Research Paper - Journal of Applied Mathematics and Statistical Applications (2018) Volume 1, Issue 2

Reliability and therapeutic decision through generalizability theory: An application in prostate cancer treatment

Introduction: Now-a-days there are several therapeutic techniques in hospitals and multiple factors that can modify the measurements for treatments. Health professionals should have objective information, in terms of reliability for therapeutic decisions. In our area exist different methods for the treatment of the prostate cancer via external radiotherapy. For each of them there are some factors that can affect the data collected to apply the treatment. The aim of this study is to use statistical advanced techniques, the Generalizability Theory, to evaluate the reliability for three image-guided radiotherapy methods to treat cancer with external radiotherapy: Electronic Portal Imaging (EPI), Cone Beam by Fiducial Markers (CBFM) and Cone Beam by Soft Tissues (CBST).

Methods: Forty patients with prostate cancer were enrolled in a prospective study. Before each daily session, EPI, CBFM and CBST images were sequentially acquired for eleven days in three positions: lateral, vertical and longitudinal. Generalizability Theory is used to analyze reliability and estimate other situations for radiotherapy application.

Results: Generalizability Theory shows high reliability for each method, one by one and among each other. We obtain high reliability also for each position but not two-by-two positions. Using only one method we obtain 0.9 reliability or more, from fifteen sessions.

Discussion: Generalizability Theory is a powerful statistic methodological tool that allows obtaining reliability coefficients in many different situations to help health professionals in therapeutic decisions.

Author(s): Carolina Lagares-Franco, Ma del Carmen Salas-Buzon, Lucia Gutierrez-Bayard, Santiago de los Reyes-Vazquez, Juan-Luis Gonzalez-Caballero, Ilaria Montagni, Jose Almenara-Barrios

Abstract Full Text PDF