Bleeding in the first trimester of pregnancy
Wieringa-de Waard, M.
Publication date
2002
Link to publication
Citation for published version (APA):
Wieringa-de Waard, M. (2002). Bleeding in the first trimester of pregnancy.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.
Expectantt management versus surgical
evacuationn in first-trimester miscarriage
health-relatedhealth-related quality of life in randomised and
non-randomisednon-randomised patients
Margreett Wieringa-de Waard, Esther E. Hartman, Willem M. Ankum, Johanness B. Reitsma, Patrick J.E. Bindels, Gouke J. Bonsel
Submittedd for publication Abstract t
OBJECTIVESS To determine health-related quality of life (HRQL) changes in women with aa miscarriage, managed either expectantly or undergoing surgical evacuation; to esti-matee the impact of treatment assignment by either randomisation or by own choice. DESIGNN Randomised controlled trial; eligible women who refused randomisation because off a preference for one of the treatment options, were managed according to their choice. SETTINGG Two Amsterdam hospitals.
PARTICIPANTSS Patients with a non-viable pregnancy or incomplete miscarriage at a gestationall age of less than 16 completed weeks.
MAINN OUTCOME MEASURES The scores of a set of general and specific HRQL measures comparedd across the two randomised and two preference groups. The MOS 36-ltem Short-Formm Health Survey, aggregated into Physical Component Summary scale (PCS) and Mental Componentt Summary scale (MCS), State-Trait Anxiety Inventory and the Perinatal Grief Scale weree offered at inclusion, and after 2, 4, 6 and 12 weeks. Longitudinal analysis was applied. RESULTSS 229 women participated in the study. Mental health of women (MCS) rando-misedd to expectant management improved more and earlier (treatment effect, p=0.003) thann mental health of women randomised to surgical evacuation. The overall difference wass 7.4 (2.4 to 12.4) points. MCS scores were significantly better in women preferring thann in women randomised to curettage (overall difference of 5 points, p=0.03). The twoo preference groups did not differ in scores of MCS.
CONCLUSIONSS Women with a miscarriage who chose their own treatment had a better health-relatedd quality of life. Therefore, women who prefer either expectant or surgical managementt should be allowed to undergo the management of their choice. Women with-outt treatment preference should be encouraged to initially follow expectant management.
Introduction n
AA woman diagnosed with a non-viable pregnancy may decide to wait for a spon-taneouss loss of pregnancy, or to undergo surgical uterine evacuation. In many countriess surgical evacuation is the treatment of choice among gynaecologists. Severall observational studies from General Practice, however, have propa-gatedd expectant management as an alternative treatment option.1'2 In two randomisedd controlled trials, including our own study, no differences were found betweenn expectant management and surgical evacuation in terms of safety andd complications.3'4
Equivalencee in terms of medical outcomes, robust as these findings may be, doess not rule out differences in health-related quality of life (HRQL) in wom-enn undergoing either expectant management or surgical evacuation. If sub-stantial,, these differences may even play a dominant role in guidelines and individuall decisions.
Inn a randomised controlled trial we measured general and specific quality of lifee in women with miscarriages, who were managed either expectantly or by surgicall evacuation. Assuming that quality of life might well be influenced by patients'' treatment preferences, these results were compared with those of womenn who, because of a treatment preference, were managed according to theirr own choice.5
Methods s
Protocol Protocol
Inn a partially randomised controlled trial we measured the HRQL of women withh a first-trimester miscarriage. The clinical outcome measures are report-edd separately.4 The study was conducted between April 1998 and September 20000 in two Amsterdam hospitals, the Academic Medical Center and the Onzee Lieve Vrouwe Gasthuis, and was approved by the ethics committees of bothh hospitals. Patients with an established diagnosis of a non-viable preg-nancyy or incomplete miscarriage at a gestational age of less than 16 complet-edd weeks participated in the HRQL study. Exclusion criteria for enrolment were:: inability to understand the Dutch or English language, severe bleeding, painn or fever necessitating immediate surgical evacuation.
Transvaginall ultrasonographic criteria for a non-viable pregnancy were: a mean gestationall sac diameter greater than 15 mm without a measurable embryonic pole,, an embryo without cardiac activity, or a gestational sac diameter less than 155 mm, not showing any growth after a seven-day interval.6 7 An incomplete miscarriagee was diagnosed in case of ultrasonographical evidence of retained productss of conception greater than 15 mm in the anteroposterior (AP) diame-ter.. All transvaginal scans were performed by trained physicians using a transvaginall 6.5 MHz sonographic probe (Hitachi Corporation, Tokyo, Japan).
Assignment Assignment
Afterr written informed consent had been obtained, women were randomly al-locatedd by the attending physician to either expectant or surgical management (randomisedd groups), using central electronic randomisation. Women who refusedd randomisation because of a preference for either one of the manage-mentt options, received the treatment of their own choice (preference groups) andd were asked for consent to participate in the HRQL measurements.
InterventionsInterventions and follow-up
Surgicall uterine evacuation was performed within a week after inclusion under locall or general anaesthesia in daytime surgery.
Alll women (randomised and preference groups) were assessed clinically and byy ultrasonography during the bi-weekly appointments at the outpatient clin-icc until a complete evacuation of the uterus was established either by surgical evacuationn or through spontaneous loss. Women who became impatient while beingg managed expectantly and requested surgical evacuation as yet, were scheduledd to undergo a curettage within a week, but were analysed in the originall treatment group.
Womenn were asked to fill out questionnaires at five time points: immediately afterr inclusion, and 2, 4, 6 and 12 weeks later. The first questionnaire was filledd out after ultrasonography. All questionnaires comprised the Medical Outcomee Study 36-Item Short-Form Health Survey (SF-36), the State-trait Anxietyy Inventory (STAI) and the Perinatal Grief Scale (PGS). At inclusion sociodemographicc questions were added.
Duringg the bi-weekly visits, questionnaires were taken in and patients recei-vedd instructions about the questionnaire for the next interval. The question-nairess were completed by the patient at home and were returned in a pre-stampedd envelope. Reminders were sent to non-responders once after each timee point.
Outcomee measures
Wee used three measures, all validated in Dutch, to analyse differences in HRQL betweenn patients: the generic SF-36, the domain-specific STAI, and a short formm of the disease-specific P G S .8 1 2 We used the 'acute' version of the SF-36, employingg a one-week time frame. The scores of the eight subscales, which rangee from zero (worst health) to 100 (best health) were conventionally aggre-gatedd into a Physical Component Summary scale (PCS) and a Mental Compo-nentt Summary scale (MCS) (mean 50, SD 10). The STAI contains two 20-item scaless covering both current (state) and background (trait) anxiety; items are ratedd on a four-point scale with sumscores ranging from 20-80, where higher scoress represent higher levels of anxiety.8-13 The PGS measures grief associat-edd with pregnancy loss. The short form used here contains 33 questions, to be answeredd on a five-point Likert scale ranging from 'Strongly Agree' to
'Stronglyy Disagree' with a neutral midpoint. The PGS consists of three sub-scales:: Active Grief, Difficulty Coping and Despair, which range from 11-55 withh higher scores indicating more grief. Active Grief is the normal grief reac-tionn that may follow a pregnancy loss. Difficulty Coping and Despair represent increasinglyy severe forms of grieving. The PGS earlier has been confirmed as reliablee and valid in the evaluation of women who miscarried.10'14
Referencee scores for the SF-36 subscales were obtained from published data off a Dutch population sample (16-40 years of age).1 2 , 1 5"1 7
SampleSample size: For the power calculation we focussed on the SF-36. Assuming
aa 5-point difference in the summary PCS or MCS measures as clinical rele-vant,, we needed 33 patients in each randomised group to detect this differ-encee (alpha = 0.05, beta = 0.2).1 5
StatisticalStatistical analysis
Wee analysed the data according to the intention-to-treat principle.
Womenn with missing data were included if at least two sets of questionnaires weree available. We assumed that women with a short period until complete evacuationn of the uterus (spontaneously or by surgical intervention) might be lesss interested to participate in the study. We therefore compared the time untill complete evacuation between included and excluded women. Differences inn patient characteristics and HRQL at inclusion between the randomised groupss and the preference groups were analysed using descriptive statistics. Thee Statistical Package of the Social Sciences (SPSS, version 10.07) was usedd for all described analyses.
Wee employed repeated measurements design to compare the HRQL changes off the randomised groups and the preference groups in more detail. A mixed linearr model was used to sort out the treatment effect and the treatment assignmentt effect, apart from time and other effects, as expressed in
observedd patterns in the various quality of life measurements. The mixed lin-earr model methodology accounts for the correlation between different mea-surementss over time within the same women, thus avoiding overestimation of specificc effects such as due to treatment. For each participant we assumed in thiss analysis that the more measurements were closer in time, the more they weree correlated. Our model included the following explanatory variables: the (initial)) treatment option (at four different levels: curettage and expectant in bothh the randomised and preference group), time (as a categorical variable withh four levels, excluding the baseline measurements), interaction between treatmentt and time, and the baseline measurement. P-values below 0.05 weree considered significant. Results were presented as differences in mean scoree with 95% confidence intervals.1 8 , 1 9
Alll mixed models were done in SAS 6.12 (syntax and data available from the authors). .
Transvaginall sonography assessment t
(nn = 1101)
Non-viablee pregnancy and incompletee miscarriage (n=449) ) Excludedd (n=652) -- viable pregnancy -- complete miscarriage -- other diagnoses Includedd (n=427) RCT(n=122) ) ExcludedExcluded (n=22) -- immediate curettage necessary y Preferencee (n=305)
+ +
Expectantt (n=64) Curettagee (n=58)
I-I-HRQL(n=46) ) Inclusionn (n=44) Weekk 2 (n=36) Weekk 4 (n=33) Weekk 6 (n=33) Weekk 12 (n=30) Excluded; ; -- language problem or noo consent (n= 120) -- <2 questionnaires (n HRQLL (n=36) Inclusionn (n=33) Weekk 2 (n=27) Weekk 4 (n=26) Weekk 6 (n=32) Weekk 12 (n=29)
4r r
Expectantt (n=126) == 78)4r 4r
Curettagee (n=179) HRQL(n=61) )4, ,
Inc c We e We e We e We e usio o ek k ek k ek k ekk 1 nn (n=57) 22 (n=36) 11 (n=42) 33 (n=41) 22 (n=45) HRQL(n=86) )4r r
Inclusionn (n=78) Weekk 2 (n=61) Weekk 4 (n=46) Weekk 6 (n=55) Weekk 12 (n=59)Figuree 5 . 1 . Flow diagram.
Results s
ParticipantParticipant flow and follow-up
Outt of 1101 women, 427 were eligible for the randomised controlled trial (Fi-guree 5.1). Of these, 122 were randomised to either expectant or surgical managementt and 305 were managed according to their own preference. For thee HRQL study, 198 women (or 46%) were excluded because of insufficient Dutchh or English language skills, because they refused to participate (n=120), orr because they returned only one questionnaire (n=78). The other 229 womenn (54%) completed two or more questionnaires and were included in thee present study. Compared to women who were excluded, the participants originatedd more often from Western-European countries (72% versus 46%), andd showed a higher degree of education. No differences in age or prior
Tablee 5 . 1 . Patient characteristics at inclusion according to treatment allocation and preference.3 3 Randomisedd groups Expectantt Curettage (n=46)) (n = 36) Preferencee groups Expectantt Curettage (nn = 61) (n=86) Age-yr,Age-yr, mean 33.1 1 33.9 9 33.5 5 32.4 4 Parity Parity 0 0 1 1 > 1 1 255 (54.3) 155 (32.6) 66 (13.0) 177 (47.2) 166 (44.4) 33 (8.3) 300 (49.2) 199 (31.3) 122 (19.7) 488 (55.8) 311 (36.0) 77 (8.1) PriorPrior experience
Noo prior curettage
orr expectant management Priorr curettage
(legall abortion or miscarriage) Priorr expectant management Priorr curettage and
expectantt management Unknown n GestationalGestational age << 8 wk 8-122 wk 12-16wk k Uncertain n NativeNative country
Western-Europee and USA Africann country
Surinamm and Antilles Otherr and unknown
EducationEducation untii-yrs << 18 >> 18 Unknown n STAI-trait,STAI-trait, mean (SD) 27 7 11 1 5 5 3 3 6 6 28 8 9 9 3 3 34 4 1 1 9 9 2 2 24 4 21 1 1 1 42 2 (58.7) ) (23.9) ) (10.9) ) (6.5) ) --(13.0) ) (60.9) ) (19.6) ) (6.5) ) (73.9) ) (2.2) ) (19.6) ) (4.3) ) (52.2) ) (45.7) ) (2.2) ) (13.0) ) 18 8 12 2 5 5 1 1 6 6 18 8 9 9 3 3 22 2 3 3 9 9 2 2 17 7 18 8 1 1 42 2 (50.0) ) (33.3) ) (13.9) ) (2.8) ) --(16.7) ) (50.0) ) (25.0) ) (8.3) ) (61.1) ) (8.3) ) (25.0) ) (5.6) ) (47.2) ) (50.0) ) (2.8) ) (14.0) ) 40 0 16 6 5 5 5 5 38 8 13 3 5 5 46 6 3 3 9 9 3 3 29 9 30 0 2 2 39 9 (65.6) ) (26.2) ) (8.2) ) --(8.2) ) (62.3) ) (21.3) ) (8.2) ) (75.4) ) (4.9) ) (14.8) ) (4.9) ) (47.5) ) (49.2) ) (3.3) ) (11.0) ) 49 9 24 4 5 5 7 7 1 1 8 8 48 8 23 3 7 7 62 2 5 5 17 7 2 2 45 5 39 9 2 2 41 1 (57.0) ) (27.9) ) (5.8) ) (8.1) ) (1.2) ) (9.3) ) (55.8) ) (26.7) ) (8.1) ) (72.1) ) (5.8) ) (19.8) ) (2.3) ) (52.3) ) (45.3) ) (2.3) ) (11.0) ) aa
experiencee with any of the management options were observed between par-ticipantss and excluded women; also the time from diagnosis until curettage or spontaneouss loss of pregnancy did not differ (data not shown).
Analysis Analysis
Baselinee characteristics of included women are presented in Table 5.1. No sig-nificantt differences in sociodemographic or clinical characteristics were pre-sentt between the randomised groups, nor between the preference groups. Also,, no differences in prior experience with one of the management options, education,, native country or anxiety (STAI-trait) were observed.
Thee course of the (unadjusted) mean values for the SF-36 mental and physical summaryy scores, STAI-state, and PGS total scores, in the four groups are illus-tratedd in Figure 5.2. At inclusion we did not find significant differences in scores betweenn the randomised groups. In contrast, women preferring expectant man-agementt showed significantly better scores on the Mental health subscale (meann score 66) than women preferring curettage (mean score 57). Similarly, att inclusion women preferring expectant management showed significantly less Activee grief (mean score 30) than women preferring curettage (mean score 35). Changess in HRQL over time in the four groups were analysed in a repeated measurementss mixed model (Table 5.2).
RandomisedRandomised groups
Theree were no significant differences between the two treatments in physical healthh (PCS) scores (treatment effect p = 0.7). Women randomised to expec-tantt management, however, had significantly better mental health (MCS) scoress (treatment effect p = 0.004). The overall differential treatment effect (differencee between the expectant management group versus the curettage group,, all time points together) was 7.4 points on the MCS in favour of expectantt management. The difference at twelve weeks was still 6.3. There weree no significant differences in changes over time between the two treat-mentt arms in PCS and MCS (time by treatment p = 0.6).
Wee found no significant differences in improvement between the two arms in state-anxietyy scores of the STAI. The negative score indicates a lower aver-agee score (less anxiety) for the expectant group, but this was not significant (treatmentt effect p = 0.09). Patients randomised to expectant management alsoo showed lower scores on the Perinatal Grief Scale-total and its subscales (lesss grief) than patients randomised to curettage, but again the difference didd not reach statistical significance (treatment effect p = 0.3).
PreferencePreference groups
Noo statistically significant differences in any of the HRQL measures was ob-servedd between women preferring expectant management and those prefer-ringg surgical evacuation.
(A A CD D i --O --O
u u
(0 0 60 0 50 0 < ret expectant AA ret curettage - O —— pref expectant '' -Is - pref curettagenormm score 4 0 --30 0 better r 00 2 4 6 8 10 weeks s
SF-366 Physical Component Summary scale
11 I 122 14 6 0 - 1 1 in in 0) ) i
--o --o
o o
(0 0 >> ret expectant AA ret curettage — o —— pref expectant-is-is - pref curettage
normm score j - w — - c m m
A A
SF-366 Mental Component Summary scale
Figuree 5.2. Mean HRQL scores of randomised and preference groups. Direction of
O O 00 0 I I
o o
(0 0 o o o o o o 0) 0) 60 0 55 5 50 0 4 5 --40 0 35 5 30 0 25 5 20 0 99 ret expectant AA ret curettage - o —— pref expectant - £ - pref curettage normm score-o o
i i
better r ~ii i i r 44 6 8 10 weeks sStatee anxiety (STAI)
12 2 14 4 100 0 90 0 m m <o o T " " I I n n </> > 0) )
o o
u u
(A A ret expectant AA ret curettage —— pref expectant -is-is - pref curettage60 0 50 0
T ^ ^
n-rssQ. n-rssQ.
40 0 30 0I I
better r 6 6 10 0 12 2 14 4 weeks s( 0 0 3 3 o o CO O > > CD D H— — CD D r r ^ ^ o o coo co üü CD —— i _ 055 o —'' CO coo " " DD CD COO o 4-1 4-1 cc co CDD CD EE = CDD CO O)) > CO O CD D EE co +-- en cc CD coo c OO ~0 CDD C Q .. CO CDD CD C C CD D CD D O O O O CD D O) ) COO -C CDD CD o-- tö _ TT o
Ü l l
C/)) — __ co COO CD ?? — \L\L co coo > *—— CD COO > Q33 * -oo O coo ®® tó ~~ Q. oo o O) COO CJ 33 c O"" CD "OO CÜ CDD — 'SS ï"55
->->- T3 .CC C COO I— -J J?oo §«« ë »
CDD CD T3 OO § CD CDD CD O ^^ ? © OO « E oi i o o 5 5 o o _> > o o 5 5 a> > E E > > co o a> > >_ _ ** * ( 0 0 t _ _ CD D > > O O * * c c CD D E E ca a a> > o o CD D o o Ü Ü ** * o o a> > *--«^ ^ CD D a> > _ 3 3 CO O > > n n >> C 0) J33 <D _3 0)) E CO ETSS > ÏZZ « o. o o o oo o o o o o o o o o o o o c o o o c oo c o o - i - m n ^- s s ^ r ^ t r ^ m - ** có có co Ó - I - C D O Ó Ó L f i r ^ c o co co ^j- c\i c o i o o o jj n co co co co co LD co h- LO *r co Ö O T ^ ÓÓ co co o c\i - t d w ^ o J d - ^ ö f^"" ^ ^ LTT ^ c o ' t c T t c r (B"^ ao ^ c S ^ ^ c o * r j - i - ^ c oo c v j i b ^ - d ó *t c\i r^ T-^ c\i c\i ^ro o o oo o o o o o o o o o o o o T-- en co co ^ r» o io (o o ^ co o i o - i - o ii c Ü T ^ ^ a J tb c\i in ó ** t\i eb T^ IS -- -a; C\I LT> ' t o c o o o co co n O C M L O C D ö r ^^ r ^ o r ^ c v i ó i b o ^ ó d -^ h -- C O - < * c o o o o o o oo co oo co a> o o o o o o 05 5 oo h- LO i -oo o o o C\JJ co r^ c\j o o o o o COO N CO N óó ó ó ó CDD N CO CO oo o O O LDD -^ r>- -r-oo ó ó ó CDD CO CD * f r oo ó ó ó CD D O ) ) CO O a> > ^--3 ^--3 O O \-\-O \-\-O CC C co o 3 3 CO O CD D > > c c CO O o o CD D CL L X X CD D H H Ü Ü CC C c c CD D C C O O o_ _ i--O i--O
o o
a a fli fli O ) ) Cl} } 3 3 o o "CD D n n co o 3 3 CD D > > * * c c co o o o CD D C l l X X CD D CD D O O c c 0 0 CD D ^ ^ 0_ _ 0 0 CO O u u co o » » c c co o o o CD D CL CL X X CD D CD D O O l_ _ CD D CD D CU U Q . . CO O 3 3 CO O CD D > > c c CO O o o CD D Q . . X X CD D H H (D D O ) ) CO O CD D 3 3 O O CD D O O c c CD D CD D CD D Q . . CO O 3 3 CO O CD D > > CD D O ) ) CO O CD D 3 3 O O H H OO O CC C CC C CD D O ) ) CO O 00 0 *— — 3 3 O O 1— — ü ü CC C co o 3 3 CO O CD D > > c c co o *-» » o o CD D O . . X X CD D 1 --ü --ü CC CC c c Q> Q> O O Q . . E E o o ü ü <T> > o> > CO O 0) ) O O "o o n n co o 3 3 CD D > > ^j j c c co o o o CD D t l l X X CD D 0 0 O O c c CD D^ ^
3 3
^ ^ CL L O O CO O U U CO O ^^ ^ c c CO O o o 0 ) ) Q . . X X CD D CD D O O L. . 0 ) ) CD D CD D O . . CO O 3 3 CO O CD D > > c c co o *-» » o o CD D Q . . X X CD D H H CD D O ) ) CO O CD D O O CD D O O c c CD D CD D Q . . CO O 3 3 CO O CD D > > CD D O ) ) CO O " " CD D D D O O 1 --OO O rr r cc cc CD D O ) ) CO O CD D 3 3 O O H H O O CC C co o D D CO O CD D > > c c m m o o CD D Q . . X X CD D 1 --ü --ü DC C o o 0> > ( 0 0 n i i O ) ) co o CD D 3 3 O O CD D n n co o 3 3 CD D > > +_, +_, C C CO O u u CD D C l l X X a i i CD D O O c c CD D ^ ^ ^ ^ CD D Q_ _ H - » » CC CD COO O ) oo iS CDD - £ CLL £ XX 3 CDD o CDD CD OO ü cc C CDD CD CDD CD *4—— CDD CD Q -- Q . COO to 33 3 coo co CDD o >> > cc ™ caa 9? OO + ; CDD % o.. Ir XX 3 CDD O 1 1 -üü o cccc cc CD D co o 'S) ) ^ ^ 3 3 O O h --O --O CC C co o 3 3 CO O CD D > > c c m m o o <D D C l l X X CD D H H ü ü CC C co o o o ( / } } 15 5 01 1 O ) ) CO O 0) ) 3 3 O O CD D n n co o 3 3 CO O CD D > > c c m m o o CD D n n X X CD D CD D O O c c CD D CD D CD D C L L u u 4 _ i i CC CD coo cn HH co CDD - t ; Q-- £ XX 3 CDD o 00 CD cc c CDD CD^^ 5
CDD CD O -- CL COO w 33 3 COO ui -- t . CDD CD >> > cc ™ coo ?? -^^ co CJJ e CDD 's; o.. ï XX 3 CDD O 1-- H oo O ccc ccÏ Ï a> > _> > a> > S S (V V > > O O 0 )) + - z pp —
«« c >
»» ® n' o>> E to izz ® Ü. oo o a>> co LOO CO 0 00 CO o o CO O CO O O ) ) o o 1 --CM M r--o r--o r». . CO O O ) ) o o O ) ) ^j" " CM M OO O i nn LO COO CM 1-1- o oo o o o CDD ^r o co < r » i - LO w o m s s C00 LO CO CM OO T ^ T ^ CO OO O O O - . -- O O T-C\ll N S ^ f óó O ö ó NN i - CO l O OO ó ó O OO O i - o O )) LO ' t CD OO O O O CU U O ) ) CD D 3 3 O O 1 --O --O CC C to o 3 3 CO O a> > > > ^^ ^ c c ca a CJ J CD D Q . . X X CD D H H O O CC CC en n O ) ) CO O CD D ^ ^ 3 3 Ü Ü CD D Q . . CO O 3 3 i— i— CD D > > 4 _ > > C C CO O CJ J CD D O O X X CD D CD D O O c c CD D CD D Q . . 4 - 1 1 cta n n tag e e 0 33 m O-- £ CDD < j CDD ei) CDD CD CDD CD CDD CD Q -- O . COO co =33 I J COO co CDD cjj >> > == i1!. «« 2* + -- ( 0 OO + -O»» ^ C LL I _ XX 3 CDD O - h-üü O CEE CC CD D O ) ) CC C CD D 3 3 O O \-\-o \-\-o cc c ( 0 0 3 3 to o CD D > > +-C" " co o ü ü CD D Q Q X X CD D 1 --u --u LX X fP P O ) ) CO O Cl) ) 3 3 CJ J CD D Q . . CO O 3 3 CD D > > r r m m o o CD D O O X X Cl) ) Cl) ) o o c c CD D <1) ) Cu u OO O O O h -- h» LO CD d o o d d COO i - co co d o d o o CC CD coo en oo 2S.2 2
XX 3 CDD 0 CDD CDëë ë
CDD CD ®® CD CDD CD O.O. o . coo C0 9,, ^ coo co CDD CD >> > «oo ü? Ü 2 2 CDD Q XX 3 CDD U 1 I -üü O CCC cc CDD i CO i -OO CO f- CM COO LO ^ CO CMM CO CO CO "<** CM CM i -r»-- ^t ^t o SS O CD CO CMM CM CM CM " ** CO CO LO CDD i - N O oo m s t coo i oo i -COO O 1 - CM COO O CO LO dd ^ d d h>-- - c o L O d o d o o CC CD COO O ) OO S XX 3 CDD o CDD CD£ë ë
CDD CD ^^ CD CDD "CD Q -- Q . CD D O ) ) CO O + J J CD D 3 3 ü ü h--o h--o cc c CD D CJ) ) CO O CD D 3 3 O O ^ ^ CD D Q . . CO O 3 3 f0 0 CD D COO ^ 33 C COO CO CDD O >> CD —— C L cc x C00 CD " oo CD QQ Ü O .. C XX CD CDD * -.. 05 == 3 »» CO CDD CD >> >oo s
ss »
Q .. i -XX 3 CDD O OO ï O O CCC CL CC CC co o E E <o o "o o co o V) V) 3 3 cö ö O O CD D £} £} -o-o - i5 5 5
XX S CDD ( OII
S
33 CD cc E CU U EE £ coo - s COO x cüü a> nn E C DD co coo « oo o CC C OO O ÜÜ O COO »RandomisedRandomised versus preference groups
Inn expectantly managed patients, no differences in HRQL measures were foundd between those treated according to randomised allocation compared to thosee managed according to their preference. Between women randomised to curettagee and women who preferred curettage a significant overall 5 point differencee (p = 0.03) was found in MCS scores. Women treated according to randomisedd allocation performed worse than those who received the same treatmentt but at their own request.
AdditionalAdditional analysis
Inn randomised women no differences were found between the HRQL scores of womenn allocated to the treatment of their choice (the treatment for which theyy expressed a slight preference shortly before randomisation) and those nott allocated to their preferred treatment. We also found no differences betweenn the proportion of high responders (four or five questionnaires) and loww responders (two or three questionnaires) in the randomised and in the preferencee groups.
Discussion n
Inn this study of women with a miscarriage we found significant differences in scoress of the SF-36 Mental Component Summary scale between women man-agedd expectantly or surgically in the RCT. During the entire three-month studyy period, women allocated to expectant management experienced better mentall health than women allocated to surgical evacuation. Even at twelve weekss a clinically relevant difference between these women still remained. In contrast,, no differences in HRQL measures were seen between the two treat-mentt options in women who were managed according to their own preference. Comparingg randomised and non-randomised patients, women preferring curettagee showed significantly better mental health scores than women under-goingg this very treatment after randomisation. Our results clearly confirm the impactt on experienced HRQL of patients' preferences and shared decisions in thee management of miscarriages.20
AA decreased motivation among women randomised to a moderately non-pre-ferredd treatment could not explain the differences between the two domisedd groups; HRQL scores did not differ between women who were ran-domisedd to their preferred treatment and women who were not. An explanation forr the differences found in mental health scores may be the absence of a treatmentt option for women randomised to curettage as opposed to women randomisedd to expectant management (who still have the option to change theirr mind and have a curettage at a later stage). In our study 50% of the womenn allocated to expectant management underwent a curettage at their ownn request after a period of waiting in vain for a spontaneous loss.
Generalisabilityy of a randomised trial may be limited, because women with ann outspoken treatment preference are likely to refuse participation in such a study;; usually they receive the treatment of their choice after having been excluded.211 Our study does not have this limitation and confirms that provid-ingg a free choice of treatment increases the perceived quality of life that comess with it.5
Strikingg was the significantly lower mental health scores at inclusion among womenn preferring surgical evacuation in comparison with all other women. Maybee this reflects the mental attitude and stability required in accepting thee uncertainties of expectant management.
Thee differential response to both treatments has received little attention so far.. To our knowledge only one study has compared anxiety and depressive reactionss in a randomised trial comparing expectant management and surgi-call evacuation.22 That study reported no anxiety or depressive reactions nor differencess between the treatment groups after a two-week period. We also didd not find any differences in anxiety after two weeks or at twelve weeks, butt we did find an overall better mental health during the entire 12 weeks observationn period in women allocated to expectant management.
Wee conclude that women with a miscarriage in the first trimester who made theirr own treatment choice had a better quality of life after 12 weeks follow-up. Womenn preferring either expectant or surgical management should be allowed theirr choice in absence of clinical considerations. Women without a treatment preferencee should be encouraged to initially follow expectant management. References s
1.. Ambulatory Sentinel Practice Network. Spontaneous abortion in primary care. J Am Boardd Fam Pract 1988; 1:15-23.
2.. Wiebe E, Janssen P. Management of spontaneous abortion in family practices and hospitals.. Fam Med 1998; 30:293-6.
3.. Nielsen S, Hahlin M. Expectant management of first-trimester spontaneous abortion. Lancett 1995;345:84-6.
4.. Wieringa-de Waard M, Vos J, Bonsel GJ, Bindels PJE, Ankum WM. Management of miscarriage:: a randomised controlled trial of expectant management versus surgical evacuation.. Submitted.
5.. Henshaw RC, Naji SA, Russell IT, Templeton AA. Comparison of medical abortion with surgicall vacuum aspiration: women's preferences and acceptability of treatment. BMJ 1993;307:714-7. .
6.. Coulam CB, Goodman C, Dorfmann A. Comparison of ultrasonographic findings in spontaneouss abortions with normal and abnormal karyotypes. Hum Reprod 1997;12:823-6. .
7.7. Deaton JL, Honore GM, Huffman CS, Bauguess P. Early transvaginal ultrasound followingg an accurately dated pregnancy: the importance of finding a yolk sac or fetal heartt motion. Hum Reprod 1997;12:2820-3.
8.. Ploeg van der HM, Defares PB, Spielberger CD. Zelfbeoordelingsvragenlijst. Handleiding.. Lisse, Swets en Zeitlinger; 1981.
9.. Toedter LJ, Lasker JN, Alhadeff J M . The Perinatal Grief Scale: development and initiall validation. Am J Orthopsychiat 1988;58:435-49.
10.. Potvin L, Larsen J, Toedter LJ. Measuring grief: a short version of the perinatal grief scale.. Journal of Psychopathology and Behavioral Assessment 1989;11:29-45. 11.. Ware JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I.
Conceptuall framework and item selection. Med Care 1992;30:473-83.
12.. Aaronson NK, Muller M, Cohen PD, et al. Translation, validation, and norming of the Dutchh language version of the SF-36 Health Survey in community and chronic disease populations.. J Clin Epidemiol 1998;51:1055-68.
13.. Spielberger CD, Gorsuch RL, Lushene RE. STAI manual for the State-Trait Anxiety Inventory.. Consulting Psychologists Press Inc.. Palo Alto, CA, USA; 1970.
14.. Cuisinier MC, Graauw de CP, Kuijpers JC. Meting van rouw na een zwanger-schapsverlies.. Eerste bevindingen met de Nederlandse versie van de Perinatal Grief Scale.. Tijdschrift voor Psychiatrie 2000;34:651-6.
15.. Ware JE, Kosinski M, Keller SD. SF-36 Physical and Mental Summary Scales: A User'ss Manual. MA: New England Medical Center, The Health Institute, Boston, 1994. 16.. Ware JE, Snow KK, Kosinski M, Gandek B. SF-36 Health Survey Manual and Interpretation
Guide.. MA: New England Medical Center, The Health Institute, Boston, 1993.
17.. Ware JE, Gandek B, Kosinski M, et al. The equivalence of SF-36 summary health scoress estimated using standard and country-specific algorithms in 10 countries: resultss from the IQOLA Project. International Quality of Life Assessment. J Clin Epidemioll 1998;51:1167-70.
18.. Verbeke G, Molenberghs G. Linear Mixed Models in Practice: An SAS-Oriented Approach.. New York: Springer, 1997.
19.. Littell R, Pendergast J, Natarajan R. Modelling covariance structure in the analysis of repeatedd measures data. Stat Med 2000;19:1793-1819.
20.. Ankum WM, Wieringa-de Waard M, Bindels PJE. Management of spontaneous mis-carriagee in the first trimester: an example of putting informed shared decision making intoo practice. BMJ 2001 ;322:1343-6.
2 1 .. Brewin CR, Bradley C. Patient preferences and randomised clinical trials. BMJ 1989;299:313-5. .
22.. Nielsen S, Hahlin M, Moller A, Granberg S. Bereavement, grieving and psychological morbidityy after first trimester spontaneous abortion: comparing expectant manage-mentt with surgical evacuation. Hum Reprod 1996;11:1767-70.
