We propose to apply item response theory (IRT) and computerized adaptive testing (CAT) to achieve major advances in the standardization of Chronic Kidney Disease (CKD) health-related quality of life (HRQOL) metrics, in the efficiency of data collection and processing technologies, and in the usefulness of clinician and patient reports. In contrast to static instruments, which ask all respondents the same questions regardless of their answers and lack the precision to detect changes over a wide range of severity levels, we will use Dynamic Health Assessment (DYNHA?) software to match questions to each patient's score and to the requirements of each application to achieve more practical and more useful assessments. In our Phase I test among hemodialysis patients, a prototype CKD-CAT application achieved comparable CAT scores and equivalent or better precision with substantial reductions in respondent burden compared to three full-length static scales (Effects of Kidney Disease, Sleep Functioning, and Bodily Pain) including items from the Kidney Disease Quality Of Life Questionnaire (KDQOL). Shortcomings of our approach were also identified: (a) the need for additional questionnaire item development, (b) limitations of sampling only hemodialysis patients to represent a wide range of CKD severity levels, and (c) problems with the interface between patients and data collection technologies. To address these concerns as well as meet the needs for assessment tools and software that are applicable across a wider range of disease severity levels, we propose a Phase II effort that matches samples and intensity/expense of data collection to the distinct requirements of each of the following aims: (1) Expand the item pool and evaluate the appropriateness of new items for CKD-specific applications; (2) Identify the most useful and feasible domains and calibrate their items by collecting and analyzing responses from large samples of adults varying across stages of CKD (Stages 3-5); (3) Create a fully operational CKD-CAT system and evaluate its performance through """"""""real data"""""""" simulation studies, to determine reliability and clinical validity and the potential for reductions in respondent burden, in comparison with full-length forms; and (4) Confirm the administrative feasibility, patient acceptance, reliability and validity of scores, and the usefulness of clinician and patient reports from actual CKD-CAT administrations to patients (CKD Stages 3-5) sampled and assessed in clinical and in non-clinical settings. By greatly shortening surveys and lowering the costs of administering them, it will be possible to monitor the quality of life effects of CKD more accurately for more patients and help clinicians and patients better understand and communicate about treatment success. ? ? ?
| Deng, Nina; Allison, Jeroan J; Fang, Hua Julia et al. (2013) Using the bootstrap to establish statistical significance for relative validity comparisons among patient-reported outcome measures. Health Qual Life Outcomes 11:89 |
