatal steroids
in reducing respiratory distress syndrome40 and antihyperten-
sives41 that might help to reduce fetal and maternal compli-
cations. Moreover, it is not possible to be certain of the
finding where only a small number of studies exist in a sub-
group, due to imprecision.
Our review has consistently observed poor performance of
uric acid in predicting various maternal and fetal outcomes,
across various studies, settings and population. The consis-
tency of such poor performance of uric acid in predicting
complications in those with pre-eclampsia cannot be ignored.
The predictive value of a positive test was particularly poor for
some fetal outcomes like stillbirths and neonatal deaths. The
confidence intervals of the pooled LRs for a positive result
extended to 1 or less than 1, suggesting the possibility of
finding the same or more number of complications in those
with normal urate levels compared with raised levels. The
same anomaly was seen for a negative test result in predicting
stillbirths and neonatal deaths (Figure 4b).
Given our results, uric acid test does not seem to be
a clinically useful test to predict maternal or fetal compli-
cations in women with pre-eclampsia. There is no strong
evidence to justify the use of therapeutic measures like
magnesium sulphate use42 or early delivery aimed at reduc-
ing maternal and fetal complications based on the levels of
uric acid.
TIPPS (Tests in Prediction of
Pre-eclampsia Severity) review group
Shakila Thangaratinam (specialist registrar), Khaled MK
Ismail (senior lecturer/consultant), Fidelma O’Mahony
(senior lecturer/consultant), Shaughn O’Brien (professor in
obstetrics and gynaecology) affiliated to Academic Unit of
Obstetrics and Gynaecology, Keele University School of Med-
icine, University Hospital of North Staffordshire, Stoke-
on-Trent, ST4 6QG, UK.
Steve Sharp (electronic information librarian) affiliated to
NeLH Specialist Library for ENT and Audiology, Radcliffe
Infirmary, Oxford, OX2 6HE, UK.
Arri Coomarasamy (clinical lecturer/specialist registrar),
Khalid S Khan (professor in obstetrics and gynaecology) affil-
iated to Education Resource Centre, Birmingham Women’s
Hospital, Birmingham, B15 2TG, UK.

376 ª RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology


Figure 4. LRs for predicting fetal outcomes using 350-mmol/l threshold level of serum uric acid. (a) Prediction of small-for-gestational-age fetus,
(b) prediction of stillbirth and neonatal death.
Contributors
K.S.K. conceived the idea of the review and developed the
protocol with K.M.K.I., F.O.’M., A.C. and S.T. S.S. searched
the electronic databases to identify the studies. K.M.K.I.,
F.O.’M. and S.O.’B. obtained funding for S.T. from University
Hospital North Staffordshire Research and Development
Department, Stoke-on-Trent, UK.
Funding
University Hospital North Staffordshire Research and Develop-
ment Department, Stoke-on-Trent, UK (Ref No. R 5177680). j
References
1 HMSO. Report on Confidential Enquiries into Maternal Deaths in the
United Kingdom. London, UK: RCOG, 1997–99. 2001.
2 World Health Organization. The Hypertensive Disorders of Pregnancy:
Report of a WHO Study Group. Switzerland: World Health Organisa-
tion Tech Rep Service. 1987;788:1–114.
ª RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology
Uric acid to predict complications of pre-eclampsia
3 Siemons JM, Bogert LJF. The uric acid content of maternal and fetal
blood. J Biol Chem 1917;32:63–7.
4 Fadel HE, Northrop G, Misenhimer HR. Hyperuricemia in pre-eclampsia.
A reappraisal. Am J Obstet Gynecol 1976;125:640–7.
5 Chesley LC, Williams LO. Renal glomerular and tubular function in
relation to the hyperuricemia of preeclampsia and eclampsia. Am J
Obstet Gynecol 1945;50:367–75.
6 Many A, Hubel CA, Roberts JM. Hyperuricemia and xanthine oxidase in
preeclampsia, revisited. Am J Obstet Gynecol 1996;174:288–91.
7 Liedholm H, Montan S, Aberg A. Risk grouping of 113 patients with
hypertensive disorders during pregnancy, with respect to serum urate,
proteinuria and time of onset of hypertension. Acta Obstet Gynecol
Scand Suppl 1984;62(Suppl 118):43–8.
8 Redman CW, Beilin L J, Bonnar J, Wilkinson RH. Plasma-urate measure-
ments in predicting fetal death in hypertensive pregnancy. Lancet
1976;1:1370–3.
9 Lancet M, Fisher IL. The value of blood uric acid levels in toxemia of
pregnancy. J Obstet Gynaecol Br Emp 1956;63:116–19.
10 McFarlane CN. An evaluation of the serum uric acid level in pregnancy.
J Obstet Gynaecol Br Commonw 1963;70:63–8.
11 Irwig LM, Tosteton AN, Gatsonis CA, Lao J, Colditz G, Chalmers TC, et al.
Guidelines for meta-analyses evaluating diagnostic tests. Ann Intern
Med 1994;120:667–76.
12 Cochrane Methods Working Group on Systematic Reviews of
Screening and Diagnostic Tests: Recommended methods. 1996.
377
 Thangaratinam et al.
13 Deeks JJ. Systematic reviews in health care: systematic reviews of
evaluations of diagnostic and screening tests. BMJ 2001;323:157–62.
14 Khan KS, Dinnes J, Kleijnen J. Systematic reviews to evaluate diagnostic
tests. Eur J Obstet Gynecol Reprod Biol 2001;95:6–11.
15 Yassaee F. Hyperuricemia and perinatal outcomes in patients with
severe preeclampsia. Iran J Med Sci 2003;28:198–9.
16 Williams KP, Galerneau F. The role of serum uric acid as a prognostic
indicator of the severity of maternal and fetal complications in hyper-
tensive pregnancies. J Obstet Gynaecol Can 2002;24:628–32.
17 D’Anna R, Baviera G, Scilipoti A, Leonardi I, Leo R. The clinical utility of
serum uric acid measurements in pre-eclampsia and transient hyper-
tension in pregnancy. Panminerva Med 2000;42:101–3.
18 Martin JN Jr, May WL, Magann EF, Terrone DA, Rinehart BK, Blake P-G.
Early risk assessment of severe preeclampsia: admission battery of symp-
toms and laboratory tests to predict likelihood of subsequent significant
maternal morbidity. Am J Obstet Gynecol 1999;180:1407–14.
19 Odendaal HJ, Pienaar ME. Are high uric acid levels in patients with
early pre-eclampsia an indication for delivery? S Afr Med J 1997;87:
213–18.
20 Shah DM, Reed G. Parameters associated with adverse perinatal out-
come in hypertensive pregnancies. J Hum Hypertens 1996;10:511–15.
21 Magann EF, Chauhan SP, Naef RW, Blake PG, Morrison JC, Martin J.
Standard parameters of preeclampsia: can the clinician depend upon
them to reliably identify the patient with the HELLP syndrome? Aust N
Z J Obstet Gynaecol 1993;33:122–6.
22 Voto LS, Illia R, Darbon-Grosso HA, Imaz FU, Margulies M. Uric acid
levels: a useful index of the severity of preeclampsia and perinatal
prognosis. J Perinat Med 1988;16:123–6.
23 Sagen N, Haram K, Nilsen ST. Serum urate as a predictor of fetal outcome
in severe pre-eclampsia. Acta Obstet Gynecol Scand 1984;63:71–5.
24 Varma TR. Serum uric acid levels as an index of fetal prognosis in
pregnancies complicated by pre-existing hypertension and pre-
eclampsia of pregnancy. Int J Gynaecol Obstet 1982;20:401–8.
25 Mathews DD, Agarwal V, Shuttleworth TP. The effect of rest and
ambulation on plasma urea and urate levels in pregnant women with
proteinuric hypertension. Br J Obstet Gynaecol 1980;87:1095–8.
26 Dequiedt P, Tacquet A, Puech F, Cotteel M, Potier A, Leroy JL, et al. The
prognostic value of raised uric acid in the blood in arterial hypertension
in pregnancy. 58 cases (author’s transl). J Gynecol Obstet Biol Reprod
(Paris) 1979;8:115–20.
27 Fadel HE, Sabour MS, Mahran M, Seif-el DD, el-Mahallawi MN. Serum
uric acid in pre-eclampsia and eclampsia. J Egypt Med Assoc 1969;
52:12–23.
378
28 Connon AF, Wadsworth RJ. An evaluation of serum uric acid estim-
ations in toxaemia of pregnancy. Aust N Z J Obstet Gynaecol 1968;8:
197–201.
29 Seitchik J. Observations on the renal tubular reabsorption of uric acid. I.
Normal pregnancy and abnormal pregnancy with and without pre-
eclampsia. Am J Obstet Gynecol 1953;65:981–5.
30 Lijmer JG, Mol BW, Heisterkamp S, Bonsel GJ, Prins MH, van der
Meulen JHP, et al. Empirical evidence of design-related bias in studies
of diagnostic tests. JAMA 1999;282:1061–6.
31 Rennie D. Improving reports of diagnostic tests: the STARD initiative.
JAMA 2003;289:89–90.
32 Petitti DB. Meta-analysis, Decision Analysis and Cost Effectiveness
Analysis. Oxford, UK: Oxford University Press, 1994.
33 Jaeschke R, Guyatt GH, Sackett DL. Users’ guide to the medical liter-
ature. III. How to use an article about a diagnostic test. B. What are the
results and will they help me in caring for my patients? JAMA 1994;
271:703–7.
34 DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin
Trials 1986;7:177–88.
35 Sutton AJ, Abrams KR, Jones DR, Sheldon TJ, Song F. Systematic
reviews of trials and other studies. Health Technol Assess 1998;2:
1–276.
36 Greenland S. Quantitative methods in the review of epidemiologic
literature. Epidemiol Rev 1987;9:1–30.
37 Brown M, Buddle ML. Hypertension in pregnancy: maternal and fetal
outcomes according to laboratory and clinical features. Med J Aust
1996;165:360–5.
38 Chappell LC, Seed PT, Briley A, Kelly FJ, Beverley J, Charnock-Jones DS,
et al. A longitudinal study of biochemical variables in women at risk of
preeclampsia. Am J Obstet Gynecol 2002;187:127–36.
39 Moses LE, Shapiro D, Littenberg B. Combining independent studies
of a diagnostic test into a summary ROC curve: data analytic
approaches and some additional considerations. Stat Med 1993;12:
1293–316.
40 Crowley P. Prophylactic corticosteroids for preterm birth. In: The
Cochrane Library. Oxford, UK: Update Software; 2001.
41 Rey E, Le Lorier J, Burgess E, Lange IR, Leduc L. Report of the Canadian
Hypertension Society Consensus Conference: 3. Pharmacologic
treatment of hypertensive disorders in pregnancy. CMAJ 1997;157:
1245–54.
42 The Magpie Trial Collaborative Group. Do women with pre-eclampsia,
and their babies, benefit from magnesium sulphate? The Magpie Trial:
a randomised placebo-controlled trial. Lancet 2002;359:1877.
ª RCOG 2006 BJOG An International Journal of Obstetrics and Gynaecology