INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. University Microfilms International A Bell & Howell Information C om pany 3 0 0 North Z e e b Road. Ann Arbor. M l 4 81 0 6 -13 4 6 USA 3 1 3 /7 6 1 -4 7 0 0 8 0 0 /5 2 1 -0 6 0 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. O rd er N u m b er 9213205 Induced elective abortion and perinatal grief Williams, G ail Barger, Ph.D . New York University, 1991 C o p y rig h t © 1992 b y W illiam s, G ail B a rg e r. A ll rig h ts re se rv e d . UMI 300 N. Zeeb Rd. Ann Arbor, M I 48106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sponsoring Committee: Professor Joanne G riffin, Chairman Professor Ardis Swanson Professor GilbertTrachtman INDUCED ELECTIVE ABORTION AND PERINATAL GRIEF Gail Barger W illiams S u b m itte d in p a rtia l fu lfillm e n t of th e re q u ire m e n ts f o r t h e d e g re e o f D o c to r o f P hilosophy in th e School o f E ducation, H e a lth , Nursing and A rts Professions, N e w Y o r k U niversity 1991 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I hereby guarantee th a t no pa rt of th e dissertation which I have submitted fo r publication has been heretofore published and (or) copyrighted in the United States o f America, except in the case o f passages quoted from other published sources; that I am the sole author and proprietor o f said dissertation; th a t the dissertation contains no m atter which, if published, w ill be libelous or otherwise injurious, or infringe in an yw a y the copyright o f any other party; and th a t I w ill defend, indem nify and hold harmless New York University against all suits and proceedings w hich may be brought and against all claims which may be made against New York University by reason o f the publication o f said dissertation. Gajj^Barger W illiams Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Dedicated to the late Elmer W. Barger Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGMENTS I wish to express my appreciation to th e follow ing people: Dr. Joanne G riffin, whose intellectual and conceptual abilities were integral to this project. Dr. Ardis Swanson, whose consistent patience and encouragement were essential to me. Dr. G ilbert Trachman, whose analytical ab ility was essential to me. Dr. Paul W ing, whose statistical consultation and support were invaluable to me. Rose McDermott, w ith o u t whose typin g expertise this project w ould not have been completed. Kathy Losure, a dear friend whose encouragement and belief in me were essential. Rita C. Kopf, a dear friend, w ho w hile com pleting her own doctoral dissertation made tim e to read my work and give me a critique. The health care professionals w ho assisted in the recruitm ent of potential participants. The women who were courageous enough to be w illin g to share th e ir feelings regarding such a sensitive and highly personal topic. Catherine M. Sanders, one o f the authors o f the GEI w ho provided consultation and support for this research endeavor. James E. W illiams, whose loving support and patient encouragement made the completion o f this research a reality. ii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES LIST OF FIGURE CHAPTER I THE PROBLEM 1 Introduction Problem Statement Research Questions D efinitions Theoretical Rationale Hypothesis Significance of th e Study Delim itations II ' REVIEW OF LITERATURE 12 Grief Perinatal Grief A bo rtion Sequelae III 12 15 16 METHOD 20 The Sample Demographic Characteristics The Procedure Obtaining the Sample Instrument Construction of th e Inventory V alidity Scales Bereavement Scales Research Scales Scoring Reliability Validity Analysis o f the Data IV 1 2 2 2 3 8 8 10 REPORT OF THE FINDINGS: ANALYSIS OF THE DATA Problem Statement Research Questions Auxiliary Findings 20 22 32 34 35 35 37 41 42 43 44 45 47 49 49 58 64 iii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V DISCUSSION OF THE FINDINGS Discussion of Fir dings on Study Hypothesis Discussion of Findings About Perinatal Grief Resolution Time Since A bortion The Presence o f Other Living Children A uxiliary Findings Weeks Pregnant at A bortion History o f Miscarriage Counseling Following A bortion Discussion of Findings Related to the Sample Support Systems The Pri mary Reason for A borti on Discussion of Methodological Factors Experimenter Bias Sociopolitical Climate VI SUMMARY, CONCLUSIONS, AND SUGGESTIONS FOR FURTHER STUDY Summary Conclusions Suggestions fo r Further Research Clinical Implications Clinical Suggestions Postscript 70 70 72 72 74 76 76 77 78 79 80 80 81 81 82 84 84 86 88 88 89 89 BIBLIOGRAPHY 91 APPENDICES 97 A LETTER SEEKING PERMISSION TO CONDUCTTHE STUDY 97 B BRIEF EXPLANATION OF THE STUDY 98 C WRITTEN INFORMED CONSENT 99 D INSTRUCTIONS FOR COMPLETION OF THE GRIEF EXPERIENCE INVENTORY 101 GRIEF EXPERIENCE INVENTORY (FORM B) 102 E iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F DEMOGRAPHIC DATASHEET 107 G SCORING KEY: GRIEF EXPERIENCE INVENTORY (FORM B) 113 H T-SCORE CONVERSION TABLE FOR GEI (FORM B) 114 I INTERCORRELATIONS OF GEI SCALES WITH THE MMPI 118 J MEAN SCORES GEI FOR SURVIVORS OF DEATH OF CHILD, SPOUSE, OR PARENT 120 PERMISSION TO USE THE GEI (FORM B) 121 K V Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Demographic Sample Characteristics 24 Comparison o f Mean T-scores of 4 Individual Women w ith Elevated A R > 70 and the Total Sample 40 Means and Standard Deviations GEI (Form B) 50 Means and Standard Deviations fo r Form B (Non-Death Version) 52 T-Scores and Standard Deviations GEI (Form B) 53 T-score Correlation Coefficients o f GEI (Form B) Scales and Time Since A bortion (TSA) 59 Mean T-Scores o f GEI (Form B) by Time Since A bortion (TSA) 60 Means, Standard Deviations and t-tests GEI (Form B) Presence o f Living Children at Abortion 63 T-Score Correlation Coefficients o f GEI (Form B) Scales and Weeks Pregnant at A bortion (WPA) 65 Means, Standard Deviations and t-tests Comparison between Miscarriage and No Miscarriage 66 Means, Standard Deviations and t-tests GEI (Form B) Pressure Regarding A bortion 68 Means, Standard Deviations and t-tests GEI (Form B) Counseling Following A bortion 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 1. GEI (Form B) Profile 55 vii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER I THE PROBLEM Introduction There is a grow ing body of lite rature w hich documents the phenomenon o f perinatal grief. Concentrated efforts to define perinatal grief began in the 1970s, spurred by the research of Kennell, Slyter and Klaus (1970). These studies indicated that death of a newborn was indeed a significant loss, accompanied by grief. M ore recently Lake, Knuppel, Murphy, and Johnson (1983) depict perinatal grief as sim ilar to th e typical grief reaction tha t ensues fo llo w in g the death o f a beloved adult. Peppers and Knapp (1980) found th a t perinatal grief occurs despite the length o f gestation. Today health care professionals provide grief interventions fo r individuals experiencing such perinatal loss as miscarriage, stillbirth, and neonatal death (Capitulo & M affia, 1985). This is in direct contrast to previous decades when perinatal losses were not recognized as "tru e " losses w ith a legitim ate concomitant grief reaction. Prior to the last decade and a half, parents were not encouraged to grieve or were encouraged to deny th e ir feelings when stillbirths or miscarriages occurred. W hen the effects of abortion became o f interest, few if any negative sequelae were documented (Osofsky, Osofsky, & Rajan, 1973). An interesting finding o f an early study was tha t women undergoing first trim ester abortions more often used the words "th e pregnancy" or "a fetus" as compared to the women undergoing second trimester abortions w h o used the words "this baby" or "a child" (Kaltreider, 1973). Since the legalization o f abortion in 1973, the number of abortions performed annually has steadily increased. Current statistics (U.S. Bureau o f the Census, 1990) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. indicate th a t there were over a m illio n and a h a lf legal abortions performed in 1985. Ninety percent o f these abortions are performed in the first trim ester o f pregnancy. The abortion rate (number o f abortions per 1,000 women 15-44 years old) has m ore than doubled in the last decade, increasing from 13.2 in 1972 to 28.0 in 1985 (U.S. Bureau o f the Census, 1990). Despite this increase, there have been fe w empirical studies undertaken to determine the presence o f psychological sequelae. Problem Statement Do women w ho have undergone i nduced elective abortion experience perinatal grief? Research Questions Is there a relationship between the intensity o f perinatal grief in women w h o have undergone induced elective abortion and the tim e since th e abortion? Is there a difference in the intensity o f perinatal grief experienced by wom en w ho have undergone induced elective abortion w ho have living children and those w ho do no t have living children? Definitions G rief. A definite syndrome w ith psychological and somatic symptoms follow ing death o r loss. This grief is a normal reaction to a distressing situation and is accompanied by somatic distress, such as deep sighing respirations, lack of muscular strength, loss o f appetite and sense o f taste.tightness in the throat, and a choking sensation accompanied by shortness o f breath. The syndrome also includes preoccupation w ith an image o f the deceased, guilt, hostile reactions, and loss o f usual patterns of conduct (Li ndemann, 1944). Induced Elective A bo rtion. The expulsion o f the products o f conception before tw en ty weeks gestation or before the fetus has reached a stage o f via bility by medical Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. intervention. In this study induced elective abortions are those abortions m eeting the above criteria and perform ed to term inate an unwanted pregnancy and reported by participants (Pilliteri, 1985; Ziegel SCranley, 1984). Intensity o f Perinatal G rief. The extreme in degree or strength to which th e symptoms o f grief in relation to a perinatal loss are felt. The intensity o f perinatal g rie f in relation to the abortion experience was measured by elevated T-scores > 50 on the scales o f the Grief Experience Inventory (GEI) (Form B) (Sanders, Mauger and Strong, 1985). Perinatal. The tim e before and after birth, defined as beginning w ith conception and extending through the 28th day o f newborn life (Reeder,Mastroianni, & M artin, 1983). Perinatal G rief. The g rie f described by Lindemann (1944) and experienced in a specific context, follow ing the death or loss th a t occurs during the perinatal period. In this study perinatal g rie f was measured by scores on the several scales o f the GEI (Form B) which represent th e multidim ensional aspects o f grief. Time Since A b o rtio n . The am ount of tim e in years th a t has elapsed since th e abortion as reported by the women. Theoretical Rationale G rief is a normal response to death which is a universal experience th a t is repeatedly encountered. When th e loss is the death o f a loved one, it threatens the individual w ith a negation o f self and all tha t is valued (Rando, 1984). The United States is said to be a death-denying culture where there is widespread refusal to confront death and consequently few e r rituals fo r recognizing it. All to o often, the dying are sent to institutions to die (Rando, 1984; Kubler-Ross, 1975). There is little or no open acknowledgm ent o f death or communication regarding it. Very often Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. society places unrealistic expectations on individuals about grie f; g rie f is openly discouraged and grievers are to ld to "be brave" (Rando, 1984). This death-denying attitude persists despite the w ork o f Lindemann (1944) which conceptualized the human experience o f th e g rie f process. Lindemann's (1944) conceptual fram ework o f grief and the basic tasks of the g rie f process are valid today. This conceptualization o f grief and grief work has been utilized by many investigators (Parkes & Weiss, 1983; Wordon, 1982)inan attem p t to describe different types of losses and grief reactions. Lindemann (1944) firs t described th e acute grief reaction tha t occurs in response to a distressing situation and described a definitive syndrome w ith psychological and somatic symptomatology. Five distinct characteristics o f grief were noted to be common to all individuals. Lindemann (1944) identified three stages o f the grief process. The first stage was considered one o f shock and disbelief, which is recognizable by the inability to accept the loss or on occasion denial that the loss has occurred. The second stage is the period o f acute mourning, characterized by acceptance o f the loss, disinterest in daily affairs, crying, feelings o f loneliness, insomnia, and loss o f appetite. There is also an intense preoccupation w ith the image o f the deceased. The third stage or resolution o f the grief process occurs as there is a gradual reentry into the activities o f daily life and a reduction in preoccupation w ith th e image o f th e deceased. Interest in the g rie f reaction to perinatal loss was stimulated by the research o f Kennell,Slyter, and Klaus (1970) w h o modeled th e ir research after th e typical grief reaction described by Lindemann (1944). They found th a t parents experienced grief follow ing the death o f a newborn. Peppers and Knapp (1980) found th a t a grief response occurs w ith stillbirths and miscarriages also. Their report indicated th a t the perinatal grief experience is sim ilar to the grief experienced fo llo w in g the death o f a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. beloved adult. More recent studies (Kellner, Donnelly & Gould, 1984; Lakeetal., 1983) not only support these findings but indicate th a t the perinatal g rie f response is as severe as the grief experienced after the death o f an ad ult w ith whom the bereaved has shared a long-term relationship. W hile it is acknowledged th a t perinatal grief is similar to that w hich follows the death of an adult, there is also some indication that perinatal grief m ight be unique. According to Furman, (1976) the g rie f follow ing the loss of a child involves not only dealing w ith the actual loss, but also dealing w ith th e loss o f parts o f oneself because parental love consists o f love for the child as w ell as self love. Peretz (1970) described categories o f loss. Loss of some aspect o f self was conceptualized as a loss tha t related to fantasies and expectations of w h a t a specific role m ight have been. Furman (1978) postulated th a t perinatal grief encompasses tw o distinct tasks or processes: identification and detachm ent. Resolution o f parental grief is facilitated when the parents have an opportunity to identify w ith the ir newborn in some meaningful way. The identification w ith th e child facilitates detachment or le ttin g go. Viewed in this lig h t, perinatal g rie f can be arduous since there is lim ited tim e or opportunity fo r identification. Furman (1978) likens perinatal loss to an am putation and advocates providing parents w ith some memory to facilitate grieving. Rando (1986) suggests th a t society's reluctance to recognize perinatal loss and the concomitant grief w ork is related to the uniqueness o f this grief work. Kirkley-Best (1981) found th a t both intrauterine and intrapartum deaths are accompanied by a grief response and that many of these were followed by chronic grief. Lakeetal. (1983) reported evidence o f a pathologic g rie f reaction in some instances o f perinatal loss which they indicated was related to the lack o f emotional Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. support fo r the grieving parents. Perinatal losses seem to be surrounded by w h at has been termed a "conspiracy o f silence" (Johnson-Sodenberg, 1981). Estok and Lehman (1983), in a fo llo w -u p study o f grieving parents, corroborated tha t parents found "validation o f the loss" most helpful in th e ir grief. W hile the acute phase o f g rie f as described by Lindemann (1944) may last only weeks or months, true resolution o f perinatal loss may last months o r years (Wong, 1980). It seems typical th a t the period o f maternal grief extends beyond the tim e afforded by society for support. W allHaas, (1985) reported th a t women w h o had first trim ester spontaneous abortions experienced grief. These women cited lack o f support and understanding by significant others as th e ir major problem in the resolution o f grief. The literature (Furman, 1978; Peppers 8t Knapp, 1980; Kirkley-Best, 1981; W ong, 1980; Johnson-Sodenberg, 1981; Estok and Lehman, 1983; Lakeetal., 1983; Miles and Crandall, 1983; Kellner, Donnelly & Gould, 1984; Wall-Haas, 1985) documents tha t perinatal grief is experienced fo llo w in g perinatal loss. Perinatal g rie f is present regardless of actual iength o f gestation, i.e.: spontaneous abortion, miscarriage, stillbirth, or neonatal death. This phenomenon o f perinatal grief seems consistent w ith attachm ent theory. A ttachm ent has been described as a developmental process which changes over tim e (Mercer, 1981). A ccordingto Kirkley-Best (1981) the tasks o f attachm ent may vary according to the trim ester o f pregnancy w ith attachment during the first trim ester representing bonding to the pregnancy. W hile attachm ent begins in utero, it may be unrelated to w a ntin g pregnancy (Tanner, 1971; K en ne ll.eta l., 1970). Cranley (1981) suggested th a ta degree o f attachm ent exists early in gestation, and tha t this attachment is heightened as th e pregnancy progresses and takes on new dimensions, as w ith the actual confirm ation o f pregnancy and occurrence o f fe ta l movements. The fantasies wom en Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. experience throughout pregnancy related to the unborn child are th o u g h t to be indicative o f the maternal fetal attachm ent process (Tulman, 1981). Buckles (1982) and Joy (1985) found th a t many women studied post-abortion were suffering from depression as a result o f unresolved grief. Viewed in this context induced elective abortion can be thought o f as having a d e finite potential for perinatal grief. Recently, Rando (1986) described loss in relation to induced elective ab ortio n in terms o f being deprived of a valued object or person. According to Lindemann's(1944) description, acute grief entails grieving not only fo r the actual individual who is lost, but fo r the hopes, dreams, fantasies, and un fu lfille d expectations th a t the griever held fo r the lost relationship. Subsequently, the griever must not only identify w hat is being lost in the present, but w h a t has been lost in th e future. In much the same way, the pregnant woman, from th e m om ent th a t she is aware of the pregnancy, starts to develop feel ings regardi ng the pregnancy and the future child and to have many dreams and expectations fo r th e relationship (Rando, 1984). Three additional studies (Katz, 1983; Rando, 1983; Redlener, 1985) explored the g rie f response related to loss o f a child using the GEI. Rando (1983) found in a study o f 54 parents th a t grieving actually intensified in the third year o f bereavement. Katz (1983) studied 40 volunteer parents w ho had lost a child suddenly; they reported a trend o f increased intensified grief. Redlener (1985) investigated variables th a t m ediate the long-term outcome o f bereavement. Redlener reported th a t correlations between tim e since last death and GEI scale variables revealed a trend tow ard dim inishm ent o f g rie f levels w ith the passage o f time. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Rando (1986), in a review o f abortion literature, reported th a t g rie f appears as th e individual becomes more fully aware th a t the loss has occurred and th a t tim e may vary considerably. Theobold (1985) hypothesized th a t w om en who had previously delivered a viable baby w ould perceive the abortion experience more negatively than women w h o had not delivered a viable baby. This was n o t supported. However, Rando (1986) reported that grief reartions have been noted at the anniversary o f the expected date o f confinement, w ith subsequent pregnancies, or w ith subsequent spontaneous abortions and in fe rtility problems. Consequently, th e presence o f other living children prior to or fo llo w in g the abortion may be related to the g rie f experience. Hypothesis Women w ho have undergone induced elective abortion w ill experience perinatal grief. Significance o f the Study Interest in perinatal grief emerged in the 1970s w ith the w ork o f Kennell, eta l., (1970) w ho were among the first to describe the phenomenon o f perinatal grief. The d e fin itio n o f perinatal grief has been expanded (Peppers & Knapp, 1980) to encompass various types o f losses including death of a newborn, stillbirth, miscarriage, and spontaneous abortion. To date, there has been little investigation to determ ine w hether perinatal grief reaction follow s an elective abortion. Perhaps the lack of investigation mirrors th e lack o f recognition and validation th a t society once afforded other types o f perinatal losses. Raphael (1980) cites tha t fo llo w in g induced elective abortion, society usually gives the woman the covert message th a t she should be relieved rather than sad. Frequently the procedure is Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. handled as a mere surgical intervention by health care professionals w ith little or n o m ention o f "products o f conception," "fetus" or "baby." W ith all the recent a tte n tio n to perinatal grief, the lack of research into the area of abortion is strikin g . It has been w ell documented that grieving follow ing spontaneous ab ortio n is common (Stack, 1984). In order fo r women to engage in th e process o f grieving fo llo w in g such loss, the loss must be acknowledged by self and others. Corney and Horton (1974) noted th a t symptoms of pathologic grief have occurred in some w om en fo llo w in g spontaneous abortion whether or not th e pregnancy was planned. Intensity o f g rie f over tim e has been studied using th e GEI (Katz, 1983; Rando, 1980,1983; Redlener, 1985). Although these three studies representthe first tha t have begun to study loss o f a child in a more systematic fashion, utilizin g a standardized measure o f g rie f w ith reported reliability and validity, the resultsare contradictory. W hile Katz (1983) and Rando (1980,1983) found th a t g rie f intensified in th e third and fo u rth year follow ing loss of a child; Relener (1985) found th a t g rie f intensity diminished w ith the passage of time. A t present nursing is actively involved in the provision o f nursing care t o individuals experiencing induced elective abortions in settings such as hospitals and ab ortio n clinics. In order to provide holistic care, nursing needs an adequate d e fin itio n o f the needs o f this specific population. Nursing research endeavors are w arranted to establish whether a grief reaction is experienced by this group. Only if a perinatal g rie f response is identified can nursing begin to address appropriate interventions such as providing permission fo r these individuals to grieve thereby fa cilita tin g th e ir grief w o rk and preventing pathologic sequelae. It is now estimated th a t approximately 1.5 million abortions are conducted annually in the United States (Peppers, 1987-1988). The research, to date, indicates Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -1 0- th a t a small, unspecified percentage o f these wom en suffer negative em otional consequences (Lodi, McGettigan &8ucy, 1985). The exact nature and extent o f this g rie f reaction is unknown. It has been emphasized th a t some wom en may be unaware th a t emotional feelings are legitim ate fo llo w in g abortion. If they are aware o f the feelings they may be reluctant to express them since th e ir loss is a socially unacceptable one (Rando, 1986). Consequently, the number o f women experiencing negative sequelae may be underreported. More data are needed regarding the factors th a t m itigate the grief response for these wom en. In addition, studies th a t utilize a measure of grief w ith reported reliability and v a lid ity and study g rief intensity over tim e are warranted. Delim itations 1. In order to control for the effect o f grief from previous abortions this investigation was lim ited to women w ho had experienced th e ir first induced elective abortio n (Theobold, 1985). 2. In order to control for grief from other concurrent losses wom en w ith a reported history of perinatal losses w ith in th e last fiv e years w ere n o t included in the study (Peppers and Knapp, 1980). This delim itation was m odified since 10w om en w ith a reported history of perinatal loss (miscarriage) in th e last five years completed th e GEI (Form B) and the Demographic Data Sheet. Their mean GEI (Form B) scores were compared to the mean scores of those women (n = 73) w ith o u t a reported history o f perinatal loss in the last five years. No statistically significant differences were found on any o f the GEI (Form B) scales so a decision was made to include them (n = 10) in the overall data analysis. 3. This investigation was lim ited to women w h o w ere able to read, w rite and comprehend English. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4. Only wom en w ith o u t a documented psychiatric history were included in the study since review o f the psychiatric lite rature indicates that psychological sequelae fo llo w in g an abortion m ight be the result o f preexisting psychiatric problems (Rando, 1986). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -12- CHAPTER I! REVIEW OF THE LITERATURE G rief Grief has been defined as psychological, social, and somatic reactions to th e perception o f loss. Grief is a normal reaction to loss and a universal experience w hich is repeatedly encountered (Rando, 1984). Lindemann (1944) developed a conceptualization o f grief and the essential tasks o f this process. In his landmark study o f the tragedy o f the Coconut Grove Fire in Boston in 1944, Lindemann described acute g rie f as a normal reaction to a distressing situation. One hundred and one survivors o f relatives w ho died in the fire were interviewed. These observations and clear descriptions o f acute grief were widely recognized as an account o f the norm al g rie f process. Results o f psychiatric interviews with these clients revealed a clinical picture w hich was similar fo r all individuals suffering from acute grief. Five d istin ct characteristics were noted: somatic distress, preoccupation w ith the im age o f th e deceased, gu ilt, hostile reactions, and loss o f patterns of conduct. The sensations o f somatic distress were described as occurring in waves lasting from tw en ty minutes to an hour. Individuals suffering from acute grief experienced a feeling o f tightness in the throa t, choking w ith shortness of breath and a need fo r sighing. These individuals also experienced an em pty feeling in the stomach, lack o f muscular pow er, and an intense subjective distress described as mental pain or tension. There seemed to be an awareness on the part o f individuals experiencing these symptoms th a t they were precipitated and/or exacerbated by mention o f the deceased, or condolence calls and expressions of sympathy. It was noted th a t individuals avoided these reminders in order to avoid the distressing symptomatology. They also described an Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -13- altered sensorium w ith a slight sense o f unreality. They were described as having an increased em otional distance from others w hile experiencing an intense preoccupation w ith an image o f the deceased. Many experienced a strong preoccupation w ith feelings of guilt. They seemed to relive th e tim e p rio r to the death o r loss in search o f a failure to do the right th in g in relation to the dead individual. These individuals also experienced w h at was called a loss o f w arm th in relationships to oth er people. They were described as having a tendency toward anger o r hostility tow ard well-intentioned members of fam ily and friends. And finally, the acutely grieved exhibited changes in th e ir pattern o f daily conduct. They exhibited restlessness, an inability to remain still. They seemed to be in a state of constant m otion, as in searching for something to do. Even ordinary, routine tasks were approached w ith great difficulty. Lindemann's (1944) study was the first published systematic investigation o f bereaved individuals. This study is the first attem pt to classify signs and symptoms o f typical grief responses based on empirical evidence. Resolution o f g rie f was reported to be dependent on how w ell th e individual accomplished his or her g rie f w ork. This g rie f w o rk is described as the achievement o f emancipation from bondage to the deceased and readjustment to the environment from which the deceased is absent. Finally, the individual begins to form new relationships. W hat Lindemann observed is th a t many individuals experience difficulty in the resolution o f grief w o rk because they attem pt to avoid "the intense distress connected w ith the grief experience and to avoid the expression of emotion necessary for it" (Lindemann, 1944, p. 143). Lindemann stressed that grief was not a medical or psychiatric disorder, but rather a normal reaction to a distressing event. Failure to achieve g rie f resolution was th o u g h t to predispose one to mental illness. Finally, it was reported th a t normal g rie f Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -14- may appear im m ediately fo llo w in g the loss, o r be delayed, exaggerated o r even absent. In 1958, Marris conducted a retrospective study o f young East London widows w h o had lost th e ir husbands w ith in the last tw o years. This study focused exclusively on loss o f a spouse. His findings indicated th a t although the m ajority o f the respondents had been w idow s fo r approximately tw o years, they fe lt th a t they had n o t yet recovered from the loss. Their lives were described as empty and futile. The study indicated th a t the widows had w ithdraw n from social ties and had experienced frequent symptoms of insomnia and w eight loss. Marris concluded tha t grief resolution m ight require tw o or more years as opposed to the several weeks reported by Lindemann. The findings were based on interviews, and no inquiry was made to the duration o f these reactions. Parkes (1965), expanded upon Lindemann's concept o f grief. He studied the reacti ons and feel ings o f widows fo r over 12 years. He matched 22 bereaved psychiatric patients to the widow s in Marris' (1958) study. He reported typical, uncomplicated g rie f confirm ing the earlier findings o f Lindemann (1944). Following this study, he (Parkes, 1971) conducted a study in w hich he examined the longitudinal effects o f grief on 22 widows in th e London area. This study revealed im portant inform ation regarding the process o f g rie f and the changes tha t occur overtim e w hich corroborated the belief th a t grief was a phasic process w ith no clear demarcation from one phase to th e next. In 1974, Glick, Weiss and Parkes studied bereaved widows and widowers using more stringent controls. Structured interviews were conducted w ith 68 individuals under the age o f 45 over a tw o to four year period at five d iffe re n t intervals. An im portant finding was the length o f tim e involved in the grieving process which Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -15- indicated th a t the duration o f grief should be measured in years rather tha n weeks o r months. Most im portant was the find in g th a t some individuals never recover from th e loss. A lim itation o f th e ir findings was th a t th e ir sample centered on th e lower end o f the socioeconomic spectrum and lim its generalizability to other socioeconomic groups. This data was reanalyzed (Parkes, 1975) in order to determ ine w h a t variables m ig h t correlate most highly w ith negative reactions one year a fte r the loss. The principal conclusions from this study are th a t if intense grief, anger, or self-reproach expressed shortly after the loss does not decline in intensity w ith in six weeks, a poor outcom e is predicted a year later. In addition, lo w socioeconomic status, lack o f preparation for the loss, a sudden or unexpected loss, the num ber and intensity o f life crises experienced before the loss, were predictors o f poor outcome. The consensus among investigators of g rie f (Lindemann, 1944; Marris, 1958; Parkes, 1965; 1971; G licketal., 1974) is tha t there is a d e fin ite observable g rief response follow ing the death of a beloved adult. The actual length o f the g rie f in term s o f resolution has been redefined as years rather than weeks. Perinatal Grief Interest in the presence of a g rie f response associated w ith loss o f a child was an extension o f the adult grief/bereavement literature. In 1970, Kennell, Slyter, and Klaus studied the grief response o f 20 mothers w h o had experienced the death of an in fa n t ranging in age from an hour to 12 days. Their study revealed no significant differences in grief indexes such as sadness, increased irrita b ility, and preoccupation w ith the deceased related to the age o f the child. Peppers and Knapp (1980) interview ed 100 bereaved mothers w ho had experienced neonatal, s tillb irth and fetal losses. Using a modified version o f The M ourning Score developed by Kennell, Slyter, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -16- and Klaus (1970), they found no significant differences among the g rie f scores fo r the three types o f perinatal losses. Study o f perinatal grief in the last decade and a half is an extension o f the existing g rief literature w ith regard to death o f a child. Sanders (1977; 1979-1980) studied the differences in grief reaction among 102 newly-bereaved parents and compared them to 107 controls. She developed th e G rief Experience Inventory (GEI) to measure experiences, feelings, and behavior during the g rie f period. The highest grief intensities were found among the bereaved parents w ho revealed more somatic complaints, greater depression, anger, g u ilt and despair, than those w h o had sustained loss o f either a parent or spouse. This is consistent w ith th e w o rk o f Arnold and Gemma (1983) and Hagan (1974) which id en tified g rie f reactions o f parents follow ing the loss o f an infant from sudden in fa n t death syndrome. A conclusion o f th e ir studies was th a t parents may never truly resolve th e ir g rie f fo llo w in g the death o f an infant. A bortion Sequelae The lite rature on abortion and abortion sequelae is at best contradictory. In general, induced elective abortion has not been considered to be a cause fo r grief. Prior to the legalization of abortion in 1973, the m ajority of studies focused on the medical effects of abortion. In 1973, Osofsky, Osofsky, and Rajan reviewed the literature w ith regard to studies done fo llo w in g legalization. They concluded th a t there was little evidence to conclude th a t abortion was associated w ith negative psychological sequelae. They reported tha t w h ile the m in ority of wom en did experience some feelings of sadness, guilt, and depression initially, these symptoms subsided w ith in a few months. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -17- In 1981, Doane and Quigley published a m ajor review o f the ab ortio n literature published since 1970. They concluded tha t approximately 5% o f patients reported feeling worse several months after th e ir abortion. The most commonly reported symptoms, accounting for 57% o f adverse effects, were depression w ith o r w ith ou t grief, g u ilt, and crying. They also indicated that when data was gathered a year or more fo llo w in g the abortions, almost no such symptoms were reported. They concluded th a t w hile the proportion o f clients w ho experience adverse psychological effects was small, abortion nonetheless can have the potential fo r more intense consequences depending upon the individual woman. Raphael (1983) indicated tha t w h ile the proportion o f wom en suffering adverse effects o f abortion is small, there may be a problem w ith id en tification o f this select population. She suggested th a t some women may employ denial in order to undergo abortion and this denial may persist In the post-abortion period. These women may mask th e ir em otional responses in later follow-up. This a u thor contends tha t fo r some women relief and sadness coexist follow ing abortion. Rando (1986) suggests tha t some individuals may be more at risk fo r emotional sequelae. A woman w ith a previous history of emotional illness or one w h o had been coerced or pressured into undergoing an abortion would be at a higher risk for adverse emotional reactions. Gilver (1987) explored the psychosocial adaptation o f 16 college wom en in response to unplanned pregnancy and abortion. The women were interviewed prior to the abortion, 2 weeks fo llo w in g the abortion, and again at 3 to 6 m onths post abortion. Half o f the wom en (8) were reported to experience continued thoughts and feelings related to the pregnancy. Four of the eight women were observed to have a marked decline in fun ction follow ing the abortion. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -18- Speckhard (1985) interviewed 30 women who were perceived to vie w the ir abortion as stressful. Immediate and long-term responses included feelings o f grief, depression, anger, g u ilt, fear, surprise, preoccupation w ith image o f the aborted child, lowered self-w orth, victimization, inability to express feelings, and discom fort w ith small children and pregnant women. They reported frequent crying, in ab ility to ta lk regarding the experience, flashbacks, suicidal ideation, and increased alcohol use. Seventy-three percent o f the w om en demonstrated suppression o f affect as a coping mechanism fo llo w in g abortion. Speckhard (1985) reported tha t for many wom en the stress of abortion was relieved only after they found social systems th a t perm itted them to express feelings concerning the abortion experience. The generalizability o f these findings is lim ited since the subjects were selected on the basis o f th e ir perceptions o f abortion as a stressful experience. Despite the num ber of publications regarding abortion only one has reported g rie f response after abortion. Peppers (1987-1988) studied 80 volunteers fo llo w in g abortion. The participants were given a three-part perinatal grief questionnaire (The M ourning Score, Kennell, Slyter, & Klaus, 1980) just prior to the procedure and then six weeks fo llo w in g the procedure. The author concluded th a t the data suggest there is a g rie f response subsequent to elective abortion. The mean postprocedure grief response scores (35.45) were low er as compared to the preprocedure scores (60.98). Peppers concluded th a t the grief is resolved rapidly. There was a w ide variation in overall g rie f scores, w hich seems to indicate th a t some women suffer tremendous em otional trauma. The author also reported th a t the longer the pregnancy continued, the more em otionally traumatic the termination, and the m ore d iffic u lt the g rie f resolution. The grief scores of abortion clients were compared to g rie f scores o f wom en w h o had experienced involuntary fetal or in fa nt loss. The data Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -19- suggested th a t the response to abortion loss was sim ilar to th a t experienced w ith miscarriage, stillbirth, and neonatal loss. Even though methodological lim itations prevented statistical analysis of some of the data, the prelim inary results lent support to the hypothesis o f this present study; th a t there is a de finite g rie f response fo llo w in g induced elective abortion. As Lodi e ta l. (1985) point out, disagreement in the lite ra tu re still exists as to w hether women's lives are disrupted by induced abortion. The lack o f data in this area appears to be because of a reluctance to bring attention to any negative consequences of abortion for fear o f being seen as providing support to the anti abortion/pro-life groups. Despite the controversy regarding abortion, there is a need t o determ ine w hether a perinatal grief response does exist. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -20- CHAPTER III METHOD This chapter describes the m ethod o f the i nvestigation. The sample, the instruments, data collection procedure, and analysis o f the data are discussed. The Sample The sample fo r this descriptive study consisted o f 83 volunteer w om en w ith a self-reported history o f one induced elective abortion. Women who m et th e study delim itations were contacted by th e ir primary health care provider and asked to participate by com pleting the materials in the research packet provided by the investigator. Sites were selected in various states in order to include w om en from a variety o f settings and backgrounds. In order to meet the delim itations o f th e study, women w ere included who had experienced only one induced elective abortion. Only wom en w ith no self-reported history o f perinatal losses w ith in the last five years and those w ith no documented psychiatric history were recruited fo r the study. In addition, only wom en who were able to read, w rite , and comprehend English were included. An im portant determ ination o f sample size is consideration of how the resulting data w ill be analyzed. In the present study descriptive statistics are utilized. It is generally recommended th a t a sample size o f at least 10, and preferably 20 to 30 be selected for each subdivision o f the data (Polit and Hungler, 1983). Castles(1987) recommends 30 subjects fo r each relationship analyzed. The main study hypothesis was tested by examination o f the meanT-scores o f all the scales o f the GEI (Form B). A variety o f descriptive statistics were utilized to describe the findings. The study questions as w ell as ancillary findings utilized correlational statistics and t-tests of significance. When using inferential statistics, a minimum o f 64 is appropriate fo r a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -21- t-test (2 tail) between tw o samples w ith the level of significance set at .05 and a medium effect size (W itte, 1980; Cohen, 1977, Tabachnick & Fidell, 1983). A to ta l of 84 women responded utilizing the research packet. The data from one woman had t o be excluded since she reported tha t she completed th e GEI (Form B) as she remembered she had fe lt 20 years ago and not a t the present tim e. O f the 83 usable responses, 10 women who responded had a history o f miscarriage. Therefore, the data was analyzed utilizing the to ta l sample (n = 83), and w ith those w ho met all the delim itations (no history of miscarriage) (n = 73), and w ith those w h o had a history of miscarriage (n = 10). The groups were similar and a decision w asm adeto include all th e sample (n = 83) in data analysis. In order to meet the delimitations o f the study, the study sample was lim ited to wom en w ho had undergone only one induced elective abortion in order to control fo r possible effects of g rie f from other abortions. All w om en were screened fo r their ab ility to read, w rite, and comprehend English. Women w ith a documented history o f treatm ent fo r mental illness such as schizophrenia were n o t included since it has been reported th a t psychological sequelae to abortion m ight be the result o f pre-existing psychiatric problems (Rado, 1986). Women were screened fo r a history o f perinatal losses w ith in the last five years. Demographic data was collected regarding age, m arital status, occupation, ethnicity, socioeconomic status, educational level, religious a ffilia tio n and participation, length o f pregnancy at tim e o f abortion, time since abortion, and presence of living children. Data regarding history o f other losses w ith in the last five years was also collected as a check fo r g rie f in relation to other significant losses. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -22- Demoaraohic Characteristics The sample consisted o f a total o f 83 women w ith a self-reported history o f one induced elective abortion. Demographic characteristics appear in Table 1. These women had a mean age of 31.9 years w ith a range o f 19-50 years of age. W hile approximately 69% (n = 57) were single at the tim e o f abortion, only 15% (n = 13) w ere single a t the time the study was conducted. Approxim ately 98% of the women (n = 81) were w hite and 2% were black (n = 2). The m ajority o f the women (78.4%; n = 65) reported some college or an undergraduate degree. Only one respondent (n = 1) had less than an eighth grade education. The women reported a wide variety of occupations w ith slightly more than h a lf o f the sample (55.4%; n = 46) presently employed at the time o f the study. Approxim ately fifty percent (50.6%; n = 42) of th e fathers o f the aborted pregnancies also had some college preparation o r held an undergraduate degree. The occupations o f the fathers varied. The m ajority o f the wom en (47.0%; n = 39) were Christian (non-Catholic), w hile 41.0% (n = 34) were Protestant. The m ajority o f the wom en (72.3%; n = 60) reported the frequency o f attendance a t religious services to be weekly. Approximately one fourth (25.9%; n = 21) of the women reported a present annual household income o f $50,000 o r greater. Over half of the women (58.5%; n = 48) reported the ir husbands as the m ajor wage earner. Approximately half o f the wom en (49.4%; n = 41) reported that they were living w ith their husbands and children at present. Sixty-four percent (n = 53) o f th e sample reported having living children at the tim e of the study, w ith 34.9% (n = 29) having no living children. The actual number o f living children from the sample ranged from 1 to 9 w ith a mean of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -23- 2 children. The m ajority (85.5%; n = 71) o f the w om en did n o t have any livin g children at the tim e o f abortion. The years during which the women reported having th e ir abortions ranged from 1964 to 1990. The total sample (n = 83) remembered the year o f the abortion; 91.6% (n = 76) remembered th e month; 51.8% (n = 43) remembered the day. The women reported a mean o f 8.7 weeks pregnant at tim e o f abortion w ith a range o f 1 to 20 weeks. The mean months were reported as 2.47 w ith a range o f 1 to 5 months gestation. Only 19.3% (n = 16) reported tha t they w ere curre ntly living w ith th e father of th e aborted pregnancy. The primary reason for the abortion varied greatly w ith th e m a jo rity (16.9%; n = 14) reporting "b e in g a fra id " astheprim ary reason. The perceived pressure to have th e abortion varied for this population w ith 27.7% (n = 23) reporting "never pressured" and 24.1 % (n = 20) reporting "very pressured." The present husbands were reported by 42.2% (n = 35) o f the wom en as the individual w ho was most supportive w ith regard t o the abortion experience, 26.5% (n = 22) reported the ir friends had been most supportive; and 19.3% (n = 16) reported "others" as being most supportive. In response to the question o f whether or n o t they received any formal counseling after the abortion, 60.2% (n = 50) reported th a t the y had not had any counseling. O f the 32 women w ho had counseling after the ab ortio n, the m ajority, 34.4% (n = 11), participated in post abortion counseling. In response to whether or not they had experienced any deaths in th e last 5years,37.0% (n = 31) reported no deaths. The remainder of th e women (n = 52) experienced a wide range o f deaths such as distant friends and relatives w ith only one woman (1.2%) reporting death o f a spouse. W ith regard to other losses Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -2 4 - Table 1 Demographic Sample Characteristics T |, . Characteristic (N " 83) No Miscarriage Sample (N = 73) Miscarriage Sample (N = 10) Current Age Mean (in years) SD Range o f Ages (in years) 31.9 33.0 31.8 6.81 6.90 6.26 19-50 19-50 2 4 -4 2 7 6 M arital Status (at tim e o f abortion) Married Single (8.4%) 57 (68.7%) (8.2%) 1 (10%) 51 (69.9%) 6 (60%) Separated 8 (9.6%) 8 (11.0%) 0 Divorced 5 (6.0%) 3 (4.1%) 2 (20%) W idowed 2 (2.4%) 2 (2.7%) 0 Unmarried/living w ith 4 (4.8%) 3 (4.1%) 1 (10%) 50 (68.5%) 1 0 ( 100%) fathe r of terminated pregnancy M arital Status (at tim e o f study) Married 60 (72.3%) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -2 5- Table 1 - C ontinued Characteristic Single Total Sample ... aJ: (N - 83) 13 (15.7%) Sample (N = 73) 13 (17.8%) Sample (N = 10) 0 Separated 2 (2.4%) 2 (2.7%) 0 Divorced 5 (6.0%) 5 (6.8%) 0 W idow ed 1 (1.2%) 1 (1.4%) 0 Unm arried/living w ith 2 (2.4%) 2 (2.7%) 0 fa th e r o f term inated pregnancy Race W h ite Black 81 (97.6%) 71 (97.3%) 10(100%) 2 (2.4%) 2 (2.7%) 0 Less than 8th grade 1 (1.2%) 1 (1.4%) 0 High School Grad/GED 7 (8.4%) 6 (8.2%) 1 (10%) Vocational School Grad 5 (6.0%) 4 (5.5%) 1 (10%) W om an's Present Educational Level Some College 33 (39.8%) 29 (39.7%) 4 (40%) College Deg (Undergrad) 32 (38.6%) 28 (38.4%) 4 (40%) Masters' Degree 5 (6.0%) 5 (6.8%) 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -2 6 - T ab le 1 - C on tin ued Total Sample ^ ^r Miscarriage Sample Sample (N = 73) Miscarriage Sample (N = 10) Yes 46 (55.4%) 39 (53.4%) 7 (70%) No 37 (44.6%) 34 (46.6%) 3 (30%) Characteristic Presently Employed Father's Educational Level (at tim e of abortion) Less than 8th grade High School Grad/GED Vocational School Grad 4 (4.8%) 24 (28.9%) 4 (4.8%) 4 (5.5%) 21 (28.8%) 4 (5.5%) 0 3 (30%) 0 Some College 25 (30.1%) 22 (30.1%) 3 (30%) College Deg (Undergrad) 17 (20.5%) 15 (20.5%) 2 (20%) Masters' Degree 4 (4.8%) 3 (4.1%) 1 (10%) Doctoral Degree 3 (3.6%) 3 (4.1%) 0 Unknown 2 (2.4%) 1 (1.4%) 1 (10%) 7 (8.4%) 6 (8.2%) 1 (10%) Religious A ffilia tio n (Present) Catholic Protestant 34 (41.0%) 30 (41.1%) 4 (40%) Christian (non-Catholic) 39 (47.0%) 34 (46.6%) 5 (50%) Jewish 1 (1.2%) 1 (1.4%) 0 None 1 (1.2%) 1 (1.4%) 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -27- Table 1 - Continued Characteristic Missing Total Sample (N = 83) 1 (1.2%) No Miscarriage Sample (N = 73) 1 (1.4%) Miscarriage Sample (N = 10) 0 Frequency o f Attendance at Religious Services (Present) Weekly 60 (72.3%) 52 (71.2%) 8 (80%) M onthly 4 (4.8%) 4 (5.5%) 0 Yearly 5 (6.0%) 5 (6.8%) 0 Other Never 13 (15.7%) 11 (15.1%) 2 (20%) 1 (1.2%) 1 (1.4%) 0 5 (6.0%) 5 (6.8%) 0 Annual Household Income (Present) $9,999 or less $10,000-$19,999 14 (16.9%) 13 (17.8%) 1 (10%) $20,000-$29,000 17 (20.5%) 17 (23.3%) 0 $30,000 - $39,000 14 (16.9%) 10 (13.7%) 4 (40%) $40,000 - $49,000 10 (12.0%) 8 (11.0%) 2 (20%) Over $50,000 21 (25.3%) 18 (24.7%) 3 (30%) Missing 2 (2.4%) 2 (2.7%) 0 Living Children (at present) Yes 53 (63.9%) 44 (60.3%) 9 (90%) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -2 8 - T able 1 - C on tin ued Characteristic No Missing Total Sample (N = 83) 29 (34.9%) 1 ( 1.2 %) No Miscarriage Sample (N = 73) 29 (39.7%) 0 Miscarriage Sample (N = 10) 0 1 ( 10%) Living Children (at tim e o f abortion) Yes 12 (14.5%) 11 (15.1%) 1 ( 10% ) No 71 (85.5%) 62 (84.9%) 9 (90%) Date o f A bo rtion Mean year SD Range o f years 1979 1979 1978 5.9 6.09 5.07 1964-1990 1964-1990 1970-1988 Months Pregnant at A bortion Mean SD Range of months 2.47 .47 .83 .81 1.0 2.0 2.5 - 5.0 1.0-5.0 1.0-3.0 Weeks Pregnant at Abortion Mean 8.71 8.9 6.9 SD 3.15 3.23 1.66 Range o f weeks 1. 0 - 2 0 .0 1. 0 - 20.0 4.0 -1 0.0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -2 9- T a b le 1 - C ontinued Characteristic _ . .c . °. ® _aPJP6 (N " 83) No Miscarriage Sample (N = 73) Miscarriage Sample (N = 10) Currently living w ith father o f term inated pregnancy Yes 16 (19.3%) 15 (20.5%) 1 ( 10%) No 66 (79.5%) 57 (78.1%) 9 (90%) Unsure 1 (1.2%) 1 (1.4%) Supports (w ith regard to abortion experience) Husband 35 (42.2%) 30 (41.1%) 5 (50%) Boyfriend 6 (7.2%) 5 (6.8%) 1 ( 10%) M other 4 (4.8%) 4 (5.5%) 0 Father 0 0 0 Friends 22 (26.5%) 20 (27.4%) 2 (20 %) Other 16 (19.3%) 14 (19.2%) 2 (20 %) Yes 32 (38.6%) 27 (37.0%) 5 (50%) No 50 (60.2%) 45 (61.6%) 5 (50%) Formal Counseling after A bortion Missing 1 (1.2%) 1 (1.4%) 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -3 0 - T able 1 - C ontinued Characteristic Total Sample r ^ N o M is c a r r ia g e Sample (N = 73) Miscarriage Sample (N = 10) Type o f Counseling after A bortion Psychiatrist 3 (9.4%) 3 (11.1%) 0 Psychotherapist 3 (9.4%) 3 (11.1%) 0 11 (34.4%) 9 (33.3%) 2 (40%) Post A bortion Counseling Planned Parenthood 1 (3.1%) 1 (3.7%) 0 Group Therapy 1 (3.1%) 1 (3.7%) 0 Counseling 7 (21.9%) 4 (14.8%) 3 (60%) Bible Study 4 (12.5%) 4 (14.8%) 0 Support Group 1 (3.1%) 1 (3.7%) 0 Church 1 (3.1%) 1 (3.7%) 0 Deaths in last 5 years Yes 52 (63%) 42 (58%) No 31 (37%) 31 (42%) 10(100%) Types of Deaths in last 5 yrs Husband 1 (1.2%) 1 (1.4%) 0 Boyfriend 1 (1.2%) 1 (1.4%) 0 M other 4 (4.8%) 4 (5.5%) 0 Father 5 (6.0%) 4 (5.5%) 1 (10%) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -31- Table 1 - Continued Characteristic Total Sample (N = 83) Sister 0 Brother 1 Friend Child No Miscarriage Sample (N = 73) 0 (1.2%) 14 (16.9%) 0 1 Miscarriage Sample (N = 10) 0 (1.4%) 13 (17.8%) 0 Other 26 (31.3%) 18 (24.7%) None 31 (37.3%) 31 (42.5%) Yes 41 (49%) 31 (42%) No 42 (51%) 42 (58%) 0 1 (10%) 0 8 (80%) O ther Losses in last 5 years 10(100%) 0 Types o f other Losses i n I ast 5 years Divorce 7 Separation 0 Relocation (8.4%) 20 (24.1%) 7 (9.6%%) 0 0 0 20 (27.4%) 0 Physical Illness 2 (2.4%) 2 (2.7%) 0 Rape 1 (1.2%) 1 (1.4%) 0 Miscarriage Missing None 10 (12.0%) 1 (1.2%) 42 (50.6%) 0 1 (1.4%) 10(100% ) 0 42 (57.5%) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -3 2 - w ith in th e last 5 years, 51.0% (n = 42) reported "n one." Seven wom en (8.4%) reported "divorce" and 24.1% (n = 20) reported "relocation." The sample for this study was a purposive sample (Dempsey & Dempsey, 1981; Castles, 1987) since the participants were included specifically fo r their abortion experience. In this descriptive study the dependent variable was perinatal g rie f and the independent variable was the abortion experience. The investigator sampled a group o f wom en w ith a self-reported history o f one abortion and described the ir experience/response in terms o f the dimensions o f perinatal g rie f in relation to the abortion experience. The research design was ex post facto si nee the investigator did not have control over th e independent variable. The Procedure The investigator sought permission (Appendix A) to conduct this study in various o u tpatie nt clinics, fam ily planning agencies, w om en’s health service agencies, and through individual psychologist, physician, nurse, and counselor referrals. The investigator either met w ith or had a phone consultation w ith the administrators at the various agencies. Once permission was obtained to collect data at a specific site the investigator either m et or had a phone consultation w ith the primary health care provider w h o assisted in the recruitm ent of potential participants. The health care providers screened and selected potential participants u tilizin g th e research packet. A t each site the health care providers only recruited wom en w h o could read and w rite English by verifying this w ith the women at the tim e th a t the y introduced the study. The health care providers were asked to id en tify wom en w ho m et all the delim itations of the study. Once the primary health care provider identified potential participants, they were provided a research packet w hich included a le tte r of introduction (Appendix B) from the investigator. This le tte r (Dear Participant) was Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -33- eith er read o r mailed by the primary health care provider. No names, addresses, or any other id e n tifyin g data about participants were released to the investigator. All client contact was through the primary health care provider. If a wom an m et the delim itations o f the study and agreed to participate, the primary health care provider gave the woman w ith copies of th e inform ed consent form (Appendix C). The inform ed consent form was read, signed, and witnessed. One copy was given to th e woman and the other returned to th e investigator in a separate stamped pre-addressed envelope. Each participant was assured of anonym ity by the primary health care provider. The health care providers were requested by the investigator to verbally reiterate to th e participants that th e ir willingness and o r refusal to participate in the study would in no way affect the health care services provided them . The participants were given a research packet in addition to th e Informed Consent w h ich contained: Instructions fo r Completion o f th e Grief Experience Inventory (Form B), (Appendix D). Grief Experience Inventory (Form B), (Appendix E). Demographic Data Sheet, (Appendix F). A stamped pre-addressed envelope fo r return of research m aterial to the investigator. A stamped pre-addressed envelope fo r separate return o f one copy o f Informed Consent. In the event th a t a participant experienced the need to discuss feelings o f grief related to answering the questions on either the GEI (Form B) o r the Demographic Data Sheet, th e phone number of the investigator was provided in th e Inform ed Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -3 4 - Consent Form. Appropriate referrals fo r g rie f counseling were made as necessary to either private practitioners or support groups that specialized in g rie f and bereavement counseling. A ppropriate practitioners and support groups th a t provided post-abortion counseling were also identified by the investigator as additional resources. No referrals w ere requested. The only telephone contact th a t occurred between the participants and th e investigator was th a t of phone contact initiated by the women if the y wished t o seek clarification from th e investigator or discuss feelings in relation to th e study. The investigator was contacted by phone by five women. O btaining the Sample A total o f 70 personal contacts were made by the investigator over a 9-month period of tim e in order to obtain th e present study sample. The investigator was denied access to prospective participants by many health professionals w ho expressed concern regarding the possible outcome o f th e study. M any health professionals indicated fear th a t the results o f the present study m ig h t lend support to the pro-life or pro-choice movement. Even though anonymity was assured, three hospitals declined th e investigator access since they did not want th e public to know th a t abortions were performed at th e ir facilities. Several abortion facilities declined access to th e ir population because they did n o t w a nt to draw atten tion to the ir facilities at a tim e when the abortion issue and legislation was under debate. A few professionals expressed concern th a t th e act o f answering the GEI (Form B) m ig ht cause the women to grieve. It is interesting to note th a t when the investigator gained access to potential participants through primary health care providers, the wom en were very responsive. In fact, five wom en initiated phone contact w ith the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -35- investigator in order to discuss the ir feelings. These five women indicated tha t it was helpful to have the op portunity to discuss their abortion experience. A to ta l o f 39 agencies o r health professionals who were contacted assisted to ob tain the present sample. A number o f wom en at each site had to be excluded since they did not meet the delim itations; many had had a history o f miscarriage(s) or had more than one abortion. The data collection took place in several states over a 9m onth period o f tim e (February through November, 1990). Instrument Perinatal grief was measured by the scores obtained on the 15 scales o f the G rief Experience Inventory (GEI) (Form B) (Sanders etal., 1985). The GEI (Form B) is a self-report inventory, which consists o f 104 true-false statements found to be frequently associated w ith grief and bereavement. The GEI was developed as an instrum ent to objectively measure the multidimensional aspects o f grief. The GEI (Form B) measures the various behaviors of individuals during the g rie f process by means o f 15 scales and consists o f six standard bereavement scales measuring varying dimensions of grief, six research scales, and three validity scales. Each scale o f th e GEI (Form B) yields a raw score which is then converted to a T score. The larger th e T score th e greater th e intensity o f the behavior measured by the scale (Sanders e ta l., 1985). Construction o f the Inventory The G rief Experience Inventory arose from the development of a Q-sort technique to quantify data from individual interviews w ith bereaved individuals. The 180 Q-sort items were culled from th e existing literature on grief and bereavement, actual statements made by i ndividuals experienci ng grief, and observations (Lindemann, 1944; Marris, 1958; Parkes, 1965). These statements were tallied and the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -36- 180 most common experiences were w ritten as Q-sort statements w hich sampled the experiences, feelings, symptoms, and behaviors o f individuals d u rin g the process o f bereavement. These Q-sort items were later revised to an objective true-false form at in order to enhance instrument clarity and the speed w ith w hich individuals could respond (Sanders, et al., 1985). A m ultistage internal consistency item analysis was perform ed on th e responses o f 135 volunteer participants each o f whom had experienced th e death o f a close relative w ith in the prior year (Sanders, et al., 1985). The items w ere grouped in to scales according to th e ir content, based upon grief and bereavement theory and empirical studies. The Denial Scale (Sanders, et al., 1985) was developed to indicate the degree o f defensiveness w ith which an individual approaches the test, m odeled after th e Lie scale o f the Minnesota M ultiphasic Inventory (Hathaway & McKinley, 1951). The first step in the item analysis procedure was to correlate each item w ith th e Denial Scale. Any item w hich correlated .30 or greater w ith the Denial Scale was deleted from the item pool. Such items were viewed as being too highly affected by asocial desirability response set. The second step in th e item analysis procedure was calculation o f item remainder coefficients fo r each item and the remainder o f the items on the scale to w hich it had been assigned (Sanders, e t al., 1985). Any item w hich correlated greater than .30 w ith th e other items on the scale was retained. Items w ith correlations less than .30 were deleted. This procedure was performed to refine the scale so th a t each scale w ould have reasonably homogeneous content. The scales are reported to be unidimensional in tha t they only measure one tra it or response class. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A final step in the item analysis procedure was the correlation o f each item w ith all o f the scales on th e inventory (Sanders, et al., 1985). Any item w hich correlated as much w ith th e score on any other scale as it did w ith the rem ainder o f the items on the scale to w hich it had been assigned was deleted. If the item correlated more highly w ith another scale and the content seemed appropriate fo r th a t scale, th e item was transferred to that scale. This was done in order to m inim ize the overlap o f the characteristics being measured by the various scales. The second step o f the item analysis was performed to produce scales which each measured only one characteristic o f grief, w hile the final step was performed to produce scales which all measure d iffe re n t characteristics o f grief. These procedures represent the application, a t the item level, o f the concepts of convergent and discrim inant validity (Campbell & Fiske, 1959). As a result o f the item analysis, 62 items were removed from the original inventory leaving a to ta l o f 128 items (Sanders, et al., 1985). The revised GEI (Form A) had seven new items added making a final total o f 135 items. The GEI (Form B), a shortened version o f the GEI, was developed from the original to o l, elim in atin g w ording tha t pertains to a specific death situation and has 104 items and is the version tha t was utilized in this study. Validity Scales These scales were designed to reflect test taking attitudes. The individual's attitud e tow ard the inventory affects the scores on the Bereavement Scales. Therefore, th e Validity Scales indicate w hether an individual's scores are interpretable. The investigator follow ed the recommendations o f the authors o f the GEI (Sanders, eta l., 1985) regarding these scales. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -3 8- The Denial Scale (Den), which consists o f 11 items, is useful to indicate individuals w ho exhibi i. hesitancy to adm it to socially undesirable weaknesses and feelings. Individual elevations can be anticipated since denial can bean adaptive measure in grief. Denial scores begin to indicate defensiveness at elevations greater than a T score of 55 (Sanders, e ta l., 1985). Profiles w ith scores above a T score o f 70 are n o t interpretable, since the individual is closed to dealing w ith feelings, symptoms, and behaviors which they see as inappropriate. Other individuals may be in such denial th a t they are truly unaware o f their feelings. The elim ination o f profiles w ith Denial T-scores o f greater than 70 is suggested in research use o f the GEI since such cases do not provide valid variance and can distort the data o f groups o f people (Sanders, e ta l., 1985). In the present study o f 83 women, the mean Denial Tscore was 45.66 w ith a range o f 37.00 to 69.00. Since none o f the Denial T-scores were greater than 70, all o f the data was utilized in this study. The Atypical Response Scale (AR), w hich consists o f 20 items, is useful in indicating an unusual response set. There may be several reasons fo r high scores on this scale. The individual may be overwhelmed by his feelings and too upset and confused to closely attend to the test items. The individual may have a reading problem and cannot comprehend the items. Other individuals may be m otivated, for some reason, to portray themselves as experiencing extreme bereavement. According to th e authors o f th e GEI, profiles w ith Atypical Response T-scores greater th a n 70 should be interpreted w ith extreme caution (Sanders, e t al., 1985). The frequency o f Atypical Response (AR) score elevations w ould be expected to be higher in samples of older, less educated or more acutely bereaved individuals. The investigator follow e d the recommendations o f the authors o f the GEI (Sanders, et al., 1985). In th e present study the mean Atypical Response T-score fo r the 83 women was 50.85 w ith a range Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -3 9 - o f 37.00 to 77.00. In the present study sample o f 83 women, 4 wom en had slightly elevated (> 7 0 ) Atypical Response (AR) T-scores. Their (AR) scores w ere 77, 74, 71 and 74. Therefore, th e ir overall GEI (Form B) mean T-scores on all scales w ere compared w ith th e overall mean T-scores fo r the total sample (Table 2). Examination o f th e women's individual and mean T-scores reveals th a t they scored higher than the to ta l group on all but tw o Validity Scales (Denial, Social Desirability) and one Bereavement /Research Scale (Dependency). Since the ir (AR) scores were only slightly elevated and the mean T-scores on the oth er scales w ere higher but similar to the to ta l sample, a decision was made to include these fo u r wom en in the overall data analysis. Similar elevations and sample inclusion are noted by oth e r researchers u tilizin g the GEI (Redlener, 1985). The Demographic Data sheets were also examined for these 4 women; the responses were sim ilar to the responses for th e overall sample. Their Demographic Data Sheets were examined fo r patterns or themes. The ages ranged from 28 to 50 years of age. The tim e since abortio n (TSA) ranged from 4 years to 20 years ago. Woman 1, whose abortion was 4 years ago, scored the same or higher than the other 3 women on 9 o f the Bereavement/Research Scales (Despair, Social Isolation, Loss of Control, Somatization, Sleep Disturbance, Loss of Appetite, Loss of Vigor, Physical Symptoms, and Optimism/Despair). No oth e r patterns were noted. A ll 4 women had the ir abortions a t 2-4 months gestation. All the wom en had living children at the tim e the study was conducted. Two o f th e wom en had living children a t the tim e o f abortion and tw o did not have any living children at the tim e of abortion. Two women responded "very pressured" w ith regard to having the abortion. One woman reported "never pressured" and one wom an reported "m oderately pressured" w ith regard to the abortion experience. None o f the women had experienced perinatal losses or other significant losses in the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -40- Table 2 Comparison of Mean T-scores o f 4 Individual W omen w ith Elevated AR > 70 and the Total Sample S ubjects Scales / " r | /r— _ _q \ VJu l V r O r fn b y M ean _ ■ (N = 4) 1 2 3 4 M ean (N = 83) Range (N = 83) D en D e n ia l 51.00 3 7 .0 0 4 6 .0 0 3 7.0 0 4 2 .7 5 4 5 .6 6 3 7 .0 0 — 6 9 .0 0 AR A ty p ic a l Response 77.00 7 4 .0 0 7 1 .0 0 7 4.0 0 7 4 .0 0 5 0.8 5 3 7 .0 0 SD S ocial D esirability 3 6.0 0 4 4 .0 0 4 4 .0 0 2 9.0 0 3 8 .2 5 51.41 2 9 .0 0 — 6 0 .0 0 Des D e sp air 74.0 0 6 8 .0 0 7 1 .0 0 74.0 0 7 1 .7 5 4 9 .3 2 3 4 .0 0 — 7 4 .0 0 AH A n g e r/H o s tility 65.0 0 7 0 .0 0 6 0 .0 0 7 0 .0 0 6 6 .2 5 5 1 .9 3 3 3 .0 0 — 7 0 .0 0 SI Social Isolation 7 4.0 0 7 4.0 0 7 4 .0 0 6 7.0 0 7 2 .2 5 5 1.9 6 3 3 .0 0 — 7 4 .0 0 LC Loss o f C ontrol 6 7.0 0 6 0 .0 0 6 7 .0 0 67 00 6 5 .2 5 5 4 .7 4 3 0 .0 0 — 7 3 .0 0 Som S o m a tiza tio n 7 7.0 0 6 5 .0 0 6 8 .0 0 58.00 6 7 .0 0 4 9 .3 8 3 3 .0 0 — 7 7 .0 0 DA D e a th A n x iety 54.00 6 3 .0 0 7 1 .0 0 7 1.0 0 6 4 .7 5 5 0.3 5 2 9 .0 0 — 7 1 .0 0 SSD S le e p Disturbance 73.00 5 9.0 0 6 4 .0 0 54.00 6 2 .5 0 4 5 .8 9 3 6 .0 0 — 8 2 .0 0 SAP Loss o f A p p e tite 6 0.0 0 3 7 .0 0 4 9 .0 0 4 9.0 0 4 8 .7 5 4 3 .2 5 3 7 .0 0 — 7 2 .0 0 SVI Loss o f Vigor 74.00 6 8 .0 0 6 8 .0 0 6 8 .0 0 6 9 .5 0 5 2 .0 0 3 6 .0 0 — 7 4 .0 0 SP Physical Sym ptom s 8 0.0 0 6 4 .0 0 7 5 .0 0 5 3.0 0 6 8 .0 0 50.13 3 6 .0 0 — 8 0 .0 0 OD O p tim ism /D e sp a ir 74.00 6 6 .0 0 6 6 .0 0 7 4.0 0 7 0 .0 0 4 8 .2 0 4 1 .0 0 — 7 4 .0 0 Dep D e pendency 35.00 4 4 .0 0 5 2 .0 0 3 5.0 0 4 1 .5 0 51.03 2 7 .0 0 — 6 8 .0 0 — 7 7 .0 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -4 1 - last 5 years. Two o f the women had participated in counseling w h ile tw o had not participated in any formal counseling fo llo w in g the abortion. The GEI (Form B) AR V a lid ity Scale, however, is useful in identifying respondents w ho m ight have an unusual response set. According to Sanders e ta l. (1985) individuals w ith elevated AR scales may be attem pting to present themselves in a specific way; for example in "extrem e grief." Sanders, et al., (1985) also notes th a t AR elevations m ig ht also be found in older more acutely bereaved subjects. It appears th a t these wom en are more acutely bereaved when considering their mean T-scores on the GEI (Form B) scales w ith th e overall sample. The Social Desirability Scale (SD), which consists o f six items, reflects th e individual's tendency to respond in a socially more desirable or acceptable manner. According to the authors of the GEI, the strength o f the socially desirable response set may be related to the need for self protection, avoidance o f criticism, social conform ity, and general approval (Sanders, e ta l., 1985). This scale provides a reliable index by which to differentiate participants w h o m ight respond in a socially desirable manner. The mean Social Desirability T-scores were all in the interpretable range fo r the sample o f 83 women. Bereavement Scales The Despair Scale (Des), w hich consists o f 17 items, measures the mood state o f the individual. It is the longest and most reliable of the bereavement scales and is reported to measure the most pervasive psychological expression o f grief. The Anger/Hostility Scale (AH), which consists of seven items, indicates a respondent's level o f irritation, anger, and feelings o f injustice. High scores on this Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -4 2 - scale have appeared more frequently follow ing th e death o f a child. o r the death o f a spouse at a young age (Sanders, et al., 1985). The Social Isolation Scale (SI) samples behaviors characterized by w ithdraw al from social contacts and responsibilities and consists o f seven items. The Loss o f Control Scale (LC) indicates an individual's inability to control his overt em otional experiences and consists o f seven items. The Somatization Scale (Som) measures the extent o f somatic complaints which occur under th e g rie f experience and consists o f 19 items. The Death Anxiety Scale (DA) measures the intensity o f one's personal death awareness and consists o f 11 items. Research Scales The Sleep Disturbance Scale (SSD), which consists o f 10 items, has proved to be effective in tapping bereavement reactions. The A pp etite Scale (SAP) indicates the respondent's desire for food and consists o f three items. The Vigor Scale (SVI) taps the perceptions o f physical strength o f the respondent and consists of six items. The Physical Symptoms Scale (SP) consists o f nine items and deals w ith bodily symptoms and concerns. Concerns and preoccupation w ith physical symptoms often dom inate and restrict the range o f activities and interpersonal relations. The Optimism vs. Despair Scale (OD) consists o f fo u r items and measures the degree o f despair and loss o f meaning fe lt by the respondent. The Dependency Scale (Dep) consists o f six items and measures th e degree o f need the respondent may indicate to lean or depend on others. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -43- Scoring A scoring key is provided fo r the G rief Experience Inventory (Form B) (Appendix G). Each scale has a number o f true and false items. The key specifies w hich o f the items are given a numerical value of 1 or 0. In this manner a numerical composite score is calculated fo r each scale. These raw scores are then converted to Tscores using a conversion table (Appendix H). The GEI scales are expressed as T-scores. A T-score is a standard score w ith a mean o f 50 and a standard deviation o f 10. The larger the T-score the greater the intensity o f the behavior measured by the scale. The actual scoring is n o t d ifficu lt but is time-consuming. This is not a disadvantage in lig h t o f the wealth o f data gained by 15 individual scores on each o f the 15 subscales. In addition, the scores may be plotted on a graph providing a pictorial representation of each subject's or group's profile. The GEI (Form B) does n o t yield a composite score but individual scores fo r each scale which represents a distinct aspect o f grief. The Grief Experience Inventory was designed to objectively and quantitatively describe the grief phenomenon. Since g rie f is tho ugh t to be multidimensional, the GEI examines the various components o f grief rather than yielding a single composite unidimensional score (Sanders e t al. 1985) The GEI has been utilized in numerous studies (Rando, 1980; Katz, 1983; Golden, 1984; Redlener, 1985; Ferrell, 1984; Ferraro, 1984; Bradshaw, 1985; Steele, 1985; Barrentine, 1986; and Bolin, 1986). A ll but three of the previous investigators (Ferraro, 1984; Rando, 1980; Barrentine, 1986) utilized comparison groups and examined differences in intensities o f grief by the various scores obtained on the GEI scales. Two investigators, (Ferraro, 1984 and Rando, 1980) utilizing the GEI made high and low grief groups and drew comparisons between the tw o groups. The raw scores upon which these tw o investigators made th e ir decision was not explicit. One study (Barrentine, 1986) utilized the GEI in a pre- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -44- and post-test manner thereby creating tw o groups. Sixty adults w h o had lost a close fam ily member in the last 3 years were studied using the GEI. Mean T-scores on the scales of th e GEI ranged from 50.60 to 54.93 and the intensity of g rie f was described as "m oderate." Since th e present study is the first study to utilize the GEI w ith o u t a comparison group (main study hypothesis) consultation was sought from C.M. Sanders (personal communication, August, 1989 and March, 1991). The author of the GEI indicated th a t T-scores > 50 indicate grief; T-scores < 50 may occur in the general population in response to conditions of everyday living. Reliability The reliability o f the Grief Experience Inventory scales has been studied in several samples (Sanders, et al., 1985). The internal consistency or homogeneity o f the scales is indicated by the values of coefficient alphas which ranged from .52 to .84 fo r the clinical and validity scales, and varied largely in relation to the number o f items in the scale. Test-retest reliability wascom puted for tw o samples. For a sample of 22 college students who had experienced g rief sometime w ith in the rather long period o f five years, test-retest coefficients following a nine week interval ranged from .52 to .87. For th e second sample o f 79, recently bereaved individuals w h o were retested 18 months later, test-retest reliabilities ranged from .18 to .69. Of the clinical and validity scales, only the Denial scale had a test-retest reliability o f less than .47. This suggests th a t test-taking attitudes are rather transitory w h ile the actual clinical scales are stable. The authors (Sanders e t al., 1985) indicate th a t these correlations are also o f comparable size to those reported for the Minnesota Multiphasic Personality Inventory over longer test-retest intervals (Dahlstrom, Welsh, & Dahlstrom, 1975). These data indicate tha t the GEI scores are more stable w hen the death was more Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -45- rem ote in tim e and the inter-test interval was brief. The low er test-retest coefficients are most likely the result o f the many real changes in the grief experience during the 18 m onth period between test administrations. A Cronbach alpha was performed on the present sample. A coefficient alpha is a measure o f internal consistency which is used in order to estimate an instrum ent's re lia b ility (Polit and Hungler, 1983). An alpha o f .8731 was obtained on th e 6 Bereavement Scales; Despair (Des), Anger/Hostility (AH), Social Isolation (SI), Loss of Control (LC), Somatization (Som), and Death Anxiety (DA). An alpha of .7926 was obtained on the 6 Research Scales; Sleep Disturbance (SSD), Loss o f A ppetite (SAP), Loss of V igor (SVI), Physical Symptoms (SP), Optimum/Despair (OD), and Dependency (Dep). V alidity The validity o f the GEI has been explored in several ways (Sanders, e ta l., 1985). The GEI has been correlated w ith other scales or inventories measuring sim ilar constructs. It has been tested in the comparison o f bereaved and non-bereaved groups o f individuals. In addition, the inventory has been tested in the comparison o f d iffe re n t types of bereavement, i.e., the death o f a child, spouse, o r parent. Construction of the items w ith in the GEI have been based upon the verbal expressions o f actually bereaved individuals and by exploration o f the relevant lite rature provide face validity. Concurrent validity was established by the analyses o f convergent and discriminant relationships w ith the MMPI. The correlations (Appendix I) suggest th a t b o th scales measure similar but not identical behaviors. Furthermore, it is reported (Sanders, 1979-1980, Sanders, et al., 1985) tha t the GEI taps aspects o f bereavement n o t assessed by the MMPI. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -4 6- The validity o f the GEI was further documented by the inventory's a b ility to distinguish bereaved individuals from non-bereaved individuals as w ell as to distinguish among individuals experiencing d iffe re n t types o f losses. In order to determ ine w hether symptoms outlined by the GEI were indeed those representative o f grief and w o uld be specific to bereaved individuals w hile n o t to non-bereaved individuals, a control group was utilized. A m odified version o f the GEI, (Form B) which elim inates 3 scales and 31 items referring to a specific death, was administered to a group o f individuals (n = 127) who had not experienced a death during the previous five years. A t-te s t between the bereaved and the non-bereaved groups yielded significant differences at the .001 level on all the scales. These data add t o the overall va lid ity o f the GEI and its usefulness as an instrument to assess bereavement reactions (Sanders, e t at., 1985). The established T-scores can be used as benchmarks upon w hich to measure grief across a variety o f grief/loss experiences. Further validity was established by comparing the scores o f those bereaved w ho had suffered the loss o f a child, spouse, or parent. When differences were plotted, using the Early Bereavement group, it was found tha t th e loss o f a child produced significantly higher scores on the GEI than did the oth e r types of losses (Sanders, e ta l., 1985) (Appendix J). The prelim inary studies (Sanders, e ta l., 1985) support th e validity o f the G rief Experience Inventory and the Grief Experience Inventory (Form B). The inventories have been shown to be sensitive to differences in th e type of bereavement experienced. In addition, it samples behaviors not sampled by general measures o f psychopathology such as the MMPI. The decision to utilize the GEI (Form B) as th e measure o f the dependent variable was made after conducting 2 p ilo t studies and review o f th e ir outcome. An Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -47- in itia l p ilo t study was conducted utilizing the M ourning Score (TMS) (Kennell, Slyter and Klaus, 1970), a too l which had been utilized in several studies o f perinatal grief. Ten wom en were sampled w ith in one m onth follow ing abortion using the TMS. None o f the women exhibited perinatal g rie f as measured by this tool. Review o f the anecdotal remarks, however, revealed a theme of denial of feelings although none of th e wom en indicated any discomfort w ith the tool. Asecond pilot study was conducted using the GEI (Form B) (Sanders, eta l., 1985) and the Response to Loss Inventory (RLI) (Deutsch, 1982). Data was collected from nine women w ho had experienced recent (w ith in the last 1-2 years) losses such as loss o f spouse, close friend, grandparent, child, and miscarriage. The women were given both tools and were inteviewed after completion. A ll nine wom en indicated th a t it was easier to com plete th e GEI (Form B) and that the w ording was more neutral since it did not deal w ith a specific death or loss. The Denial Scale o f the GEI (Form B) was deemed preferable by th e investigator since use o f this tool allow ed identification o f one individual w h o was in denial. Examination o f this individual's scores on the RLI revealed a lo w level o f g rie f b u t w ith ou t th e benefit o f detecting denial. Therefore, a decision was made to u tilize the GEI (Form B) since it is a standardized grief instrument w ith neutral w o rdin g and the ab ility to detect denial. Permission was obtained to use the GEI (Form B) in the present study (Appendix K). Analysis of the Data The Grief Experience Inventory (Form B) was scored as previously described. Individual raw scores were obtained and T-scores calculated fo r each individual participant on the 15 scales representing the various dimensions o f the grief experience. The study hypothesis was tested by examination o f the T-scores (means) on the scales o f the GEI (Form B ). Correlation coefficients were performed on each of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -4 8 - 15 scales to examine the first research question. A correlation coefficient is an index tha t summarizes th e degree o f relationship between tw o variables (Polit and Hungler, 1983). The second research question was tested by a tw o ta il t-test. A t-te s tis a parametric statistical test used fo r analyzing the difference between tw o means (Polit and Hungler, 1983). Descriptive statistics, such as frequencies, means, and standard deviations were performed to describe and summarize the demographic variables. SPSS (1986), a program for advanced statistics was utilized to analyze all o f the data. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -49- CHAPTER IV REPORT OF THE FINDINGS ANALYSIS OF THE DATA Problem Statement Do women w h o have undergone induced elective abortion experience perinatal grief? Mean raw scores and mean T-scores were calculated for th e to ta l sample (n = 83) on the various scales o f the GEI (Form B). Of the total sample (n = 83), 79 of the women had scores below the T-score of 70 on both the Denial (DenT) and Atypical Response (ART). Therefore, th e ir scores were interpreted as suggested by th e author (Sanders, et al., 1985). The authors o f the GEI (Form B) suggest th a t profiles greater than 70 should be interpreted cautiously. Atypical Response T-scores (ART) higher than 70 may indicate th a t the individual responding is m otivated to present herself as experiencing exaggerated grief. The frequency o f the elevation is expected to be somewhat higher in study populations w ho are older, less educated or more acutely bereaved (Sanders, e t al. 1985). Four wom en had slightly elevated Atypical Response (ART) scores; 77,74, 71,74. Since the scores were only slightly elevated, they were included in th e overall data analysis as previously discussed. The mean raw scores fo r the sample o f 83 women on th e GEI (Form B) scales are presented in Table 3. It must be noted th a t although in the present study th e en tire sample was included, the actual num ber (n) varied fo r each scale. This is a ttrib u te d to the fact that some respondents le ft an item or items blank. O ther respondents answered both “ tru e " and "false" to some items and therefore th e ir data could n o t be utilized in the overall data analysis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -50- Table 3 Means and Standard Deviations GEI (Form B) Means Standard Deviations n Scales (80) Den 1.9 1.65 (76) AR 5.46 3.45 (78) SD 3.92 1.01 (78) Des 5.44 5.06 (79) AH 3.55 2.15 (80) SI 2.81 2.01 (78) LC 4.09 1.62 (81) Som 5.16 3.45 (77) DA 5.03 2.67 (78) SSD 2.17 2.45 (81) SAP 0.53 0.82 (82) SVI 2.52 2.08 (82) SP 2.57 2.04 (82) OD 0.87 1.31 (82) Dep 2.90 1.34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -5 1 - The GEI (Form B) has 3 V alid ity Scales, 6 Bereavement Scales and 6 Research Scales. The scales and th e ir acronyms are as follows: Scales V alidity Scales Bereavement Scales Research Scales Scale Acronyms T-scores Denial (Den) (DenT) Atypical Response (AR) (ART) Social Desirability (SD) (SDT) Despair (Des) (DesT) Anger/Hostility (AH) (AHT) Social Isolation (SI) (SIT) Loss of Control (LC) (LCT) Somatization (Som) (SomT) Death Anxiety (DA) (DAT) Sleep Disturbance (SSD) (SSDT) Loss of Appetite (SAP) (SAPT) Loss of Vigor (SVI) (SVIT) Physical Symptoms (SP) (SPT) Optimism/Despair (OD) (ODT) Dependency (Dep) (DepT) Raw scores, means and standard deviations fo r GEI (Form B) previously reported by Sanders e ta l. (1985) are presented in Table 4. These scores are sim ilar to those found in the present study. T-scores for all the scales o f the GEI (Form B) were calculated fo r th e 83 wom en. These T-scores are presented in Table 5. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -52- Table 4 Means and Standard Deviations fo r Form B (Non-Death Version) rScales i Means Standard „ . .. Deviations Den 2.83 2.14 AR 5.15 3.22 SD 3.76 1.29 Des 5.62 3.55 AH 3.19 1.87 SI 2.51 1.45 LC 3.32 1.62 Som 5.32 3.19 DA 4.96 2.39 SSD 3.07 2.16 SAP 1.09 0.87 SVI 2.29 1.54 SP 2.54 1.80 OD 1.10 1.21 Dep 2.76 1.21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -53- Table 5 T-Scores and Standard Deviations GEI (Form B) GEI (Form B) T-Scores Range Standard Deviations DenT 45.66 37.00 — 69.00 7.69 ART 50.85* 37.00 — 77.00 10.66 SDT 51.41* 29.00 — 60.00 8.05 DesT 49.32 34.00 — 74.00 13.97 AHT 51.93* 33.00 — 70.00 11.52 SIT 51.96* 33.00 — 74.00 13.58 LCT 54.74* 30.00 — 73.00 10.03 SomT 49.38 33.00 — 77.00 10.63 DAT 50.35* 29.00 — 71.00 11.25 SSDT 45.89 36.00 — 82.00 11.32 SAPT 43.25 37.00 — 72.00 9.61 SVIT 52.0* 36.00 — 74.00 13.30 SPT 50.13* 36.00 — 80.00 11.21 ODT 48.20 41.00 — 74.00 10.90 DepT 51.03* 27.00 — 68.00 11.05 * T-scores > 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -54- T-scores on the bereavement/research scales o f the GEI (Form B) o f greater than 50 are indicative o f g rief as per C. M. Sanders (personal communication, August, 1989 and March, 1991). The GEI (Form B) T-scores fo r the present sample o f 83 wom en were shared w ith one o f the authors o f the GEI. According to C. M. Sanders (personal communication, March, 1991), the elevation in T-scores (> 50) observed in the present study may be interpreted as minimal elevations o f grief. For this sample, the wom en were shown to have minimal levels o f g rie f intensity on slightly over half (7) o f the 12 scales; specifically in terms o f Anger (AHT), Social Isolation (SIT), Loss o f Control (LCT), Death Anxiety (DAT), Vigor (SVIT), Physical Symptoms (SPT) and Dependency (Dept). The GEI (Form B) does not yield a composite score fo r either individuals o r groups. The GEI (Form B) utilizes standardized scales which represent the separate components o f grief. A GEI profile was constructed fo r the 83 women using th e mean raw scores and mean T-scores to provide a pictorial representation o f the study sample (Figure). Examination of the data in th e figu re indicates th a t this sample o f 83 wom en w h o had undergone induced elective abortion an average o f 11 years ago experienced minimal symptoms of perinatal grief as indicated by elevated T-scores ( > 50) on 7 of the 12 scales. Therefore the hypothesis was supported in that these wom en experienced perinatal grief specifically in terms of persistence o f several of the multidimensional aspects o f g rie f as measured by the GEI (Form B). These wom en had elevated anger or hostility (AH), w ithdraw al from social contacts and responsibilities (SI), inability to control overt em otional experiences (LC), heightened personal death awareness (DA), loss or lack o f physical strength (SVI), preoccupation w ith bodily symptoms (SP) and the need to depend on others (Dep). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -55- FIGURE: GEI (FORM B) PROFILE 80 5 0 .8 5 51.41 4 5.6 6 VALIDITY SCALES 20 Den R a w Scores T Scores AR SD 1.9 5 .4 6 3 .9 2 4 5 .6 6 5 0 .8 5 51.41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -56- FIG. CONTIN U ED 85 80 75 70 65 60 55 5 1.9 3 5 1 .9 6 4 9 .3 8 4 9 .3 2 50 45 40 35 30 CLINICAL SCALES 25 20 Des R a w Scores T S cores AH SOM SI LC 5 .4 4 3.55 2.81 4 .0 9 5.1 6 5.03 4 9 .3 2 5 1.9 3 5 1 .9 6 5 4 .7 4 4 9 .3 8 5 0.3 5 DA Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -57- FIG. CONTINUED 80 5 2.0 5 1 .0 3 4 8 .2 0 5 0.1 3 4 5 .8 9 43.2 5 40 RESEARCH SCALES R a w Scores SSD SAP 0 53 T Scores 45 8 9 43 2 5 SVI 2.52 5 2 .0 SP OD DEP 2 .5 7 0 .8 7 2 .9 0 5 0 .1 3 4 8 .2 0 5 1 .0 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -58- Research Questions 1. Is there a relationship between the intensity o f perinatal grief in wom en w ho have undergone induced elective abortion and the tim e since abortion? Time since ab o rtio n (TSA) was calculated for th e entire sample o f 83 women. The mean TSA was 11.0 years w ith a range o f less than 1 year to 26 years. Correlations w ith TSA were calculated fo r all o f the scales on GEI (Form B) (Table 6). Review o f the correlations indicates th a t all o f the bereavement and research scales were negatively correlated w ith tim e since abortion (TSA), indicating th a t intensity o f grief lessens overtim e. Two scales. Loss of Control (LC) and Loss o f Vigor (SVI) were significantly (negatively) correlated w ith tim e since abortion (TSA); respectively p = .005 and p = .010. The answer to the research question is th a t intensity o f grief diminishes w ith passage of time. All o f the relationships on the dimensions of grief as measured by the 12 scales are negatively correlated w ith TSA. Only tw o dimensions, Loss o f Control (LC) and Loss o f V igor (SVI) were shown to be significantly (negatively) correlated w ith tim e since abortion (TSA). Therefore, the relationship between tim e since abortion (TSA) and Loss o f Control (LC) and Loss o f Vigor (SVI) are more pronounced. Time since abortion (TSA) was further examined by construction o f 5 subsets of the entire sample. Mean T-scores fo r all the scales of the GEI (Form B) were examined for those wom en who had experienced their abortion less than 1 year ago, those w h o had their abortion 1-5 years ago, those who had their abortion 5-10 years ago, those w ho had th e ir abortion 10-15 years ago, and those who had th e ir abortion m ore than 15 years ago (Table 7). It is o f note tha t 2 Research/Bereavement Scales (Despair and Loss o f Vigor) mean T-scores decreased w ith the passage o f tim e as expected. The other 10 Research/Bereavement Scales failed to evidence such a decrease w ith the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -59- Table 6 T-Scores Correlation Coefficients o f GEI (Form B) Scales and Time Since A bortion (TSA) Correlation Coefficients N (N um be rof Respondents) . PValue DenT -.0714 80 .529 ART -.0955 76 .412 SDT .0861 78 .453 DesT -.2044 78 .073 AHT -.0854 79 .454 SIT -.1482 80 .190 LCT -.3126 78 .005** SomT -.1781 81 .112 DAT -.1838 77 .110 SSDT -.1280 78 .264 SAPT -.1103 81 .327 SVIT -.2838 82 .010* SPT -.1779 82 .110 ODT -.0833 82 .457 DepT -.2053 82 .064 GEI (Form B) *P < .05 **P [...]... undergone i nduced elective abortion experience perinatal grief? Research Questions Is there a relationship between the intensity o f perinatal grief in women w h o have undergone induced elective abortion and the tim e since th e abortion? Is there a difference in the intensity o f perinatal grief experienced by wom en w ho have undergone induced elective abortion w ho have living children and those w ho... attachm ent process (Tulman, 1981) Buckles (1982) and Joy (1985) found th a t many women studied post -abortion were suffering from depression as a result o f unresolved grief Viewed in this context induced elective abortion can be thought o f as having a d e finite potential for perinatal grief Recently, Rando (1986) described loss in relation to induced elective ab ortio n in terms o f being deprived... experience perinatal grief Significance o f the Study Interest in perinatal grief emerged in the 1970s w ith the w ork o f Kennell, eta l., (1970) w ho were among the first to describe the phenomenon o f perinatal grief The d e fin itio n o f perinatal grief has been expanded (Peppers & Knapp, 1980) to encompass various types o f losses including death of a newborn, stillbirth, miscarriage, and spontaneous abortion. .. Johnson-Sodenberg, 1981; Estok and Lehman, 1983; Lakeetal., 1983; Miles and Crandall, 1983; Kellner, Donnelly & Gould, 1984; Wall-Haas, 1985) documents tha t perinatal grief is experienced fo llo w in g perinatal loss Perinatal g rie f is present regardless of actual iength o f gestation, i.e.: spontaneous abortion, miscarriage, stillbirth, or neonatal death This phenomenon o f perinatal grief seems consistent... documents the phenomenon o f perinatal grief Concentrated efforts to define perinatal grief began in the 1970s, spurred by the research of Kennell, Slyter and Klaus (1970) These studies indicated that death of a newborn was indeed a significant loss, accompanied by grief M ore recently Lake, Knuppel, Murphy, and Johnson (1983) depict perinatal grief as sim ilar to th e typical grief reaction tha t ensues... Elevated A R > 70 and the Total Sample 40 Means and Standard Deviations GEI (Form B) 50 Means and Standard Deviations fo r Form B (Non-Death Version) 52 T-Scores and Standard Deviations GEI (Form B) 53 T-score Correlation Coefficients o f GEI (Form B) Scales and Time Since A bortion (TSA) 59 Mean T-Scores o f GEI (Form B) by Time Since A bortion (TSA) 60 Means, Standard Deviations and t-tests GEI (Form... death o f an infant A bortion Sequelae The lite rature on abortion and abortion sequelae is at best contradictory In general, induced elective abortion has not been considered to be a cause fo r grief Prior to the legalization of abortion in 1973, the m ajority of studies focused on the medical effects of abortion In 1973, Osofsky, Osofsky, and Rajan reviewed the literature w ith regard to studies... GEI (Form B) Presence o f Living Children at Abortion 63 T-Score Correlation Coefficients o f GEI (Form B) Scales and Weeks Pregnant at A bortion (WPA) 65 Means, Standard Deviations and t-tests Comparison between Miscarriage and No Miscarriage 66 Means, Standard Deviations and t-tests GEI (Form B) Pressure Regarding A bortion 68 Means, Standard Deviations and t-tests GEI (Form B) Counseling Following... abortion To date, there has been little investigation to determ ine w hether perinatal grief reaction follow s an elective abortion Perhaps the lack of investigation mirrors th e lack o f recognition and validation th a t society once afforded other types o f perinatal losses Raphael (1980) cites tha t fo llo w in g induced elective abortion, society usually gives the woman the covert message th a t she... (Form B) (Sanders, Mauger and Strong, 1985) Perinatal The tim e before and after birth, defined as beginning w ith conception and extending through the 28th day o f newborn life (Reeder,Mastroianni, & M artin, 1983) Perinatal G rief The g rie f described by Lindemann (1944) and experienced in a specific context, follow ing the death or loss th a t occurs during the perinatal period In this study perinatal ... nduced elective abortion experience perinatal grief? Research Questions Is there a relationship between the intensity o f perinatal grief in women w h o have undergone induced elective abortion and. .. rature on abortion and abortion sequelae is at best contradictory In general, induced elective abortion has not been considered to be a cause fo r grief Prior to the legalization of abortion in... tim e since th e abortion? Is there a difference in the intensity o f perinatal grief experienced by wom en w ho have undergone induced elective abortion w ho have living children and those w ho