Format

Send to

Choose Destination
Psychol Assess. 2013 Jun;25(2):456-66. doi: 10.1037/a0031392. Epub 2013 Jan 28.

Quantifying diagnostic uncertainty using item response theory: the Posterior Probability of Diagnosis Index.

Author information

1
Department of Psychiatry, University of Pittsburgh School of Medicine, Pittsburgh, PA15213, USA. lindhiemoj@upmc.edu

Abstract

Using traditional Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision (American Psychiatric Association, 2000) diagnostic criteria, clinicians are forced to make categorical decisions (diagnosis vs. no diagnosis). This forced choice implies that mental and behavioral health disorders are categorical and does not fully characterize varying degrees of uncertainty associated with a particular diagnosis. Using an item response theory (latent trait model) framework, we describe the development of the Posterior Probability of Diagnosis (PPOD) Index, which answers the question: What is the likelihood that a patient meets or exceeds the latent trait threshold for a diagnosis? The PPOD Index is based on the posterior distribution of θ (latent trait score) for each patient's profile of symptoms. The PPOD Index allows clinicians to quantify and communicate the degree of uncertainty associated with each diagnosis in probabilistic terms. We illustrate the advantages of the PPOD Index in a clinical sample (N = 321) of children and adolescents with oppositional defiant disorder.

PMID:
23356682
PMCID:
PMC3790577
DOI:
10.1037/a0031392
[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for American Psychological Association Icon for PubMed Central
Loading ...
Support Center