Automated Survival Prediction in Metastatic Cancer Patients Using High-Dimensional Electronic Medical Record Data

Abstract

Background: Oncologists use patients' life expectancy to guide decisions and may benefit from a tool that accurately predicts prognosis. Existing prognostic models generally use only a few predictor variables. We used an electronic medical record dataset to train a prognostic model for patients with metastatic cancer. Methods: The model was trained and tested using 12 588 patients treated for metastatic cancer in the Stanford Health Care system from 2008 to 2017. Data sources included provider note text, labs, vital signs, procedures, medication orders, and diagnosis codes. Patients were divided randomly into a training set used to fit the model coefficients and a test set used to evaluate model performance (80%/20% split). A regularized Cox model with 4126 predictor variables was used. A landmarking approach was used due to the multiple observations per patient, with t 0 set to the time of metastatic cancer diagnosis. Performance was also evaluated using 399 palliative radiation courses in test set patients. Results: The C-index for overall survival was 0.786 in the test set (averaged across landmark times). For palliative radiation courses, the C-index was 0.745 (95% confidence interval [CI] ¼ 0.715 to 0.775) compared with 0.635 (95% CI ¼ 0.601 to 0.669) for a published model using performance status, primary tumor site, and treated site (two-sided P < .001). Our model's predictions were well-calibrated. Conclusions: The model showed high predictive performance, which will need to be validated using external data. Because it is fully automated, the model can be used to examine providers' practice patterns and could be deployed in a decision support tool to help improve quality of care.

SOURCE: A Registry-Based Prediction Model for Overall Survival in Patients with Metastatic Oesophageal or Gastric Cancer

“…Indeed, the SOURCE model showed a fairly discriminative ability, with a c-index of approximately 0.71 for oesophageal cancer and 0.68 for gastric cancer. Although certain other models were able to discriminate better between patients, it must be noted that our dataset was relatively homogenous, including patients with metastatic oesophagogastric cancer only [17,18]. Differentiating between survival outcomes of a rather homogeneous group of patients is more complex than differentiating between survival outcomes of patients with cancers from various primary origins and known large differences in survival.…”

Breast Cancer Prognosis Using a Machine Learning Approach

“…As reported above, ML has started to take hold across the oncology community to develop prognostic classifications models of BC progression and survivability [9]. In this regard, the possibility to perform an automated survival prediction in metastatic cancer patients using high-dimensional electronic health records (EHR) data has been recently highlighted [23]. By using an ML approach on EHR-derived predictor variables (clustered into categories), Gensheimer et al, in fact, devised an AI system, with a better c-statistic than previously reported prognostic models, which could be deployed in a DSS to help improve quality of care in the metastatic setting [23].…”

“…In this regard, the possibility to perform an automated survival prediction in metastatic cancer patients using high-dimensional electronic health records (EHR) data has been recently highlighted [23]. By using an ML approach on EHR-derived predictor variables (clustered into categories), Gensheimer et al, in fact, devised an AI system, with a better c-statistic than previously reported prognostic models, which could be deployed in a DSS to help improve quality of care in the metastatic setting [23]. More recently, four major nonlinear ML methods (integrating multiple clinicopathological features and genomic data) were used to compare survival predictions in a large cohort of BC patients [24].…”

Natural language processing for abstraction of cancer treatment toxicities: accuracy versus human experts

“… 20 Separate from semantically extracting information from notes, other recent studies have focused on the use of aggregate data for outcome prediction. 8 , 21 …”

“…Accurate extraction of clinical elements from the free text may offer refined and rational features for predictive models, building on studies where aggregate text provided utility in predicting clinical outcomes. 8 , 21 Our team recently completed one of the first prospective, randomized studies of machine learning, utilizing EHR data to generate accurate predictions of acute care, and direct supportive care. 12 NLP offers an additional source of insights from routine clinical data that may augment its performance for clinical decision support.…”