Predicting financial toxicity in cancer patients using machine learning: a Twitter or X-based approach
DOI:
https://doi.org/10.18203/2394-6040.ijcmph20241164Keywords:
Cancer, Financial toxicity, Machine learning, Naïve bayes, Synthetic data, TwitterAbstract
Background: Financial strain resulting from cancer treatment correlates with reduced quality of life, treatment nonadherence, bankruptcy, and maladaptive behaviours. This study aims to explore the potential of a supervised machine learning algorithm in predicting financial toxicity in cancer patients based on their Tweets.
Methods: A dataset of Tweets related to cancer and financial toxicity was constructed using Twitter's API. The dataset was curated, and synthetic Tweets were generated to augment the final dataset. A supervised machine learning algorithm, specifically Multinomial Naïve Bayes, was trained and tested to predict financial toxicity in cancer patients.
Results: The model demonstrated high accuracy (0.97), precision (0.95), recall (0.99), specificity (0.96), F-1 score (0.97) and area-under-the-receiver-operating-characteristics (0.98) in predicting financial toxicity from Tweets. Wordcloud visualizations illustrated distinct linguistic patterns between Tweets related to financial toxicity and those unrelated to financial toxicity. The study also outlined potential proactive strategies for leveraging social media platforms like Twitter to identify and support cancer patients experiencing financial toxicity.
Conclusions: This study marks the first attempt to construct a dataset of Tweets related to financial toxicity in cancer patients and to evaluate a predictive model trained on this dataset. The findings highlight the predictive capabilities of the model and its potential utility in guiding health systems and cancer center financial navigators to alleviate economic burdens associated with cancer treatment.
Metrics
References
Yadav J, Menon GR, John D. Disease-specific out-of-pocket payments, catastrophic health expenditure and impoverishment effects in India: an analysis of national health survey data. Appl Heal Econom Health Policy. 2021;19:769-82.
Mariotto AB, Enewold L, Zhao J, Zeruto CA, Yabroff KR. Medical care costs associated with cancer survivorship in the United States. Cancer Epidemiol Biomar Prevent. 2020;29(7):1304-12.
Gilligan AM, Alberts DS, Roe DJ, Skrepnek GH. Death or debt? National estimates of financial toxicity in persons with newly-diagnosed cancer. Ame J Med. 2018;131(10):1187-99.
Banegas MP, Guy GP, de Moor JS, Ekwueme DU, Virgo KS, Kent EE, et al. For working-age cancer survivors, medical debt and bankruptcy create financial hardships. Health Affairs. 2016;35(1):54-61.
Zafar SY, Peppercorn JM, Schrag D, Taylor DH, Goetzinger AM, Zhong X, et al. The financial toxicity of cancer treatment: a pilot study assessing out‐of‐pocket expenses and the insured cancer patient's experience. Oncolog. 2013;18(4):381-90.
Bestvina CM, Zullig LL, Rushing C, Chino F, Samsa GP, Altomare I, et al. Patient-oncologist cost communication, financial distress, and medication adherence. J Oncol Pract. 2014;10(3):162-7.
Dowling EC, Chawla N, Forsythe LP, de Moor J, McNeel T, Rozjabek HM, et al. Lost productivity and burden of illness in cancer survivors with and without other chronic conditions. Cancer. 2013;119(18):3393-401.
Ramsey S, Blough D, Kirchhoff A, Kreizenbeck K, Fedorenko C, Snell K, et al. Washington State cancer patients found to be at greater risk for bankruptcy than people without a cancer diagnosis. Health affairs. 2013;32(6):1143-52.
Ramsey SD, Bansal A, Fedorenko CR, Blough DK, Overstreet KA, Shankaran V, et al. Financial insolvency as a risk factor for early mortality among patients with cancer. J Clin Oncol. 2016;34(9):980.
Gordon LG, Merollini KM, Lowe A, Chan RJ. A systematic review of financial toxicity among cancer survivors: we can’t pay the co-pay. Patient-Patient-Cent Outc Res. 2017;10:295-309.
Slevin ML, Nichols SE, Downer SM, Wilson P, Lister TA, Arnott S, et al. Emotional support for cancer patients: what do patients really want?. Brit J Can. 1996;74(8):1275-9.
Murthy D, Gross A, Longwell S. Twitter and e-health: A case study of visualizing cancer networks on Twitter. In: Proceedings of Information Society (i-Society). International Conference. London, UK; 2011.
Sugawara Y, Narimatsu H, Hozawa A, Shao L, Otani K, Fukao A. Cancer patients on Twitter: a novel patient community on social media. BMC Res Not. 2012;5:1-9.
Myrick JG, Holton AE, Himelboim I, Love B. # Stupidcancer: exploring a typology of social support and the role of emotional expression in a social media community. Health Communi. 2016;31(5):596-605.
Vermeulen A, Vandebosch H, Heirman W. # Smiling,# venting, or both? Adolescents’ social sharing of emotions on social media. Comput Hum Behav. 2018;84:211-9.
Wang J, Wei L. Fear and hope, bitter and sweet: Emotion sharing of cancer community on twitter. Soci Med Soci. 2020;6(1):2056305119897319.
Paul M, Dredze M. You are what you tweet: Analyzing twitter for public health. InProceedings of the International AAAI Conference on Web and Social Media 2011;5(1):265-72.
Padrez KA, Ungar L, Schwartz HA, Smith RJ, Hill S, Antanavicius T, et al. Linking social media and medical record data: a study of adults presenting to an academic, urban emergency department. BMJ Qual Safe. 2016;25(6):414-23.
Fisher J, Clayton M. Who gives a tweet: assessing patients’ interest in the use of social media for health care. World Evidence‐Based Nurs. 2012;9(2):100-8.
Huesch MD, Currid-Halkett E, Doctor JN. Public hospital quality report awareness: evidence from National and Californian Internet searches and social media mentions, 2012. BMJ open. 2014;4(3):e004417.
Wong CA, Sap M, Schwartz A, Town R, Baker T, Ungar L, Merchant RM. Twitter sentiment predicts Affordable Care Act marketplace enrollment. J Med Inter Res. 2015;17(2):e51.
Wallace BC, Paul MJ, Sarkar U, Trikalinos TA, Dredze M. A large-scale quantitative analysis of latent factors and sentiment in online doctor reviews. J Ame Medi Informat Associat. 2014;21(6):1098-103.
Greaves F, Laverty AA, Cano DR, Moilanen K, Pulman S, Darzi A, et al. Tweets about hospital quality: a mixed methods study. BMJ Qual Safe. 2014;23(10):838-46.
Sidey-Gibbons C, Pfob A, Asaad M, Boukovalas S, Lin YL, Selber JC, et al. Development of machine learning algorithms for the prediction of financial toxicity in localized breast cancer following surgical treatment. JCO. 2021;5(5):338-47.
Liu Y, Chen PH, Krause J, Peng L. How to read articles that use machine learning: users’ guides to the medical literature. Jama. 2019;322(18):1806-16.
Alpaydin E. Introduction to Machine Learning. 4th ed. Cambridge, MA, The MIT Press;2020.
Sidey-Gibbons JA, Sidey-Gibbons CJ. Machine learning in medicine: a practical introduction. BMC Med Res Methodol. 2019;19:1-8.
Gilardi F, Alizadeh M, Kubli M. ChatGPT outperforms crowd workers for text-annotation tasks. Proceed Nat Acad Sci. 2023;120(30):e2305016120.
Josifoski M, Sakota M, Peyrard M, West R. Exploiting asymmetry for synthetic training data generation: Synthie and the case of information extraction. arXiv preprint arXiv:2303.04132; 2023.
Ziems C, Held W, Shaikh O, Chen J, Zhang Z, Yang D. Can large language models transform computational social science?. Computat Linguist. 2024:1-55.
Abaskohi A, Rasouli A, Zeraati T, Bahrak B. Utnlp at semeval-2022 task 6: A comparative analysis of sarcasm detection using generative-based and mutation-based data augmentation. arXiv preprint arXiv:2204.08198; 2022.
Sennrich R, Haddow B, Birch A. Improving neural machine translation models with monolingual data. arXiv preprint arXiv:1511.06709; 2015.
Maqsud U. Synthetic text generation for sentiment analysis. InProceedings of the 6th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis 2015; pp. 156-161.
Feng SY, Gangal V, Wei J, Chandar S, Vosoughi S, Mitamura T, et al. A survey of data augmentation approaches for NLP. arXiv preprint arXiv:2105.03075. 2021 May 7.
Patki, N.; Wedge, R.; Veeramachaneni, K. The Synthetic Data Vault. 2016 IEEE International Conference on Data Science and Advanced Analytics (DSAA), Montreal, QC, Canada. 2016:399-410.
De Souza JA, Yap BJ, Wroblewski K, Blinder V, Araújo FS, Hlubocky FJ, et al. Measuring financial toxicity as a clinically relevant patient‐reported outcome: the validation of the COmprehensive Score for financial Toxicity (COST). Cancer. 2017;123(3):476-84.
De Souza JA, Yap BJ, Hlubocky FJ, Wroblewski K, Ratain MJ, Cella D, et al. The development of a financial toxicity patient‐reported outcome in cancer: the COST measure. Cancer. 2014;120(20):3245-53.
Williams C, Meyer L, Kawam O, Leventakos K, DeMartino ES. The faces of financial toxicity: a qualitative interview study of financial toxicity in advanced cancer patients in phase i oncology trials. Mayo Cli Proceed Innovat Qual Out. 2023;7(6):524-33.
Chakrabarti S, Roy S, Soundalgekar MV. Fast and accurate text classification via multiple linear discriminant projections. VLDB J. 2003;12:170-85.
Ting SL, Ip WH, Tsang AH. Is Naive Bayes a good classifier for document classification. Int J Softw Engin Applicat. 2011;5(3):37-46.
Gamallo P, Garcia M. Citius: A naive-bayes strategy for sentiment analysis on english tweets. InProceedings of the 8th international Workshop on Semantic Evaluation (SemEval 2014). 2014; 171-175.
Wang SI, Manning CD. Baselines and bigrams: Simple, good sentiment and topic classification. InProceedings of the 50th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers). 2012; 90-94.
Gilligan T, Coyle N, Frankel RM, Berry DL, Bohlke K, Epstein RM, et al. Patient-clinician communication: American Society of Clinical Oncology consensus guideline. Obstet Gynecol Surv. 2018;73(2):96-7.
Downloads
Published
How to Cite
Issue
Section
