There is increasing pressure for integration of the patient’s perspective in clinical research. An indication of this growth is the establishment of the Patient-Centered Outcomes Research Institute (PCORI). Patient-reported outcomes (PROs) offer the patient’s perspective. There is ample evidence for not relying solely on clinician reporting of patients’ subjective experiences. There is also increasing evidence that PROs can provide more information about toxicity and symptoms than physician or adverse event (AE) reports. These findings support the need for tools to collect patient-reported symptoms and Health-Related Quality of Life (HRQL). The need to reduce the number of items administered in measuring PROs has long been a concern in health outcomes research. This concern has led researchers toward emerging theories such as item response theory (IRT) and computer adaptive testing (CAT). By only administering questions targeted to the subject’s trait level, CAT models can cut testing times by half while improving overall test reliability. Patients can complete a PRO CAT in an average of five questions with reliability similar to a typical 15–25-item survey measure. With this lower level of overall burden, it becomes much easier to administer more comprehensive assessments to patients. Funded by the National Institutes of Health (NIH), the Patient Reported Outcomes Measurement Information System (PROMIS) initiative developed a broad array of high-quality, state-of-the-art PRO measures using IRT/CAT. The Department of Medical Social Sciences at Northwestern University developed Assessment Center, which is the software that administers PROMIS CATs. Assessment Center is a web-based research management software system that allows researchers to create and administer PRO studies. PROMIS is unidimensional in that there is a separate item bank and therefore a separate set of questions administered for each measured trait. This has been a good initial step in addressing the issue of patient burden, but there is a distinct and achievable avenue through which we can substantially reduce this burden further. That is, there are correlations between related traits. For example, depression and anxiety are closely related, which means many questions that ascertain level of depression can also ascertain level of anxiety. A multidimensional CAT (MCAT) selects items from an item pool such that the items selected maximize information provided on several correlated traits. Such a model would significantly reduce the number of items administered in measuring PROs and thereby further ameliorate the patient’s burden. We propose to develop and test MCAT using existing PROMIS item pools. We will show that MCAT is an improvement over fixed-time measures and unidimensional CAT. We will use an API to distribute the MCAT algorithm and item banks developed under this project. Being situated at Northwestern University, the home of the PROMIS project, and having on our team Michael Bass, the lead software developer for Assessment Center, and Seung Choi, the principal developer of the PROMIS CAT engine, we are in a unique position to conduct this research.
|Effective start/end date||6/15/15 → 6/14/19|
- National Library of Medicine (5R01LM011962-04)
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.