Postpartum depression is an illness that is frequently unreported and undetected for a variety of reasons and may be potentially devastating for the mother affected as well as her family. Routine screening of postpartum women enables health practitioners to detect symptoms of PPD early and provides an opportunity for early intervention which may improve the outcome and increase the mother’s chances of an earlier recovery. It is therefore important that reliable and convenient screening tools are available to health practitioners who have contact with postpartum women. The primary objective of this research was to make an Afrikaans version of an existing screening scale available – the Postpartum Depression Screening Scale (PDSS), designed specifically to encompass the multifaceted phenomenon of PPD. In accordance with this objective, the validity and reliability of the PDSS and its Afrikaans version was investigated in English- and Afrikaans-speaking South African mothers. A further objective of this study was to compare the performance of the PDSS with the Edinburgh Postnatal Depression Scale (EPDS) and the Quick Inventory of Depressive Symptomatology (QIDS-SR16). Various factors have been reported to be associated with the development of PPD. The final objective of this study was to explore the relationship between known risk factors for PPD and high scores on the PDSS amongst women in South African. A total of 365 South African mothers, between 4 and 16 weeks postpartum participated in this study. English-speaking mothers (n = 187) completed the PDSS, EPDS, QIDS, and a demographic and psychosocial questionnaire, while Afrikaans speaking mothers (n = 178) completed the respective Afrikaans versions of these questionnaires. A multiple translation method – Brislin’s back-translation method and the committee approach – was used to translate the PDSS and the QIDS into Afrikaans. An item response theory (IRT), Rasch analysis, was used to examine dimensionality, item difficulty, differential item functioning, and category functioning of the PDSS and the Afrikaans PDSS. Results reveal excellent person reliability estimates for the Afrikaans PDSS as well as for the PDSS in a South African sample. Both language versions performed reasonably well and the majority of items in the PDSS dimensions and the Afrikaans PDSS dimensions demonstrated fit statistics that supported the underlying constructs of each dimension. Some items were identified as problematic, namely Item 2, Item 25, Item 28, and Item 30. The item person construct maps show reasonably good spread of items. There were, however, persons that scored higher than the items could measure and an overrepresentation of items at the mean level. The Likert response categories proved to be effective for all the Afrikaans PDSS items and almost all the PDSS items. Results indicate that 49.7% of mothers screened positive for major PPD using the PDSS. A further 17.3% of mothers obtained scores indicating the presence of significant symptoms of PPD. Statistically significant correlations were obtained between total scores on the PDSS, the EPDS, and the QIDS-SR16. Stepwise multiple regression analysis identified 11 variables that were significantly associated with a high PDSS total score. These were a history of psychiatric illness, postpartum blues, feeling negative or ambivalent about expecting this baby, fearful of childbirth, infant temperament, antenatal depression in recent pregnancy, lack of support from the baby’s father, concern about health related issues regarding the infant, lack of support from friends, difficulty conceiving, and life stress. / Thesis (PhD)--University of Pretoria, 2012. / Psychology / unrestricted
Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:up/oai:repository.up.ac.za:2263/25576 |
Date | 16 June 2012 |
Creators | Struik, Melony |
Contributors | Prof D J F Maree, mel.struik@gmail.com |
Source Sets | South African National ETD Portal |
Detected Language | English |
Type | Thesis |
Rights | © 2011 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria. |
Page generated in 0.002 seconds