JCO Clinical Cancer Informatics最新文献

Purpose: Our study explores how attitudes of patients with cancer toward smartphone-based commercial health care apps affect their use and identifies the influencing factors.

Materials and methods: Of the 960 patients with cancer who participated in a randomized controlled trial for a smartphone-based commercial health care app, only 264 participants, who completed a survey on app usage experiences conducted between May and August 2022, were included in this study. Participants were categorized into three groups: Positive Persistence (PP), Negative Nonpersistence (NN), and Neutral (NE) on the basis of their attitude and willingness to use smartphone-based commercial health care apps. The Health-Related Quality of Life (QOL) Instrument (8 Items), European QOL (5 Dimensions; 5 Levels), The Human Interaction and Motivation questionnaire, and open-ended questionnaires were used to examine factors potentially influencing extended utilization of digital interventions.

Results: Despite demographic similarities among the three groups, only the PP and NE groups showed similar app usage compared with the NN group. The combined group (positive persistence and neutral) exhibited significant improvement in depression (P = .02), anxiety (P = .03), and visual analog scale scores (P = .02) compared with the NN group. In addition, patient interaction (P < .01) and the presence of a chatbot/information feature on the app (P < .01) demonstrated a significant difference across the three groups, with the most favorable response observed among the PP group. Patients were primarily motivated to use the app owing to its health management functions, while the personal challenges they encountered during app usage acted as deterrents.

Conclusion: These findings suggest that maintaining a non-negative attitude toward smartphone-based commercial health care apps could lead to an improvement in psychological distress. In addition, the social aspect of apps could contribute to extending patient's utilization of digital interventions.

Purpose: The magnitude of cardiorespiratory fitness (CRF) impairment during anticancer treatment and CRF response to aerobic exercise training (AT) are highly variable. The aim of this ancillary analysis was to leverage machine learning approaches to identify patients at high risk of impaired CRF and poor CRF response to AT.

Methods: We evaluated heterogeneity in CRF among 64 women with metastatic breast cancer randomly assigned to 12 weeks of highly structured AT (n = 33) or control (n = 31). Unsupervised hierarchical cluster analyses were used to identify representative variables from multidimensional prerandomization (baseline) data, and to categorize patients into mutually exclusive subgroups (ie, phenogroups). Logistic and linear regression evaluated the association between phenogroups and impaired CRF (ie, ≤16 mL O₂·kg^-1·min^-1) and CRF response.

Results: Baseline CRF ranged from 10.2 to 38.8 mL O₂·kg^-1·min^-1; CRF response ranged from -15.7 to 4.1 mL O₂·kg^-1·min^-1. Of the n = 120 candidate baseline variables, n = 32 representative variables were identified. Patients were categorized into two phenogroups. Compared with phenogroup 1 (n = 27), phenogroup 2 (n = 37) contained a higher number of patients with none or >three lines of previous anticancer therapy for metastatic disease and had lower resting left ventricular systolic and diastolic function, cardiac output reserve, hematocrit, lymphocyte count, patient-reported outcomes, and CRF (P < .05) at baseline. Among patients allocated to AT (phenogroup 1, n = 12; 44%; phenogroup 2, n = 21; 57%), CRF response (-1.94 ± 3.80 mL O₂·kg^-1·min^-1 v 0.70 ± 2.22 mL O₂·kg^-1·min^-1) was blunted in phenogroup 2 compared with phenogroup 1.

Conclusion: Phenotypic clustering identified two subgroups with unique baseline characteristics and CRF outcomes. The identification of CRF phenogroups could help improve cardiovascular risk stratification and guide investigation of targeted exercise interventions among patients with cancer.

Purpose: A major barrier to the incorporation of biometric data into clinical practice is the lack of device integration with electronic medical records (EMRs). We developed infrastructure to transmit biometric data from an Apple Watch into the EMR for physician review. The study objective was to test feasibility of using this infrastructure for patients undergoing radiotherapy.

Methods: The study included patients with breast or prostate cancer receiving ≥3 weeks of radiotherapy who reported owning an Apple Watch. Daily resting heart rate (HR), HR variability, step count, and exercise minutes were automatically transferred to our EMR using a custom app installed on each patient's iPhone. Biometric data were presented to the treating radiation oncologist for review on a weekly basis during creation of the on-treatment note. Feasibility was defined a priori as physician review of biometric data for at least 90% of patients. Time trends in biometric data were tested using the Jonckheere-Terpstra test. Patient satisfaction was assessed using the System Usability Scale (SUS), with scores above 80 considered above-average user experience.

Results: Of the 20 patients enrolled, biometric data were successfully transmitted to the EMR and reviewed by the radiation oncologist for 95% (n = 19) of patients, thus meeting the a priori feasibility threshold. For patients with radiation courses ≥4 weeks, exercise minutes decreased over time (P = .01) and daily mean HR variability increased over time (P = .02). The median SUS was 82.5 (IQR, 70-87.5).

Conclusion: Our study demonstrates the feasibility of real-time integration of biometric data collected from an Apple Watch into the EMR with subsequent physician review. The high rates of physician review and patient satisfaction provide support for further development of large-scale collection of wearable device data.

Smartphone biometrics in the EMR: is the 5th vital sign here? @JCOCCI_ASCO commentary by Sankaran and @RahulBanerjeeMD here.

Purpose: Large language model (LLM) artificial intelligences may help physicians appeal insurer denials of prescribed medical services, a task that delays patient care and contributes to burnout. We evaluated LLM performance at this task for denials of radiotherapy services.

Methods: We evaluated generative pretrained transformer 3.5 (GPT-3.5; OpenAI, San Francisco, CA), GPT-4, GPT-4 with internet search functionality (GPT-4web), and GPT-3.5ft. The latter was developed by fine-tuning GPT-3.5 via an OpenAI application programming interface with 53 examples of appeal letters written by radiation oncologists. Twenty test prompts with simulated patient histories were programmatically presented to the LLMs, and output appeal letters were scored by three blinded radiation oncologists for language representation, clinical detail inclusion, clinical reasoning validity, literature citations, and overall readiness for insurer submission.

Results: Interobserver agreement between radiation oncologists' scores was moderate or better for all domains (Cohen's kappa coefficients: 0.41-0.91). GPT-3.5, GPT-4, and GPT-4web wrote letters that were on average linguistically clear, summarized provided clinical histories without confabulation, reasoned appropriately, and were scored useful to expedite the insurance appeal process. GPT-4 and GPT-4web letters demonstrated superior clinical reasoning and were readier for submission than GPT-3.5 letters (P < .001). Fine-tuning increased GPT-3.5ft confabulation and compromised performance compared with other LLMs across all domains (P < .001). All LLMs, including GPT-4web, were poor at supporting clinical assertions with existing, relevant, and appropriately cited primary literature.

Conclusion: When prompted appropriately, three commercially available LLMs drafted letters that physicians deemed would expedite appealing insurer denials of radiotherapy services. LLMs may decrease this task's clerical workload on providers. However, LLM performance worsened when fine-tuned with a task-specific, small training data set.

Purpose: Limited studies have used natural language processing (NLP) in the context of non-small cell lung cancer (NSCLC). This study aimed to validate the application of an NLP model to an NSCLC cohort by extracting NSCLC concepts from free-text medical notes and converting them to structured, interpretable data.

Methods: Patients with a lung neoplasm, NSCLC histology, and treatment information in their notes were selected from a repository of over 27 million patients. From these, 200 were randomly selected for this study with the longest and the most recent note included for each patient. An NLP model developed and validated on a large solid and blood cancer oncology cohort was applied to this NSCLC cohort. Two certified tumor registrars and a curator abstracted concepts from the notes: neoplasm, histology, stage, TNM values, and metastasis sites. This manually abstracted gold standard was compared with the NLP model output. Precision and recall scores were calculated.

Results: The NLP model extracted the NSCLC concepts with excellent precision and recall with the following scores, respectively: Lung neoplasm 100% and 100%, NSCLC histology 99% and 88%, histology correctly linked to neoplasm 98% and 79%, stage value 98.8% and 92%, stage TNM value 93% and 98%, and metastasis site 97% and 89%. High precision is related to a low number of false positives, and therefore, extracted concepts are likely accurate. High recall indicates that the model captured most of the desired concepts.

Conclusion: This study validates that Optum's oncology NLP model has high precision and recall with clinical real-world data and is a reliable model to support research studies and clinical trials. This validation study shows that our nonspecific solid tumor and blood cancer oncology model is generalizable to successfully extract clinical information from specific cancer cohorts.

Purpose: Categorizing patients with cancer by their disease stage can be an important tool when conducting administrative claims-based studies. As claims databases frequently do not capture this information, algorithms are increasingly used to define disease stage. To our knowledge, to date, no study has used an algorithm to categorize patients with bladder cancer (BC) by disease stage (non-muscle-invasive BC [NMIBC], muscle-invasive BC [MIBC], or locally advanced/metastatic urothelial carcinoma [la/mUC]) in a US-based health care claims database.

Methods: A claims-based algorithm was developed to categorize patients by disease stage on the basis of the administrative claims portion of the SEER-Medicare linked data. The algorithm was validated against a reference SEER registry, and the algorithm's parameters were iteratively modified to improve its performance. Patients were included if they had an initial diagnosis of BC between January 2016 and December 2017 recorded in SEER registry data. Medicare claims data were available for these patients until December 31, 2019. The algorithm was evaluated by assessing percentage agreement, Cohen's kappa (κ), specificity, positive predictive value (PPV), and negative predictive value (NPV) against the SEER categorization.

Results: A total of 15,484 patients with SEER-confirmed BC were included: 10,991 (71.0%) with NMIBC, 3,645 (23.5%) with MIBC, and 848 (5.5%) with la/mUC. After multiple rounds of algorithm optimization, the final algorithm had an agreement of 82.5% with SEER, with a κ of 0.58, a PPV of 87.0% for NMIBC, and 76.8% for MIBC and a high NPV for la/mUC of 98.0%.

Conclusion: This claims-based algorithm could be a useful approach for researchers conducting claims-based studies categorizing patients with BC at diagnosis.

Purpose: Data on end-of-life care (EOLC) quality, assessed through evidence-based quality measures (QMs), are difficult to obtain. Natural language processing (NLP) enables efficient quality measurement and is not yet used for children with serious illness. We sought to validate a pediatric-specific EOLC-QM keyword library and evaluate EOLC-QM attainment among childhood cancer decedents.

Methods: In a single-center cohort of children with cancer who died between 2014 and 2022, we piloted a rule-based NLP approach to examine the content of clinical notes in the last 6 months of life. We identified documented discussions of five EOLC-QMs: goals of care, limitations to life-sustaining treatments (LLST), hospice, palliative care consultation, and preferred location of death. We assessed performance of NLP methods, compared with gold standard manual chart review. We then used NLP to characterize proportions of decedents with documented EOLC-QM discussions and timing of first documentation relative to death.

Results: Among 101 decedents, nearly half were minorities (Hispanic/Latinx [24%], non-Hispanic Black/African American [20%]), female (48%), or diagnosed with solid tumors (43%). Through iterative refinement, our keyword library achieved robust performance statistics (for all EOLC-QMs, F1 score = 1.0). Most decedents had documented discussions regarding goals of care (83%), LLST (83%), and hospice (74%). Fewer decedents had documented discussions regarding palliative care consultation (49%) or preferred location of death (36%). For all five EOLC-QMs, first documentation occurred, on average, >30 days before death.

Conclusion: A high proportion of decedents attained specified EOLC-QMs more than 30 days before death. Our findings indicate that NLP is a feasible approach to measuring quality of care for children with cancer at the end of life and is ripe for multi-center research and quality improvement.

Purpose: Emerging evidence suggests that the use of artificial intelligence can assist in the timely detection and optimization of therapeutic approach in patients with prostate cancer. The conventional perspective on radiomics encompassing segmentation and the extraction of radiomic features considers it as an independent and sequential process. However, it is not necessary to adhere to this viewpoint. In this study, we show that besides generating masks from which radiomic features can be extracted, prostate segmentation and reconstruction models provide valuable information in their feature space, which can improve the quality of radiomic signatures models for disease aggressiveness classification.

Materials and methods: We perform 2,244 experiments with deep learning features extracted from 13 different models trained using different anatomic zones and characterize how modeling decisions, such as deep feature aggregation and dimensionality reduction, affect performance.

Results: While models using deep features from full gland and radiomic features consistently lead to improved disease aggressiveness prediction performance, others are detrimental. Our results suggest that the use of deep features can be beneficial, but an appropriate and comprehensive assessment is necessary to ensure that their inclusion does not harm predictive performance.

Conclusion: The study findings reveal that incorporating deep features derived from autoencoder models trained to reconstruct the full prostate gland (both zonal models show worse performance than radiomics only models), combined with radiomic features, often lead to a statistically significant increase in model performance for disease aggressiveness classification. Additionally, the results also demonstrate that the choice of feature selection is key to achieving good performance, with principal component analysis (PCA) and PCA + relief being the best approaches and that there is no clear difference between the three proposed latent representation extraction techniques.