Predicting the Survival of Patients With Cancer From Their Initial Oncology Consultation Document Using Natural Language Processing

Abstract

ImportancePredicting short- and long-term survival of patients with cancer may improve their care. Prior predictive models either use data with limited availability or predict the outcome of only 1 type of cancer.ObjectiveTo investigate whether natural language processing can predict survival of patients with general cancer from a patient’s initial oncologist consultation document.Design, Setting, and ParticipantsThis retrospective prognostic study used data from 47 625 of 59 800 patients who started cancer care at any of the 6 BC Cancer sites located in the province of British Columbia between April 1, 2011, and December 31, 2016. Mortality data were updated until April 6, 2022, and data were analyzed from update until September 30, 2022. All patients with a medical or radiation oncologist consultation document generated within 180 days of diagnosis were included; patients seen for multiple cancers were excluded.ExposuresInitial oncologist consultation documents were analyzed using traditional and neural language models.Main Outcomes and MeasuresThe primary outcome was the performance of the predictive models, including balanced accuracy and receiver operating characteristics area under the curve (AUC). The secondary outcome was investigating what words the models used.ResultsOf the 47 625 patients in the sample, 25 428 (53.4%) were female and 22 197 (46.6%) were male, with a mean (SD) age of 64.9 (13.7) years. A total of 41 447 patients (87.0%) survived 6 months, 31 143 (65.4%) survived 36 months, and 27 880 (58.5%) survived 60 months, calculated from their initial oncologist consultation. The best models achieved a balanced accuracy of 0.856 (AUC, 0.928) for predicting 6-month survival, 0.842 (AUC, 0.918) for 36-month survival, and 0.837 (AUC, 0.918) for 60-month survival, on a holdout test set. Differences in what words were important for predicting 6- vs 60-month survival were found.Conclusions and RelevanceThese findings suggest that models performed comparably with or better than previous models predicting cancer survival and that they may be able to predict survival using readily available data without focusing on 1 cancer type.

Predicting which patients with cancer will see a psychiatrist or counsellor from their initial oncology consultation document using natural language processing

“…Our patient selection was the same as in prior work 55 . Of the 59,800 BC Cancer patients, we excluded 2784 due to starting cancer care multiple times, and 9391 due to not having a medical or radiation oncology consultation within 180 days of their cancer diagnosis.…”

“…Two types of neural models, CNN and LSTM, outperformed the simpler BoW models. This suggests these predictions may benefit from a more complex understanding of language made possible by neural networks, in contrast to related work using similar data and techniques to predict the survival of patients with cancer survival 55 . While we could not find similar work to which we could compare these results, these metrics are comparable to or better than other applications of ML for predicting future events in psychiatry, such as predicting whether a patient’s depression will respond to an antidepressant 56 , 57 , whether someone will complete or attempt suicide 58 , or if a child will later develop a bipolar disorder 59 .…”

“…We did not find NLP literature predicting psychosocial outcomes from non-psychiatric medical documents, or find prior work using neural NLP to predict future psychiatric outcomes. There has been more NLP work in oncology 51 – 54 , including our recent work predicting survival from oncologist consultations 55 .…”

Predictive Modeling in Medicine

“…Upon conducting a prognostic investigation on 47,625 individuals with cancer, it was found that natural language processing can proficiently estimate their survival rates through traditional and neural models [36]. The outcomes were as good as or superior to previous research, signifying its viability for practical usage in predicting the endurance of patients with cancer.…”

Pseudo-grading of tumor subpopulations from single-cell transcriptomic data using Phenotype Algebra

“…A growing number of NLP-based methods have recently been gaining popularity for their ability to predict patients' survival based on their clinical records and for facilitating efficient analyses of highdimensional scRNA-seq data [16][17][18][19] . For instance, Nunez et al 16 employ language models to predict the survival outcomes of breast cancer patients based on their initial oncologist consultation documents. Similarly, in the domain of scRNA-seq analysis, scETM 18 uses topic models to infer biologically relevant cellular states from single-cell gene expression data.…”