The impact of opioid use associated with curative-intent cancer surgery on safe opioid prescribing practice among veterans: An observational study.

Background: Opioid exposure during cancer therapy may increase long-term unsafe opioid prescribing. This study sought to determine the rates of coprescription of benzodiazepine and opioid medications and new persistent opioid use after surgical treatment of early-stage cancer.

Methods: A retrospective cohort study was conducted among a US veteran population via the Veterans Affairs Corporate Data Warehouse database.

Performance Drift in a Nationally Deployed Population Health Risk Algorithm in the US Veterans Health Administration.

Importance: Clinical risk algorithms inform clinical decision support and system-level quality metrics. However, algorithm performance can drift over time and possibly promote misinformed decision-making and resource allocation. The Veterans Health Administration (VA) Care Assessment Needs (CAN) algorithm is a nationally deployed population risk algorithm used to predict risk of 90-day hospitalization and/or mortality and to allocate resources for more than 5 million veterans annually.

Derivation and External Validation of Objective Performance Status Among Patients With Metastatic Cancer.

Purpose: Performance status (PS) assessment is used to determine clinical trial eligibility among patients with cancer, but may be inaccurately assessed by oncology clinicians. Wearable accelerometers may allow objective assessment of physical activity, a proxy for PS. In this analysis of two prospective studies, we derive and externally validate objective PS (OPS) by measuring the association between daily physical activity and overall survival among patients with metastatic cancer.

The Impact of Machine Learning Mortality Risk Prediction on Clinician Prognostic Accuracy and Decision Support: A Randomized Vignette Study.

BackgroundMachine learning (ML) algorithms may improve the prognosis for serious illnesses such as cancer, identifying patients who may benefit from earlier palliative care (PC) or advance care planning (ACP). We evaluated the impact of various presentation strategies of a hypothetical ML algorithm on clinician prognostic accuracy and decision making.MethodsThis was a randomized clinical vignette survey study among medical oncologists who treat metastatic non-small-cell lung cancer (mNSCLC).

Specificity of Food and Drug Administration postmarketing requirements and associations with timely submissions and regulatory decisions for oncology accelerated approvals, 2011-2023: a cross-sectional analysis.

Objectives: To assess the specificity of postmarketing requirement (PMR) statements and associations between PMR statement specificity and PMR study characteristics, timeliness and regulatory decisions.

Methods And Analysis: This was a cross-sectional analysis of publicly available Food and Drug Administration (FDA) databases to characterise PMR statements for oncology accelerated approvals (AAs) between January 2011 and July 2023. Characteristics of trials supporting AA and PMR studies were identified from product labels on the Drugs@FDA database and ClinicalTrials.

Driving Knowledge to Action: Building a Better Future With Artificial Intelligence-Enabled Multidisciplinary Oncology.

Artificial intelligence (AI) is transforming multidisciplinary oncology at an unprecedented pace, redefining how clinicians detect, classify, and treat cancer. From earlier and more accurate diagnoses to personalized treatment planning, AI's impact is evident across radiology, pathology, radiation oncology, and medical oncology. By leveraging vast and diverse data-including imaging, genomic, clinical, and real-world evidence-AI algorithms can uncover complex patterns, accelerate drug discovery, and help identify optimal treatment regimens for each patient.

Cancer Informatics: Novel Methods and Applications of Artificial Intelligence in Cancer Care Delivery.

Objectives: To summarize significant research contributions on cancer informatics published in 2023, an extensive search using PubMed/MEDLINE was conducted to identify the scientific contributions published in 2023 that address topics in cancer. The selection process comprised three steps: (i) ten candidate best papers were first selected by the two section editors, (ii) external reviewers from internationally renowned research teams reviewed each candidate best paper, and (iii) the final selection of three best papers was conducted by the editorial board of the Yearbook.

Results: The two selected papers demonstrate advances in the clinical implementation of cancer informatics methodologies.

Primary Care Provider Preferences Regarding Artificial Intelligence in Point-of-Care Cancer Screening.

Unlabelled: It is unclear how to optimize the user interface and user experience of cancer screening artificial intelligence (AI) tools for clinical decision-making in primary care. We developed an electronic survey for US primary care clinicians to assess 1) general attitudes toward AI in cancer screening and 2) preferences for various aspects of AI model deployment in the context of colorectal, breast, and lung cancer screening. We descriptively analyzed the responses.

Algorithm-Based Palliative Care in Patients With Cancer: A Cluster Randomized Clinical Trial.

Importance: Among patients with advanced solid malignant tumors, early specialty palliative care (PC) is guideline recommended, but strategies to increase PC access and effectiveness in community oncology are lacking.

Objective: To test whether algorithm-based defaults with opting out and accountable justification embedded in the electronic health record (EHR) increase completed PC visits.

Design, Setting, And Participants: This 2-arm cluster randomized clinical trial was conducted from November 1, 2022, to December 31, 2023.

Assessing Algorithm Fairness Requires Adjustment for Risk Distribution Differences: Re-Considering the Equal Opportunity Criterion.

The proliferation of algorithm-assisted decision making has prompted calls for careful assessment of algorithm fairness. One popular fairness metric, equal opportunity, demands parity in true positive rates (TPRs) across different population subgroups. However, we highlight a critical but overlooked weakness in this measure: at a given decision threshold, TPRs vary when the underlying risk distribution varies across subgroups, even if the model equally captures the underlying risks.

Inpatient Immunotherapy Outcomes Study: A Multicenter Retrospective Analysis.

Purpose: Immune checkpoint inhibitors (ICIs) have revolutionized the care of patients with cancer, but use among hospitalized patients is controversial as a result of questionable benefit and high costs. To evaluate the role of ICIs in the inpatient (IP) setting, we conducted the Inpatient Immunotherapy Outcomes Study (IIOS) to describe characteristics and outcomes of patients who received IP ICIs.

Methods: IIOS is a retrospective study of patients treated with ICIs during hospitalization between 2012 and 2021 at five academic institutions.

Evaluating generalizability of oncology trial results to real-world patients using machine learning-based trial emulations.

Randomized controlled trials (RCTs) evaluating anti-cancer agents often lack generalizability to real-world oncology patients. Although restrictive eligibility criteria contribute to this issue, the role of selection bias related to prognostic risk remains unclear. In this study, we developed TrialTranslator, a framework designed to systematically evaluate the generalizability of RCTs for oncology therapies.

Explainable Machine Learning to Predict Treatment Response in Advanced Non-Small Cell Lung Cancer.

Purpose: Immune checkpoint inhibitors (ICIs) have demonstrated promise in the treatment of various cancers. Single-drug ICI therapy (immuno-oncology [IO] monotherapy) that targets PD-L1 is the standard of care in patients with advanced non-small cell lung cancer (NSCLC) with PD-L1 expression ≥50%. We sought to find out if a machine learning (ML) algorithm can perform better as a predictive biomarker than PD-L1 alone.

Real-World and Clinical Trial Validation of a Deep Learning Radiomic Biomarker for PD-(L)1 Immune Checkpoint Inhibitor Response in Advanced Non-Small Cell Lung Cancer.

Purpose: This study developed and validated a novel deep learning radiomic biomarker to estimate response to immune checkpoint inhibitor (ICI) therapy in advanced non-small cell lung cancer (NSCLC) using real-world data (RWD) and clinical trial data.

Materials And Methods: Retrospective RWD of 1,829 patients with advanced NSCLC treated with PD-(L)1 ICIs were collected from 10 academic and community institutions in the United States and Europe. The RWD included data sets for discovery (Data Set A-Discovery, n = 1,173) and independent test (Data Set B, n = 458).

Real-world impact of the platinum chemotherapy shortage on US patients with advanced cancer.

The recent cisplatin and carboplatin ("platinum") chemotherapy shortage, first announced on February 10, 2023, has impacted cancer patients nationwide. Here, we quantify the extent to which the shortage affected platinum chemotherapy prescribing and short-term mortality. This cohort study included 11 797 adults with advanced solid cancers who initiated first-line therapy during the 1-year period before (February 1, 2022-February 9, 2023) or during (February 10, 2023-January 31, 2024) the platinum shortage.

Racial Bias in Clinical and Population Health Algorithms: A Critical Review of Current Debates.

Among health care researchers, there is increasing debate over how best to assess and ensure the fairness of algorithms used for clinical decision support and population health, particularly concerning potential racial bias. Here we first distill concerns over the fairness of health care algorithms into four broad categories: () the explicit inclusion (or, conversely, the exclusion) of race and ethnicity in algorithms, () unequal algorithm decision rates across groups, () unequal error rates across groups, and () potential bias in the target variable used in prediction. With this taxonomy, we critically examine seven prominent and controversial health care algorithms.

Using the Translational Science Benefits Model to assess the impact of the Penn Implementation Science Center in Cancer Control.

Traditional approaches for evaluating the impact of scientific research - mainly scholarship (i.e., publications, presentations) and grant funding - fail to capture the full extent of contributions that come from larger scientific initiatives.

"Start low, go slow," a strategy to tailor treatment dosing in older or vulnerable adults with advanced solid cancer: A systematic review and meta-analysis.

Introduction: The use of standard-dose cancer treatment can result in a decline in the functional abilities of older adults with cancer. The "start-low, go-slow" (SLGS) strategy involves initiating cancer treatment at lower-than-standard doses in selected patients who are vulnerable to excess toxicity and escalating based on tolerance. We performed a systematic review and meta-analysis to assess the available data and the effectiveness of the SLGS strategy in the treatment of cancer in older adults with incurable solid cancer.

Phenotypic evaluation of deep learning models for classifying germline variant pathogenicity.

Deep learning models for predicting variant pathogenicity have not been thoroughly evaluated on real-world clinical phenotypes. Here, we apply state-of-the-art pathogenicity prediction models to hereditary breast cancer gene variants in UK Biobank participants. Model predictions for missense variants in BRCA1, BRCA2 and PALB2, but not ATM and CHEK2, were associated with breast cancer risk.

Association of Participation in Medicare's Oncology Care Model With Spending, Utilization, and Quality Outcomes Among Commercially Insured and Medicare Advantage Members.

Purpose: The Oncology Care Model (OCM), a value-based payment model for traditional Medicare beneficiaries with cancer, yielded total spending reductions that were outweighed by incentive payments, resulting in net losses to the Centers for Medicare & Medicaid Services. We studied whether the OCM yielded spillover effects in total episode spending, utilization, and quality among commercially insured and Medicare Advantage (MA) members, who were not targeted by the program.

Patients And Methods: This observational study used administrative claims from a large national payer, yielding 157,189 total patients with commercial insurance or MA with solid malignancies who initiated 229,376 systemic anticancer therapy episodes before (2012-2015) and during (2016-2021) the OCM at 125 OCM-participating practices (a subset of total OCM practices) and a 1:10 propensity-matched set of 860 non-OCM practices.

DISCRET: Synthesizing Faithful Explanations For Treatment Effect Estimation.

Designing faithful yet accurate AI models is challenging, particularly in the field of individual treatment effect estimation (ITE). ITE prediction models deployed in critical settings such as healthcare should ideally be (i) accurate, and (ii) provide faithful explanations. However, current solutions are inadequate: state-of-the-art black-box models do not supply explanations, post-hoc explainers for black-box models lack faithfulness guarantees, and self-interpretable models greatly compromise accuracy.