What are we missing? The costs versus benefits of skip rule designs

Authors

  • Sonja A. Swanson,

    1. Department of Epidemiology, Harvard School of Public Health, Boston, MA, USA
    Search for more papers by this author
  • Tiffany A. Brown,

    1. Department of Psychology, Florida State University, Tallahassee, FL, USA
    Search for more papers by this author
  • Ross D. Crosby,

    1. Department of Clinical Neuroscience, University of North Dakota School of Medicine and Health Sciences, Fargo, ND, USA
    2. Neuropsychiatric Research Institute, Fargo, ND, USA
    Search for more papers by this author
  • Pamela K. Keel

    Corresponding author
    1. Department of Psychology, Florida State University, Tallahassee, FL, USA
    • Correspondence: Pamela K. Keel, Department of Psychology, Florida State University, 1107 W. Call Street, Tallahassee, FL 32306, USA.

      Telephone (+1) 850-645-9140

      Fax (+1) 850-644-7739

      Email: keel@psy.fsu.edu

    Search for more papers by this author

Abstract

Many research diagnostic interviews employ skip rules, such that some questions are only asked based on answers to prior questions. In the context of large-scale epidemiological studies, skip rules are important to study feasibility by reducing the time, money, and participant burden required for assessment. However, less is understood about information lost when questions are skipped. This study examines the relative prevalence, clinical significance, and additional time required to assess eating disorder symptom patterns skipped in the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I) to understand the costs and benefits of following skip rules. Data come from the second stage of a two-stage cohort sample (N = 400) in which the SCID-I eating disorders module was administered without following skip rules. Results were weighted to correct for the sampling framework. Over a third of subjects endorsed symptoms that would have been missed had skip rules been followed. Uncaptured symptom patterns were associated with increased psychosocial impairment, and the additional time required to assess all symptoms averaged 1.8 minutes per participant. Clinically significant symptom patterns are missed by the SCID-I and similar diagnostic tools, suggesting that epidemiologic studies using such instruments under-estimate the prevalence and public health impact of mental disorders. Copyright © 2013 John Wiley & Sons, Ltd.

Ancillary