Development of the Perinatal Depression Inventory (PDI)-14 using item response theory: a comparison of the BDI-II, EPDS, PDI, and PHQ-9

Abstract

The objective of this study is to develop a simple, brief, self-report perinatal depression inventory that accurately measures severity in a number of populations. Our team developed 159 Likert-scale perinatal depression items using simple sentences with a fifth-grade reading level. Based on iterative cognitive interviewing (CI), an expert panel improved and winnowed the item pool based on pre-determined criteria. The resulting 67 items were administered to a sample of 628 pregnant and 251 postpartum women with different levels of depression at private and public sector obstetrics clinics, together with the Beck Depression Inventory (BDI-II), Edinburg Postpartum Depression Scale (EPDS), and the Patient Health Questionnaire (PHQ-9), as well as Module A of the Structured Clinical Interview for DSM-IV Diagnoses (SCID). Responses were evaluated using Item Response Theory (IRT). The Perinatal Depression Inventory (PDI)-14 items are highly informative regarding depression severity and function similarly and informatively across pregnant/postpartum, white/non-white, and private-clinic/public-clinic populations. PDI-14 scores correlate well with the PHQ-9, EPDS, and BDI-II, but the PDI-14 provides a more precise measure of severity using far fewer words. The PDI-14 is a brief depression assessment that excels at accurately measuring depression severity across a wide range of severity and perinatal populations.

Keywords: Depression; IRT; PDI-14; Perinatal; Postpartum.

Fig. 1

The test information function and standard error curves of the PDI-14, the BDI-II, the PHQ, and the EPDS