Skip to main content

    Licensed instrument

    OCI-12: scoring, cutoffs & interpretation

    A 12-item OCD symptom measure covering checking, ordering, washing and obsessing. Items are copyright-restricted, so it is not scorable here.

    02 - The clinician's brief

    Last reviewed: · Reviewed by Lotte Kjær Svalberg

    01 - What it measures

    What this scale measures.

    The OCI-12 is a self-report measure of obsessive-compulsive symptom severity across the four dimensions that define the modern conception of OCD: checking, ordering, washing and obsessing. Each dimension gets three items rated for how much distress it caused in the past month. It is a shortened, syndromally cleaner revision of the 18-item OCI-R, removing the hoarding items (hoarding is a separate DSM-5 diagnosis) and the neutralising items (which loaded poorly and blurred with obsessing).

    The result is a total score that tracks OCD as currently defined, rather than OCD plus its former neighbours. It suits screening and progress monitoring; it is not a diagnostic instrument.

    02 - Origin & purpose

    Where it comes from.

    Amitai Abramovitch, Jonathan Abramowitz and Dean McKay published the OCI-12 in 2021, derived from Foa and colleagues' OCI-R (2002), itself a shortening of the original 42-item OCI (1998). The motivation was DSM-5: hoarding disorder left the OCD chapter in 2013, yet the OCI-R kept scoring it, inflating totals for a symptom no longer part of the syndrome.

    The validation drew on 1,040 OCD patients, 423 patients with anxiety-related disorders and 1,194 non-clinical controls across Rogers Behavioral Health, UNC and the Mayo Clinic. The OCI-12 matched or slightly beat the parent OCI-R on diagnostic accuracy (AUC 0.91 vs 0.88 against controls) while dropping a third of its items.

    03 - Scoring & cutoffs

    How scoring works.

    Each of the 12 items is rated 0 ('Not at all') to 4 ('Extremely') for distress over the past month, giving a total of 0-48 across four 3-item subscales: checking, ordering, washing and obsessing. In the 2021 validation, a total score of 11 or above best separated OCD patients from non-clinical controls (sensitivity ~83%, specificity ~81%); distinguishing OCD from other anxiety-related disorders required a higher cutoff of 14 (sensitivity ~72%, specificity ~73%). Severity benchmarks: 0-12 mild, 13-21 moderate, 22-48 severe. The authors are explicit that the cutoff must not be used on its own to diagnose OCD.

    04 - Validation evidence

    How well it performs.

    The 2021 validation drew on 1,040 OCD patients, 423 patients with anxiety-related disorders and 1,194 non-clinical controls across Rogers Behavioral Health, UNC and the Mayo Clinic. At the cutoff of 11 or above against non-clinical controls, sensitivity was 83% and specificity 81%; the stricter cutoff of 14 was needed to separate OCD from other anxiety-related disorders (~72% / ~73%). The total score discriminated OCD from non-clinical controls at AUC 0.91, slightly better than the parent OCI-R (AUC 0.88).

    Internal consistency (alpha) ranged 0.71-0.89 across samples and subscales. The OCI-12 correlated with its parent OCI-R at r = 0.92-0.97. Twelve-week test-retest ICC was 0.85 (n = 212, students). Treatment sensitivity was comparable to the Y-BOCS: across exposure-based treatment the effect size was d = 1.89, and OCI-12 change correlated at r = 0.42 with Y-BOCS change.

    83%
    SENSITIVITY (CUTOFF >=11, OCD VS CONTROLS)
    81%
    SPECIFICITY (CUTOFF >=11, OCD VS CONTROLS)
    0.85
    TEST-RETEST ICC (12 WK, n=212)
    0.91
    AUC (OCD VS NON-CLINICAL)

    05 - How it compares

    How it compares to the alternatives.

    Instrument
    Items
    Time
    When to reach for it
    OCI-12
    12
    ~3 min
    12-item self-report, 0-48 across four 3-item subscales (checking, ordering, washing, obsessing). Copyright-unverified; not scorable here.
    18
    ~5 min
    18 items, ~5 min. The parent measure; includes hoarding and neutralising, so totals can be inflated relative to DSM-5 OCD. Scorable in this library.
    10
    ~30 min
    10 clinician-rated items, ~30 min with checklist. The severity gold standard for treatment trials; scorable in this library.
    20
    ~5 min
    20 items, ~5 min. Dimensional severity regardless of symptom theme; copyright-restricted (see its page).
    10
    ~20 min
    Clinician-rated, child/adolescent. The paediatric equivalent of the Y-BOCS.
    OCI-4
    4
    ~1 min
    4 items, ~1 min. Ultra-brief progress-monitoring offshoot; same item-copyright position.

    06 - When to use it

    Right tool, wrong tool.

    The OCI-12 is a severity and progress-monitoring instrument for confirmed OCD, not a diagnostic one. Its items are drawn from the copyrighted OCI-R, and permission to reproduce them has not been verified, so it cannot be self-scored here. Use the Y-BOCS in this library for a scorable OCD severity measure, and the DOCS page for a further alternative.

    Reach for it when

    • -Brief repeated progress monitoring in confirmed OCD where a licence or permission to reproduce items is in place.
    • -Research on the four core symptom dimensions (checking, ordering, washing, obsessing).
    • -When hoarding contamination of totals must be avoided, because the OCI-12 excludes it by design.

    Reach for something else when

    • -Diagnosis: a clinical interview is required; the OCI-12 screens and monitors severity.
    • -Clinician-rated severity for treatment trials: use the Y-BOCS.
    • -Hoarding assessment: the OCI-12 deliberately measures none.
    • -Any setting without verified permission to reproduce the items - use the Y-BOCS here instead.

    07 - Confidence & precision

    Reading the score with care.

    Twelve-week test-retest ICC was 0.85 in students; the moderate underlying stability (r = 0.74) means single-point changes are noise. As a rough guide, the difference between the mild and moderate band midpoints (~8 points) reflects a clearly meaningful shift; no formal MCID has been published yet.

    08 - Limitations

    What it cannot tell you.

    No published MCID. The cutoff of 11 overlaps the 'mild' severity band (0-12), which can confuse interpretation. The ordering subscale discriminates poorly (AUC 0.60). Norms are US clinic and student samples. Self-report with no insight assessment. Item copyright unverified, so it cannot be self-scored here.

    09 - Licensing, explained

    How licensing works.

    The OCI-12's items are drawn directly from the OCI-R (Foa et al. 2002); we have not verified separate permission to reproduce them in this shortened form, so we do not host the questionnaire yet. The parent OCI-R is scorable in this library, and the Y-BOCS offers clinician-rated severity.

    Common questions from clinicians

    FAQ

    Scale without compromise

    See how Aisel removes friction where it costs most. A 20-minute walkthrough tailored to your clinic.