Scoring system for breakthrough-pain likelihood in cancer patients

Scoring system for breakthrough-pain likelihood in cancer patients Submission No: 2044 Submission Type: Poster Abstract Submission Aim of Investigation: Breakthrough pain (BTP) shows variable prevalence in different clinical contexts of cancer patients. BTP diagnostic tools with demonstrated reliability, validation and prognostic capability are lacking. We report the preliminary results of an ongoing impact study following the development of a diagnostic/prognostic tool, the IQ-BTP, for BTP recognition and its likelihood among cancer patients. Methods: A sample of 120 chronic pain patients was used as a learning set. An univariate analysis was conducted in order to select the variables associated with the variable of interest (likelihood level of BTP); these have been employed in the planning of a Naive Bayes classifier; a set of scores for each predictor was computed, by extending the approach usually used for binary classification to the case of three likelihood classes of BTP (High, Low, None). The developed BTP-likelihood scoring system, under the form of a questionnaire (IQ-BTP), recently validated, is now used in a multicenter impact study. The study was authorized by the hospital Ethics Committee, and conducted according to the Helsinki declaration and IASP's guidelines for pain research in animals and humans. Results: The scoring system analysis showed that two thirds of the patients may be correctly classified as potentially having (or not having) BTP; proportions of correct classification for BTP likelihood classes "High" and "No BTP" was greater than 80%: most of the patients satisfying the criteria for having BTP were correctly recognized. Conclusions: The IQ-BTP with its peculiar scoring system and with adequate feasibility may enable, in cancer patients, the detection of potential-BTP and its likelihood. The latter may have significant relevance to BTP epidemiology and management.