Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ipob20
Download by: [Erasmus University] Date: 21 February 2017, At: 06:25
Journal of Psychosomatic Obstetrics & Gynecology
ISSN: 0167-482X (Print) 1743-8942 (Online) Journal homepage: http://www.tandfonline.com/loi/ipob20
Major obstetric hemorrhage: a follow-up survey on
quality of life of women and their partners
Giel van Stralen, Laurine L. M. Ruijten, Ad A. Kaptein, Ron Wolterbeek & Jos
van Roosmalen
To cite this article: Giel van Stralen, Laurine L. M. Ruijten, Ad A. Kaptein, Ron Wolterbeek
& Jos van Roosmalen (2017): Major obstetric hemorrhage: a follow-up survey on quality of
life of women and their partners, Journal of Psychosomatic Obstetrics & Gynecology, DOI:
10.1080/0167482X.2017.1286640
To link to this article: http://dx.doi.org/10.1080/0167482X.2017.1286640
Published online: 09 Feb 2017.
Submit your article to this journal
Article views: 8
View related articles
View Crossmark data
ORIGINAL ARTICLE
Major obstetric hemorrhage: a follow-up survey on quality of life of women
and their partners
Giel van Stralen
a
, Laurine L. M. Ruijten
b
, Ad A. Kaptein
c
, Ron Wolterbeek
d
and Jos van Roosmalen
a
a
Department of Obstetrics, Leiden University Medical Centre, Leiden, the Netherlands;
b
Institute of Psychology, Erasmus University
Rotterdam, Rotterdam, the Netherlands;
c
Section of Medical Psychology, Leiden University Medical Centre (LUMC), Leiden, the
Netherlands;
d
Department of Medical Statistics, Leiden University Medical Centre (LUMC), Leiden, the Netherlands
ABSTRACT
Introduction: A description is given of the quality of life (QOL) of women who were treated
with peripartum embolization or hysterectomy for major obstetric hemorrhage (MOH).
Methods: Questionnaires assessing QOL, combined with questionnaires and drawings assessing
illness perceptions, were used to compare women with MOH to reference groups provided by
the questionnaires.
Results: Women who experienced MOH have similar scores on QOL questionnaires compared to
reference groups. Women treated with arterial embolization scored better than women treated
with hysterectomy. Partners of women with MOH scored better on QOL questionnaires than ref-
erence groups. Drawings reflect the major emotional impact of MOH.
Discussion: MOH does not seem to have a negative effect on QOL 6–8 years after the event,
although drawings and verbalizations indicate major emotional impact. More extensive follow-up
is advised for early recognition of the need for psychological help. Women after embolization
seem to have better QOL compared to women after hysterectomy.
ARTICLE HISTORY
Received 15 April 2015
Revised 23 November 2016
Accepted 15 January 2017
KEYWORDS
Childbirth; mental health;
obstetrics; psychological
well-being; quality of life
Introduction
Major obstetric hemorrhage (MOH) contributes to sig-
nificant maternal morbidity and mortality [1]. Recent
reports indicate that the incidence of postpartum
hemorrhage (PPH) is increasing worldwide [ 2–13]. The
physical impact of MOH has been studied extensively
[14]; however hardly any data are available on psycho-
logical consequences. In 2010, a single unit study
addressing quality of life (QOL) after uterine arterial
embolization because of MOH was published [15].
Women in that study showed good QOL which most
likely could be explained by “benefit finding”: people
who have experienced a traumatic event have the
capacity to find benefits from the situation [16,17]. The
psychological impact on partners of the studied
women was underestimated, illustrated by a partner’s
comment: “I can still smell the blood” [15].
QOL is a multidimensional concept and describes
the individual or group’s well-being on several
domains: physical, functional, psychological and social
[18]. QOL fluctuates over time and focuses on patient’s
experiences and is therefore subjective. There is exten-
sive evidence showing that illness perceptions
influence QOL [19,20]. Illness perceptions consist of
multiple components, which basically come down to
the following questions: What is it? What causes it?
What can I do about it? What can the physician do
about it and how long will it last? [21 ].
The Common Sense Model (CSM) of self-regulation
provides a theoretically based explanation for the rela-
tion between illness perceptions and QOL [22]. This
model assumes that people’s behavior is goal-directed
and constantly monitored and adapted by the individ-
ual in order to reach these goals. Input for illness per-
ceptions can be derived not only from individual
experience but also from the media, doctors, or other-
wise [20].
The current study was conducted to add QOL data
to existing data on the psychological impact of MOH
in a nationwide cohort. In addition, partners of the
women were asked to participate by answering ques-
tions about their QOL.
Methods
This study is part of a 2-year nationwide cohort study
that assessed severe acute maternal morbidity and
CONTACT Giel van Stralen gielvs@gmail.com Department of Obstetrics, Leiden University Medical Centre, P.O. Box 9600, 2300 RC Leiden,
the Netherlands
ß 2017 Informa UK Limited, trading as Taylor & Francis Group
JOURNAL OF PSYCHOSOMATIC OBSTETRICS & GYNECOLOGY, 2017
http://dx.doi.org/10.1080/0167482X.2017.1286640
mortality during pregnancy, delivery, and puerperium
in the Netherlands [23]. Women were included nation-
wide from 98 hospitals with a maternity unit from
1 August 2004 until 1 August 2006. Detailed informa-
tion about the data collection has been described pre-
viously [23]. An attempt to contact all women who
experienced MOH (n ¼ 189), defined as peripartum
hysterectomy or – embolization after a minimum ges-
tational age of 24 completed weeks, was made
between June 2012 and August 2013. All women and
partners received a questionnaire at their home
address via regular mail.
Questionnaires
The RAND-36, also known as the Short Form Health
Survey (SF-36), is a questionnaire measuring QOL on a
0–100 scale in which a higher score reflects better out-
come [24,25]. It consists of 36 items, classified in nine
dimensions. Average scores of groups of respondents
can be compared to the provided healthy reference
groups which are subdivided by age and by gender.
The Brief Illness Perception Questionnaire (B-IPQ) is
a nine-item questionnaire measuring patient’s percep-
tions of a disease or medical condition [26,27]. The
results of the questionnaire can be compared to five
reference groups which are provided by the B-IPQ. We
chose patients recovering from myocardial infarction
because this is a comparable condition (short and
potential life-threatening) and patients suffering from
asthma because of comparable average age and better
representation of women. Because of the shortcom-
ings of the provided reference groups, Dutch patients
with SLE (Systemic Lupus Erythematosus) were added
as a reference group as described by Daleboudt et al.
[28]. Furthermore, a patient survey was added,
addressing women’s and partners’ experiences of the
event and of hospital care and aspects of somatic
recovery such as: lactation (problems) and recurrence
of menstruational cycle and successive pregnancies
(if applicable).
Drawings
Illness perceptions are usually assessed with question-
naires. A relatively new method to assess illness per-
ceptions is through patients’ drawings of affected
organ(s) [21,28–33]. It has the advantage of circum-
venting possible social desirability issues and provides
a unique and value free illustration of a person’s illness
perceptions [31]. All women were asked to retrospect-
ively draw images of her uterus and pelvic area during
the event (Drawing 1) and at the time of returning the
questionnaires (Drawing 2). The area of the uterus and
(remaining) internal genitals were measured by hand
using graphical notebook paper. In addition, two
authors (G. S. and L. R.) independently scored the
drawings on degree of blood loss, degree of damage
to the uterus, and psychological impact, using a
3-point scale. Congruity between scores was assessed
by intraclass correlation coefficients. Mean scores of
drawing 1 and drawing 2 were calculated and com-
pared. This approach to illness perceptions scoring has
been described earlier by Tiemensma et al. [31,33].
Statistics
Data were analyzed using SPSS 20.0 (IBM, Armonk, NY).
Descriptive statistics were used to describe clinical
parameters. Differences between scores of women and
reference groups and scores of partners and reference
groups were calculated and analyzed by using the one-
sample t-tests. Differences between means of two
groups were analyzed with the independent-samples t-
tests. Differences between dichotomous variables were
analyzed with chi-square tests. Calculating intraclass
correlation coefficients in a linear mixed model (one-
way random ANOVA) assessed congruity between
scores of G. S. and L. R. To examine the influence of ill-
ness perceptions on QOL, correlations between ques-
tionnaire scores RAND-36 and B-IPQ were analyzed with
Pearson’s correlation coefficient. Closed survey ques-
tions were analyzed using descriptive statistics and the
independent samples t-test. Open survey questions
were analyzed and interpreted by G. S. and L. R. No cor-
rections for multiple testing have been applied.
Details of ethics approval
Medical Ethics Commission of the Leiden University
Medical Centre (LUMC) was obtained on 28 July 2010.
Reference: CME-V 005.
Results
Questionnaires were filled in between June 2012 and
August 2013, the range of the interval between the
event and the completion of the questionnaires was
6–9 years. Table 1 describes characteristics of the 58
included women. Figure 1 describes selection and
inclusion participants.
RAND-36 (Table 2, Figure 2)
Included women scored similar to gender- and age-
specific reference groups except for four dimensions.
2 G. VAN STRALEN ET AL.
Compared to the gender-specific reference group,
women scored better on dimension “physical
functioning” and “pain” and worse on dimension
“vitality”. Compared to the age-specific reference
group, women scored better on dimension “physical
functioning” and “pain” and worse on dimension
“vitality” and “social functioning”.
Women who were treated with embolization scored
better compared to women who were treated with
hysterectomy on all dimensions and significantly
better on dimensions: “pain” and “role limitations (due
to physical problems)” (Figure 2).
Partners
Mean age of the partners was 34.5 years (spread:
29–41 years). Compared to their age-specific and gen-
der-specific reference group, partners scored signifi-
cantly better on dimensions: “physical functioning”,
“pain”, “role limitations (due to emotional problems)”,
“role limitations (due to physical problems)”, and
“social functioning”. Compared to the gender-specific
reference group, partners scored better on dimension
“general health perception”. There were no significant
differences between partners of women after embol-
ization and partners of women after hysterectomy.
B-IPQ (Table 3)
The reported impact of the event reflected by dimen-
sion “consequences” is higher than the impact
reported by patients of reference groups. In addition,
women expected a shorter duration and experienced
less personal control over the event than reference
groups, as reflected by dimensions “timeline” and
“personal control”. Women in our study experienced
less disease-related symptoms (“identity”) and felt less
concerned compared to patients with asthma or SLE.
Women treated with embolization scored signifi-
cantly lower on dimensions “consequences” and
Table 1. Characteristics of included women and women lost to follow-up (LTF).
Included embolization
n ¼ 24
Included hysterectomy
n ¼ 34 p Values
LTF embolization n ¼ 66 LTF hysterectomy n ¼ 65
Mean maternal age
(years) (range)
33.1 (26–40) 33.8 (22–42) 0.49 31.0 (20–45) 34.8 (23–44)
BMI mean (range) 22.5 (20.1–25.7) 24.4 (18–46) 0.26 23.4 (17–35) 24.4 (17–43)
Weeks of gestation
(range)
37 þ 6 (31 þ 6to40þ 6) 38 þ 3 (26 þ 3to42þ 1) 0.41 38 þ 6 (28 þ 2to42þ 1) 38 þ 1 (26 þ 5to42þ 2)
Prematurity <37 weeks 4 (17) 7 (20) 0.75 10 (15) 15 (23)
Nulliparity 16 (67) 9 (26) 0.01 40 (61) 11 (17)
Multiple pregnancy 4 (17) 3 (9) 0.37 8 (12) 5 (8)
Previous cesarean 2 (8) 16 (47) 0.02 6 (9) 27 (41)
Uterine rupture 0 7 (20) 0.02 1 (1) 8 (12)
Placenta previa 2 (8) 4 (12) 0.67 1 (1) 9 (14)
Placenta accr/incr/
percr
2 (8) 9 (26) 0.08 6 (9) 14 (21)
Mode of delivery
Spontaneous 10 (42) 9 (26) 0.60 36 (54) 17 (26)
Ventouse 4 (17) 5 (15) 8 (12) 3 (5)
Planned cesarean 6 (25) 11 (32) 8 (12) 16 (25)
Emergency cesarean 4 (17) 9 (26) 14 (22) 29 (45)
Mean total blood loss
(ml) (range)
5797 (1500–20,000) 7478 (1500–20,000) 0.23 5050 6305
Mean no of blood
products (range)
21 (6–70) 26 (3–70) 0.36 18 (0–66) 24 (0–66)
Hysterectomy after
embolization
410
The values represent n
(%) unless otherwise stated.
Comparing included women after embolization with included women after hysterectomy.
accr/incr/percr: accreta, increta, and percreta.
Informed consent denied
(n = 2)
Eligible
n = 189
Informed consent but
questionnaire not returned
by patient (n = 17)
Included
n = 58
n = 75
n = 77
No reply from patient
(n = 112)
Figure 1. Selection and inclusion of women.
JOURNAL OF PSYCHOSOMATIC OBSTETRICS & GYNECOLOGY 3
“timeline” and significantly higher on “treatment con-
trol” compared to women treated with hysterectomy.
Partners
B-IPQ scores indicated that partners of women who
were treated with embolization felt less concerned
and had more trust in embolization as a treatment
modality compared to partners of women who were
treated with hysterectomy.
Drawings (Figure 3)
The 46 peripartum uterus drawings (Drawing 1) had
an average surface of 29.1 square cm. This dropped
significantly (p < 0.01) to 14.3 square cm in the follow-
up drawings (Drawing 2, n ¼ 38), indicating a reduc-
tion in emotional impact of the event over time. The
score on “degree of blood loss”, “degree of damage to
the uterus”, and “psychological impact”, ranging from
0 to 12, dropped from 6.1 (Drawing 1) to 2.3 (Drawing
2) for women after embolization. In the hysterectomy
group, the score dropped from 5.9 to 2.6. These results
are consistent with results from other drawing studies
in that the size of the drawn organs is reduced after
the upsetting event has taken place [21,29–33]. The
intraclass correlation coefficient between G. S. and
L. R. on Drawing 1 and Drawing 2 was 0.82 and 0.66,
respectively.
Neither drawing scores and B-IPQ scores nor draw-
ing surfaces and B-IPQ scores correlated significantly.
Survey
The majority of women and partners (75%) were satis-
fied with the hospital care provided. In general,
women and partners highly appreciated personal
attention, compassionate care, and time for questions
and explanations. However, not all women appeared
satisfied with the provided care as one woman
explained: “Some health-care workers are born to this
work, they are compassionate, sweet and understand-
ing. Others are just making a living”.
Several women advised more extensive follow-up
(i.e. offering outpatient consultation 6 months after
the event). Partners put into words that they missed
information and attention during the event and felt
lonely, insecure, and powerless.
On the basis of the responses, both patients and
partners recognized the severity of the event; 58% of
the women thought they were going to die compared
to 72% of their partners. One woman illustrated:
“I looked death in the eye”. We found contradictory
expressions of “benefit finding” as reflected by the
Table 2. RAND-36 results of women (embolization and hysterectomy combined) compared to reference groups.
Included women
(n ¼ 58) mean (SD)
Reference groups
Included partners
(n ¼ 47–49) mean (SD)
Gender Age (35–44 years) Age (45–54 years)
Ref value p values Ref value p Values Ref Value p Values
General health perception 70.1 (23.7) 71.5 0.23 74.0 0.21 71.6 0.63 79.0 (19.6)
Physical functioning 94.5 (8.8) 80.7 0.000 90.0 0.000 79.9 0.000 98.2 (5.1)
Change in health 52.6 (17.3) 53.4 0.72 55.4 0.22 51.9 0.69 53.6 (15.3)
Mental health 75.0 (17.4) 75.5 0.85 76.9 0.43 76.7 0.48 80.6 (14.7)
Pain 91.9 (17.6) 80.0 0.000 83.8 0.000 80.5 0.000 96.0 (9.1)
Role limitations (emotional) 87.4 (29.8) 82.5 0.22 83.6 0.34 82.2 0.19 98.0 (8.1)
Role limitations (physical) 82.3 (33.1) 78.3 0.36 82.9 0.90 78.9 0.43 96.4 (16.1)
Social functioning 81.7 (23.0) 86.1 0.15 88.0 0.04 86.1 0.15 96.2 (8.9)
Vitality 60.5 (20.9) 66.3 0.04 67.5 0.01 67.1 0.02 70.6 (15.0)
A higher score reflects a better outcome.
Reference values are only applicable for included women, not for partners.
Table 3. B-IPQ results of women (embolization and hysterectomy combined) compared to reference groups.
Included women
(n ¼ 58) mean (SD)
Reference groups
Included partners
(n ¼ 45–49)
mean (SD)
MI Asthma SLE
Ref value p Values Ref value p Values Ref value p Values
Consequences 7.5 (2.6) 4.1 0.000 3.5 0.000 6.5 0.000 5.7 (2.8)
Timeline 4.7 (2.8) 7.2 0.000 8.8 0.000 9.2 0.000 3.9 (2.6)
Personal control 3.4 (3.1) 7.7 0.000 6.7 0.000 5.6 0.000 3.7 (3.2)
Treatment control 8.2 2.3) 8.8 0.06 7.9 0.40 8.4 (0.50) 0.50 7.7 (2.8)
Identity 3.2 (3.1) 3.1 0.79 4.5 0.002 6.0 0.000 N/A
Concern 3.5 (3.2) 6.2 0.000 4.6 0.013 5.8 0.000 1.8 (2.3)
Emotional response 4.1 (3.3) 4.2 0.86 3.5 0.16 5.8 0.000 2.6 (2.8)
Understanding 7.2 (3.2) 8.0 0.05 6.5 0.13 6.8 0.40 7.2 (2.6)
MI: myocardial infarction; SLE: systemic lupus erythematosus.
A lower score reflects a better outcome.
Reference values are only applicable for included women, not for partners.
4 G. VAN STRALEN ET AL.