Severe oral mucositis pain afflicts most patients who undergo bone marrow transplantation (BMT). Because the investigators can measure both oral mucositis severity and pain, and demonstrate a relationship between them, they have a unique opportunity to develop new pain measurement methodology. The investigators propose to develop competing multidimensional pain measurement methods derived from psychophysics and psychometrics and to compare these methods with their existing unidimensional method as well as with one another. The psychophysical approach identifies pain as a sensory message representing tissue trauma in consciousness. The power law approach identifies pain as a sensory message representing tissue trauma in consciousness. The power law defines the psychophysical relationship between measured oral mucositis severity and pain report. The psychometric approach defines pain as a multidimensional hypothetical construct quantifiable indirectly through statistical modeling. While the psychophysicist assumes that a patient's report is accurate, the psychometrician holds that such reports contain measurement error, some of which derives from potentially measurable sources. In pilot work both methods improve upon raw pain report. In a controlled study, each of two patient groups will use a different technology to report pain: a) the investigators' current noninteractive method (control condition); or b) a new interactive, multidimensional method. The investigators will derive pain scores in the second group in two ways: via psychophysics and psychometrics. They will form three clinical databases and maintain them separately (one is the control group, and the other two are alternative forms derived from the single patient sample in the second group). They will evaluate the methods by comparing features of their respective data bases. The outcome variables for this study are: prediction accuracy for pain scores (the best method will yield the most accurate predictions of future scores from past ones), variability over days in individual case oral pain measures, rate of censored observations in respective databases, physician time for patient consultation, and user satisfaction. It is hypothesized that: 1) the interactive, multidimensional computer tool will differ from the conventional pain interview tool in rate of censored observations, mean physician time for pain assessment interview, as well as patient and staff satisfaction; and 2) the psychophysical and psychometric approaches to pain scaling will differ in pain prediction accuracy and magnitude of day-to-day variation in patient pain report.
| Donaldson, Gary (2008) Patient-reported outcomes and the mandate of measurement. Qual Life Res 17:1303-13 |
| Chapman, C R; Gavrin, J (1999) Suffering: the contributions of persistent pain. Lancet 353:2233-7 |
