We propose to use Item Response Theory (IRT), the mathematical study of computerized adaptive testing, to construct a brief, highly branched mental health screening and outcomes tracking instrument. This instrument will use electronic survey technology to administer surveys while continuously adapting their content and structure in real time, allowing patients to see only the most individually relevant questions. The final Phase II product will be the first IRT-based instrument in mental health, and will screen and track patients with depression, anxiety, alcohol or substance abuse, and coexisting impairment in social or occupational function. It is intended for use by clinicians and researchers within primary care and mental health settings. It will be highly sensitive to change; have a low floor and high ceiling; use a constant 5-point Likert scale; and screen for six domains in under eight minutes. In Phase I, we will assess the feasibility of using IRT to create the instrument by pilot testing a long unbranched version on 300 mental health patients of varying symptom severity using touch-screen PCs at two Kaiser Permanente outpatient clinics. We hope the increased efficiency in tracking outcomes will help improve individual patient care and help researchers improve the effectiveness of healthcare interventions.
The final Phase II product will be a brief (8-minute) IRT-based mental health screening and outcomes-tracking instrument for use by clinicians and mental health researchers within primary care and mental health settings. It will be highly sensitive to changes with a high ceiling and low floor. The instrument will contain four separable symptom domains and 2 separable function domains. Its increased efficiency in tracking outcomes will help clinicians and mental health researchers improve the effectiveness of healthcare interventions.
