Validation of a Brief Measure for Complicated Grief Specific to Reproductive Loss

Abstract

Objective Complicated grief reactions follow some pregnancy outcomes, like miscarriage, stillbirth, neonatal death, infant death, selective reduction, or termination of pregnancy. Stigma can delay treatment and worsen outcomes. Screening tools such as the Edinburgh Postnatal Depression Scale detect complicated grief poorly, and specific tools for prolonged or complicated grief after a reproductive loss are cumbersome. In this study, a five-item questionnaire to detect complicated grief after reproductive loss of any type was designed and preliminary validated. Methods A questionnaire patterned after the extensively validated Brief Grief Questionnaire (BGQ) was created by a group of physicians and lay advocates to employ non-traumatic but specific language related to grief after miscarriage, stillbirth, neonatal death, infant death, selective reduction, or termination of pregnancy. One hundred and forty women at a large academic center were recruited in person and via social media to validate the questionnaire with well-studied instruments for anxiety (7-item Panic Disorder Severity Scale, PDSS), trauma (22-item Impact of Events Scale), and reproductive grief and depressive symptoms (33-item Perinatal Grief Scale [PGS]). Results The response rate was 74.9%. Of the 140 participants, 18 (12.8%) experienced their loss during high-risk pregnancies, and 65 (46.4%) were recruited via social media. Seventy-one (51%) respondents had a score > 4, a positive screen for the BGQ. On average, women experienced their loss 2 years prior to participation (IQR 1-5 years). Cronbach's alpha was 0.77 (95% CI: 0.69-0.83). The goodness of fit indices of the model met Fornell and Larker criteria (RMSEA = 0.167, CFI = 0.89, SRMR = 0.06). The AVE was 0.42 and the CR 0.78. Conclusions This investigator-created screening tool is internally consistent and meets preliminary criteria for discriminant validity. This tool can be refined prior to testing for sensitivity and specificity in screening for complicated grief after a reproductive loss.

Keywords: abortion; miscarriage; neonatal death; persistent complex bereavement disorder; postpartum depression; pregnancy loss; prolonged grief; screening; stillbirth; survey.

Figure 1. Distribution of timing in 113 losses which occurred during pregnancy

Postnatal losses are not depicted.

Figure 2. Distribution of total Reproductive Grief Screen score (n = 140)

Figure 3. Crossover between IESR and RGS items

All items labeled “IESR” belong to a trauma scale but cross-load onto the Complicated Grief Factor (lower left) with items labeled “RGS,” which correctly load. Ideally, the items (questions) in the trauma survey (IESR) would "load" onto the trauma factor, and the items in the reproductive grief screen (RGS) would load onto the complicated grief factor. This would indicate that the two surveys measure different symptom pools. However, multiple IESR items "cross-load" (load onto another factor) onto the complicated grief factor. The cross-loading items relate to vigilance, dreams, panic symptoms, concentration, symptoms coming in waves, effects on sleep, reliving experiences, being easily startled, avoidance, and irritability. This suggests that these symptoms in particular may overlap with complicated grief, or that the RGS may be overly sensitive for trauma symptoms.