evaluation of the perinatal grief intensity scale in the subsequent pregnancy after perinatal loss

JOGNN R E S E A R C H

Evaluation of the Perinatal GriefIntensity Scale in the SubsequentPregnancy After Perinatal LossMarianne H. Hutti, Deborah S. Armstrong, and John Myers

CorrespondenceMarianne H. Hutti, PhD,WHNP-BC,555 S. Floyd St.,Louisville, KY [email protected]

Keywordsperinatal losssubsequent pregnancyperinatal loss instrumentmiscarriagestillbirthneonatal death

ABSTRACT

Objectives: To evaluate the reliability and validity of the Perinatal Grief Intensity Scale (PGIS) for identifying a woman’s

grief intensity in the immediate subsequent pregnancy after a miscarriage, stillbirth, or neonatal death.

Design/Setting/Participants: A web-based approach was used to collect data from 227 pregnant women after each

woman had experienced a perinatal loss in her previous pregnancy.

Methods: Participants completed a demographic information form and the 14-item PGIS.

Results: Cronbach’s alphas for the PGIS total scale and subscales were high: 0.75 (PGIS total), 0.80 (Reality), 0.82

(Confront Others), and 0.80 (Congruence), which indicated good internal consistency reliability. Validity was supported

by factor analysis of the PGIS, which accounted for 66.94% of the total variance. Mothers in the neonatal death group

experienced more intense grief, as measured by the PGIS, when compared with mothers in the miscarriage or stillbirth

groups.

Conclusions: Data from this study provided initial support for the reliability and validity of the PGIS in women in

their immediate subsequent pregnancies after perinatal loss as well as the concepts of the grief intensity theoretical

framework.

JOGNN, 42, 697-706; 2013. DOI: 10.1111/1552-6909.12249

Accepted July 2013

Marianne H. Hutti, PhD,WHNP-BC, is a professorand coordinator of theWomen’s Health-FamilyNurse Practitioner DoubleMajor Program, Universityof Louisville School ofNursing, Louisville, KY.

Deborah S. Armstrong,PhD, RN, is an associateprofessor in the Universityof Louisville School ofNursing, Louisville, KY.

John Myers, PhD, MSPH,is an associate professor inthe Child Health ServicesResearch Unit, Departmentof Pediatrics, University ofLouisville School ofMedicine, Louisville, KY.

The purpose of this study was to evaluate thereliability and validity of the Perinatal Grief In-

tensity Scale (PGIS) for identifying the intensity ofa woman’s grief in the subsequent pregnancy aftera miscarriage, stillbirth, or neonatal death. Miscar-riage, stillbirth, and neonatal death represent earlypregnancy, late pregnancy, and postpregnancylosses, respectively. They occur in approximately25% of all pregnancies, and when consideredtogether they are called perinatal loss (Hutti,Armstrong, & Myers, 2011). The mental healthconsequences of perinatal loss are significant,common, and well documented (Lang, Fleiszer,Duhamel, Sword, Gilbert, & Corsini-Munt, 2011;Woods-Giscombe, Lobel, & Crandell, 2010). Be-yond expected normal grieving, mothers may ex-perience high levels of anxiety (Armstrong, Hutti,& Myers, 2009; Gaudet, Sejourne, Camborieux,Rogers, & Chabrol, 2010; Woods-Giscombeet al.), depression (Gaudet et al.; Gausia, Moran,Ali, Ryder, Fisher, & Koblinsky, 2011; Swanson,Hsien-Tzu, Graham, Wojnar, & Petras, 2009; Su-tan et al., 2010), posttraumatic stress (Armstrong

et al., 2009; Jind, 2003), and marital discordand divorce (Gausia et al.; Gold, Sen, & Hay-ward, 2010; Lang et al.). For some grieving par-ents, the mental health consequences of peri-natal loss may last many years (Schaap et al.,1997).

In the only study of conception rates after a pre-vious loss, Cuisinier, Janssen, deGraaw, Bakker,and Hoogduin (1996) found 86% of parents be-came pregnant again within 18 months. However,the mental health consequences of a previousperinatal loss often do not resolve with the birthof a subsequent healthy infant (Blackmore et al.,2011). Many parents will continue to grieve pre-vious losses during the next pregnancies andwill continue to experience high levels of anxi-ety (Armstrong et al., 2009; Couto et al., 2009;Hutti et al., 2011; Woods-Giscombe, et al., 2010),depression (Couto et al.; Armstrong et al., 2009;Hutti et al., 2011) and posttraumatic stress(Armstrong, et al., 2009; O’Leary, 2005). In-creased use of health care resources during the

The authors report no con-flict of interest or relevantfinancial relationships.

http://jognn.awhonn.org C© 2013 AWHONN, the Association of Women’s Health, Obstetric and Neonatal Nurses 697

R E S E A R C H Perinatal Grief Intensity in the Subsequent Pregnancy After Perinatal Loss

Identification of parents who may experience intense grief andneed for follow-up cannot be accurately determined by parental

behavior soon after a perinatal loss.

subsequent pregnancy (Hutti et al.; Robson,Leader, Bennett, & Dear, 2010) and increasedpostpartum depression after delivery of the subse-quent healthy infant (Armstrong, 2007; Blackmoreet al.) are also common.

However, not all parents who have a perinatal losswill experience a grief reaction (Hutti, 1992). Thepresence of grief is mediated by the presence ofprenatal attachment (Cote-Arsenault & Dombeck,2001; Hutti; Uren & Wastell, 2002). According toHutti, attachment occurs initially as the idea of thepregnancy becomes a reality for the parents; oncethat occurs, the fetus becomes real to them. Cote-Arsenault and Dombeck (2001) designated this asa process of assigning personhood to the fetus.Parents become prenatally attached over time asthey perceive specific characteristics and a per-sonality in their unborn infants (Hutti). Mothers andfathers often accomplish this process at differentrates. The mother has early pregnancy symptomsthat usually help make the pregnancy and fetusa reality for her sooner than for the father. Oncean attachment has formed and a loss occurs, agrief reaction will follow (Hutti; Shreffler, Greil, &McQuillan, 2011). If parents have not yet becomeattached to the fetus and a loss occurs, they maynot experience a grief reaction (Hutti). Therefore,the loss may be perceived as something otherthan the death of a child, and the typical hospi-tal interventions used by maternity nurses aftera perinatal loss may be inappropriate (e.g., sug-gesting a funeral or memorial service; naming thedeceased fetus).

Accurately identifying parents who are grievingand need follow-up from those who are not griev-ing may be difficult in the early hours or days afterperinatal loss. Parents respond to the diagnosis ofa perinatal loss in ways that vary widely. Some areimmediately responsive and may grieve openly.Others may experience little or no grief but maybe very frightened about the bleeding and painthey or their partner is experiencing. Still othersmay grieve and experience a phase of shock andnumbness in the first hours to days after a perina-tal loss, during which they may not demonstrate orexpress their grief to health care providers (Hutti,1992; Rowlands & Lee, 2010). They may not moveout of this phase of shock and numbness until after

hospital discharge, which makes it difficult to usetheir behavior as a guide for need for follow-upby health care providers. An instrument that couldidentify parents at greatest risk of intense grief af-ter a perinatal loss and therefore, with greatestneed for follow-up, would be very helpful to ob-stetric providers.

Theoretical Framework for thePerinatal Grief Intensity ScaleThe theoretical framework for the PGIS was basedon Dougherty’s model of cognitive representa-tion (1984) and developed for use in a perinatalloss population based on a qualitative study con-ducted by Hutti (1992). The theoretical frameworkassists clinicians to better anticipate which par-ents may be most likely to experience intense griefand need for follow-up after perinatal loss.

The theoretical framework explains the relation-ship between the parents’ perceptions of the eventof perinatal loss and their behavior subsequent tothe loss. The framework is based on the assump-tion that the most critical influence of parents’ be-havior and actions after a perinatal loss is howthe mother or father perceived the experience ofthe loss, rather than the actual facts surround-ing the event. In the development of the theo-retical framework, Hutti (1992) found three fac-tors influenced grief intensity after pregnancy loss:(a) perceived reality of the pregnancy and infant,(b)congruence between the experience of the ac-tual loss and the parents’ standard of the desirable(or perceived ideal against which the experienceis being evaluated), and (c) the ability of parentsto make decisions or act in ways to increase thiscongruence.

Reality of the Pregnancy and InfantAccording to Hutti’s framework (1992), prenatalattachment begins when the pregnancy and fetusbecome real to the parents. If parents experiencea miscarriage, stillbirth, or neonatal death after at-tachment to the fetus or infant has occurred, anintense, long-lasting grief response will result. Ifparents have become attached to the pregnancy,but the fetus is still perceived as generic, withoutspecific personality traits and identity, a relativelyshort, less intense grief reaction will follow. If aloss occurs before the parents have become at-tached to the pregnancy, fetus, or infant, little orno grief response will ensue, but parents may bevery distressed regarding the physical symptoms

698 JOGNN, 42, 697-706; 2013. DOI: 10.1111/1552-6909.12249 http://jognn.awhonn.org

Hutti, M. H., Armstrong, D. S., and Myers, J. R E S E A R C H

of cramping and bleeding that are often part of theexperience of perinatal loss (Hutti).

Congruence between the Experience ofthe Actual Loss and the Parents’Standard of the DesirableThe standard of the desirable is the perceivedideal against which parents evaluate their expe-rience of perinatal loss. It is the individual mea-suring stick each parent uses to determine if theexperience of the loss is unfolding in the least trau-matic way possible. Unless parents have had aprevious perinatal loss, they usually do not havea standard of the desirable already developed forthe experience of the loss. Therefore, the stan-dard of the desirable generally develops as theexperience unfolds, often in response to nega-tive events. An example of the development ofa parent’s standard of the desirable is when anurse avoids contact with a bereaved parent, andthe parent responds by thinking, “I don’t like itwhen the nurse does that; I wish she would talkto me and just say something like ‘I’m sorry foryour loss’ instead” (Hutti, 1992; Hutti et al., 1998;Hutti, 2005).

Even if parents do not experience intense griefwith their perinatal loss, the experience is alwaysdifficult. A loss often is accompanied by bleed-ing and pain, fear of the unknown, guilt, worry,and interaction with family members, friends, andhealth care providers who may not know how toappropriately comfort and support the parents.When the experience of the actual loss is congru-ent with the parents’ standard of the desirable, itindicates that family members, friends, and healthcare providers are finding words and actions thatparents perceive as supportive in the midst of thedifficult experience. In this situation, the parentsoften will feel satisfied with their interactions withothers and generally will feel supported in theirexperience of the perinatal loss. When the expe-rience of the actual loss is widely divergent fromthe standard of the desirable, it is perceived asunfolding in a way that parents find unsupportiveand unacceptable, and parents often respond withanger and feelings of victimization (Hutti, 1992;Hutti et al., 1998; Hutti, 2005).

The ability of parents to make decisionsor act in ways to Increase congruence.If the experience of the actual loss is developingin a way that is widely divergent from the parents’standard of the desirable, parents may choose to

be assertive and confront others about their un-supportive words and/or behaviors. If parents areable to tell others what they want and need, oftentheir experience of the actual loss may be modifiedso that it more closely aligns with their standard ofthe desirable. Another option is for nurses to findout what parents want and need and then serve asadvocates in getting these needs met. If parentsare unable to be assertive, and if nurses are unwill-ing to serve as their advocates, then parents mayfeel even more traumatized by the experience ofthe perinatal loss, thus adding to their grief inten-sity (Hutti, 1992; Hutti et al., 1998; Hutti, 2005).

In summary, grief after perinatal loss may rangefrom no grief to an intense, long-lasting response.Hutti (1992) found the most intense grief responsewas associated with parents who perceived thepregnancy and fetus or infant as real, perceivedthe experience of the actual loss as widely diver-gent from the parents’ perceived ideal, and per-ceived themselves as unable to do anything aboutit (see Figure 1).

Purpose. The purpose of this study was to evaluatethe reliability and validity of the PGIS for identifyinga woman’s grief intensity in the subsequent preg-nancy after a miscarriage, stillbirth, or neonataldeath.

The specific aims of the study were to

1. Evaluate the dimensionality of the PGIS inwomen who are in the immediate subsequentpregnancy after a perinatal loss.

2. Evaluate the internal consistency of the over-all PGIS scale and its subscales.

3. Determine whether grief intensity measuredby the PGIS varies by type of loss in orderto evaluate the construct validity of the mea-sure.

Hypothesis 1: Women who experience miscar-riage will have a significantly lower mean PGISscore than those with neonatal death.

Hypothesis 2: Women who experience stillbirthwill have a significantly lower mean PGIS scorethan those with neonatal death.

MethodsDesignA descriptive, correlational research design wasused in this cross-sectional, web-based study.

JOGNN 2013; Vol. 42, Issue 6 699


Percep on of reality of pregnancy/

baby

Asser ve/Confront others to

congruence

Congruence in loss

experience

High intensity grieving

Percep on of reality of pregnancy/

baby

Congruence in loss

experience

Asser ve/Confront others to

congruence

Low intensity grieving

Figure 1. Perinatal Grief Intensity Theoretical Framework.

SampleTo enroll in the study, participants had to becurrently pregnant and greater than 20 weeksgestation; had to have experienced a previousmiscarriage, stillbirth, or neonatal death in theirimmediate past pregnancy; and had to be olderthan age 18, and speak English. Participants were477 women who met the study inclusion criteria,accessed the web site over an 11-month period,and attempted to answer at least one question.The total sample size was based on the number ofparticipants who answered all of the PGIS ques-tions. A total of 227 women (48%) completed thedata form and composed the sample for this study.This sample size afforded sufficient power (greaterthan 80%) to test the hypotheses. No Bonferronicorrections were performed because every out-come was viewed as important in isolation. Al-though the study was not restricted by gender,only women chose to participate.

MeasuresThe PGIS is a 14-item self-report questionnairethat was tested initially in a sample of 186 womenwho had experienced a miscarriage. Scoring ofthe PGIS is based on a 4-point Likert-type scale(1 = strongly agree, 4 = strongly disagree). Items1, 2, 3, and 6 are reversed before obtaining itemscores. The mean score for each subscale is com-puted by averaging subscale item scores and mayrange from 1–4. The total PGIS score is obtainedby the following equation: Total PGIS score = 3.08+ (.41 x mean Reality subscale score) – (0.2 xmean Confront Others subscale score) – (.15 xmean Congruence subscale score). Bold num-bers are constants in this equation. Higher totalscores reflect more intense grief (Hutti et al., 1998).

After the initial factor analysis (Hutti et al., 1998),three factors were extracted as predicted andwere named: (a) realty of the pregnancy and fetus(Reality), (b) congruence between the experienceof the actual loss and the standard of the de-sirable (Congruence), and (c) ability to confrontothers (Confront Others). The Reality subscaleconsisted of six items, which asked bereaved par-ents about their perceptions of how real the preg-nancy or infant were to them at the time of the loss.Understanding how mothers perceive the reality oftheir pregnancies and infants is critical for healthcare providers to grasp the significance of the lossfor the parent, and whether to treat it as a death.An example of a reality item is “I felt I had lostmy son or my daughter, not just my pregnancy.”The eigenvalue for the Reality subscale was 4.53,Cronbach’s alpha was .89, and the subscaleaccounted for 32.25% of the total variance (Huttiet al.).

The Congruence subscale consisted of four itemsthat asked bereaved parents about their percep-tions of the congruence between the experienceof the actual loss and their perceived ideal orstandard of the desirable. An example of a con-gruence item is “During and after my perinatalloss, I was satisfied with the way my loss expe-rience unfolded, given that I had to go through it.”The eigenvalue for the Congruence subscale was3.11, Cronbach’s alpha was .71, and the subscaleaccounted for 10.4% of the total variance (Huttiet al., 1998).

The Confront Others subscale consisted of fouritems that asked bereaved parents about theirperceived ability to make decisions or act



assertively to increase the congruence betweenthe actual loss they were experiencing versus theirperceived ideal. An example of a Confront Othersitem is “In the first hours and days after my loss,if people said or did things that made me feelbad, I was able to ask them to stop.” The eigen-value for the Confront Others subscale was 1.46,Cronbach’s alpha was .84, and the subscale ac-counted for 22.2% of the total variance.

The three subscales together accounted for 65%of the total variance, and had a Cronbach’s alphaof .82. The factor correlation matrix demonstratedlow correlations between factors, which indicatedthat although a relationship existed among the fac-tors, they measured different facets of grieving(Hutti et al., 1998).

ProcedureThe study was reviewed and approved by the In-stitutional Review Board of a southern metropoli-tan university. Participants were recruited throughthe Internet from messages and links on women’shealth, pregnancy, perinatal loss, and perinatalloss support web sites. Participants were directedto the study web site and were asked to read acover letter on the web site that introduced the in-vestigators and described the study purposes inEnglish at a fourth-grade reading level to increasereadability and suitability. The cover letter was fol-lowed up by a consent letter in which participantsprovided consent before they could complete thestudy instruments. If they did not click I agree toparticipate, then study instruments remained un-available. When they agreed to participate, studyinstruments were presented with a demographicdata form first, followed by the PGIS (Hutti et al.,1998).

The study instruments and data form were com-pleted in a point-and-click fashion. In addition, acounting system kept track of the number of peo-ple who logged onto the web site, visited the studyintroduction, completed the consent page, andcompleted all or part of the survey instrument.The web site also was set up so that once theparticipants completed the survey they could notre-access the web site from the same IP address.

Data AnalysisThree loss groups were identified by participantswho had experienced a previous perinatal loss: (a)before 20 weeks of gestation (miscarriage), (b) af-ter 20 weeks of gestation but before a live birth(stillbirth), and (c) after a live birth but within the

newborn’s first 28 days of life (neonatal death).Chi-squared analysis was used to evaluate differ-ences among the three loss groups for categoricalsocio-demographic variables. Differences amongthe groups in continuous sociodemographic vari-ables were tested using one-way ANOVA. Beforethe use of ANOVA, an exploratory analysis indi-cated that the assumption of normality was metfor all variables examined. When extreme outlierswere found by examining box plots, they were re-moved from the analysis. Exploratory factor analy-sis was conducted to examine the dimensionalityof the PGIS. Cronbach’s alpha was used to eval-uate the internal consistency reliability of the totalPGIS and its subscales.

A one-way ANOVA was performed to evaluatewhether differences in scores of the PGIS existedbased on the type of perinatal loss participantsexperienced in the previous pregnancy (miscar-riage, stillbirth, or neonatal death). Tukey’s posthoc pairwise comparison tests were performedto determine where the difference between thegroups occurred. All data were downloaded in en-crypted files from the web site and then analyzedusing SPSS 19.

ResultsParticipantsThe mean age of the women in this study was31 years (SD = 5.9), the majority were White, andthree fourths lived in the United States. More thanone half of the participants had a college degreeor higher and an income of $50,000 or more. Thevast majority was married and had experiencedeither one or two previous perinatal losses (seeTable 1).

Internal consistency reliability. The Cronbach’s al-phas for the PGIS total scale and subscales werehigh: 0.75 (PGIS total), 0.80 (Reality), 0.82 (Con-front Others), and 0.80 (Congruence), indicatinggood reliability. Cronbach’s alphas stratified byloss group were also high, which indicated goodinternal consistency reliability (see Table 2).

Dimensionality. Exploratory factor analysis (EFA)was conducted to identify the latent structure ofthe 14-item PGIS. Before conducting the EFA,we tested several of the statistical assumptionsnecessary for such an analysis. The Kaiser–Meyer–Olkin index of sampling adequacy was0.70, which indicated that partial correlationswere small and that the matrix was suitablefor factor analysis (Tabachnick & Fidell, 2001).



Table 1: Sociodemographic Characteristicsof the Study Sample (N = 227)

Variable/Category n (%)

Female 227 (100%)

Race

White 177 (78.0%)

Other 50 (22.0%)

Education

<4 years College 68 (30.0%)

College Degree 92 (40.5%)

Master’s Degree or Above 38 (16.7%)

Missing 29 (12.8%)

Income

<$49, 999 68 (30.0%)

$50,000 – $75, 999 51 (22.5%)

>$80, 000 80 (35.2%)

Missing 28 (12.3%)

Married

Yes 204 (89.9%)

No 20 (8.8%)

Missing 3 (1.3%)

Gestational Age Loss Group

Miscarriage 96 (42.3%)

Stillbirth 82 (36.1%)

Neonatal Death 49 (21.6%)

Location

USA 171 (75.3%)

Central America 33 (14.5%)

Other 17 (7.6%)

Missing 6 (2.6%)

Mean Age (SD) 31.0 (5.9)

Note. SD = standard deviation.

Bartlett’s test of sphericity was statistically signifi-cant (χ2 = 203.16, df = 91, p < 0.001), and no ev-idence of multicollinearity or singularity was found(Tabachnick & Fidell). These results showed thatthe EFA could be adequately performed. In orderto determine the number of factors to retain, twomethods were used: (a) the eigenvalues greaterthan 1 rule, and (b) the scree plot to identify thenumber of factors above the elbow to retain androtate (Cattell, 1966).

Table 2: Cronbach’s Alphas Stratified byGroup

Miscarriage Stillbirth Neonatal Death

Group Group Group

PGIS Total 0.77 0.79 0.68

Reality 0.79 0.86 0.71

Confront Others 0.84 0.87 0.75

Congruence 0.83 0.86 0.72

Note. PGIS = Perinatal Grief Intensity Scale.

For initial examination of the items, principal com-ponents analysis with varimax rotation was usedto maximize variance. This revealed three factorswith an eigenvalue greater than one. In addition,the scree plot results suggested a three-factor so-lution similar to the original factor analysis in whichthree factors were retained (Hutti, dePacheco, &Smith, 1998). Because there were no items witha factor loading less than 0.40 and none cross-loaded, all of the original 14 PGIS items wereretained.

These three factors accounted for 66.94% of thetotal variance (see Table 3). Judged by item con-tent, the first factor comprised a combination ofall the Congruence and Confront Others items asfound in the original factor analysis (Hutti et al.,1998). This first factor was labeled Satisfactionand accounted for 34.62% of the variance. Thesecond factor focused on all items associatedwith reality of the pregnancy (Q1, Q5-Q6) andwas labeled Reality of the Pregnancy; it accountedfor 22.13% of the variance. The third factor con-tained all items associated with reality of the fe-tus/newborn (Q2-Q4) and was labeled Reality ofthe Baby; it accounted for 10.19% of the variance.

Variations in grief intensity by type of loss. ThePGIS total scores varied by type of loss in the sub-sequent pregnancy after a perinatal loss, F(474,2)] = 14.58, P = .002. Post hoc comparisonsdemonstrated a significant mean difference of1.16 between the stillbirth group (X = 2.81) andthe neonatal death group (X = 3.97, P = .02),which indicated that the neonatal death groupparticipants reported more intense grief in thesubsequent pregnancy than the stillbirth group.In addition, post hoc comparisons indicated asignificant mean difference of 0.13 between themiscarriage group (X = 3.84) and the neona-tal death group (X = 3.97, P = .007), which



Table 3: Varimax Rotation of Three Factors Using Exploratory Factor Analysis for the Peri-natal Grief Intensity Scale

Factors

Items I II III

I felt satisfied with the way my loss experience unfolded, given that I had to

go through it.

.63

I felt satisfied with my interactions with my family. .85

I felt satisfied with my interactions with my nurses. .56

I felt satisfied with my interactions with my friends. .78

In the first hours and days after my pregnancy loss/baby’s death I was able

to ask people who said or did things that made me feel bad to stop.

.71

In the first hours and days after my pregnancy loss/baby’s death I was able

to resolve problems if something happened that I did not like.

.76

In later weeks after my pregnancy loss/baby’s death I was able to ask people

who said or did things that made me feel bad to stop.

.79

In later weeks after my pregnancy loss/baby’s death I was able to resolve

problems if something happened that I did not like.

.70

At the time my pregnancy loss/baby’s death the pregnancy did not seem real

to me.

.85

At the time my pregnancy loss/baby’s death occurred, my pregnancy and

baby seemed real to me.

.93

At the time my pregnancy loss/baby’s death it seemed like the loss of

pregnancy not the loss of a baby.

.50

At the time my pregnancy loss/baby’s death I did not think of the baby as a

person.

.85

At the time my pregnancy loss/baby’s death I did not think of the baby as

having a personality yet.

.67

At the time my pregnancy loss/baby’s death I felt that I had lost my son or my

daughter.

.87

Eigenvalue 4.84 3.10 1.43

Explained Variance 34.62% 22.13% 10.19%

Note. aOnly factor loadings of .40 or above are shown. bFactors: I = Satisfaction; II = Reality of Pregnancy; III = Reality of Baby.

indicated that the neonatal death group experi-enced more intense grief when compared with themiscarriage group. The mean difference of 0.08 onthe PGIS between the miscarriage and stillbirthgroups was not significantly different. No otherpairwise comparisons achieved significance.

DiscussionThe PGIS demonstrated good beginning totaland subscale internal consistency reliability whenused in subsequent pregnancy after all types ofperinatal loss, including miscarriage, stillbirth, and

neonatal death. The original 14 items of the PGISwere retained in this second factor analysis andalso loaded onto three factors. However, the itemsloaded into the three factors somewhat differentlyin this second analysis compared with the first(Hutti, dePacheco, & Smith, 1998). In the firstfactor analysis, items loaded onto concepts de-scribed as reality of the pregnancy and fetus (Re-ality), congruence between the experience of theactual loss and the parents’ standard of the de-sirable or perceived ideal against which the ex-perience is being evaluated (Congruence), andthe ability of parents to make decisions or act



Gestational age of the fetus is not a reliable predictor of griefintensity and need for follow-up after perinatal loss.

in ways to increase this congruence (ConfrontOthers) (Hutti, 1992; Hutti, 2005).

In the current factor analysis, the first factor con-sisted of all of the Congruence and Confront Oth-ers items, which were loaded together into a singlefactor. The second factor consisted of all Realityitems associated with the reality of the pregnancy.The third factor consisted of all Reality items as-sociated with the reality of the infant. It is not sur-prising that the Congruence and Confront Othersitems were loaded together. In our clinical experi-ence, the congruence between the experience ofthe actual loss and the parents’ standard of the de-sirable is constantly being moderated by the moth-ers’ ability or inability to be assertive about theirneeds and desires during a perinatal loss, andto confront others when necessary—-or to havean advocate who will do so for them. Together,the concepts of Congruence and Confront Oth-ers form an overall satisfaction with the care andsupport received by significant others (e.g., fam-ily, friends, health care providers) during the ex-perience of the loss. When a mother experienceslow congruence with her standard of the desirable(e.g., nothing is occurring as the mother believesit should) and the mother perceives herself as un-able to confront others about this incongruity, shewill often feel angry and unsupported in the expe-rience of the loss. When a mother who feels angryand unsupported also experiences a pregnancyof high reality, that is, the pregnancy and the infantare real to the parent, intense grief is likely to occur(↑Reality + ↓ Congruence + ↓ Confront Others =Intense Grief).

When the mother experiences high congruencewith her standard of the desirable (e.g., almosteverything is occurring as the mother believes itshould) and she is able to confront others when in-congruity exists, the mother often will feel a senseof satisfaction with care and support received dur-ing the experience of the perinatal loss. If themother generally feels satisfied with her care andsupport also experiences a pregnancy of low re-ality, that is, neither pregnancy nor fetus or infantare real to the parent, little or no grief is likely tooccur (↓ Reality + ↑ Congruence + ↑ ConfrontOthers = Little or No Grief). Thus, the conceptsof Congruence and Confront Others form the ba-sis of a mother’s satisfaction with her care. This

satisfaction, coupled with how she perceives thereality of the pregnancy and infant, form the basisfor whether she will experience intense grief afterperinatal loss.

From a theoretical perspective, this study sup-ports the concepts of the grief intensity theoreticalframework of reality of the pregnancy and fetus,congruence between the experience of the actualloss and the mothers’ standard of the desirable,and the ability of mothers to make decisions or actin ways to increase this congruence. The conceptsof Congruence and Confront Others work togetherto influence the mother’s overall satisfaction withcare and support received by family members,friends, and health care providers during the ex-perience of the perinatal loss.

Grief intensity varied significantly by type of loss,with the neonatal death group exhibiting the mostintense grief compared with the stillbirth and mis-carriage groups, thereby supporting the constructvalidity of the PGIS. Although grief intensity scoreswere higher in the stillbirth group than the mis-carriage group, the difference was not statisticallysignificant. It has been long known that seeingand holding an infant after birth will significantlyenhance attachment (Kennell & Klaus, 1998). Thisenhanced attachment should translate into moreintense grieving in the event of a neonatal death.

If a fetus should die during pregnancy, it is un-likely that attachment to the pregnancy and fetusis a matter of only having an increased gestationalperiod (Shreffler et al., 2011). In this study, no dif-ferences in grief intensity were found in the miscar-riage and stillbirth groups. Health care providersoften expect that a parent who experiences alonger gestation and has a stillbirth always willgrieve longer and more intensely when comparedwith a parent who has been pregnant for a fewweeks or months and then experiences a miscar-riage. Although it is common for parents of still-born neonates to grieve intensely, numerous stud-ies have documented that women and men whoexperience the loss of an early pregnancy mayand often do experience significant grief as well(Hutti et al., 1998; Swanson et al., 2009; Swanson,Connor, Jolley, Pettinato, & Wang, 2007). Gesta-tional age of the fetus is not a reliable predictorof grief intensity after perinatal loss. This is an-other reason why it is so difficult for health careproviders to accurately identify parents in need ofgrief follow-up after a perinatal loss. We hope thatthe PGIS may eventually be able to assist healthcare providers with the difficult clinical decisions



about care that rest upon fully understanding thesignificance the loss has for the mother, as wellas supporting a bereavement experience that ismore congruent with the mother’s needs.

LimitationsBecause this sample was not random, the abil-ity to generalize to an entire population is cur-tailed. This was an international sample, but thedegree of English fluency of those participants forwhom English was not a primary language is un-known and may have affected understanding ofthe instrument. Despite the international nature ofthe sample, the race of most of the participantswas Caucasian, and responses from women ofother races may have differed significantly fromthose reported here. In addition, these respon-dents sought out this web site and study; there-fore, their grief and its consequences may havebeen more intense than that of other women whoexperienced similar perinatal losses.

ConclusionsThe perinatal loss literature has documented ad-verse mental health consequences and health ef-fects for parents and their subsequent children,and represents a tremendous burden of diseasefor the health of the nation. Improving a mother’smental health also may improve the health andwell-being of her family. The findings from thisstudy provided support for the reliability and con-struct validity of the PGIS, which may be usedto identify women who have experienced miscar-riage, stillbirth, or neonatal death and need inter-vention from a health care provider for the crisis ofperinatal loss.

The PGIS initially was developed to be used in thefirst hours and days after a perinatal loss to helpidentify parents at risk of intense grief and in needof follow-up care by health care providers. It nowalso has been used in the subsequent pregnancyafter a perinatal loss to identify women who are ex-periencing the most intense grief. There may beother uses for the instrument that are still not identi-fied. This instrument can identify women at risk ofhigh-intensity versus low-intensity grief reactionsafter perinatal loss. It needs further evaluation todetermine cut-off scores for identifying need forfollow-up by health care providers and its validityfor predicting intense grief reactions when usednear the time of the loss. Further evaluation of thescoring system also is warranted to determine ifthe PGIS needs to be simplified for clinical use,

It is imperative that reliable and valid clinical instruments bedeveloped so that health care providers direct scarce resources

to those most in need.

and if so, how that can be reliably accomplished.These are all areas of research that merit furtherstudy.

Researchers have been studying the psychologi-cal consequences of perinatal loss for more than40 years. However, very few randomized, con-trolled nursing intervention trials have been con-ducted to begin testing theory and frameworksin the area of perinatal loss (Swanson, 1999a,1999b; Swanson et al., 2009). The need for the-oretical, evidenced-based practice is significanteverywhere in nursing practice, but there is adearth of evidence surrounding interventions as-sociated with nursing practice and perinatal loss.

As the researchers of this study found, it is com-mon for mothers who experience a neonatal deathto grieve intensely. Identifying grief intensity af-ter miscarriage or stillbirth may be more difficult.Not all mothers grieve after perinatal loss, andfor those who do, neither the gestational age ofthe fetus nor parental behavior are reliable pre-dictors of current or future grief intensity. Healthcare providers want to help grieving families, butthe first step in doing so is accurately identify-ing those who will most likely need follow-up andsupport. It is imperative that reliable and validclinical instruments be developed so that healthcare providers may direct scarce resources tothose most in need. It is time to use our accumu-lated knowledge to also test nursing interventionsand perinatal grief clinical instruments to help pro-mote healing and personal growth, foster couplerelationships, and improve the quality of life forfamilies who have experienced these devastatinglosses.

AcknowledgementFunded by the Norton Hospital Foundation,Louisville, KY.

REFERENCESArmstrong, D. S. (2007). Subsequent pregnancy after the birth of a

healthy infant. Advances in Neonatal Care, 7(4), 200–206.

Armstrong, D. S., Hutti, M. H., & Myers, J. (2009). The influence of prior

perinatal loss on parents’ psychological distress after the birth

of a subsequent healthy infant. Journal of Obstetric, Gyneco-

logic, & Neonatal Nursing, 38(6), 654–666. doi:10.1111/j.1552-

6909.2009.01069.x



Blackmore, E. R., Cote-Arsenault, D., Tang, W., Glover, V., Evans, J.,

Golding, J., & O’Connor, T. G. (2011). Previous perinatal loss as

a predictor of perinatal depression and anxiety. British Journal

of Psychiatry, 198, 373–378. doi:10.1192/bjp.bp.110.083105

Cattell, R. B. (1966). The scree test for the number of factors.

Multivariate Behavioral Research, 1, 245–276. doi:10.1207/

s15327906mbr0102_10

Cote-Arsenault, D., & Dombeck, M. T. B. (2001). Maternal assignment

of fetal personhood to a previous loss: Relationship to anxiety in

the current pregnancy. Health Care for Women International, 22,

649–665.

Couto, E. R., Couto, E., Vian, B., Gregoria, Z., Nomura, M. L., Zaccharia,

R. & Passini, R. (2009). Quality of life, depression, and anxiety

among pregnant women with previous adverse pregnancy out-

comes. Sao Paulo Medical Journal, 127(4), 185–189.

Cuisinier, M., Janssen, H., de Graauw, C., Bakker, S., & Hoogduin, C.

(1996). Pregnancy following miscarriage: Course of grief and

some determining factors. Journal of Psychosomatic Obstetrics

and Gynecology, 17, 168–174.

Dougherty, E. (1984). Cognitive models of face-to-face interaction: A

semiotic approach (Unpublished doctoral dissertation). Buffalo,

NY: State University of New York at Buffalo.

Gaudet, C., Sejourne, N., Camborieux, L., Rogers, R., & Chabrol, H.

(2010). Pregnancy after perinatal loss: Association of grief, anx-

iety, and attachment. Journal of Reproductive and Infant Psy-

chology, 28(3), 240–251. doi:10.1080/02646830903487342

Gausia, K., Moran, A. C., Ali, M., Ryder, D., Fisher, C., & Koblinsky, M.

(2011). Psychological and social consequences among moth-

ers suffering from perinatal loss: Perspective from a low income

country. BioMed Central Public Health, 11, 451–460. Retrieved

from http://www.biomedcentral.com/1471-2458/11/451

Gold, K. J., Sen, A., & Hayward, R. A. (2010). Marriage and cohabitation

outcomes after pregnancy loss. Pediatrics, 125(5), 1202–1207.

doi:10.1542/peds.2009–3081.

Hutti, M. H. (1992). Parents’ perceptions of the miscarriage experience.

Death Studies, 16(5), 401–415.

Hutti, M. H. (2005). Social and professional support needs of fami-

lies after perinatal loss. Journal of Obstetric, Gynecologic, and

Neonatal Nursing, 34(5), 630–638.

Hutti, M. H., Armstrong, D. S., & Myers, J. (2011). Healthcare uti-

lization in the pregnancy following a perinatal loss. MCN: The

American Journal of Maternal/Child Nursing, 36(2), 104–111.

doi:10.1097/NMC.0b013e3182057335

Hutti, M. H., dePacheco, M. R., & Smith, M. (1998). Development and

validation of the Perinatal Grief Intensity Scale. Journal of Ob-

stetric, Gynecologic, and Neonatal Nursing, 27(5), 547–555.

Jind, L. (2003). Parents’ adjustment to late abortion, stillbirth, or infant

death: The role of causal attributions. Scandinavian Journal of

Psychology, 44, 383–394.

Kennell, J. H., & Klaus, M. H. (1998). Bonding: Recent observa-

tions that alter perinatal care. Pediatrics in Review, 19, 4–12.

doi:10.1542/pir.19-1-4

Lang, A., Fleiszer, A. R., Duhamel, F., Sword, W., Gilbert, K. R.,

& Corsini-Munt, S. (2011). Perinatal loss and perinatal grief:

The challenge of ambiguity and disenfranchised grief. Omega,

63(2), 183–196. doi:10.2190/OM.63.2.e

O’Leary, J. (2005). The trauma of ultrasound during a pregnancy fol-

lowing a perinatal loss. Journal of Trauma and Loss, 10,183–

204.

Robson, S. J., Leader, L. R., Bennett, M. J., & Dear, K. B. G. (2010).

Do women’s perceptions of care at the time of unexplained

stillbirth influence their wishes for management in subsequent

pregnancy? An Internet-based empirical study. Obstetrics and

Gynaecology Research, 36(1), 108–114. doi:10.1111/j.1447-

0756.2009.01113.x

Rowlands, I. J., & Lee, C. (2010). The silence was deafening:

Social and health service support after miscarriage. Jour-

nal of Reproductive and Infant Psychology, 28(3), 274–286.

doi:10.1080/02646831003587346

Schaap, A. H. P., Wolf, H., Bruinse, H. W., Barkhof-van de Lande, S., &

Treffers, P. E. (1997). Long-term impact of perinatal bereave-

ment: Comparison of grief reactions after intrauterine versus

neonatal death. European Journal of Obstetrics & Gynecology

and Reproductive Biology, 75, 161–167.

Shreffler, K., Greil, A., & McQuillan, J. (2011). Pregnancy loss and

distress among US women. Family Relations, 60, 342–355.

doi:10.1111/j.1741–3729.2011.00647.x

Sutan, R., Amin, R. M., Ariffin, K. B., Teng, T. Z., Kamal, M. F., & Rusli, R.

Z. (2010). Psychosocial impact of mothers with perinatal loss and

its contributing factors: An insight. Journal of Zhejiang University,

11(3), 209–217.

Swanson, K. M. (1999a). The effects of caring, measurement, and time

on miscarriage impact and women’s well being in the first year

after loss. Nursing Research, 48, 288–298.

Swanson, K. M. (1999b). Research-based practice with women who

have had miscarriages. Image, 31, 339–345.

Swanson, K. N., Connor, S., Jolley, S. N., Pettinato, M, & Wang, T.

(2007). Contexts and evolution of women’s responses to mis-

carriage during the first year after loss. Research in Nursing &

Health, 30, 2–16. doi:10.1002/nur.20175

Swanson, K. M., Hsien-Tzu, C., Graham, J. C., Wojnar, D. M., & Petras,

A. (2009). Resolution of depression and grief during the first

year after miscarriage: A randomized controlled clinical trial of

couples-focused intervention. Journal of Women’s Health, 18(9),

1245–1257. doi:10.1089/jwh.2008.1202.

Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics.

Needham Heights, MA: Allyn & Bacon.

Uren, T. H. & Wastell, C. A. (2002). Attachment and meaning-

making in perinatal bereavement. Death Studies, 26,

279–308.

Woods-Giscombe, C. L., Lobel, M., & Crandell, J. L. (2010). The

impact of miscarriage and parity on patterns of maternal

distress in pregnancy. Research in Nursing & Health, 33,

316–328.


evaluation of the perinatal grief intensity scale in the subsequent pregnancy after perinatal loss

Documents