Search

Purpose: Machine learning (ML) algorithms that incorporate routinely collected patient-reported outcomes (PROs) alongside electronic health record (EHR) variables may improve prediction of short-term mortality and facilitate earlier supportive and palliative care for patients with cancer., Methods: We trained and validated two-phase ML algorithms that incorporated standard PRO assessments alongside approximately 200 routinely collected EHR variables, among patients with medical oncology encounters at a tertiary academic oncology and a community oncology practice., Results: Among 12,350 patients, 5,870 (47.5%) completed PRO assessments. Compared with EHR- and PRO-only algorithms, the EHR + PRO model improved predictive performance in both tertiary oncology (EHR + PRO v EHR v PRO: area under the curve [AUC] 0.86 [0.85-0.87] v 0.82 [0.81-0.83] v 0.74 [0.74-0.74]) and community oncology (area under the curve 0.89 [0.88-0.90] v 0.86 [0.85-0.88] v 0.77 [0.76-0.79]) practices., Conclusion: Routinely collected PROs contain added prognostic information not captured by an EHR-based ML mortality risk algorithm. Augmenting an EHR-based algorithm with PROs resulted in a more accurate and clinically relevant model, which can facilitate earlier and targeted supportive care for patients with cancer., Competing Interests: Ravi B. ParikhStock and Other Ownership Interests: Merck, Google, GNS Healthcare, Onc.AIConsulting or Advisory Role: GNS Healthcare, Cancer Study Group, Onc.AI, Thyme Care, Humana, NanOlogy, MerckResearch Funding: HumanaPatents, Royalties, Other Intellectual Property: Technology to integrate patient-reported outcomes into electronic health record algorithmsTravel, Accommodations, Expenses: The Oncology Institute of Hope and Innovation Jill S. HaslerPatents, Royalties, Other Intellectual Property: Patent currently pending for machine learning systems using electronic health record data and patient-reported outcomes William FerrellResearch Funding: Humana Peter E. GabrielTravel, Accommodations, Expenses: Varian Medical Systems Justin E. BekelmanStock and Other Ownership Interests: Reimagine CareHonoraria: National Comprehensive Cancer NetworkConsulting or Advisory Role: UnitedHealthcare, Reimagine CareNo other potential conflicts of interest were reported.

Background:  The electronic health record (EHR) has become increasingly ubiquitous. At the same time, health professionals have been turning to this resource for access to data that is needed for the delivery of health care and for clinical research. There is little doubt that the EHR has made both of these functions easier than earlier days when we relied on paper-based clinical records. Coupled with modern database and data warehouse systems, high-speed networks, and the ability to share clinical data with others are large number of challenges that arguably limit the optimal use of the EHR OBJECTIVES:  Our goal was to provide an exhaustive reference for those who use the EHR in clinical and research contexts, but also for health information systems professionals as they design, implement, and maintain EHR systems., Methods:  This study includes a panel of 24 biomedical informatics researchers, information technology professionals, and clinicians, all of whom have extensive experience in design, implementation, and maintenance of EHR systems, or in using the EHR as clinicians or researchers. All members of the panel are affiliated with Penn Medicine at the University of Pennsylvania and have experience with a variety of different EHR platforms and systems and how they have evolved over time., Results:  Each of the authors has shared their knowledge and experience in using the EHR in a suite of 20 short essays, each representing a specific challenge and classified according to a functional hierarchy of interlocking facets such as usability and usefulness, data quality, standards, governance, data integration, clinical care, and clinical research., Conclusion:  We provide here a set of perspectives on the challenges posed by the EHR to clinical and research users., Competing Interests: T.S.C. reports grants from ACRIN, NIH, ACR, and RSNA, as well as royalties from the Osler Institute for lectures in 2013, outside the submitted work. D.S.H. reports grants and nonfinancial support from Roche Diagnostics, outside the submitted work. R.A.H. reports grants from Johnson & Johnson, Merck, and Pfizer, outside the submitted work. Q.L. reports grants from NIH, during the conduct of the study; grants from Pfizer and Bayer, outside the submitted work., (Thieme. All rights reserved.)

Electronic health records (EHR) have been implemented successfully in a majority of United States healthcare systems in some form. There has been a rise in secondary uses of EHR, especially for research. EHR data is large, heterogenous, incomplete, noisy, and primarily created for purposes other than research. This presents many challenges, many of which are beginning to be overcome with the application of computer science artificial intelligence techniques, such as natural language processing and machine learning. EHR are gradually being redesigned to facilitate future research, though we are still far from a "complete EHR.", (Copyright © 2019 Elsevier Inc. All rights reserved.)

Purpose: The purpose of this study was to compare the effectiveness of ensemble methods (e.g., random forests) and single-model methods (e.g., logistic regression and decision trees) in predictive modeling of post-RT treatment failure and adverse events (AEs) for breast cancer patients using automatically extracted EMR data., Methods: Data from 1967 consecutive breast radiotherapy (RT) courses at one institution between 2008 and 2015 were automatically extracted from EMRs and oncology information systems using extraction software. Over 230 variables were extracted spanning the following variable segments: patient demographics, medical/surgical history, tumor characteristics, RT treatment history, and AEs tracked using CTCAEv4.0. Treatment failure was extracted algorithmically by searching posttreatment encounters for evidence of local, nodal, or distant failure. Individual models were trained using decision trees, logistic regression, random forests, and boosted decision trees to predict treatment failures and AEs. Models were fit on 75% of the data and evaluated for probability calibration and area under the ROC curve (AUC) on the remaining test set. The impact of each variable segment was assessed by retraining without the segment and measuring change in AUC (ΔAUC)., Results: All AUC values were statistically significant (P < 0.05). Ensemble methods outperformed single-model methods across all outcomes. The best ensemble method outperformed decision trees and logistic regression by an average AUC of 0.053 and 0.034, respectively. Model probabilities were well calibrated as evidenced by calibration curves. Excluding the patient medical history variable segment led to the largest AUC reduction in all models (Average ΔAUC = -0.025), followed by RT treatment history (-0.021) and tumor information (-0.015)., Conclusion: In this largest such study in breast cancer performed to date, automatically extracted EMR data provided a basis for reliable outcome predictions across multiple statistical methods. Ensemble methods provided substantial advantages over single-model methods. Patient medical history contributed the most to prediction quality., (© 2018 American Association of Physicists in Medicine.)

Purpose: Survivorship care plans (SCPs) are recommended for all cancer survivors. Myriad barriers to implementation exist. This study was performed to evaluate the feasibility of interface development between an SCP and an electronic medical record (EMR)., Methods: An information technology application was developed to extract data from the EMR in use at our center (Epic). Data were transferred to autopopulate an Internet-based tool for creation of SCPs (LIVESTRONG Care Plan) that had been previously used for the creation of more than 35,000 plans., Results: Data (demographic characteristics, surgeries, chemotherapy drugs, radiation site) were extracted from the EMR and transferred to the care plan platform, without transfer of protected health information. Care plans were created and transferred back to the EMR. During clinical testing, SCPs were created by nurse practitioners during scheduled clinic visits for 146 sequential, eligible patients (67% breast cancer, 33% colorectal cancer). All patients received completed care for a single cancer diagnosis at our institution. All data points that were automatically populated were reviewed by practitioners, and missing/blank data fields were populated manually when necessary. Data entered into generated care plans were accurate in 97% of audited cases, and the process of care plan generation could be completed in < 1 minute., Conclusion: This is a feasible solution for the autopopulation of SCPs from the EMR. It represents a future methodology through which widespread implementation of SCPs may be undertaken. Future directions include further clinical testing, assessment of provider-perceived usefulness, and integration into routine clinical care., (Copyright © 2016 by American Society of Clinical Oncology.)

PURPOSE: Current guidelines recommend molecular genotyping for patients newly diagnosed with metastatic nonsquamous (mNSq) non–small-cell lung cancer (NSCLC). The association between availability of molecular genotyping before first line (1L) therapy and overall survival (OS) is not known. METHODS: We conducted a real-world cohort study using electronic health records in patients newly diagnosed with mNSq NSCLC. Cox proportional-hazards multivariable regression models were constructed to examine the association between OS and test result availability before 1L therapy, adjusting for covariates. Additional analyses were conducted to assess the consistency and strength of the relationship. Multivariable logistic regression models were used to examine the association between concurrent tissue and plasma testing (v tissue alone) and result availability. RESULTS: Three hundred twenty-six patients were included, 80% (261/326) with results available before 1L (available testing group), and 20% (65/326) without results available (unavailable testing group). With 14.2-month median follow-up, patients in the available testing group had significantly longer OS relative to the unavailable testing group (adjusted hazard ratio, 0.43; 95% CI, 0.30 to 0.62; P <.0001). The adjusted odds of availability of results before 1L therapy was higher with concurrent tissue and plasma testing (v tissue testing alone; adjusted odds ratio, 2.06; 95% CI, 1.09 to 3.90; P =.026). CONCLUSION: Among patients with mNSq NSCLC in a real-world cohort, availability of molecular genotyping results before 1L therapy was associated with significantly better OS. Concurrent tissue and plasma testing was associated with a higher odds of availability of results before 1L therapy. These findings warrant renewed attention to the completion of molecular genotyping before 1L therapy. Availability of NGS results before 1L therapy for patients with mNSCLC is associated with better overall survival. [ABSTRACT FROM AUTHOR]

PURPOSE Routine collection of patient-reported outcomes (PROs) for patients with advanced solid malignancies is an evidence-based practice and critical component of high-quality cancer care, but real-world adherence is poorly characterized. We sought to describe real-world adherence to PRO monitoring and its potential predictors. METHODS We conducted a retrospective cross-sectional study using deidentified electronic health record data from a National Cancer Institute Cancer Center, encompassing one academic and two community sites. Participants included individuals with lung cancer receiving systemic therapy from January 1 to December 31, 2019. The primary outcome was patient-level adherence, defined as the proportion of treatment visits during which a PRO questionnaire (spanning symptoms, functional status, and global quality-of-life domains) was completed within 30 days. Practice-level performance was calculated as unadjusted mean patient-level adherence. We modeled patient-level adherence using multivariable ordinary least squares regression and identified covariates associated with adherence using a significance threshold of P < .05. RESULTS In 2019, there were 18,604 encounters for 1,105 patients with lung cancer (mean [standard deviation] age 65.8 [10.2] years; 621 [56.2%] female; 216 [19.6%] Black) receiving systemic therapy. The mean patient-level PRO adherence ranged from 27.2% to 70.0% across sites and was 49.4% overall. Advanced age (≥ 65 years) and Black or African American race were negatively associated with PRO adherence (P < .01). CONCLUSION Across this real-world cohort of patients undergoing treatment for lung cancer, adherence to PRO monitoring lagged that achieved in seminal clinical trials, with potential age- and race-based disparities, demonstrating an implementation gap that could be addressed with standardized reporting of an adherence-based quality metric. [ABSTRACT FROM AUTHOR]

Background: Serious illness conversations (SICs) are an evidence-based approach to eliciting patients' values, goals, and care preferences that improve patient outcomes. However, most patients with cancer die without a documented SIC. Clinician-directed implementation strategies informed by behavioral economics ("nudges") that identify high-risk patients have shown promise in increasing SIC documentation among clinicians. It is unknown whether patient-directed nudges that normalize and prime patients towards SIC completion-either alone or in combination with clinician nudges that additionally compare performance relative to peers-may improve on this approach. Our objective is to test the effect of clinician- and patient-directed nudges as implementation strategies for increasing SIC completion among patients with cancer.Methods: We will conduct a 2 × 2 factorial, cluster randomized pragmatic trial to test the effect of nudges to clinicians, patients, or both, compared to usual care, on SIC completion. Participants will include 166 medical and gynecologic oncology clinicians practicing at ten sites within a large academic health system and their approximately 5500 patients at high risk of predicted 6-month mortality based on a validated machine-learning prognostic algorithm. Data will be obtained via the electronic medical record, clinician survey, and semi-structured interviews with clinicians and patients. The primary outcome will be time to SIC documentation among high-risk patients. Secondary outcomes will include time to SIC documentation among all patients (assessing spillover effects), palliative care referral among high-risk patients, and aggressive end-of-life care utilization (composite of chemotherapy within 14 days before death, hospitalization within 30 days before death, or admission to hospice within 3 days before death) among high-risk decedents. We will assess moderators of the effect of implementation strategies and conduct semi-structured interviews with a subset of clinicians and patients to assess contextual factors that shape the effectiveness of nudges with an eye towards health equity.Discussion: This will be the first pragmatic trial to evaluate clinician- and patient-directed nudges to promote SIC completion for patients with cancer. We expect the study to yield insights into the effectiveness of clinician and patient nudges as implementation strategies to improve SIC rates, and to uncover multilevel contextual factors that drive response to these strategies.Trial Registration: ClinicalTrials.gov , NCT04867850 . Registered on April 30, 2021.Funding: National Cancer Institute P50CA244690. [ABSTRACT FROM AUTHOR]

There are emerging data that patients <50 years are diagnosed with esophageal adenocarcinoma (EAC) more frequently, suggesting that the age threshold for screening should be revisited. This study aimed to determine the age distribution, outcomes, and clinical features of EAC over time. The pathology database at the Hospital of the University of Pennsylvania was reviewed from 1991 to 2018. The electronic health records and pathology were reviewed for age of diagnosis, pathology grade, race, and gender for a cohort of 630 patients with biopsy proven EAC. For the patients diagnosed from 2009 to 2018, the Penn Abramson Cancer Center Registry was reviewed for survival and TNM stage. Of the 630 patients, 10.3% (65 patients) were <50 years old [median 43 years, range 16–49]. There was no increase in the number of patients <50 years diagnosed with EAC (R  = 0.133, P  = 0.05). Characteristics of those <50 years versus >50 years showed no difference in tumor grade. Among the 179 eligible patients in the cancer registry, there was no significant difference in clinical or pathological stage for patients <50 years (P value = 0.18). There was no association between diagnosis age and survival (P  = 0.24). A substantial subset of patients with EAC is diagnosed at <50 years. There was no increasing trend of EAC in younger cohorts from 1991 to 2018. We could not identify more advanced stage tumors in the younger cohort. There was no significant association between diagnosis age and survival. [ABSTRACT FROM AUTHOR]

The article discusses the benefits of capturing disease-specific structured data elements in electronic health records. Topics include the impact on efficiency and expressivity of clinical documentation, the importance of adhering to data standards when available, and the potential of electronic health records to improve health-care quality.

Introduction: Widespread adoption of Electronic Health Records (EHR) increased the number of reported disease association studies, or Phenome-Wide Association Studies (PheWAS). Traditional PheWAS studies ignore visit type (i.e., department/service conducting the visit). In this study, we investigate the role of visit type on disease association results in the first Visit-Wide Association Study or 'VisitWAS'. Results: We studied this visit type effect on association results using EHR data from the University of Pennsylvania. Penn EHR data comes from 1,048 different departments and clinics. We analyzed differences between cancer and obstetrics/gynecologist (Ob/Gyn) visits. Some findings were expected (i.e., increase of neoplasm diagnoses among cancer visits), but others were surprising, including an increase in infectious disease conditions among those visiting the Ob/Gyn. Conclusion: We conclude that assessing visit type is important for EHR studies because different medical centers have different visit type distributions. To increase reproducibility among EHR data mining algorithms, we recommend that researchers report visit type in studies. [ABSTRACT FROM AUTHOR]