the grief of late pregnancy loss

Patient Education and Counseling 31 (1997) 57–64

The grief of late pregnancy loss

a , b c*J.A.M. Hunfeld , J.W. Wladimiroff , J. PasschieraInstitute Medical Psychology & Psychotherapy, Erasmus University, P.O. Box 1738, Cf 310, 3000 DR Rotterdam,

The NetherlandsbDepartment of Obstetrics & Gynaecology, Division of Prenatal Diagnosis, University Hospital Dijkzigt, Rotterdam,

The NetherlandscInstitute Medical Psychology & Psychotherapy, Erasmus University, P.O. Box 1738, Cf 216, 3000 DR Rotterdam,

The Netherlands

Received 20 October 1995; revised 15 May 1996; accepted 5 July 1996

Abstract

We studied 46 women who had an ultrasound diagnosis of a lethal fetal anomaly (gestational age S 24 weeks).Shortly after the diagnosis, 45% of these 46 women showed severe psychological instability established by aconsensus diagnosis. Three months later, this percentage had diminished significantly to 22%. The total GHQ-28score revealed that after 4 years, 11 out of the 29 remaining participants (38%) had a score of 5 or more, whichindicated a clinically significant degree of general psychological distress. Depression and despair measured with thePerinatal Grief scale, did not decrease significantly over the 4-year period. Women with a strong disposition towardsfeelings of inadequacy or ‘neuroticism’, measured with the Dutch Personality Questionnaire, displayed significantlymore intense grief reactions than women without such a strong disposition. The implications of our study are that inthe face of (threatened) late pregnancy loss, medical care should include (i) paying attention to the need formedical information and emotional support and (ii) performing psychosocial screening of women identified asshowing signs of inadequacy.

Keywords: Late pregnancy loss; Course of grief; Need for assistance; Feelings of inadequacy

1. Introduction This can particularly be attributed to medicaladvances in prenatal diagnosis that have made it

Coping with pregnancy loss has been receiving possible to detect many fetal anomalies beforeincreasing attention over the past 15 years [1–9]. the baby is born. Based on these studies a

distinction can be made between reactions whichoccur after losing an older loved one or after*Corresponding author. Department of Paediatric Surgery,losing a baby. The latter is also referred to asSophia’s Children’s Hospital, Erasmus University, Rotter-perinatal grief and signifies a specific situation indam. Tel.: 1 31 10 4087803; fax: 1 31 10 4363981; e-mail:

[email protected]. which someone does not grieve for a consciously-

0738-3991/97/$17.00 Copyright u 1997 Elsevier Science Ireland Ltd. All rights reservedPII S0738-3991( 97 )01008-2

58 J.A.M. Hunfeld et al. / Patient Education and Counseling 31 (1997) 57 –64

known person, but for a fantasised child with 2.2. Procedurewhom a bond has been formed even before hisor her birth. In addition, specific emotions are To assess psychological instability 2–6 weeksbelieved to play a central role in coping with late (first measurement) after the women had beenpregnancy loss, such as guilt (‘it was my fault’) informed of the presence of a severe or lethaland loss of self-esteem (‘other women have fetal malformation a medical psychologist (JH)managed to bring healthy babies into the paid a home visit to carry out an audiotapedworld’). interview. Medium-term grief and psychological

Scarce knowledge and insight exist regarding instability were measured during a second inter-the reactions to late (threatened) pregnancy loss view 3 months after delivery (second measure-(S 24 weeks) due to congenital anomalies. Par- ment). The Perinatal Grief Scale (PGS) and theticularly prospective and longitudinal studies are Impact of Event Scale (IES) were completed atlacking. Also because of a lack of personality the first and second measurements. To assessassessments it is not known which women are at long-term grief and general psychological dis-risk for pathological mourning (see also [10]). tress, a third audiotaped interview was held byTherefore, this study was performed on women telephone 4 years after the loss (third measure-in whom at a late stage (S 24 weeks) ultrasound ment). Again, the PGS and the IES were com-examination had revealed that their pregnancy pleted and additionally, the General Healthwas complicated by a severe or lethal fetal Questionnaire-28 (GHQ-28). Feelings ofanomaly. The following questions were ad- inadequacy were measured with the Dutch Per-dressed: (1) What is the prevalence of severe sonality Questionnaire (DPQ) at the second andpsychological instability in women 2–6 weeks third measurements only.after an unfavourable ultrasound diagnosis and 3months after the loss, and what is the prevalence 2.3. Instrumentsof psychological distress 4 years later? (2) Whatis the course of stress and grief up to 4 years Severe psychological instability was measuredafter the loss? 3) How is the personality trait by a semi-structured interview and was consid-inadequacy related to stress, grief and psycholog- ered to be reflected by symptoms which led toical distress? 4) What is the need for assistance? long-term disruption of various aspects of daily

functioning. These symptoms could be presenton a physical, psychological and social level, and

2. Materials and methods included extreme eating and/or sleeping disor-ders, panic or fear, neglecting household ac-

2.1. Subjects tivities, neglecting the children, medicine, drug oralcohol abuse and social isolation. Four years

Fifty-five patients were approached who were after the loss, the GHQ-28 [11] was used insteadreferred to the Division of Prenatal Diagnosis in of a clinical evaluation to detect clinically signifi-the period between January 1990–August 1991 cant psychological distress. Specific grief re-for an anomaly scan. Inclusion criteria were: (i) a actions were measured using the PGS [12] withpregnancy of 24 weeks or more and (ii) the the subscales for active grief, difficulty copingpresence of a fetal malformation which was not and despair. Active grief registers normal emo-compatible with extrauterine life or would at tional reactions following a traumatic event, suchleast result in severe mental and/or physical as sadness and missing the baby. More compli-handicap(s). Excluded were women with a previ- cated emotional reactions were measured underously known risk for a congenital fetal structural the heading difficulty coping, which includedmalformation and women who could not express items about social isolation and problems withtheir experience sufficiently in the Dutch lan- day to day functioning. Despair items, such asguage. ‘the best part of me died with the baby’, reflected

J.A.M. Hunfeld et al. / Patient Education and Counseling 31 (1997) 57 –64 59

potential serious long-term effects of the loss. second and third measurement. Pearson’s Prod-The PGS was adapted for pregnant women. The uct Moment correlations were used to assessreliability and validity of the scale are good for social and personal inadequacy in relation to theboth the American and Dutch version [6,13]. intensity of stress, grief and general psychological

The IES measures subjective stress after a distress.traumatic event [14]. It consists of two subscales:(i) intrusion (i.e. unbidden thoughts and imagesabout the loss) and (ii) avoidance (i.e. denial of 3. Resultsmeaning and consequences of the event). Thereliability and validity of the scale are good. The 3.1. Subjectstotal scores on the PGS and IES consist of thesummation of each subscale score. Feelings of A total of 46 pregnant women agreed toinadequacy were measured with the scales of participate in the study. Nine of the 55 womensocial inadequacy (SI) and personal inadequacy approached refused to participate because they(PI) which are subscales from the DPQ. High did not wish to be confronted with the traumaticscores for social inadequacy indicate incompe- event of the loss. At the second measurement,tence in social contacts. These persons feel in- five women withdrew from the study becausehibited and shy. ‘Personal inadequacy’ items they did not wish to be confronted with the loss.include questions about depressed mood, anxie- Four years later, nine women declined furtherty, feelings of insufficiency or low self-esteem. participation out of fear of becoming upset byBoth subscales can be considered to measure discussing the loss. In three cases the (severelyinadequacy or neuroticism. The scales have good mentally or physically handicapped) infant wastest-retest reliability and validity research has still alive; these women were excluded fromconfirmed the intended content of the subscales further analyses. Therefore, the final sample[15]. consisted of 29 women. Maternal age in the

original sample (n 5 46) ranged between 19 and2.4. Data reduction and analysis 44 years (median: 30 years) and gestational age

varied between 24 and 38 weeks (median: 31The content of each audiotaped interview was weeks). At the first measurement, 31 out of the

transcribed into a report. Based on this report, original 46 women had delivered. Twenty-one ofthree clinical psychologists independently judged these infants died before or during delivery andthe emotional reactions of the women regarding seven infants died afterwards (six infants T 28the presence of severe psychological instability days and one infant S 28 days). Three women(SPI). Inter-judge agreement was achieved by had given birth to a live infant and fifteen womenconsensus diagnosis, which means that any differ- were still carrying a live baby. All the womenences between the assessments were discussed had delivered by the second measurement. In theuntil consensus was achieved. This procedure is total sample of delivered infants, 36 died withingenerally used by the clinical staff to monitor 28 days and one infant died 28 days after birth. Apsychological guidance programmes. The stan- total of five infants were alive, but were sufferingdard cut-off score of 5 was used for clinically from severe physical and/or mental handicap(s),significant general psychological distress mea- including central nervous system anomalies, car-sured on the GHQ-28 [16]. The course of stress diovascular anomalies, diaphragmatic hernia.and grief was assessed by t-tests for pairedobservations. Due to drop-outs and due to the 3.2. Psychological instabilitynecessity of including the same subjects in t-testsfor paired observations, these were 41 cases for Clinical evaluation 2–6 weeks after the un-comparison between the first and second mea- favourable prenatal diagnosis revealed severesurement and 29 cases for comparison between psychological instability in 45% of the 46 women.


Table 1Emotional reactions derived from the questionnaires at the first, second and third measurements and the P-values of thedifferences (S.D. between parentheses)

Psychological 1st measurement 2nd measurement 3rd measurement P-value of the P-value of themeasures 2–6 weeks after 3 months after 4 years after difference 1st vs. difference 2nd vs.

the ultrasound delivery (n 5 41) pregnancy loss (n 5 29) 2nd measurement 3rd measurementdiagnosis (n 5 46)

Grief (PGS)Active grief 38.0 (9.5) 35.6 (10.3) 28.0 (8.4) ns 0.000Difficulty coping 23.6 (8.3) 24.7 (9.8) 22.1 (8.3) .03 nsDespair 21.7 (9.1) 22.0 (10.9) 20.2 (7.6) ns ns

Total score 83.2 (24.8) 82.5 (29.9) 70.3 (22.8) ns 0.009Stress (IES)

Intrusion 20.7 (4.3) 19.9 (5.8) 16.5 (5.5) ns 0.005Avoidance 13.6 (4.6) 13.9 (4.7) 12.3 (5.1) ns 0.026

Total score 33.8 (8.0) 33.5 (9.5) 28.8 (9.6) ns 0.006Psychological distress (GHQ)

Somatic complaints — — 2.5 (1.3) — —Social dysfunction — — 1.2 (2.1) — —Anxiety / insomnia — — 1.6 (2.1) — —Severe depression — — 0.69 (1.6) — —

Total score — — 5.7 (6.9) — —

Three months after the infant had been deliv- would have been able to do with the baby’.ered, there was a significant decrease in the t-Tests revealed a significant decline in overallpercentage of women who showed severe psy- grief, active grief (normal grief reactions) andchological instability (22%). Four years later, 11 intrusive images between 3 months and 4 yearsout of the 29 remaining participants (38%) after the loss. Difficulty coping and despair didshowed a clinically significant level of general not show a significant decline during this period.psychological distress measured with the GeneralHealth Questionnaire. 3.4. Inadequacy related to stress, grief and

psychological distress3.3. The course of stress and grief

Table 2 shows the disposition for feelings ofContrary to the significant decrease in the inadequacy of the 29 women who participated in

number of severely psychologically unstable all measurements obtained at the second mea-women 3 months after delivery, the intensity of surement in correlation with stress, grief (secondstress and grief measured with the questionnaires and third measurements) and general psychologi-had not decreased in this period, while difficulty cal distress (third measurement).coping had even increased significantly (Table The correlations between personal inadequacy

11) . (i.e. feeling labile, depressed, anxious and havingThe latter was expressed during our interviews low self-esteem) and stress and grief were signifi-

in terms of: ‘I fell into a very deep hole’ and ‘I cant 3 months after the loss and — althoughkeep on thinking about all the things that I somewhat diminished — 4 years after the loss. A

significant positive relationship was also presentbetween personal inadequacy and general psy-

1In Table 1, means are presented for those cases present in chological distress. Feelings of social inadequacythe concerning measurement. The two paired t-tests refer to

were only positively associated with despair andthe cases present in each of the two measurements compared.difficulties related to coping with late pregnancyThe reason for presenting the means of the larger sample is a

higher power of the estimate. loss at 3 months after the loss: women with a


Table 2Pearson’s product moment correlations to assess inadequacy (obtained at the 2nd measurement) in relation to stress and grief atthe second and third measurements and general psychological distress at the third measurement

Psychological measures Determinants

Correlations 2nd measurement Correlations 3rd measurement

Social inadequacy Personal inadequacy Social inadequacy Personal inadequacy

Grief (PGS)Active grief ns 0.62**** ns 0.33*Difficulty coping 0.50**** 0.75**** ns 0.49***Despair 0.43*** 0.76**** ns 0.41**

Total 0.42 0.75**** ns 0.44**Stress (IES)

Avoidance ns 0.61**** 0.32* 0.44**intrusion ns 0.54**** ns ns

Total ns 0.63**** ns 0.38*Psychological distress (GHQ-28) — — ns 0.46**

*P 5 0.05; **P 5 -0.01; ***P 5 0.005; ****P 5 0.001.

strong disposition displayed significantly more support from their own social environment.intense grief reactions. Three months after delivery there was a greater

need for support from outside one’s own en-3.5. The need for assistance vironment (n 5 10/41, 24%), particularly in

terms of exchanging experiences with fellowIn our study, the need for assistance could be sufferers. In this period, one woman was referred

divided into a self-reported need for chiefly to a psychiatrist by her gynaecologist. Between 3information and a self-reported need for chiefly months after delivery and 4 years after the loss ofemotional support. the baby, about one third of the women (n 5 10/

We observed that shortly after hearing the 29, 34%) experienced a need for emotionaldiagnosis and after delivery, there was a need for support. Four of them had received professionalmedical information about the nature, prognosis assistance from a psychologist or psychiatrist.and cause of the baby’s anomaly, and aboutwhen and how delivery would take place (n 5 18/46, 40%); this need had almost disappeared after 4. Discussion3 months (n 5 1/41, 2%). However, in the inter-vening period (up to 4 years after the loss) the The percentages of women with severe psycho-need for medical information returned (n 5 10/ logical instability (45% at the first measurement29, 34%). In this period over one third of the and 22% at the second measurement), werewomen required information about the risk of consistent with research data from other coun-the anomaly being repeated in a new pregnancy tries. Percentages of 48 and 41% have beenor about additional medical examination (i.e. reported 1 month after pregnancy loss [17,18],prenatal diagnosis). while 6 and 12 months later the percentages had

Shortly after the unfavourable diagnosis and dropped to about 20% [18]. We presume that ourafter delivery a small number of women (n 5 5/ percentage of unstable women 3 months after the46, 11%) required extra emotional support from loss was an underestimation, because all fivethe hospital staff, particularly in the form of women who withdrew from further follow-upunderstanding for their emotions of shock and were being diagnosed as seriously psychological-bewilderment. Most women stated that in this ly unstable shortly after receiving the unfavour-period they had received sufficient emotional able news about the baby. Moreover, their mo-


tive for withdrawing was fear of experiencing — both genetically and through lessons in theemotions related to the loss by discussing the past — copes with major life events.subject. Concerning the course of stress andgrief, we observed no differences in grief scoresbetween the first and the second measurement. 5. Practice implicationsThis may be due to the possibility that a decreaseamong the women who had delivered at the first Our results confirm the formal standpoint ofmeasurement may be equalized by an increase the Dutch Association for Obstetricians andamong the women who had delivered at the Gynaecologists [21] that initially, the treatingsecond measurement. However, t-tests revealed gynaecologist, in cooperation with the generalno difference on the PGS at the first measure- practitioner, should take on the guidance andment and at the second measurement between after-care of parents with (threatened) late preg-these both groups of women. It was striking that nancy termination. On the basis of recent re-even after 4 years, difficulty coping and despair search [7–9,22] and the results of our own study,had not decreased significantly. In addition, in- the guidelines of the note [21] can be sup-trusion was still at a high level of psychological plemented as follows (Table 3):distress in 24% of the women. According to Care before delivery in the case of threatenedHorowitz [19] this indicates a need for further (late) pregnancy loss. The gynaecologist respon-diagnostic tests and treatment. With regard to the sible for informing the parents of the diagnosis ofassociation between inadequacy and grief, our a severe or lethal fetal anomaly, will do this in afinding is consistent with data from other re- ‘bad news’ consultation [23]. Characteristic for asearch into coping with pregnancy loss [6,17,18], bad-news consultation is that the doctor divulgesand also with studies on coping after various the most important aspect of the bad news, theforms of trauma [20]. This indicates that coping fact that the baby has a severe or lethal anomaly,with loss is not only determined by specific at the beginning of the consultation and inproblems, but also by the way in which a person language which the parents can understand. In

Table 3Recommendations for hospital care concerning (threatened) late pregnancy loss

Care before delivery Care after delivery

Tell the parents the bad news Give the parents time to look atalmost directly, in simple the baby and hold him or hertermsCome back later with details, Encourage the parents to give the baby aadvice and suggestions name, to take photographs, say goodbye

and arrange the funeral (cremation)Inform the general practitioner Give the parents a summary of theand treating gynaecologist postmortem (formulated in simple terms)and, if necessary, arrange and offer to discuss itsupport via them or via amedical social worker

General practitioner can provideinformation about coping with the lossand discuss any problems related witha new pregnancyGeneral practitioner can refer the parentsto a medical social worker, psychologistor bring them in contact with fellow-sufferersMonitor psychopathology


2addition, he does not answer the denial re- several years. We found that even after 4 years,actions of the parents (‘you are wrong, it can’t be many of the women in our study group still felttrue’) with arguments in favour of the opposite. the need to talk about the loss. The opportunityAt this stage, many of the parents’ questions will to do so could be offered in the form of, forrepresent a search to find out what the news will example, a sympathetic gesture on the anniver-mean to their lives. It is better for the doctor to sary of the day the baby died. For it is aspostpone giving more detailed information until Shakespeare noted: ‘The grief that does notthe parents have had more time to come to terms speak, whispers the o’er-fraught heart and bids itwith the news and similarly to postpone giving break’ (from Macbeth).advice and suggesting solutions until the parentsshow that they have accepted the news and arethemselves ready to consider what should bedone next. In this way, the doctor offers the Referencesparents the opportunity to bring up and reviewpast information at a later stage or to ask new [1] Peppers LG, Knapp RJ. Motherhood and mourning.

New York: Praeger Publishers, 1980.questions about the baby and/or delivery.[2] Leppert PC, Pahlka BS. Grieving characteristics afterCare after delivery in the case of (late) preg-

spontaneous abortion: a management approach. Obstetnancy loss. The treating gynaecologist and/or the Gynaecol 1984; 64: 19–22.midwife who supervised the delivery or termina- [3] Hohenbruck BG, De Kleine MJK, Kollee LAA, Rob-tion of pregnancy, offer the parents the oppor- broeckx LMH. Rouwverwerking en begeleiding bij het

overlijden van pasgeborenen. Ned Tijdschr Geneeskdtunity to preserve evidence of the birth and brief1985; 129: 1582–1585.life of the baby. In concrete terms this means

[4] Taner Leff P. Here I am, ma: the emotional impact ofthat the parents are given plenty of time to seepregnancy loss on parents and health care professionals.

the baby and to hold it, to take photographs, to Fam Sys Med 1987; 5: 105–114.give the baby a name (if necessary they can be [5] Theut SK, Pedersen FA, Zaslow MJ, Cain RL,encouraged to do so) and to make arrangements Rabinovich BA, Morishisa JM. Perinatal loss and paren-

tal bereavement. Am J Psychiatr 1989; 146: 635–639.for the funeral or cremation. The GP or treat-[6] Toedter LJ, Lasker JN, Alhadeff JM. The Perinataling gynaecologist can also offer to discuss the

Grief Scale: development and initial validation. Am Jautopsy report. Orthopsychiatr 1988; 58: 435–449.The GP will have been informed about the [7] Keirse EAGC. Eerste opvang bij perinatale sterfte.

situation as soon as possible after pregnancy Dissertatie. Leuven: Acco, 1989.[8] Timmers PJ, Kanhai HHH, Geerinck-Vercammen CR,termination and if necessary he can refer the

Keirse MJN. Hulpverlening bij doodgeboorte; het oor-parents to appropriate care providers, such as adeel van de moeders. Ned Tijdschr Geneeskd 1990; 134:medical social worker, a medical psychologist or2391–2395.

patient associations for fellow-sufferers. The GP [9] Van Spijker HG, Korendomp M, Iedema-Kuiper HR,can also contribute to providing support for the Bergsma J, Christiaens GCML. Opvang na

¨parents by informing them about grief reactions zwangerschapsbeeindiging op genetische indicatie: er-varingen van de betrokken vrouwen en hun partners.that can be expected, by discussing their worriesNed Tijdschr Geneeskd 1992; 136: 477–481.about a new pregnancy and by detecting psycho-

[10] Zeanah CH. Adaptation following perinatal loss: apathology at an early stage. critical review. J Am Acad Child Adolesc PsychiatrIt is important that the doctor, the family, 1989; 28: 467–480.

friends and acquaintances of the bereaved [11] Koeter MWJ, Ormel J. General Health QuestionnaireManual, Dutch adaptation. Lisse: Swets & Zeitlinger bv,mother, and often also the father, realise that the1991.desire to talk about the loss can continue for

[12] Potvin L, Lasker JN, Toedter LJ. Measuring grief: ashort version of the Perinatal Grief Scale. J Psycho-pathol Behav Assess 1989; 11: 29–45.

2Where ‘he’ is used ‘she’ is also relevant, but for con- [13] Hunfeld JAM, Wladimiroff JW, Passchier J, Uniken-venience it has been left out. Venema M, Frets PG, Verhage F. Reliability and validity


of the PGS for women who experienced late pregnancy [19] Horowitz MJ, Krupnick J, Kaltreider N, Wilner N,loss. Br J Med Psychol 1993; 66: 295–298. Leong A, Marmar C. Initial psychological response to

[14] Horowitz M, Wilner BA, Alvarez W. Impact of Event parental death. Arch Gen Psychiatr 1981; 38: 316–323.Scale: a measure of subjective stress. Psychosom Med [20] Hovens JEJM. Research into the psychodiagnostics of1979; 41: 209–218. Posttraumatic Stress Disorders. Thesis. Amsterdam:

[15] Luteijn F, Starren J, Van Dijk H. Guide to the Dutch Vrije Universiteit, 1994.Personality Questionnaire. Lisse: Swets & Zeitlinger bv, [21] Note Late Pregnancy Termination. Offical standpoint of1985. the Dutch Association of Obstetrists and Gynaecolog-

[16] Goldberg DP, Hillier VF. A scaled version of the ists, 1994.General Health Questionnaire. Psychol Med 1979; 9: [22] Cuisinier MJC, Janssen H, Timmers L, Hoogduin CAL.139–145. Verliesverwerking en ervaren steun bij miskraam en

[17] Friedman T, Gath D. The psychiatric consequences of doodgeboorte. Ned Tijdschr Geneeskd 1990; 134: 2395–spontaneous abortion. Br J Psychiatr 1989; 155: 810–813. 2399.

[18] Iles S, Gath D. Psychiatric outcome of termination of [23] Schouten JAM. Anamnese en advies. Nieuwe richtlijnen¨pregnancy for foetal abnormality. Psychol Med 1993; 23: voor de informatieuitwisseling tussen arts en patient.

407–413. Alphen aan den Rijn: Stafleu, 1982.

the grief of late pregnancy loss

Documents