Rheumatology 2017;56:1378–1385

RHEUMATOLOGY doi:10.1093/rheumatology/kex168
Advance Access publication 27 April 2017

Original article
Postpartum complications in new mothers with

Downloaded from https://academic.oup.com/rheumatology/article/56/8/1378/3782591 by Korea national university of transportation user on 30 October 2020
juvenile idiopathic arthritis: a population-based
cohort study
Debbie Ehrmann Feldman1,2,3,4, Évelyne Vinet5, Marie-Pierre Sylvestre1,6,
Beth Hazel5, Ciarán Duffy7,8, Anick Bérard9,10, Garbis Meshefedjian11 and
Sasha Bernatsky5

Abstract
Objective. The aim was to evaluate the prevalence of postpartum complications, including depression, in
new mothers who had juvenile idiopathic arthritis (JIA) and to assess whether these differ from mothers
who never had JIA.
Methods. Our cohort study used data from physician billing and hospitalizations covering Quebec,
Canada. We identified females with JIA with a first-time birth between 1 January 1983 and
31 December 2010 and assembled a control cohort of first-time mothers without JIA from the same
administrative data, matching 4:1 for date of first birth, maternal age and area of residence. We compared
the following postpartum complications: major puerperal infection, anaesthetic complications, postpartum
haemorrhage, thromboembolism, obstetrical trauma, complications of obstetrical surgical wounds and
maternal depression in the first year after delivery, in the JIA vs non-JIA groups, using bivariate analysis
and multiple logistic regression.
Results. The mean age at delivery was 24.7 years in the JIA group (n = 1681) and 25.0 years for the non-
JIA group (n = 6724). Mothers with JIA were more likely to experience complications attributable to an-
aesthetic [adjusted risk ratio (aRR) 2.17, 95% CI; 1.05, 4.48], postpartum haemorrhage (aRR = 2.75, 95%
CI: 2.42, 3.11) and thromboembolism (aRR = 5.27, 95% CI: 1.83, 15.17) but were at lower risk for
CLINICAL
SCIENCE

obstetrical trauma (aRR = 0.78, 95% CI: 0.64, 0.95) or newly to develop depression in the first year
postpartum (aRR = 0.52, 95% CI: 0.40, 0.68).
Conclusion. Mothers with JIA appear to be at higher risk for complications attributable to anaesthesia,
postpartum haemorrhage and thromboembolism. Prevention strategies for postpartum haemorrhage and
thromboembolism may be especially important in this population.
Key words: juvenile arthritis, pregnancy, postpartum haemorrhage, postpartum depression, cohort study, post-
partum complications, postpartum thromboembolism, administrative data, obstetrical complications, risk

Rheumatology key messages

. Mothers with JIA are at risk for complications from anaesthesia, postpartum hemorrhage and thromboembolism.
. Prevention for postpartum haemorrhage and thromboembolism may be important for mothers with JIA.
. There was no increase in postpartum depression among mothers with JIA.

10
1 Research Centre CHU Ste-Justine and 11Department of Public
School of Rehabilitation, Faculty of Medicine, 2Department of Social
Health City of Montreal, Montreal, Quebec, Canada
and Preventive Medicine, School of Public Health, 3Institute of
Research in Public Health, Université de Montréal, 4Centre for
Submitted 23 November 2016; revised version accepted
Interdisciplinary Research in Rehabilitation, 5Department of Medicine,
23 March 2017
Division of Rheumatology, McGill University Health Centre, 6Research
Centre of the Université de Montréal Hospital Centre, Montreal, Correspondence to: Debbie Ehrmann Feldman, Université de
Quebec, 7Department of Paediatrics, Children’s Hospital of Eastern Montréal, École de réadaptation, Pavillon 7077 du Parc, C.P. 6128,
Ontario, 8Department of Paediatrics, Faculty of Medicine, University of Succ. Centre-Ville, Montréal, Québec, Canada H3C 3J7.
Ottawa, Ottawa, Ontario, 9Faculty of Pharmacy, University of Montreal, E-mail: debbie.feldman@umontreal.ca

! The Author 2017. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com
 Postpartum outcomes in JIA

Introduction assess whether women with ongoing disease were more

likely to experience postpartum complications. In order to
Studies on pregnancy and birth outcomes in women with do so, we compared mothers with JIA who had seen a
rheumatological conditions, including juvenile idiopathic rheumatologist in the year before delivery with those with
arthritis (JIA), have focused for the most part on pregnancy JIA who had not seen a rheumatologist.
outcomes (e.g. gestational hypertension and diabetes) and

Downloaded from https://academic.oup.com/rheumatology/article/56/8/1378/3782591 by Korea national university of transportation user on 30 October 2020
birth outcomes (e.g. congenital malformations, small for
gestational age, prematurity). Reports confirm increased Methods
prevalence of prematurity, small for gestational age, con- Population
genital malformations and maternal hypertension in women
with inflammatory arthritis, including JIA [1–5]. There is little We conducted a retrospective cohort study, identifying all
information regarding maternal postpartum complications women with JIA who gave birth for the first time between
in women with inflammatory arthritis conditions and only 1 January 1983 and 31 December 2010. The data were
one on persons with JIA. In that study, Chen et al. [6] re- obtained from three linked administrative databases that
ported higher rates of postpartum haemorrhage and severe cover the entire population of residents of the province of
maternal morbidity among mothers with a history of JIA Quebec, Canada. These databases were: the physician
compared with the general population. billing database, which records information on physician
Gayed and Gordon [7] reviewed pregnancy and the visits, diagnosis and procedures; the hospitalization dis-
rheumatic diseases, noting that postpartum flare was charge database, which records hospitalization data,
common in all rheumatic diseases (although JIA was not including gestational age; and the demographic events
included in this review). A few studies addressed postpar- database, which provides demographic information on
tum outcomes in women with specific inflammatory arth- the mother and father as well as the birth weight for live
ritic conditions. Women with lupus appear to be at risk for births and stillbirths. We considered women to have ju-
postpartum haemorrhage as well as for thromboembolic venile arthritis if they had at least three physician visits
events [8]. In AS, there was no increase in complications with the International Classification of Diseases (ICD) 9
attributable to epidural analgesia, but these women were code 714 and ICD10 codes M08, M09 between 1
much more likely to undergo Caesarean section than con- January 1983 and 31 December 2010 and were 16
trols [9]. In a study of 49 women in Sweden (10 with RA years of age or younger at the time of the first JIA diag-
and 10 with AS and 29 age-matched healthy controls) nosis. We followed them until 31 December 2010 in order
[10], none of the women in this small sample developed to identify those who gave birth for the first time (stillbirth
postpartum depression within 24 weeks post-delivery. In or live birth). We compiled a comparison cohort of women
contrast, women with all types of disabilities had a greater who had no physician visits with the ICD9 or 10 codes
likelihood of developing postpartum depression in com- listed above, matching four first-time mothers without
parison with those without a disability [11]. Factors asso- JIA for each woman with juvenile arthritis who gave
ciated with postpartum depression include adverse birth birth, on date of delivery (±3 months), age (±5 years) and
outcome (prematurity, low birth weight, infant illness/dis- region of residence (using the first three digits of the six-
ability), diabetes [12], younger maternal age [13], having a digit postal code).
previous history of depression and experiencing stressful The outcomes of interest for this study were the follow-
life events during pregnancy and, to a lesser extent, low ing postpartum complications: major puerperal infection
socio-economic status [14]. (ICD9 670, 673.3; ICD10 O8689, O8832), anaesthetic
The association between Caesarean section and post- complications (ICD9 668; ICD10 O741, O8909, O742,
partum depression is unclear, with conflicting results O891, O743, O748, O749, O898, O899), postpartum
[15, 16]. However, Caesarean delivery is associated with haemorrhage (ICD9 666 and procedural codes
a higher incidence of major puerperal infection, thrombo- 99.03–99.09; ICD10 O720, O721, O722, O723 and pro-
embolic events, anaesthetic complications and obstetrical cedural codes 30233, 30243, 30253, 30263), thrombo-
surgical wound complications [17]. Also, having a baby embolism (ICD9 451, 452, 453, 671.3, 671.4, 671.5;
with high birth weight (macrosomia) may be associated 415.1, 673.2, 673.8 or ICD10 I80, I81, I822, I823, O225,
with obstetrical trauma and other adverse outcomes O228, O871, O873; I269, O882, O888), obstetrical trauma
[18, 19]. Other factors that may be associated with com- (ICD9 665.3, 665.4, 665.5, 665.6, 665.8, 665.9; ICD10
plications include hypertension and heart disease, O713, O714, O715, O716, O717, O7189, O719), compli-
eclampsia, maternal demographics and access/use of cations of obstetrical surgical wounds (ICD9 674.3; ICD10
prenatal care [20–22]. O860, O902) and postpartum depression in the first year
Our objective was to evaluate the prevalence of postpar- following delivery (ICD9 684.4, 296, 300, 309, 311; ICD10
tum complications in first-time mothers with JIA in com- O90.6, O99.34, O99.345, F53, F32.9, F30–33, F34.1,
parison with first-time mothers who never had JIA. F41.9, F43.2). The codes for complications except for
Specifically, we looked at major puerperal infection, anaes- thromboembolism and maternal depression were adapted
thetic complications, postpartum haemorrhage, thrombo- from a previous study [17]. Thromboembolism codes used
embolism, obstetrical trauma, complications of obstetrical in this study were the ones described by Liu et al. [23] in
surgical wounds and depression in the first year after de- a study on the general Canadian population and
livery. A secondary (more exploratory) objective was to included deep vein thrombosis and pulmonary embolism.

www.rheumatology.oxfordjournals.org 1379
Debbie Ehrmann Feldman et al.

For maternal depression, we included the specific codes have complications attributable to anaesthetic, postpar-
for depression in the immediate postpartum period and tum haemorrhage and thromboembolism but less likely
also adapted the definitions by Brownell et al. [24] and to have complications of obstetrical surgical wounds
Yang et al. [16] for depression within the year after and depression in the first year after delivery (Table 2).
childbirth. The results of the logistic regression models for each
outcome are described in Table 3. Logistic regression

Downloaded from https://academic.oup.com/rheumatology/article/56/8/1378/3782591 by Korea national university of transportation user on 30 October 2020
Analysis confirmed that mothers with JIA were more likely to ex-
The analysis consisted of comparisons for each of the perience complications attributable to anaesthetic [ad-
outcomes, by exposure status (having juvenile arthritis justed risk ratio (aRR) = 2.17, 95% CI: 1.05, 4.48] or
or not), using bivariate 2 analysis and multiple logistic postpartum haemorrhage (aRR = 2.75, 95% CI: 2.42,
regression. For each of the logistic regression models, 3.11) but were at lower risk for obstetrical trauma (aRR
we adjusted for covariates. These included Caesarean = 0.78, 95% CI: 0.64, 0.95), postpartum depression in the
section, socio-economic status (using the proxy variable first year (aRR = 0.49, 95% CI: 0.39, 0.63) or newly to
of maternal education), hypertension, diabetes, adverse develop depression in the first year postpartum (aRR =
birth outcome (at least one of the following: stillbirth, pre- 0.52, 95% CI: 0.40, 0.68). Owing to small numbers for
maturity, small for gestational age or congenital malforma- thromboembolism, we calculated separate models that
tion) and giving birth to a baby with macrosomia, defined included JIA and one other covariate (adjusting for pre-
as having a birth weight 54000 g [25]. For thromboembol- eclampsia, Caesarean section or age). Results of these
ism, we adjusted for age, Caesarean section and pre- bivariate logistic regression models indicated aRRs
eclampsia [26]. For maternal depression, we calculated for JIA of 5.27 (95% CI: 1.83, 15.17) adjusting for pre-
two models: having a new diagnosis of depression eclampsia, 5.29 (95% CI: 1.84, 15.21) adjusting for
(i.e. no diagnosis within 2 years before delivery); and any Caesarean section and 5.36 (95% CI: 1.63, 15.43) adjust-
diagnosis of depression after delivery, with adjustment for ing for age. Caesarean section was associated with a
depression in the 2 years before delivery in addition to the lower risk for haemorrhage or obstetrical trauma, and
other covariates. macrosomia was associated with a higher risk of haem-
For our secondary objective, we identified mothers with orrhage. Having a history of depression was strongly
JIA who had seen a rheumatologist within 1 year before associated with development of depression in the
delivery. The rationale was that women who were being 12 months postpartum (aRR = 3.56, 95% CI: 2.99, 4.23).
followed by a rheumatologist would be more likely to have Among the 1681 mothers with JIA, 107 (6.4%) saw a
symptoms or problems related to their arthritis that rheumatologist within 12 months of giving birth, possibly
required treatment and follow-up. We compared the out- indicative of ongoing disease or treatment for JIA. Those
comes between mothers with JIA who had seen a who saw a rheumatologist within a year of giving birth
rheumatologist within 12 months of delivery with those were older than those who had not seen a rheumatologist
with JIA who had not visited a rheumatologist within that [mean (S.D.) age 26.2 (4.9) vs 24.6 (4.2) years, P < 0.01],
time period, using bivariate comparisons, and if the num- more likely to have hypertension or heart disease (14.9 vs
bers were large enough, we also conducted multiple lo- 8.1%, P = 0.01) and to deliver via Caesarean section (15.9
gistic regression, adjusting for the same covariates used vs 8.1%, P = 0.01). We found a higher proportion of anaes-
in the primary analysis. thetic complications, postpartum haemorrhage and
We received ethics approval from the Quebec commis- thromboembolism among mothers with JIA who had
sion for access to information as well as the CERES,  The
seen a rheumatologist in the 12 months before delivery
Health Research Ethics Committee at the Université de vs those who did not see a rheumatologist (Table 4).
Montréal. In multivariate analysis, adjusting for covariates,
mothers with JIA who had seen a rheumatologist within
Results 12 months of delivery were more likely to experience post-
partum haemorrhage (aRR = 1.99, 95% CI: 1.53, 2.60)
Our cohort consisted of 8405 women who experienced a
compared with mothers with JIA who had not seen a
first birth: 1681 women with JIA and 6724 women who
rheumatologist within 12 months of delivery. We were
never had JIA. For the entire cohort, the mean (S.D.) age
unable to perform multivariate analysis for anaesthetic
at delivery was 24.9 (4.4) years and the age range at
complications and thromboembolism because of the
delivery was 16–46 years. There was a higher proportion
smaller number of events for these outcomes.
of hypertension and heart disease among the JIA group
(8.5%, 95% CI: 7.3, 9.9) than the non-JIA group (4.6%,
95% CI: 4.2, 5.2), and there were more adverse neonatal Discussion
outcomes in the JIA group (28.8%, 95% CI: 26.6, 31.0) vs
the non-JIA group (18.9%, 95% CI: 18.0, 19.8). The rates Women with JIA were at higher risk for several postpartum
of Caesarean section were similar for both groups. complications, including anaesthetic complications, post-
Descriptive characteristics of the JIA (n = 1681) and non- partum haemorrhage and thromboembolism. Risks for
JIA groups (6724) are listed in Table 1. these outcomes were higher for women who had visited
The outcomes in the JIA and non-JIA groups are re- their rheumatologist in the 12 months before giving birth.
ported in Table 2. Women with JIA were more likely to Mothers with JIA were at lower risk for obstetrical trauma

1380 www.rheumatology.oxfordjournals.org
 Postpartum outcomes in JIA

TABLE 1 Characteristics of the cohort (n = 8405)

Exposed (JIA) Non-exposed (non-JIA)

(n = 1681) (n = 6724) P-value

Age at delivery, years 0.04

Downloaded from https://academic.oup.com/rheumatology/article/56/8/1378/3782591 by Korea national university of transportation user on 30 October 2020
Mean (S.D.) 24.7 (4.3) 25.0 (4.5)
Median (range) 24.0 (16.0–40.0) 25.0 (16.0–46.0)
Length of stay, days 0.70
Mean (S.D.) 3.2 (1.8) 3.2 (2.2)
Median (range) 3.0 (0.0–30.0) 3.0 (0.0–74.0)
Female baby, n (%) 849 (50.5) 3207 (47.7) 0.04
Lower level of education,a n (%) 873 (51.9) 3424 (50.9) 0.46
Hypertension/heart disease,b n (%) 143 (8.5) 311 (4.6) <0.001
Diabetes,b n (%) 15 (0.9) 38 (0.6) 0.13
Caesarean section, n (%) 144 (8.6) 561 (8.4) 0.77
Macrosomia, n (%) 99 (5.9) 383 (5.7) 0.66
Adverse neonatal outcome, n (%) 485 (28.8) 1269 (18.9) <0.001

a
Less than 14 years. bBefore or during pregnancy.

TABLE 2 Postpartum complications in the JIA and non-JIA groups

JIA, % (95% CI) Non-JIA, % (95% CI)

Complication (n = 1681) (n = 6724) P-value

Major puerperal infection 3.15 (2.32, 3.99) 2.99 (2.58, 3.40) 0.73
Thromboembolic events 0.48 (0.15, 0.80) 0.09 (0.02, 0.16) 0.001
Anaesthetic complications 0.65 (0.27, 1.04) 0.33 (0.19, 0.46) 0.05
Postpartum haemorrhage 21.06 (19.11, 23.01) 7.66 (7.02, 8.29) <0.001
Obstetrical trauma 6.72 (5.53, 7.92) 8.69 (8.01, 9.36) 0.01
Complications of obstetrical surgical wounds 1.49 (0.91, 2.07) 1.55 (1.25, 1.84) 0.86
Depression in the first year after delivery 4.16 (3.21, 5.12) 8.25 (7.60, 8.91) <0.0001
New depression in first year after deliverya 3.45 (2.58, 4.32) 6.50 (5.91, 7.09) <0.0001

a
No depression in the 2 years before delivery.

and for developing depression in the period 1 year post- reported by Roberts et al. [28] in New South Wales,
partum compared with mothers without JIA. Australia (6.8%), and in Canada excluding Quebec
Anaesthetic complications in our study include pulmon- (6.1%) [29]. Possible mechanisms may be related to use
ary, cardiac, nervous system and other complications of NSAIDs during pregnancy, which can be linked to ex-
related to administration of anaesthetic or sedation in cessive bleeding [30].
labour or delivery. There was a higher proportion of Incidence of pregnancy-related thromboembolism
mothers with JIA (0.65%) than non-JIA mothers (0.33%) (deep vein thrombosis and pulmonary embolism) in
who had these complications. In a population-based Canada were reported as 12.1 and 5.4/10 000 births, re-
study in California, the proportion of mothers with these spectively [23]. In our study, the rates for deep vein throm-
complications was 0.31% [27], which is only slightly lower bosis and pulmonary embolism were comparable at
than the proportion for our entire sample, which was 13 and 4/10 000. Although there were few thrombo-
0.39%. embolic events, mothers with JIA were much more likely
Our results on postpartum haemorrhage among women to suffer these outcomes compared with mothers who did
with JIA (21.06%) are slightly higher than those reported not have JIA. Mothers with lupus have an increased risk
by Chen et al. [6] (18%); our aRR of 2.75 is similar to their for thromboembolism, but that study included events
adjusted odds ratio of 2.45. Mothers with lupus were also during pregnancy as well as in the postpartum period
more likely to have postpartum haemorrhage compared [20]. Two recent studies, one in the UK and one in
with mothers who did not have lupus, and the adjusted Taiwan, found that persons (mean age of 58 and
odds ratio of 2.7 (95% CI: 1.8, 4.0) was almost identical to 52 years, respectively) with a diagnosis of RA were at a
what we found [8]. The rate of postpartum haemorrhage in 2- to 3-fold increased risk for deep vein thrombosis or
our non-JIA group (7.6%) was slightly higher than that pulmonary embolism, although there is no information

www.rheumatology.oxfordjournals.org 1381
Debbie Ehrmann Feldman et al.

regarding the effect of pregnancy and childbirth [31, 32].

0.68)
1.00)
1.08)
1.57)
1.66)
1.38)
New case of

postpartum
depression
Possible mechanisms include the influence of inflamma-

(0.69,
(0.74,
(0.87,
(0.84,
(0.91,
(0.40,
tion on blood coagulation and also endothelial dysfunction

–
–
[33, 34].

0.83
0.89
1.17
1.18
1.12
0.52
There was less obstetrical trauma among the JIA group
(6.7%) vs the non-JIA group (8.7%). The proportion of

Downloaded from https://academic.oup.com/rheumatology/article/56/8/1378/3782591 by Korea national university of transportation user on 30 October 2020
0.63)

4.23)
1.10)
1.14)
1.39)
1.42)
1.39)
obstetrical trauma is higher than the 6.1% found in the
Depression in
the first year
postpartum
general Ohio population 1991–96 [17]. Factors associated

(0.79,
(0.82,
(0.83,
(0.77,
(0.97,
(0.39,

3.56 (2.99,
with obstetrical trauma include speed at which the head is

–
0.93 delivered, fetal size, use of forceps or vacuum extraction
0.97
1.07
1.04
1.16
0.49

and episiotomy [35]. We did not have information on many

of these factors that could potentially explain this finding.
surgical wound

3.28)
1.50)
1.64)
1.16)

2.07)
1.77)
1.36)
Interestingly, we found that women with JIA are at con-
complication
Obstetrical

siderably lower risk for developing postpartum depres-

(1.24,
(0.63,
(0.76,
(0.54,

(0.50,
(0.77,
(0.19,
– sion. There may be under-reporting of this diagnosis in
administrative databases; the prevalence in our cohort
0.97
1.12
0.79

1.01
1.17
0.50
2.02

was 5.9%, whereas Canadian studies that used the

Edinburgh Postnatal Depression Scale to assess postpar-
0.95)
0.91)

0.22)
1.14)

1.26)
1.10)
1.28)

tum depression indicate higher frequencies between 7.5

Obstetrical

and 8.7% [36, 37]. Although not addressing postpartum

trauma

0.97 (0.83,

0.91 (0.66,
0.92 (0.77,
0.92 (0.66,
0.78 (0.64,
0.78 (0.67,

0.10 (0.04,

depression, Packham et al. [38] reported lower depression

–

among their cohort of adults with JIA (mean age 35.4

years) compared with the general population. In other
rheumatic diseases, there have been no reports of
3.11)

1.30)
0.93)

1.64)
1.08)

1.25)
1.10)
haemorrhage

increased postpartum depression either, although there

Postpartum

is evidence of increased depression in persons with in-

0.94 (0.82,

0.98 (0.77,
0.95 (0.81,
2.75 (2.42,

1.14 (1.00,
0.72 (0.55,

1.31 (1.04,

flammatory arthritis [39, 40]. In fact, findings from a

–

study on 20 pregnant patients (10 with RA and 10 with

TABLE 3 Factors associated with postpartum complications: adjusted risk ratio (95% CIs)

AS) indicated that mental health scores on the SF-36 re-

mained stable throughout pregnancy and up to 24 weeks
4.48)
2.61)
2.38)
4.74)
3.37)
1.42)
3.67)
complication

postpartum, even among those with postpartum disease

Anaesthetic

flares [10]. This finding was confirmed in women with RA

(0.56,
(0.53,
(0.70,
(0.06,
(0.17,
(0.21,
(1.05,

and JIA in a Norwegian study [41].

Our results may also imply that women with ongoing
1.21
1.13
1.82
0.46
0.49
0.87
2.17

disease (defined in our study as those who saw a rheuma-

tologist in the year before giving birth) may be at higher
Thromboembolic

5.29 (1.84, 15.21)

5.99 (2.02, 17.79)

Values in bold are statistically significant (the relative risks exclude 1).

risk for anaesthetic complications and postpartum haem-

orrhage. It is possible that ongoing inflammation might
event

increase the risk for these outcomes.

–
–

–
–
–

There are several limitations to this retrospective admin-

istrative data-based study. Although we are confident of
the validity of our diagnosis, it is based on administrative
1.78)

2.23)

2.52)
1.44)

1.24)

1.49)
1.21)

physician visits for a diagnosis of inflammatory arthritis.

puerperal

We have previously used a similar algorithm, and mean

infection
Major

(1.03,

(1.10,

(1.13,
(0.79,

(0.73,

(0.51,
(0.64,

age at diagnosis and sex distribution statistics were simi-

–

lar to those reported for JIA in general, which may support

1.07

0.96

0.88
0.88
1.36

1.57

1.68

the validity of our approach to case definition [42, 43].

In this study, we were even stricter in our definition of a
Depression 2 years pre-delivery

case by including only those who had at least three coded

visits (as opposed to at least two coded visits) as cases.
Adverse neonatal outcome
Hypertension or diabetes

Validity of JIA diagnosis improves with subsequent arth-

ritis codes [44]. Also, our outcomes are based on diag-
Education 514 years
Factor

nostic and procedural codes from the administrative

Caesarean section

databases. Errors in diagnosis are likely to be non-differ-

Age 525 years

ential between those with JIA and those with no JIA.

Macrosomia

Furthermore, our results (e.g. postpartum haemorrhage,

anaesthetic complications, thromboembolism) are similar
JIA

to those reported in other studies and jurisdictions

[6, 23, 27, 28].

1382 www.rheumatology.oxfordjournals.org
 Postpartum outcomes in JIA

TABLE 4 Postpartum complications in mothers with and without rheumatologist visit within 12 months of giving birth

Rheumatology visit, RR No rheumatology visit, RR

(95% CI) (95% CI)
Complication (n = 107) (n = 1574) P-value

Downloaded from https://academic.oup.com/rheumatology/article/56/8/1378/3782591 by Korea national university of transportation user on 30 October 2020
Major puerperal infection 1.87 (0.00, 4.44) 3.24 (2.37, 4.11) 0.43
Thromboembolic events 1.87 (0.00, 4.44) 0.38 (0.08, 0.69) 0.031
Anaesthetic complications 5.61 (1.25, 9.97) 0.32 (0.04, 0.60) <0.0001
Postpartum haemorrhage 39.25 (30.00, 48.50) 19.82 (17.85, 21.79) <0.0001
Obstetrical trauma 4.67 (0.67, 8.67) 6.86 (5.61, 8.11) 0.38
Complications of obstetrical surgical wounds 0 1.59 (0.97, 2.21) 0.19
Depression in first year post-delivery 3.74 (0.14, 7.33) 4.19 (3.20, 5.18) 0.82
New depression in first year post- deliverya 3.74 (0.14, 7.33) 3.43 (2.53, 4.33) 0.87

a
No depression in the 2 years before delivery. RR: risk ratio.

As is the case with retrospective studies, we did not Conclusion

have information on some variables that are not available Mothers with JIA appear to be at higher risk for compli-
in the administrative databases, notably obesity, physical cations from anaesthesia, postpartum haemorrhage and
activity or inactivity and lifestyle factors, such as diet and thromboembolism. Prevention strategies, especially for
smoking. However, we controlled for maternal education, postpartum haemorrhage and also for thromboembolism,
which is associated with lifestyle [45]. Furthermore, per- may be important for women with JIA who are giving birth.
sons with JIA may be less likely to smoke [46]. Also, al- Physicians need to be aware of the higher risk for these
though our study includes the largest sample size to date adverse events in the postpartum period among mothers
on JIA and pregnancy outcomes, the numbers for some of with JIA.
the adverse outcomes were small (e.g. thromboembolic
events, anaesthetic complications). Funding: This study was funded by the Canadian Initiative
Finally, we did not have information on medications, for Outcomes in Rheumatology Care.
which may be associated with several outcomes. Also,
we have no indication of the clinical severity of JIA. In an Disclosure statement: The authors have declared no
attempt to explore the aspect of women having ongoing conflicts of interest.
problems attributable to their disease, we compared those
women who had visited a rheumatologist in the 12 months
before delivery with those who did not, finding a higher rate References
of adverse outcomes among those who presumably had 1 Nili F, McLeod L, O’Connell C, Sutton E, McMillan D.
ongoing problems attributable to their arthritis. Maternal and neonatal outcomes in pregnancies compli-
Our findings may have clinical implications for preven- cated by systemic lupus erythematosus: a population-
tion. The risk for anaesthetic complications is small, but based study. J Obstet Gynaecol Can 2013;35:323–8.
nevertheless, being aware of the fact that women with JIA 2 Nørgaard M, Larsson H, Pedersen L et al. Rheumatoid
are at increased risk for anaesthetic complications may be arthritis and birth outcomes: a Danish and Swedish na-
important for anaesthetists and obstetricians. For preven- tionwide prevalence study. J Intern Med 2010;268:329–37.
tion of postpartum haemorrhage, the current recommen- 3 Wallenius M, Salvesen KÅ, Daltveit AK, Skomsvoll JF.
dation by the World Health Organization is that women Rheumatoid arthritis and outcomes in first and subsequent
should be offered uterotonics (e.g. oxytocin 10 IU i.m. or births based on data from a national birth registry. Acta
i.v.) during the third stage of labour and also, possibly, Obstet Gynecol Scand 2014;93:302–7.
controlled cord traction [47]. General recommendations 4 Ehrmann FD, Vinet É, Bernatsky S et al. Birth outcomes in
for reducing the risk of thromboembolism are to encour- women with a history of juvenile idiopathic arthritis.
age early mobilization and prevent dehydration [48]. J Rheumatol 2016;43:804–9.
Pharmacological prophylaxis recommendations are 5 Feldman DE, Vinet E, Bérard A et al. Heart disease,
based for the most part on expert consensus and relate hypertension, gestational diabetes and pre-eclampsia/
to women who have known risk factors (e.g. known circu- eclampsia in mothers with juvenile arthritis: a nested case-
latory disease, APS, inflammatory conditions; although control study. Arthritis Care Res 2017;69:306–9.
JIA is not specified) or those who are undergoing surgery, 6 Chen JS, Ford JB, Roberts CL, Simpson JM, March LM.
including Caesarean section. The preventive strategies for Pregnancy outcomes in women with juvenile idiopathic
postpartum haemorrhage and for thromboembolism may arthritis: a population-based study. Rheumatology
be especially important for women with JIA, and there are 2013;52:1119–25.
implications for research regarding specific parameters in 7 Gayed M, Gordon C. Pregnancy and rheumatic diseases.
women with inflammatory conditions. Rheumatology 2007;46:1634–40.

www.rheumatology.oxfordjournals.org 1383
Debbie Ehrmann Feldman et al.

8 Yasmeen S, Wilkins EE, Field NT, Sheikh RA, Gilbert WM. 24 Brownell M, Santos R, Kozyrskyj A et al. Next steps in the
Pregnancy outcomes in women with systemic lupus ery- provincial evaluation of the BabyFirst Program: measuring
thematosus. J Matern Fetal Med 2001;10:91–6. early impacts on outcomes associated with child
9 Jakobsson GL, Stephansson O, Askling J, Jacobsson LT. maltreatment. Winnipeg, MB: Manitoba Centre for Health
Pregnancy outcomes in patients with ankylosing spon- Policy, 2007.
dylitis: a nationwide register study. Ann Rheum Dis 25 Mayo Clinic Staff. Fetal Macrosomia. Mayo Clinic 2015.

Downloaded from https://academic.oup.com/rheumatology/article/56/8/1378/3782591 by Korea national university of transportation user on 30 October 2020
2016;75:1838–42. URL:http://www.mayoclinic.org/diseases-conditions/
10 Förger F, Østensen M, Schumacher A, Villiger PM. Impact fetal-macrosomia/basics/definition/con-20035423 (30
of pregnancy on health related quality of life evaluated July 2016, date last accessed).
prospectively in pregnant women with rheumatic diseases 26 Tepper NK, Boulet SL, Whiteman MK et al. Postpartum
by the SF-36 health survey. Ann Rheum Dis venous thromboembolism: incidence and risk factors.
2005;64:1494–9. Obstet Gynecol 123:987–96.
11 Mitra M, Iezzoni LI, Zhang J et al. Prevalence and risk 27 El Haj Ibrahim S, Fridman M, Korst LM, Gregory KD.
factors for postpartum depression symptoms among Anesthesia complications as a childbirth patient safety
women with disabilities. Matern Child Health J indicator. Anesth Analg 2014;119:911–7.
2015;19:362–72. 28 Roberts CL, Ford JB, Algert CS et al. Trends in adverse
12 Kozhimannil KB, Pereira MA, Harlow BL. Association be- maternal outcomes during childbirth: a population-based
tween diabetes and perinatal depression among low- study of severe maternal morbidity. BMC Pregnancy
income mothers. JAMA 2009;301:842–7. Childbirth 9:7.
13 Rich-Edwards JW, Kleinman K, Abrams A et al. 29 Mehrabadi A, Liu S, Bartholomew S et al. Temporal trends
Sociodemographic predictors of antenatal and postpar- in postpartum hemorrhage and severe postpartum hem-
tum depressive symptoms among women in a medical orrhage in Canada from 2003 to 2010. J Obstet Gynaecol
group practice. J Epidemiol Community Health 60:221–7. Can 36:21–33.
14 Robertson E, Grace S, Wallington T, Stewart DE. 30 Risser A, Donovan D, Heintzman J, Page T. NSAID pre-
Antenatal risk factors for postpartum depression: a syn- scribing precautions. Am Fam Physician 80:1371–8.
thesis of recent literature. Gen Hosp Psychiatry 31 Choi HK, Rho YH, Zhu Y et al. The risk of pulmonary
26:289–95. embolism and deep vein thrombosis in rheumatoid arth-
15 Carter FA, Frampton CM, Mulder RT. Cesarean section ritis: a UK population-based outpatient cohort study. Ann
and postpartum depression: a review of the evidence Rheum Dis 72:1182–7.
examining the link. Psychosom Med 68:321–30. 32 Chung WS, Peng CL, Lin CL et al. Rheumatoid arthritis
16 Yang SN, Shen LJ, Ping T, Wang YC, Chien CW. The increases the risk of deep vein thrombosis and pulmonary
delivery mode and seasonal variation are associated with thromboembolism: a nationwide cohort study. Ann Rheum
the development of postpartum depression. J Affect Dis 73:1774–80.
Disord 132:158–64. 33 Borensztajn KS, von der Thusen JH, Spek CA. The role of
17 Koroukian SM. Relative risk of postpartum complications coagulation in chronic inflammatory disorders: a jack of all
in the Ohio Medicaid population: vaginal versus cesarean trades. Curr Pharm Des 2011;17:9–16.
delivery. Med Care Res Rev 2004;61:203–24. 34 Zoller B, Li X, Sundquist J, Sundquist K. Risk of pulmonary
18 Stotland NE, Caughey AB, Breed EM, Escobar GJ. Risk embolism in patients with autoimmune disorders: a na-
factors and obstetric complications associated with tionwide follow-up study from Sweden. Lancet
macrosomia. Int J Gynaecol Obstet 87:220–6. 2012;379:244–9.
19 Vendittelli F, Rivière O, Bréart G; physicians of the 35 McLeod NL, Gilmour DT, Joseph KS, Farrell SA, Luther
AUDIPOG Sentinel Network. Is prenatal identification of ER. Trends in major risk factors for anal sphincter lacer-
fetal macrosomia useful? Eur J Obstet Gynecol Reprod ations: a 10-year study. J Obstet Gynaecol Can
Biol 2012;161:170–6. 2003;25:586–93.
20 James AH, Jamison MG, Brancazio LR, Myers ER. Venous 36 Vigod SN, Tarasoff LA, Bryja B et al. Relation between
thromboembolism during pregnancy and the postpartum place of residence and postpartum depression. CMAJ
period: incidence, risk factors, and mortality. Am J Obstet 2013;185:1129–35.
Gynecol 194:1311–5. 37 Lanes A, Kuk JL, Tamim H. Prevalence and characteristics
21 Byrd TL, Mullen PD, Selwyn BJ, Lorimor R. Initiation of of postpartum depression symptomatology among
prenatal care by low-income Hispanic women in Houston. Canadian women: a cross-sectional study. BMC Public
Public Health Rep 111:536–40. Health 2011;11:302.

22 von Schmidt Auf Altenstadt JF, Hukkelhoven CW, van 38 Packham JC, Hall MA, Pimm TJ. Long-term follow-up of
Roosmalen J, Bloemenkamp KW. Pre-eclampsia in- 246 adults with juvenile idiopathic arthritis: predictive
creases the risk of postpartum haemorrhage: a nationwide factors for mood and pain. Rheumatology
cohort study in the Netherlands. PLoS One 2002;41:1444–9.
2013;8:e81959. 39 Margaretten M, Julian L, Katz P, Yelin E. Depression in
23 Liu S, Rouleau J, Joseph KS et al. Epidemiology of preg- patients with rheumatoid arthritis: description, causes and
nancy-associated venous thromboembolism: a popula- mechanisms. Int J Clin Rheumtol 2011;6:617–23.
tion-based study in Canada. J Obstet Gynaecol Can 40 Shen CC, Hu LY, Yang AC et al. Risk of psychiatric dis-
31:611–20. orders following ankylosing spondylitis: a nationwide

1384 www.rheumatology.oxfordjournals.org
 Postpartum outcomes in JIA

population-based retrospective cohort study. maternal smoking during pregnancy. Am J Public Health
J Rheumatol 43:625–31. 98:448–53.
41 Jakobsen B. Health related quality of life before, during 46 Coulson EJ, Ng WF, Goff I, Foster HE. Cardiovascular risk
and after pregnancy in norwegian women with rheumatoid in juvenile idiopathic arthritis. Rheumatology
arthritis and juvenile idiopathic arthritis. 2013. Thesis/ 2013;52:1163–71.
Dissertation. https://brage.bibsys.no/xmlui/bitstream/
47 World Health Organization Department of Reproductive

Downloaded from https://academic.oup.com/rheumatology/article/56/8/1378/3782591 by Korea national university of transportation user on 30 October 2020
handle/11250/264547/643472_FULLTEXT01.pdf?se-
Health and Research. WHO recommendations for the
quence=1 (30 July 2016, date last accessed).
prevention and treatment of postpartum haemorrhage.
42 Ehrmann FD, Bernatsky S, Abrahamowicz M et al. World Health Organization. 2012. 1. www.who.int/repro-
Consultation with an arthritis specialist for children ductivehealth/publications/maternal_perinatal_health/
with suspected juvenile rheumatoid arthritis: a population- 9789241548502/en/ (30 July 2016, date last accessed).
based study. Arch Pediatr Adolesc Med 162:538–43.
48 Royal College of Physicians (UK). National Clinical
43 Ehrmann FD, Bernatsky S, Houde M. The incidence of Guideline Centre - Acute and Chronic conditions (UK).
juvenile rheumatoid arthritis in Quebec: a population data- Venous thromboembolism: reducing the risk of venous
based study. Pediatr Rheumatol Online J 2009;7:20. thromboembolism (deep vein thrombosis and pulmonary
44 Thomas SL, Edwards CJ, Smeeth L, Cooper C, Hall AJ. embolism) in patients admitted to hospital. (NICE Clinical
How accurate are diagnoses for rheumatoid arthritis and Guidelines, No. 92.) 30, Pregnancy and up to 6 weeks post
juvenile idiopathic arthritis in the general practice research partum. National Clinical Guidelines. London: NICE
database? Arthritis Rheum 59:1314–21. Clinical Guidelines Centre; 2010. http://www.ncbi.nlm.nih.
45 Gilman SE, Breslau J, Subramanian SV, Hitsman B, gov/books/NBK116521 (30 July 2016, date last
Koenen KC. Social factors, psychopathology, and accessed).

www.rheumatology.oxfordjournals.org 1385