Abstract
Abstract Introduction: There is a growing interest in understanding the concerns of patients undergoing cancer therapy. Many patients share their experiences on online support forums, which contain millions of freely shared messages that can be used to analyze patient concerns. Unfortunately, these data are unstructured, which makes them difficult to analyze. In this project we organize the data on these forums using methods from Big Data Science (BDS), and then analyze these data by creating a Decision Support System (DSS): an interactive interface that can be used by both patients and providers to understand patient concerns about their cancer therapies. Method: We collected approximately 10 million unique messages from 20 unrestricted breast cancer forums that provide information about diagnoses, treatments, side effects, supportive therapies, and specific experiences. After using domain knowledge of breast cancer to build custom ontologies for regimens, side effects, and supportive therapy, we use the following techniques from BDS in order to create our DSS: • Topic Modeling to find keywords that best represent a given theme • Information Retrieval to filter for messages that are related to this theme • Natural Language Processing to extract the relevant data from these messages • Token Windows and Co-occurrence-based Algorithms to associate regimens with their side effects and supportive therapies. To use the DSS, a user provides disease-related parameters and the treatment. The DSS then gives the percentage of messages discussing side effects for a similar cohort of patients and the percentage of messages that discuss supportive therapies for each of these side effects. Results: We retrieved 84938 messages from patients receiving adjuvant chemotherapy with the regimens listed below, and then analyzed the percentage of people mentioning each side effect. The results are summarized in the following table. Side-Effect DataSide-EffectAC/T, %C/T, %TCH, %AC, %AC/Taxotere, %FEC/Taxotere, %Pain56.651.45051.55756.4Neuropathy6.44.66.17.54.55.7Nausea12.611.213.414.51014.6Alopecia2.53.63.22.622.5Nail changes6.87.19.375.56.4Heart related17.820.4171817.416.4Fatigue8.69.88.610.57.99.2Swelling55.77.256.66.8Rash itching7.39.68.998.17.6Total count2109628788136971105872193080 These statistics reflect patient concern about a particular side effect, instead of the true incidence of that side effect. For example, although about 90% of patients receiving the above regimens experience alopecia, only between 2% and 4% of messages on online forums mention alopecia. Our system can also associate drugs to their side effects and suggest supportive therapies. For instance, 5428 (20.6%) of 26370 messages on Neulasta, mention it as the cause of bone pain. Out of these, 1262 (23.2%) mention Loratidine in context with Neulasta as a suggestion to alleviate bone pain (p < 0.00001). Conclusion: Using methods from BDS, our DSS reliably associates side effects to a particular drug or regimen and suggests a supportive therapy. Our results reflect the concerns of patients undergoing cancer therapy, which might help the medical community identify areas of resource allocation and unmet needs. Citation Format: Aggarwal S, Liu M, Sharma R, Gupta A, Singh D, Sharma R, Yang F, Basak A, Aggarwal A. Voice of cancer patients: Analysis of concerns of patients receiving adjuvant chemotherapy for breast cancer. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P1-10-21.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.