Publication
ICWSM 2017
Conference paper

The power of the patient voice: Learning indicators of treatment adherence from an online breast cancer forum

Abstract

Social media platforms have become popular online environments for patients seeking and sharing treatment experiences. These platforms enable us to move beyond traditional sources of clinical information for learning about a patient's long-term adherence to treatment. While adherence has been studied using data derived from medical records and structured surveys, these approaches are limited in that they are often 1) time consuming, 2) limited in scale, or 3) lack selfreported patient experiences. In this paper, we investigate treatment adherence through a patient's self-reported information in online discussion forums. Specifically, we consider hormonal therapy treatment adherence (HTA) for hormone receptor positive breast cancer, a disease subtype that comprises 75% of all breast cancer cases. We focus on the inferred emotions and personality traits from the posts created by the members of a large online breast cancer community. These factors have been neglected in traditional adherence research due to a lack of information. We study over 130,000 posts from the forum, spanning 10,000 patients over 9 years. We assess emotion and personality traits with respect to three types of adherence behaviors: 1) currently on a regimen, 2) an interruption (due to discontinuing, pausing, or switching a medication regimen before five years) and 3) the completion of a five-year protocol. We find statistically significant differences in emotions across adherence behaviors. We further show that specific personality traits, including self-discipline, are associated with HTA, but in the opposite direction than what traditional research studies have shown. Finally, we illustrate that there is potential for predicting future interruption behaviors based on an individual's posts. We anticipate that our methodology can be applied to study treatment adherence for other diseases using online self-reported information.

Date

15 May 2017

Publication

ICWSM 2017

Authors

Share