Browsing by Author "Mare, Karen Thea"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Open Access
    Perinatal suicidality: prevalence and correlates in a South African cohort
    (2021) Mare, Karen Thea; Stein, Dan J
    Background – Suicide caused 1% of maternal deaths in South Africa in recent years. Suicidal ideation and behaviour (SIB) in the perinatal period is common and is associated with completed suicide. Data regarding perinatal SIB in low and middle-income countries (LMICs) are lacking, and available work is limited by reliance on self-report scales. There are even fewer data on risk categories of SIB, risk factors for SIB severity and the longitudinal course of SIB. We collected cross-sectional and longitudinal data on SIB, SIB severity, and potential correlates from peri-urban South African women who were enrolled in a birth cohort. A well validated screening measure for perinatal depression, the Edinburgh Postnatal depression scale (EPDS) has unknown sensitivity and specificity when assessing perinatal SIB (using question 10: EPDS-Q10) compared to a gold standard diagnosis. Objectives – First, to determine the perinatal prevalence of SIB and its sociodemographic, psychosocial and psychiatric correlates, using a validated clinician-administered assessment tool in a broadly representative sample of South African women. Second, to analyse correlates for low, medium and high risk categories of SIB separately. Third, to determine the longitudinal course and correlates of SIB in the perinatal period. Fourth, to compare the sensitivity and specificity of the self-report EPDS – Q10 to a gold standard diagnostic measure. Methods – Data were collected from participants in the Drakenstein Child Health Study (DCHS) cohort during pregnancy and at 6 months post-partum (PP). SIB was measured using the MINI International Neuropsychiatric Interview (MINI). Self-report and clinician administered questionnaires assessed sociodemographic, psychosocial, and psychiatric risk factors. Multivariable analysis determined cross sectional risk and protective factors. To determine predictors of the risk categories of SIB (None = 0, Low = 1-8, Medium = 9-16, and High ≥ 17), multinomial regressions were used. Selecting women who never reported SIB as a reference group, multinomial regression analysis yielded information about risk and protective factors for reporting sustained or new SIB, as well as cessation of SIB. The ROC curve determined an optimum cut-off score and the area under the curve (AUC) indicated maximum sensitivity and specificity of the EPDSQ10 when compared to the MINI - section B diagnosis of perinatal SIB in the same women. Results – The antenatal prevalence of SIB was 19.9% and the post-partum prevalence was 22.6%. Factors independently predicting the presence of SIB included younger age antenatally and PTSD post-partum with depression positively associated with SIB at both time points. Depression and PTSD predicted belonging to the high risk SIB group while membership of the low risk SIB group was associated with both these mental illnesses as well as recent intimate partner violence (IPV), food insecurity and childhood trauma. Of the 121 women assessed at both time points, 71% women never reported SIB, 10.7% reported sustained SIB (i.e., SIB before and after the birth), 9.1% reported new SIB (i.e., no SIB antenatally but SIB post-partum) and 9.1% reported cessation of SIB (i.e., SIB antenatally but no SIB post-partum) Low participant numbers prohibited inferences regarding risk factors for sustained SIB or new onset SIB. The AUC of the EPDS-Q10 was 0.605 denoting low clinical utility for SIB screening.