Human Reproduction vol 11 no 8 pp 1767-1770, 1996
Bereavement, grieving and psychological morbidity after
first trimester spontaneous abortion: comparing
expectant management with surgical evacuation
S.Nielsen1'2, M.Hahlin1, A.Moller1 and S.Granberg1
'Department of Obstetrics and Gynaecology, SabJgrenska
University Hospital, Goteborg University, GOteborg, Sweden
2
To whom correspondence should be addressed
Early pregnancy loss is a profound adverse life event for
many women, and increased psychiatric morbidity has
been shown to occur after spontaneous abortion. Dilatation
and curettage (D&C) has been the cornerstone in the
treatment of first trimester spontaneous abortion over the
last few decades. During recent years the possibility of
conservative management has, however, been increasingly
discussed. In a prospective randomized trial, we compared
psychological reactions and morbidity, after either expectant management or D&C, for miscarriages of <13 weeks
gestation in which a transvaginal ultrasound examination
showed intrauterine tissue and/or blood clots with an
antero-posterior diameter of between 15 and 50 mm. Of
the 86 patients included, 58 were randomized to expectant
management and 28 to primary D&C. In patients randomized to expectant management, pregnancy products shown
by transvaginal ultrasound disappeared within 3 days in
43 cases (74%), whereas 15 patients (26%) underwent
D&C owing to retained products of conception after 3
days. At 2 weeks after inclusion, all patients answered
self-administered questionnaires, including visual analogue
scales, concerning their experience of the pregnancy loss,
the present situation and concerns about the future. A
brief anxiety status inventory was included. This study
showed no increase in anxiety or depressive reactions 2
weeks after a first trimester spontaneous abortion when
these patients were compared with non-pregnant healthy
working females 19-39 years of age. Moreover, there were
no significant differences in psychological reactions between
patients managed either expectantly or by D&C.
Key words: anxiety/dilatation and curettage/expectant management/psychological morbidity/spontaneous abortion
Introduction
The emotional and psychological consequences of miscarriages
within the first few weeks of gestation have received relatively
little attention. The paucity of such research is surprising because
one-third of all women experience a miscarriage at some time
during their reproductive years (Laferla, 1986). Some investigators have described reactions of grief following miscarriages,
showing that early pregnancy loss is a profound adverse life
event for many women, and an increased rate of psychiatric
© European Society for Human Reproduction and Embryology
morbidity has been shown to occur after spontaneous abortion
(Comey and Horton, 1974; Seibel and Graves, 1980; Friedman
and Gath, 1989; Turner et al., 1991; Thapar and Thapar, 1992;
Prettymanefa/., 1993).
Current management assumes that potential complications
of retained placental tissue justify curettage, and for 50 years
dilatation and curettage (D&C) has been the cornerstone of the
management of miscarriages in industrialized nations (Hertig
and Livingstone, 1944;McKeeefai., 1992). Expectant management of spontaneous abortions does, however, have a similar
outcome to immediate D&C in terms of bleeding, pain, convalescence and infections (Nielsen and Hahlin, 1995). To our knowledge, the psychological effects of expectant management of
early spontaneous abortion have not been investigated.
The aim of this prospective randomized investigation was to
compare bereavement, grieving and psychological morbidity
following first trimester spontaneous abortion managed either
expectantly (defined as no medical or surgical treatment) or
using D&C.
Materials and methods
Patients
Patients attending the outpatient clinic at Sahlgrenska University Hospital (Goteborg, Sweden) with vaginal bleeding and/or pain in the
presence of a positive urinary pregnancy test were registered on a
standard form, and information about gestational age, symptoms, clinical findings and the results of a transvaginal ultrasound examination
(Bruel & Kjaer, Naerum, Denmark, 6.5 MHz) were obtained Patients
were included in the study if they were in good clinical condition with
a normal blood count, had an estimated gestational age of < 13 weeks,
and a clinical examination including transvaginal ultrasound showed
an inevitable or incomplete spontaneous abortion with intrautenne
tissue with an antero-posterior diameter of between 15 and 50 mm.
(The lower limit was chosen because pregnancy residues with a diameter
of <15 mm would not have been considered for D&C as a routine
procedure in our department; the upper limit was arbitrarily chosen and
pregnancies with retained tissue of >50 mm underwent D&C ) Only
patients who could be seen at the time of entry by one of the authors
(S.N.) were included. Only Swedish-speaking women were included.
The patients were then randomized to expectant management or D&C
in a rauo of 2 1 by drawing a closed letter from a box The patients
included in the investigation were also included in a previously published comparison of clinical results between the two treatment groups
(Nielsen and Hahlin, 1995). Of 87 patients who fulfilled the entry
cntena and were informed about the study, 86 agreed to take part in the
investigation
Protocol
Patients randomized to expectant management (n = 58) were asked to
avoid sexual intercourse for 2 weeks. They were informed that they
1767
S.Nielsen et aL
Table I. Patient characteristics, gestational age, intrautenne volume and
hormonal values at inclusion
Expectant
management
(n = 58)
Age [years, mean (SD)]
Parity [mean (range)]
Gestational age [days, mean (SD)]
Married (%)
Urban population (%)
Pregnancy planned (%)
Intrautenne volume
[ml; mean (SD)]
Serum human chononic
gonadotrophin
[TU/1; mean (SD)]
Serum progesterone
[nmol/1; mean (SD)]
Dilatation and
cure tt age
(n •= 28)
3 1 7 ( 5 7)
0.8 (0-3)
64(13)
93
91
88
13 3 (4.1)
32.5 (5.8)
0.7 (0-2)
68 (13)
92
88
76
12.7(3 2)
6994 (5834)
7597 (6128)
12.4 (8 3)
16.1 (126
Table IL Clinical results following expectant management and primary
dilatation and curettage in patients with first trimester inevitable or
incomplete spontaneous abortion
Expectant
management
(n = 58)
Bleeding [days, mean
(SD)]
Convalescence tune
[days, mean (SD)]
Pain [days, mean (SD)]
Complications
(no of patients)
Change in packed cell
volume [%; mean (SD)]
Day 3
Day 14
Dilatation and
curettage
(/. = 28)
P value
9 4 (2 9)
7 9 (3 2)
0 0526
3 0 (3.1)
1 8 (2 6)
0 1293
16(17)
2*
1 4 (1 6)
3*
>0 30
-1 4 (2 0)
-0 9 (2.3)
-1 9 (3 7)
- 0 5 ( 2 3)
>0 3O
>0 30
"Pelvic inflammatory disease.
Table i n . Results of state anxiety inventory (Spielberger) [values are means
(SD)]
Patients randomized to
expectant management
(n = 58)
Patients randomized to
dilatation and curettage
(n=28)
P value
57.5 (12 4)
57.5 (14.0)
>0 30
Patients randomized to
expectant management,
complete spontaneous
abortion within 3 days
(n = 43)
Patients randomized to
expectant management,
dilatation and curettage,
within 3 days (n = 15)
P value
56 1 (12.3)
61 6 (12 3)
0046
might expect some bleeding and pain. In the case of pain, they were
recommended to use paracetamol alone or in combination with codeine.
No prophylactic antibiotics were given. Patients who were rhesus negative were given 625 IU anti-D immunoglobulin. After 3 and 14 days,
each patient returned for a gynaecological examination, including a
transvaginal ultrasound. If the ultrasound examination 3 days after
inclusion showed retained products of conception with a diameter > 15
mm, the patient underwent D&C. If the patient experienced unac1768
ceptable pain and/or bleeding, she was advised to return to the clinic
and the physician on duty was instructed to perform a D&C.
Patients randomized to D&C (n = 28) were registered at the hospital
and premedicated with 0.5 mg atropine and 25 (ly fentanyl (Janssen
Pharmaceutica, Belgium). Curettage was performed under general
anaesthesia with diprivan (Propofol; Zeneca Ltd, UK). The cervical
canal was dilated if necessary, to a maximum of Hegar 11. No prophylactic antibiotics were given. Patients who were rhesus negative were
given 625 IU anti-D immunoglobulin. The patients were released from
hospital after an observation period of 2-4 h As with the expectant
management group, in the case of pain they were recommended to use
paracetamol alone or in combination with codeine. After 3 and 14 days,
each patient returned for a gynaecological examination.
The patient groups were compared for the number of days with pain
that required analgesics, the number of days with vaginal bleeding
necessitating sanitary protection, convalescence time (i.e. the number
of days away from work), the change in packed cell volume (PCV) 3
and 14 days after inclusion, and complications. Serum progesterone
concentrations were analysed using a commercially available fluoroimmunoassay (Delfia; Pharmacia Diagnostics, Uppsala, Sweden).
Human chononic gonadotrophin (HCG) concentrations were analysed
using a commercially available enzyme immunoassay (Abbot Diagnostics, Abbot Park, IL, USA) PCV was measured using standard
laboratory procedures.
All patients included completed an anonymous self-administered
questionnaire at the hospital immediately after the follow-up visit 14
days after inclusion. The questionnaire included forced-choice questions relating to past obstetric history, family and social background,
and whether or not the pregnancy was planned. The questionnaire also
included visual analogue scales (VAS) scored from 0 to 10 about the
experience of the miscarriage, the present situation, the medical care
and worries about future pregnancies In addition, a brief anxiety status
inventory, Spielberger State Anxiety Inventory (form x), was included
(Spielberger, 1983) It consisted of a list of 30 adjectives or descriptions
of affective states. The patient was asked to state which adjectives or
statements described her present feelings best. Each answer was given
a weighted score of 1-4, where a rating of 4 indicated the presence of
a high level of anxiety, e g. 'I feel frightened'. The scoring weights for
the 'anxiety absent' items were reversed, e.g. 'I feel calm'. The scores
were added together to give a minimum of 30 and a maximum of 120.
The means and SD for 210 healthy working females aged 19-39 years
have been shown to be 54 ± 12 (Spielberger, 1983).
The patient group randomized to expectant management (n = 58)
was compared with the patient group randomized to primary D&C
Moreover, patients randomized to expectant management who had an
empty uterine cavity following 3 days of expectancy (n = 43) were
compared with patients randomized to expectant management who had
to undergo D&C within 3 days of expectancy in = 15) Comparisons
between patient groups were made using Fisher's permutation test
(Bradley, 1968; Oden and Wedel, 1975) Two-tailed tests were used
Results
There were no significant differences between patients randomized to expectant management and primary D&C with regard to
patient characteristics, gestational age, estimated intrauterine
volume or hormonal values at inclusion (Table I).
None of the following variables differed significantly between
the two randomized patient groups: convalescence time, the time
during which the patients experienced bleeding, the time during
which the patients experienced pain, the change in PCV 3 and
14 days after inclusion, or the rate of complications (Table II).
Psychological distress following miscarriages
D
RANDOMIZED TO EXPECTANCY (n = 58)
RANDOMIZED TO DSC (n=28)
E 60 -
o
w
iu 40
o
3 30
20 <
i
I
2
3
4
5
6
QUESTION NUMBER
7
8
I
9
Figure 1. Answers to visual analogue scales (0-100 mm) in response to the following questions. Means are given. (1) Having a miscarriage is a
disappointment to me that is: not too serious/unbearable distress. (2) Having a miscarriage is: no personal failure to me/a great personal failure
(3) Having a miscarriage means to me that my identity as a woman: has not been affected/I do not feel like a real woman. (4) How would you
describe your view of life right now'' positive/negative (5) How do you experience your situation right now?- positive/negative (6) How do you
think your partner experiences the situation nght now?, not at all trying/very trying. (7) How did you experience the care you received at the
clinic9 warm and empathetic/cold, unsympathetic and indifferent. (8) How did you experience the medical interventions in relation to the
miscarriage'' positively/negatively. (9) Are you worried about not being able to have a child in the future9 not at all/very much
60 50 -
EH RANDOMIZED TO EXPECTANCY. COMPLETE
SPONTANEOUS ABORTION AFTER 3 DAYS
OF EXPECTANCY (n—43)
ES RANDOMIZED TO EXPECTANCY, D»C
WITHIN 3 DAYS OF EXPECTANCY (r>-1S)
1
2
3
4
5
6
QUESTION NUMBER
7
Figure 2. Answers to visual analogue scales (0-100 mm) in response to the following questions. Means are given (1) Having a miscarriage is a
disappointment to me that is: not too serious/unbearable distress. (2) Having a miscarriage is. no personal failure to me/a great personal failure
(3) Having a miscarriage means to me that my identity as a woman: has not been affected/I do not feel like a real woman. (4) How would you
describe your view of hfe nght now9: positive/negative. (5) How do you expenence your situation nght now?: positive/negative (6) How do you
think your partner experiences the situation right now?: not at all trying/very trying. (7) How did you expenence the care you received at the
clinic?- warm and empathetic/cold, unsympathetic and indifferent (8) How did you expenence the medical interventions in relation to the
miscarriage?: positively/negatively (9) Are you womed about not being able to have a child in the future9, not at all/very much
Figure 1 shows the scores on the VAS for patients randomized to expectant management or primary D&C. No statistically
significant differences were found between the two groups in
any of the items However, in nine out of 10 items, patients
randomized to expectant management showed less psychological distress than patients randomized to primary D&C.
Table EQ shows the results from the state anxiety inventory
for the same patient groups. No significant differences were
found between patients randomized to expectant management
and primary D&C.
In the 58 patients randomized to expectant management,
products of conception shown by transvaginal ultrasound
disappeared within 3 days in 43 cases. In all, 15 patients
had D&C 3 days after inclusion owing to retained products
of conception with a diameter >15 mm on ultrasound.
There was a tendency for less psychological distress, as
shown by the scores on the VAS (Figure 2), and significantly
less psychological distress in the state anxiety inventory
(Table EH) in those 43 patients who had an empty uterine
cavity after 3 days of expectancy than in those 15 patients
1769
S.Nielsen et al
who had to undergo D&C for mtrautenne pregnancy residues
>15 mm after 3 days of expectant management.
Discussion
This study seeks to emphasize the psychological implications
of a spontaneous abortion. The patients included in the
investigation showed only a limited increase in their scores
for anxiety and depressive reactions compared with the nonpregnant healthy females aged 19-39 years described by
Spielberger (1983). Our results indicate less psychological
distress than reported previously following spontaneous
abortion (Seibel and Graves, 1980; Turner et aL, 1991;
Thapar and Thapar, 1992). Possible reasons for this
discrepancy could be differences in the cultural and rehgious
environments, study populations or study designs. In our
investigation, patients were evaluated immediately after the
follow-up visit 14 days after their miscarriage. The fact that
none of the patients had any remaining somatic problems,
and all patients at this time were assured that they had an
empty uterine cavity, may clearly have contributed to the
absence of psychological distress. Moreover, in the present
investigation, 86 out of 87 longitudinally invited patients
agreed to participate in the study. Most studies investigating
psychological reactions and morbidity following early pregnancy loss have a much lower response rate.
Various factors contribute to the psychological distress
experienced by women who sustain a miscarriage. Hypothetically, some of them could differ depending on whether the
patient was managed expectantly or underwent primary
D&C. The experience of the miscarriage may be one of
pain and bleeding. Hospital admission and surgery are often
experienced as stressful (Levy, 1987). The passage of fetal
tissue at home, as well as the possibility of seeing the
remains, has been shown to strongly affect the women
experiencing a miscarriage (Oakley, 1986; Thapar and
Thapar, 1992; Lasker and Toeder, 1994). Our investigation
does not, however, show any significant difference in the
psychological reactions between patients randomized to
expectant management and primary D&C.
Interestingly, those 43 patients who experienced a successful complete spontaneous abortion widiin 3 days of expectant
management showed a tendency towards less psychological
distress than those 15 patients who had to undergo D&C
because of retained conception products following 3 days
of expectant management. The latter group showed the
highest score for anxiety and depressive reactions, as
indicated by the trait anxiety inventory.
Our study has two major shortcomings. Psychological
reactions and morbidity were evaluated on only one occasion,
i.e. 14 days after the miscarriage. Obviously, a longer
follow-up period would be of value. Secondly, for obvious
practical and ethical reasons, we have no information about
psychiatric and psychological morbidity prior to the patient's
miscarriage. In our opinion, however, the study design and
characteristics of the patients included make it possible to
draw some general conclusions from our results.
1770
It has been shown previously by us and others (Kurtz
et aL, 1991; Mansur, 1992; Dillon et aL, 1993; Haines
et aL, 1994; Nielsen and Hahlin, 1995) that the expectant
management of selected cases of first trimester spontaneous
abortion is a safe procedure with a similar clinical outcome
to that of primary D&C. Before expectant management of
early spontaneous abortion can be recommended, however,
it is also important to investigate the psychological effects
of such a procedure. The most important finding of our
investigation was that, from a psychological point of view,
expectant management of spontaneous abortion had the same
short-term outcome as primary D&C.
Acknowledgements
This study was supported by grants from the Swedish Medical
Research Council (B95-17X-11237-01A), the Swedish Medical
Society and the Merchant Hjalmar Svensson Foundation.
References
Bradley, J W. (1968) Distnbution-free Statistical Tests Prentice-Hall,
London, UK, pp. 68-86.
Corney, R.T and Horton, FT. (1974) Pathological gnef following
spontaneous abortion Am. J. Psychiatry, 131, 825-827
Dillon, E H, Case, C Q., Ramon, I.M et aL (1993) Endovaginal US and
Doppler findings after first-tnmester abortion Radiology, 186, 87-91
Friedman, T. and Gath, D (1989) The psychiatric consequences of
spontaneous abortion Br J Psychiatry, 155, 810-813.
Haines, C H., Chung, T and Leung, D.YL (1994) Transvaginal sonography
and the conservative management of spontaneous abortion GynecoL
Obstet Invest, 37, 14-17.
Hertig, AT and Livingstone, R G (1944) Spontaneous, threatened, and
habitual abortion their pathogenesis and treatment N Engl J. Med.,
26, 797-806
Kurtz, A B , Shlansky-Goldberg, R.D. and Choi, H Y (1991) Detection of
retained products of conception following spontaneous abortion in the
first trimester J Ultrasound Med, 10, 387-395
Laferla, JJ (1986) Spontaneous abortion Clin. Obstet GynecoL, 13,
105-114
Lasker, J N and Toeder, LJ. (1994) Satisfaction with hospital care and
interventions after pregnancy loss. Death Studies, 18, 41-46.
Levy, V (1987) The maternity blues in postpartum and postoperative
women Br J. Psychiatry, 151, 368-372
Mansur, MM (1992) Ultrasound diagnosis of complete abortion can reduce
need for curettage Eur Obstet. GynecoL Reprod. BioL, 44, 65-69.
McKee, M, Pnest, P., Ginzlet, M. and Black, N (1992) Can out-of-hours
operating in gynecology be reduced? Arch. Emerg Med., 9, 290-298
Nielsen, S. and Hahlin, M. (1995) Expectant management of first-tnmester
spontaneous abortion. Lancet, 345, 66-68.
Oakley^A. (1986) Miscarriage and its implications. Midwife Health Visitor
Community Nurse, 22, 123-126.
Oden, A. and Wedel, H (1975) Arguments for Fisher's permutation test
Ann. Statist, 3, 518-520
Prettyman, RJ., Cordle, C J and Cook, G D. (1993) A three-month followup of psychological morbidity after early miscarriage. Br J Med
Psychol, 66, 363-372.
Seibel, M and Graves, W L (1980) The psychological implications of
spontaneous abortion. / Reprod. Med., 25, 161-165.
Spielberger, C D (1983) Manual for the State-Trait Anxiety Inventory
(Form y) Consulting Psychologist Press Inc , Palo Alto, CA, USA
Thapar, A. and Thapar, A (1992) Psychological sequelae of miscarriage:
a controlled study using the general health questionnaire and the hospital
anxiety and depression scale. Br. J Gen. Pract, 42, 94-%.
Turner, M J , Flannely, G.M., Wingfield, M et aL (1991) The miscarriage
clinic: an audit of the first year. Br. J Obstet. Gynaecol., 98, 306-308
Received on February 14, 1996; accepted on May 23, 1996