Computer-adapted testing (CAT) allows for selection of only the most sensitive and discriminating items to be used to assess presence and severity of a wide range of health and mental health domains, thus greatly facilitating the assessment process, and at the same time reducing the burden of assessment by including the smallest possible number of items. Through CAT it is possible to obtain highly reliable and valid assessments with as few as 6 -8 questions. Paper and pencil questionnaires, some of which contain up to 100 items, and structured interviews, are far lengthier and not necessarily more accurate. With increasing numbers of military personnel returning from deployment in Iraq and Afghanistan, it is valuable and cost-effective to employ the most efficient and effective means possible to assess the health and mental health status of returning veterans.
The primary objectives of this project are 1) to conduct the developmental work needed to create a computer-adapted test (CAT) to screen for, assess severity of and monitor course of illness and treatment outcomes for PTSD; and 2) to lay the groundwork and obtain the preliminary data needed to prepare an investigator-initiated research proposal (IIR) to complete development of a PTSD CAT.
Using systematic review methodology, the investigators searched Medline, PsycInfo and PILOTS from the National Center for PTSD using the following MeSH terms: screen*, instrument*, Stress Disorders/Post-Traumatic/, PTSD/, valid*, diagnosis?*. Studies were included if they were published between 1980 and 2008, involved adults, used an English language self-report PTSD assessment instrument and reliability and validity data were published,. Kappas were calculated on a random sample of 20 studies to ensure inter-rater reliability between two investigators. Research team members categorized items into subdomains, revised poorly worded or confusing items and eliminated items that were fully redundant with other items to reduce the potential item pool to a manageable number of items. The subdomains constituted a conceptual framework for PTSD. The Lexile Framework was used to determine the reading level of each item and items above an 8th grade reading level were revised so that no items exceeded an 8th grade reading level. Seven cognitive interviews were then conducted with veterans who had a history of PTSD to determine whether items were clearly understood and viewed as important to the assessment of PTSD.
865 abstracts representing 480 unique studies were identified. Kappas for each inclusion criterion ranged from 0.86 to 1.0. 275 studies representing 41 PTSD instruments and 993 items met inclusion criteria. Item evaluation based on consensus of 3 research team members eliminated 872 items, leaving 121 items for cognitive testing. Using Lexile analysis items with higher than an 8th grade reading level were revised. Analysis of the cognitive interviews is currently in progress
This developmental work will contribute to facilitation and enhancement of VAs capacity to effectively and efficiently screen for and assess severity of PTSD. The proposed work will complement and supplement current work funded by an NIH cooperative agreement to develop computer-adapted tests for a range of other health conditions including pain, fatigue, social functioning and emotional distress. Computer-adapted health assessments will be transferable to existing VA initiatives such as MyHealtheVet.
External Links for this Project
None at this time.