The OCI-R: Validation of the subscales in a clinical sample

https://doi.org/10.1016/j.janxdis.2006.05.006Get rights and content

Abstract

The psychometric properties of the Obsessive–Compulsive Inventory-Revised (OCI-R) subscales have not been validated in a clinical sample of individuals diagnosed with obsessive–compulsive disorder (OCD). Data were collected on 186 patients diagnosed with OCD and 17 patients diagnosed with generalized anxiety disorder (GAD) using the OCI-R and the Yale-Brown Obsessive–Compulsive Scale (Y-BOCS). Confirmatory factor analysis revealed an acceptable factor structure. Patients with a given primary symptom subtype were elevated on the corresponding subscale of the OCI-R compared to other OCD patients and patients with GAD. In addition, patients who acknowledged that symptom subtype as present but not primary on the Y-BOCS had elevated OCI-R scores on that scale compared to patients who did not endorse that symptom subtype and patients with GAD. Results indicate that the subscales of the OCI-R are valid measures of six symptom subtypes of OCD. The OCI-R is a psychometrically sound, brief instrument. The current data combined with previous efforts suggest that it is appropriate for clinical and non-clinical populations, and for clinical and research purposes. Further research should examine the sensitivity of the specific subscales to treatment effects, and the potential for adding more items to account for other symptom domains of OCD.

Section snippets

Participants

Data were collected from 186 patients with OCD and 17 patients with GAD. OCD patients’ data were derived from two sources. One hundred and one patients sought treatment for OCD through the Center for the Treatment and Study of Anxiety (CTSA) fee-for-service clinic. Eighty-five patients participated in a randomized clinical trial examining cognitive–behavioral treatment augmentation for partial responders to SRIs either at the CTSA or at the New York State Psychiatric Institute. Patients with

Results

The average age of the 186 patients with OCD was 35 (S.D. = 13), and 45% were female. The large majority (93%) was Caucasian, while 1% was African American, 2% were Asian or Asian American, and 3% were Latino. The GAD sample had similar characteristics: average age was 31 (S.D. = 11), 53% female, and 94% Caucasian. Means and standard deviations for the OCI-R total and subscale scores are presented in Table 1. Tests of equality of variance between the patients with OCD and patients with GAD

CFA

The six-factor structure of the OCI-R was confirmed, using criteria recommended by Hu and Bentler (1999). The model had a significant Chi-square (χ2(121) = 216.4, p < .0001), a Goodness of Fix Index (GFI) of .89, a Comparative Fit Index (CFI) of .95, a root mean square residual (RMR) of .099, and a root mean square error of approximation (RMSEA) of .06 (90% confidence interval .05–.08). All of these values indicate an adequate fit for the model except the Chi-square. The fit was significantly

Convergent and discriminant validity of the subscales: differences between groups

For all six categories of symptoms, the overall ANOVAs were significant (F's(3, 172) > 16, p's < .001). For each subscale on the OCI-R, Tukey's comparisons generally showed that patients identified by the Y-BOCS to have that symptom as primary had higher scores on the corresponding OCI-R subscale than patients who had the symptom present but not primary; the latter patients had higher OCI-R subscale scores than both those OCD patients who did not endorse the symptom on the Y-BOCS checklist and

Discussion

The current study was designed to examine the psychometric properties of the subscales of the OCI-R in patients diagnosed with OCD, and to examine the convergent validity of the six subscales of the OCI-R and the corresponding six OCD subtypes derived from the Y-BOCS checklist. In addition, we further examined the discriminant validity of the OCI-R subscales to OCD by comparing the scores of OCD and GAD patients.

Adequate to excellent model fit and internal consistency were found for the six

Acknowledgements

Supported by Grants K23MH064491 to Dr. Huppert, R01MH045404 to Dr. Foa and R01MH045436 to Dr. Liebowitz from the National Institute of Mental Health, Bethesda, MD. Thanks to Shawn P. Cahill for comments on an earlier draft of the manuscript.

References (21)

There are more references available in the full text version of this article.

Cited by (206)

View all citing articles on Scopus
View full text