NIMH
ID: 1R43MH108178-01
PI: MELISSA DEROSIER, OLIVER LINDHIEM
TERM: 08/15 – 01/16
Unlike most medical conditions, mental health disorders have no definitive diagnostic or screening tools (e.g., blood tests, x-rays) to verify their presence or absence. In primary care settings, pediatricians and family medicine doctors often lack confidence in their ability to screen for significant mental health problems such as pediatric bipolar disorder, and refer out for most mental and behavioral health concerns. Unfortunately, many patients either do not follow through on referrals or do not receive appropriate referrals. As a result, both under-referrals and over-referrals are all too common in primary care settings. Over-referrals cause inefficiency while under-referrals result in prolonged patient suffering, burden on family members, and a huge cost to society.
The Posterior Probability of Diagnosis (PPOD) Index has recently been proposed as a unique step towards solving these problems. The PPOD Index is a Bayesian method that quantifies the probability that a patient meets or exceeds a diagnostic threshold for a particular disorder. The method has two advantages over traditional screening approaches. First, the PPOD Index is based on a patient’s individual pattern of symptoms rather than on traditional symptom counts. Second, the method quantifies the likelihood that a patient meets criteria for a particular diagnosis in probabilistic terms (0-100%).
The goal of this SBIR project is to develop and test a Bayesian screening tool using PPOD Index software for tablets. In Phase I, we will focus prototype development and testing on pediatric (ages 5-17) bipolar disorder, a disorder that is particularly likely to be misdiagnosed, resulting in tremendous costs for the person and society. The proposed software system will enhance screening to improve the precision and quality of referrals for pediatric bipolar disorder in primary-care settings. The following three specific aims will be accomplished during Phase I: 1) create a fully functioning working prototype of a Bayesian screening tool for pediatric bipolar disorder using PPOD Index software for tablets to run on a tablet PC, 2) perform usability testing with physicians and families, and 3) assess the product value, innovation, feasibility, acceptability, and quality.
The technology will have far-reaching generalizability to a variety of patient populations. We envision a final product (Phase II) that will include modules for other mental health disorders such as ADHD, Autism, and PTSD. By improving the efficiency and screening of mental health problems in primary-care settings, we can significantly improve the mental health outcomes of children and their families, alleviating suffering, burden on families, and cost to society.
