Search

Purpose: The advancement of natural language processing (NLP) has promoted the use of detailed textual data in electronic health records (EHRs) to support cancer research and to facilitate patient care. In this review, we aim to assess EHR for cancer research and patient care by using the Minimal Common Oncology Data Elements (mCODE), which is a community-driven effort to define a minimal set of data elements for cancer research and practice. Specifically, we aim to assess the alignment of NLP-extracted data elements with mCODE and review existing NLP methodologies for extracting said data elements., Methods: Published literature studies were searched to retrieve cancer-related NLP articles that were written in English and published between January 2010 and September 2020 from main literature databases. After the retrieval, articles with EHRs as the data source were manually identified. A charting form was developed for relevant study analysis and used to categorize data including four main topics: metadata, EHR data and targeted cancer types, NLP methodology, and oncology data elements and standards., Results: A total of 123 publications were selected finally and included in our analysis. We found that cancer research and patient care require some data elements beyond mCODE as expected. Transparency and reproductivity are not sufficient in NLP methods, and inconsistency in NLP evaluation exists., Conclusion: We conducted a comprehensive review of cancer NLP for research and patient care using EHRs data. Issues and barriers for wide adoption of cancer NLP were identified and discussed., Competing Interests: Irbaz B. RiazThis author is a member of the JCO Clinical Cancer Informatics Editorial Board. Journal policy recused the author from having any role in the peer review of this manuscript. Hua XuEmployment: Melax Technologies IncStock and Other Ownership Interests: Melax Technologies IncConsulting or Advisory Role: More Health Inc, Hebta LLC, Melax Technologies IncPatents, Royalties, Other Intellectual Property: Receive royalties from software license from UTHealth Jeremy L. WarnerThis author is an Associate Editor for JCO Clinical Cancer Informatics. Journal policy recused the author from having any role in the peer review of this manuscript.Stock and Other Ownership Interests: HemOnc.orgConsulting or Advisory Role: Westat, Roche, Flatiron Health, Melax TechNo other potential conflicts of interest were reported.

Obtaining medication use and response information is essential for both care providers and researchers to understand patients' medication use and long-term treatment patterns. While unstructured clinical notes contain such information, they have rarely been analyzed for this purpose on a large scale due to the demands of expensive manual reviews. Here, we aimed to extract and analyze medication use patterns from clinical notes for a population of breast cancer patients at an academic medical center using unsupervised topic modeling techniques. Notably, we proposed a two-stage modeling process that was built upon correlated topic modeling (CTM) and structural topic modeling (STM) to capture nuanced information about medication behavior, including drug-disease relationships as well as medication schedules. The STM-derived topics show longitudinal prevalence patterns that may reflect changing patient needs and behaviors after the diagnosis of a severe disease. The patterns also show promise as a predictor for medication-taking behavior., (©2022 AMIA - All rights reserved.)

Hormonal therapy (HT) reduces the risk of cancer recurrence and the mortality rate for patients with hormone-receptor-positive breast cancer. However, it is estimated that half of the patients fail to complete the standard 5-year adjuvant treatment protocol. We investigate the extent to which certain types of structured data in electronic medical records (EMRs), namely conditions, drugs, laboratory tests and procedures, as well as when such data is entered EMRs, can forecast HT discontinuation. Our experiments with EMR data from 2,251 patients showed that machine learning models based on these data types achieve fair performance (AUC of 0.65). More importantly, the performance was not statistically significantly different when fitting a model using all or only one feature type, suggesting that the model is robust to missing information in the EMR., (©2022 AMIA - All rights reserved.)

Patients with B-lymphoid malignancies have been consistently identified as a population at high risk of severe COVID-19. Whether this is exclusively due to cancer-related deficits in humoral and cellular immunity, or whether risk of severe COVID-19 is increased by anticancer therapy, is uncertain. Using data derived from the COVID-19 and Cancer Consortium (CCC19), we show that patients treated for B-lymphoid malignancies have an increased risk of severe COVID-19 compared with control populations of patients with non-B-lymphoid malignancies. Among patients with B-lymphoid malignancies, those who received anticancer therapy within 12 months of COVID-19 diagnosis experienced increased COVID-19 severity compared with patients with non-recently treated B-lymphoid malignancies, after adjustment for cancer status and several other prognostic factors. Our findings suggest that patients recently treated for a B-lymphoid malignancy are at uniquely high risk for severe COVID-19., Significance: Our study suggests that recent therapy for a B-lymphoid malignancy is an independent risk factor for COVID-19 severity. These findings provide rationale to develop mitigation strategies targeted at the uniquely high-risk population of patients with recently treated B-lymphoid malignancies. This article is highlighted in the In This Issue feature, p. 171., (©2022 American Association for Cancer Research.)

Background: Older age is associated with poorer outcomes of SARS-CoV-2 infection, although the heterogeneity of ageing results in some older adults being at greater risk than others. The objective of this study was to quantify the association of a novel geriatric risk index, comprising age, modified Charlson comorbidity index, and Eastern Cooperative Oncology Group performance status, with COVID-19 severity and 30-day mortality among older adults with cancer., Methods: In this cohort study, we enrolled patients aged 60 years and older with a current or previous cancer diagnosis (excluding those with non-invasive cancers and premalignant or non-malignant conditions) and a current or previous laboratory-confirmed COVID-19 diagnosis who reported to the COVID-19 and Cancer Consortium (CCC19) multinational, multicentre, registry between March 17, 2020, and June 6, 2021. Patients were also excluded for unknown age, missing data resulting in unknown geriatric risk measure, inadequate data quality, or incomplete follow-up resulting in unknown COVID-19 severity. The exposure of interest was the CCC19 geriatric risk index. The primary outcome was COVID-19 severity and the secondary outcome was 30-day all-cause mortality; both were assessed in the full dataset. Adjusted odds ratios (ORs) and 95% CIs were estimated from ordinal and binary logistic regression models., Findings: 5671 patients with cancer and COVID-19 were included in the analysis. Median follow-up time was 56 days (IQR 22-120), and median age was 72 years (IQR 66-79). The CCC19 geriatric risk index identified 2365 (41·7%) patients as standard risk, 2217 (39·1%) patients as intermediate risk, and 1089 (19·2%) as high risk. 36 (0·6%) patients were excluded due to non-calculable geriatric risk index. Compared with standard-risk patients, high-risk patients had significantly higher COVID-19 severity (adjusted OR 7·24; 95% CI 6·20-8·45). 920 (16·2%) of 5671 patients died within 30 days of a COVID-19 diagnosis, including 161 (6·8%) of 2365 standard-risk patients, 409 (18·5%) of 2217 intermediate-risk patients, and 350 (32·1%) of 1089 high-risk patients. High-risk patients had higher adjusted odds of 30-day mortality (adjusted OR 10·7; 95% CI 8·54-13·5) than standard-risk patients., Interpretation: The CCC19 geriatric risk index was strongly associated with COVID-19 severity and 30-day mortality. Our CCC19 geriatric risk index, based on readily available clinical factors, might provide clinicians with an easy-to-use risk stratification method to identify older adults most at risk for severe COVID-19 as well as mortality., Funding: US National Institutes of Health National Cancer Institute Cancer Center., Competing Interests: The following authors declare competing interests not related to the current work: ZB reports grants from Genentech/imCORE, non-financial support from Bristol Myers Squibb, and personal fees from UpToDate. AE reports salary support from the Canadian Institute of Health Research, the Detweiler Travelling Fellowship (Royal College of Physicians and Surgeons of Canada), and the Henry R Shibata Fellowship (Cedar's Cancer Foundation). CRF reports grants from Merck Foundation, NCCN/Pfizer, and National Cancer Institute, and other support from National Cancer Institute and the Patient-Centered Outcomes Research Institute. PGri reports personal fees and non-financial support from AstraZeneca, personal fees from Astellas Pharma, personal fees from Bayer, grants and personal fees from Bristol Myers Squibb, grants and non-financial support from Clovis Oncology, personal fees from Dyania Health, grants and personal fees from EMD Serono, personal fees from Exelixis, personal fees from Foundation Medicine, personal fees from Genentech/Roche, personal fees from Genzyme, grants and personal fees from GlaxoSmithKline, personal fees from Guardant Health, grants and personal fees from Immunomedics/Gilead, personal fees from Infinity Pharmaceuticals, personal fees from Janssen, grants and personal fees from Merck, grants and personal fees from Mirati Therapeutics, grants and personal fees from Pfizer, grants and personal fees from QED Therapeutics, personal fees from Regeneron Pharmaceuticals, personal fees from Seattle Genetics, personal fees from 4D Pharma, personal fees from UroGen, grants from Bavarian Nordic, and grants from Debiopharm. SGup reports grants and personal fees from Bristol Myers Squibb, personal fees from Merck, Janssen, Seattle Genetis, EMD Sorono, and Pfizer, and grants from Astellas and BMS. CHw reports grants from Merck, Bayer, and AstraZeneca, and personal fees from Tempus and EMD Sorono, and other support from Johnson and Johnson. AK reports other support from TESARO, Fibrogen, Geistlich Pharma, Astellas Pharma, Rafael Pharmaceuticals, and Novocure. ARK reports other support from Merck and Sanofi, personal fees from OncLive, grants from ASCO Conquer Cancer Foundation, and grants from Bladder Cancer Advocacy Network. HK reports position on advisory board of Sanofi Genzyme NSCLC Northeast. NMK reports personal fees from BMS, Janssen, Seattle Genetics, Celldex, Sandoz, Invitae, Beyond Spring, Spectrum G1 Therapeutics, and Total Health, and grants from Amgen, Jazz Therapeutics, G1 Therapeutics, and Samsung. CL reports grants from imCORE/Genentech. RRM reports serving on advisory board or as a consultant for Astrazeneca, Aveo, Bayer, BMS, Caris, Dendreon, Exelixis, Janssen, Merck, Myovant, Novartis, Pfizer, Sanofi, Sorrento Therapeutics, and Tempus. RRM received institutional research funding from Pfizer, Bayer, Tempus. SM reports personal fees from National Geographic. OAP reports grants from National Institutes of Health (NIH) and Agency for Healthcare Research and Quality, and personal fees from International Consulting Associates. NAP reports personal fees from AstraZeneca, Merck, Pfizer, Eli Lilly, Genentech, BMS, Amgen, Inivata, G1 Therapeutics, Xencor, Mirati, Janssen, Boehringer Ingelheim, and Sanofi-Genzyme-Regeneron. RPR reports grants from BMS and Janssen, and personal fees from BMS, Janssen, Dova, and Inari. MAT reports personal fees from VIA Oncology, GSK, and Adaptive Advisory Board, other support from Syapse, UpToDate, Takeda, Celgene, Doximity, AbbVie, BMS, CRAB CTC, Denovo, Hoosier Research Network, Lilly, LynxBio, Strata Oncology, and TG Therapeutics. JLW reports personal fees from Roche, Westat, Flatiron Health, Melax Tech, and IBM Watson Health, other support from HemOnc and Janssen, and grants from AACR. TMW reports personal fees from Carevive, and personal fees from Sanofi, and Seattle Genetics. TMW-D reports grants from BMS, Merck, Janssen, and GSK/Tesaro, personal fees from Exicure, Shattuck Labs, Merck, Caris Life Science, and SITC, and other support from High Enroll. EW-B reports grants from Pfizer Global Medical Grants, personal fees from Astellas, Aveo Oncology Bristol Myers Squibb, Exelixis, and Janssen, and other support from Immunomedics and Nektar. The following authors declare competing interests during the conduct of the study: SM reports grants and other support from National Cancer Institute and from the International Association for the Study of Lung Cancer. DPS reports grants from American Cancer Society and Hope Foundation for Cancer Research and from NIH. LT reports grants from NIH. JLW reports grants from NIH. All other authors declare no competing interests., (© 2022 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY-NC-ND 4.0 license.)

Background: The frequency of coinfections and their association with outcomes have not been adequately studied among patients with cancer and coronavirus disease 2019 (COVID-19), a high-risk group for coinfection., Methods: We included adult (≥18 years) patients with active or prior hematologic or invasive solid malignancies and laboratory-confirmed severe acute respiratory syndrome coronavirus 2 (SARS-COV-2) infection, using data from the COVID-19 and Cancer Consortium (CCC19, NCT04354701). We captured coinfections within ±2 weeks from diagnosis of COVID-19, identified factors cross-sectionally associated with risk of coinfection, and quantified the association of coinfections with 30-day mortality., Results: Among 8765 patients (hospitalized or not; median age, 65 years; 47.4% male), 16.6% developed coinfections: 12.1% bacterial, 2.1% viral, 0.9% fungal. An additional 6.4% only had clinical diagnosis of a coinfection. The adjusted risk of any coinfection was positively associated with age >50 years, male sex, cardiovascular, pulmonary, and renal comorbidities, diabetes, hematologic malignancy, multiple malignancies, Eastern Cooperative Oncology Group Performance Status, progressing cancer, recent cytotoxic chemotherapy, and baseline corticosteroids; the adjusted risk of superinfection was positively associated with tocilizumab administration. Among hospitalized patients, high neutrophil count and C-reactive protein were positively associated with bacterial coinfection risk, and high or low neutrophil count with fungal coinfection risk. Adjusted mortality rates were significantly higher among patients with bacterial (odds ratio [OR], 1.61; 95% CI, 1.33-1.95) and fungal (OR, 2.20; 95% CI, 1.28-3.76) coinfections., Conclusions: Viral and fungal coinfections are infrequent among patients with cancer and COVID-19, with the latter associated with very high mortality rates. Clinical and laboratory parameters can be used to guide early empiric antimicrobial therapy, which may improve clinical outcomes., (© The Author(s) 2022. Published by Oxford University Press on behalf of Infectious Diseases Society of America.)

The ongoing coronavirus disease 2019 (COVID-19) pandemic has left patients with current or past history of cancer facing disparate consequences at every stage of the cancer trajectory. This comprehensive review offers a landscape analysis of the current state of the literature on COVID-19 and cancer, including the immune response to COVID-19, risk factors for severe disease, and impact of anticancer therapies. We also review the latest data on treatment of COVID-19 and vaccination safety and efficacy in patients with cancer, as well as the impact of the pandemic on cancer care, including the urgent need for rapid evidence generation and real-world study designs. SIGNIFICANCE: Patients with cancer have faced severe consequences at every stage of the cancer journey due to the COVID-19 pandemic. This comprehensive review offers a landscape analysis of the current state of the field regarding COVID-19 and cancer. We cover the immune response, risk factors for severe disease, and implications for vaccination in patients with cancer, as well as the impact of the COVID-19 pandemic on cancer care delivery. Overall, this review provides an in-depth summary of the key issues facing patients with cancer during this unprecedented health crisis., (©2021 American Association for Cancer Research.)

Importance: The COVID-19 pandemic has had a distinct spatiotemporal pattern in the United States. Patients with cancer are at higher risk of severe complications from COVID-19, but it is not well known whether COVID-19 outcomes in this patient population were associated with geography., Objective: To quantify spatiotemporal variation in COVID-19 outcomes among patients with cancer., Design, Setting, and Participants: This registry-based retrospective cohort study included patients with a historical diagnosis of invasive malignant neoplasm and laboratory-confirmed SARS-CoV-2 infection between March and November 2020. Data were collected from cancer care delivery centers in the United States., Exposures: Patient residence was categorized into 9 US census divisions. Cancer center characteristics included academic or community classification, rural-urban continuum code (RUCC), and social vulnerability index., Main Outcomes and Measures: The primary outcome was 30-day all-cause mortality. The secondary composite outcome consisted of receipt of mechanical ventilation, intensive care unit admission, and all-cause death. Multilevel mixed-effects models estimated associations of center-level and census division-level exposures with outcomes after adjustment for patient-level risk factors and quantified variation in adjusted outcomes across centers, census divisions, and calendar time., Results: Data for 4749 patients (median [IQR] age, 66 [56-76] years; 2439 [51.4%] female individuals, 1079 [22.7%] non-Hispanic Black individuals, and 690 [14.5%] Hispanic individuals) were reported from 83 centers in the Northeast (1564 patients [32.9%]), Midwest (1638 [34.5%]), South (894 [18.8%]), and West (653 [13.8%]). After adjustment for patient characteristics, including month of COVID-19 diagnosis, estimated 30-day mortality rates ranged from 5.2% to 26.6% across centers. Patients from centers located in metropolitan areas with population less than 250 000 (RUCC 3) had lower odds of 30-day mortality compared with patients from centers in metropolitan areas with population at least 1 million (RUCC 1) (adjusted odds ratio [aOR], 0.31; 95% CI, 0.11-0.84). The type of center was not significantly associated with primary or secondary outcomes. There were no statistically significant differences in outcome rates across the 9 census divisions, but adjusted mortality rates significantly improved over time (eg, September to November vs March to May: aOR, 0.32; 95% CI, 0.17-0.58)., Conclusions and Relevance: In this registry-based cohort study, significant differences in COVID-19 outcomes across US census divisions were not observed. However, substantial heterogeneity in COVID-19 outcomes across cancer care delivery centers was found. Attention to implementing standardized guidelines for the care of patients with cancer and COVID-19 could improve outcomes for these vulnerable patients.

Background: Hospitalized patients with COVID-19 have increased risks of venous (VTE) and arterial thromboembolism (ATE). Active cancer diagnosis and treatment are well-known risk factors; however, a risk assessment model (RAM) for VTE in patients with both cancer and COVID-19 is lacking., Objectives: To assess the incidence of and risk factors for thrombosis in hospitalized patients with cancer and COVID-19., Methods: Among patients with cancer in the COVID-19 and Cancer Consortium registry (CCC19) cohort study, we assessed the incidence of VTE and ATE within 90 days of COVID-19-associated hospitalization. A multivariable logistic regression model specifically for VTE was built using a priori determined clinical risk factors. A simplified RAM was derived and internally validated using bootstrap., Results: From March 17, 2020 to November 30, 2020, 2804 hospitalized patients were analyzed. The incidence of VTE and ATE was 7.6% and 3.9%, respectively. The incidence of VTE, but not ATE, was higher in patients receiving recent anti-cancer therapy. A simplified RAM for VTE was derived and named CoVID-TE (Cancer subtype high to very-high risk by original Khorana score +1, VTE history +2, ICU admission +2, D-dimer elevation +1, recent systemic anti-cancer Therapy +1, and non-Hispanic Ethnicity +1). The RAM stratified patients into two cohorts (low-risk, 0-2 points, n = 1423 vs. high-risk, 3+ points, n = 1034) where VTE occurred in 4.1% low-risk and 11.3% high-risk patients (c statistic 0.67, 95% confidence interval 0.63-0.71). The RAM performed similarly well in subgroups of patients not on anticoagulant prior to admission and moderately ill patients not requiring direct ICU admission., Conclusions: Hospitalized patients with cancer and COVID-19 have elevated thrombotic risks. The CoVID-TE RAM for VTE prediction may help real-time data-driven decisions in this vulnerable population., (© 2021 The Authors. Journal of Thrombosis and Haemostasis published by Wiley Periodicals LLC on behalf of International Society on Thrombosis and Haemostasis.)

Background: Accurate and rapid clinical decisions based on real-world evidence are essential for patients with cancer. However, the complexity of chemotherapy regimens for cancer impedes retrospective research that uses observational health databases., Objective: The aim of this study is to compare the anticancer treatment trajectories and patterns of clinical events according to regimen type using the chemotherapy episodes determined by an algorithm., Methods: We developed an algorithm to extract the regimen-level abstracted chemotherapy episodes from medication records in a conventional Observational Medical Outcomes Partnership (OMOP) common data model (CDM) database. The algorithm was validated on the Ajou University School Of Medicine (AUSOM) database by manual review of clinical notes. Using the algorithm, we extracted episodes of chemotherapy from patients in the EHR database and the claims database. We also developed an application software for visualizing the chemotherapy treatment patterns based on the treatment episodes in the OMOP-CDM database. Using this software, we generated the trends in the types of regimen used in the institutions, the patterns of the iterative chemotherapy use, and the trajectories of cancer treatment in two EHR-based OMOP-CDM databases. As a pilot study, the time of onset of chemotherapy-induced neutropenia according to regimen was measured using the AUSOM database. The anticancer treatment trajectories for patients with COVID-19 were also visualized based on the nationwide claims database., Results: We generated 178,360 treatment episodes for patients with colorectal, breast, and lung cancer for 85 different regimens. The algorithm precisely identified the type of chemotherapy regimen in 400 patients (average positive predictive value >98%). The trends in the use of routine clinical chemotherapy regimens from 2008-2018 were identified for 8236 patients. For a total of 12 regimens (those administered to the largest proportion of patients), the number of repeated treatments was concordant with the protocols for standard chemotherapy regimens for certain cases. In addition, the anticancer treatment trajectories for 8315 patients were shown, including 62 patients with COVID-19. A comparative analysis of neutropenia showed that its onset in colorectal cancer regimens tended to cluster between days 9-15, whereas it tended to cluster between days 2-8 for certain regimens for breast cancer or lung cancer., Conclusions: We propose a method for generating chemotherapy episodes for introduction into the oncology extension module of the OMOP-CDM databases. These proof-of-concept studies demonstrated the usability, scalability, and interoperability of the proposed framework through a distributed research network., (©Hokyun Jeon, Seng Chan You, Seok Yun Kang, Seung In Seo, Jeremy L Warner, Rimma Belenkaya, Rae Woong Park. Originally published in JMIR Medical Informatics (http://medinform.jmir.org), 06.04.2021.)

Clinical trials establish the standard of cancer care, yet the evolution and characteristics of the social dynamics between the people conducting this work remain understudied. We performed a social network analysis of authors publishing chemotherapy-based prospective trials from 1946 to 2018 to understand how social influences, including the role of gender, have influenced the growth and development of this network, which has expanded exponentially from fewer than 50 authors in 1946 to 29,197 in 2018. While 99.4% of authors were directly or indirectly connected by 2018, our results indicate a tendency to predominantly connect with others in the same or similar fields, as well as an increasing disparity in author impact and number of connections. Scale-free effects were evident, with small numbers of individuals having disproportionate impact. Women were under-represented and likelier to have lower impact, shorter productive periods (P < 0.001 for both comparisons), less centrality, and a greater proportion of co-authors in their same subspecialty. The past 30 years were characterized by a trend towards increased authorship by women, with new author parity anticipated in 2032. The network of cancer clinical trialists is best characterized as strategic or mixed-motive, with cooperative and competitive elements influencing its appearance. Network effects such as low centrality, which may limit access to high-profile individuals, likely contribute to the observed disparities.

Background: Data on patients with COVID-19 who have cancer are lacking. Here we characterise the outcomes of a cohort of patients with cancer and COVID-19 and identify potential prognostic factors for mortality and severe illness., Methods: In this cohort study, we collected de-identified data on patients with active or previous malignancy, aged 18 years and older, with confirmed severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection from the USA, Canada, and Spain from the COVID-19 and Cancer Consortium (CCC19) database for whom baseline data were added between March 17 and April 16, 2020. We collected data on baseline clinical conditions, medications, cancer diagnosis and treatment, and COVID-19 disease course. The primary endpoint was all-cause mortality within 30 days of diagnosis of COVID-19. We assessed the association between the outcome and potential prognostic variables using logistic regression analyses, partially adjusted for age, sex, smoking status, and obesity. This study is registered with ClinicalTrials.gov, NCT04354701, and is ongoing., Findings: Of 1035 records entered into the CCC19 database during the study period, 928 patients met inclusion criteria for our analysis. Median age was 66 years (IQR 57-76), 279 (30%) were aged 75 years or older, and 468 (50%) patients were male. The most prevalent malignancies were breast (191 [21%]) and prostate (152 [16%]). 366 (39%) patients were on active anticancer treatment, and 396 (43%) had active (measurable) cancer. At analysis (May 7, 2020), 121 (13%) patients had died. In logistic regression analysis, independent factors associated with increased 30-day mortality, after partial adjustment, were: increased age (per 10 years; partially adjusted odds ratio 1·84, 95% CI 1·53-2·21), male sex (1·63, 1·07-2·48), smoking status (former smoker vs never smoked: 1·60, 1·03-2·47), number of comorbidities (two vs none: 4·50, 1·33-15·28), Eastern Cooperative Oncology Group performance status of 2 or higher (status of 2 vs 0 or 1: 3·89, 2·11-7·18), active cancer (progressing vs remission: 5·20, 2·77-9·77), and receipt of azithromycin plus hydroxychloroquine (vs treatment with neither: 2·93, 1·79-4·79; confounding by indication cannot be excluded). Compared with residence in the US-Northeast, residence in Canada (0·24, 0·07-0·84) or the US-Midwest (0·50, 0·28-0·90) were associated with decreased 30-day all-cause mortality. Race and ethnicity, obesity status, cancer type, type of anticancer therapy, and recent surgery were not associated with mortality., Interpretation: Among patients with cancer and COVID-19, 30-day all-cause mortality was high and associated with general risk factors and risk factors unique to patients with cancer. Longer follow-up is needed to better understand the effect of COVID-19 on outcomes in patients with cancer, including the ability to continue specific cancer treatments., Funding: American Cancer Society, National Institutes of Health, and Hope Foundation for Cancer Research., (Copyright © 2020 Elsevier Ltd. All rights reserved.)

The development of Fast Healthcare Interoperability Resources (FHIR) Genomics, a feasible and efficient method for exchanging complex clinical genomic data and interpretations, is described. FHIR Genomics is a subset of the emerging Health Level 7 FHIR standard and targets data from increasingly available technologies such as next-generation sequencing. Much care and integration of feedback have been taken to ease implementation, facilitate wide-scale interoperability, and enable modern app development toward a complete precision medicine standard. A new use case, the integration of the Variant Interpretation for Cancer Consortium (VICC) "meta-knowledgebase" into a third-party application, is described., Competing Interests: Competing interestsThe authors declare no competing interests., (© The Author(s) 2020.)

Hormonal therapy is an effective, but challenging, long-term treatment for patients with hormone-receptor-positive breast cancer. Raising the rate of patients who initiate therapy may be possible by characterizing the factors that influence a patient's decision. We hypothesized that online patient portal messages convey such factors. To investigate this hypothesis, we focused on breast cancer patients who were prescribed hormonal therapy at Vanderbilt University Medical Center and sent messages through the portal between diagnosis and therapy initiation. We first conducted a topic modeling analysis to generate the main themes of portal messages. We subsequently applied survival analysis to learn the association between the factors conveyed in messages, in term of semantic word groups, and the time elapsed from diagnosis to therapy initiation. We found that consulting with healthcare providers increased the probability of therapy initiation, while mentions of symptoms or negative emotions exhibited a reduced probability., (©2019 AMIA - 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.)

Natural language processing (NLP) technologies have been successfully applied to cancer research by enabling automated phenotypic information extraction from narratives in electronic health records (EHRs) such as pathology reports; however, developing customized NLP solutions requires substantial effort. To facilitate the adoption of NLP in cancer research, we have developed a set of customizable modules for extracting comprehensive types of cancer-related information in pathology reports (e.g., tumor size, tumor stage, and biomarkers), by leveraging the existing CLAMP system, which provides user-friendly interfaces for building customized NLP solutions for individual needs. Evaluation using annotated data at Vanderbilt University Medical Center showed that CLAMP-Cancer could extract diverse types of cancer information with good F-measures (0.80-0.98). We then applied CLAMP-Cancer to an information extraction task at Mayo Clinic and showed that we can quickly build a customized NLP system with comparable performance with an existing system at Mayo Clinic. CLAMP-Cancer is freely available for academic use.

Electronic health records (EHRs) have become ubiquitous tools and represent the standard of care for 96% of hospitals and 86% of ambulatory physicians in the United States. With adoption of EHRs came the promise of improved efficiency, higher-quality care, and lower costs. Unfortunately, some clinicians are now spending twice as much time on documentation as they spend seeing patients, and the documentation paradigm of problem-oriented medical records is contributing to this imbalance. It is time to consider new innovations. The collaborative wiki format offers many opportunities to ease the burden of documentation as well as to increase the usefulness of the recorded clinical data. Wikis support multiple authorship, have built-in features to track edits and changes, allow for contextual linkages (e.g., linking medical problems to their treatment), and support new technologies such as application programming interfaces, which allow for safe and secure exchange of information. In this Perspective, the authors describe the rationale for considering this approach to clinical documentation and propose a pilot to learn about its effectiveness. They believe wiki-based documentation will become increasingly attractive, especially as new legislation and directives from policymakers seek to reduce the crushing documentation burden and as the U.S. health care system transitions from an episode-based payment structure to a value-based, outcomes-focused system.

Hormonal therapy is an effective yet challenging long-term treatment for patients with hormone receptor positive breast cancer. Understanding what factors indicate discontinuation of a recommended hormonal therapy medication can help improve treatment experience. To date, studies on medication discontinuation have focused on patient information gathered through questionnaires, structured electronic medical records and online discussion boards. However, there has been little investigation into the communications between healthcare providers, which may provide additional indicators of patients' medication discontinuation, particularly from a clinical perspective. In this paper, we investigate the relation between such communications and hormonal therapy medication discontinuation. We studied a cohort of 2,579 patients on hormonal therapy at the Vanderbilt University Medical Center over a 16-year period. We adopt a data-driven approach to investigate the clinical messages communicated by their healthcare providers, the messaging patterns, topics they communicated, and the extent to which these factors are affiliated with discontinuation to a recommended 5-year treatment protocol. Our findings suggest that notification of unread messages, plans for clinical trials and the occurrence of treatment-related complications are affiliated with an increased risk of medication discontinuation. By contrast, ordering prescriptions, making appointments, using positive communication verbs, and noting patients' stable health conditions are affiliated with a decreased risk of medication discontinuation.

Metastatic patterns of spread at the time of cancer recurrence are one of the most important prognostic factors in estimation of clinical course and survival of the patient. This information is not easily accessible since it's rarely recorded in a structured format. This paper describes a system for categorization of pathology reports by specimen site and the detection of metastatic status within the report. A clinical NLP pipeline was developed using sentence boundary detection, tokenization, section identification, part-of-speech tagger, and chunker with some rule based methods to extract metastasis site and status in combination with five types of information related to tumor metastases: histological type, grade, specimen site, metastatic status indicators and the procedure. The system achieved a recall of 0.84 and 0.88 precision for metastatic status detection, and 0.89 recall and 0.93 precision for metastasis site detection. This study demonstrates the feasibility of applying NLP technologies to extract valuable metastases information from pathology reports and we believe that it will greatly benefit studies on cancer metastases that utilize EHRs.

Tumor registries are held to a very high standard for identifying and reporting new analytic cancer cases. However, current approaches to new case detection are often inefficient and costly. Efficient and effective detection of new cancer cases has the potential to maintain a high accuracy of reporting while reducing costs, increasing timeliness of reporting, and ultimately advancing cancer research. We describe the development, implementation, and evaluation of an informatics tool that integrates multiple data sources to support the workflow of new case identification at the Vanderbilt University Medical Center (VUMC) tumor registry office. The new system reduced the total number of potential cases to analyze from roughly 13,000 to 2,500 records per month. This resulted in an efficiency gain of roughly 80 man hours per month with a respective annual savings of approximately 50,000 dollars. Further iterative refinement of this approach along with support for case abstraction could result in further efficiencies.

Robust electronic medical records (EMR's) have made large-scale phenome-based analysis feasible. The context-dependent phenome of a large ICU-based EMR database (MIMIC II) was explored, as a function of a clinical feature: white blood cell count (WBC). Phenome visualization led to the discovery that peak WBC in the range 15-45 K/μl was highly associated with the diagnoses of Clostridium difficile and bacterial sepsis; thus, it is conceivable that clinicians might delay ordering targeted antimicrobials towards C. difficile for patients with peak WBC in this range. This hypothesis was confirmed, with significant delays in this group (median 135 vs. 85 hours, p = 0.002). These delays could be associated with adverse effects on patient health and high hospitalization costs (e.g. an additional $3,000,000 for the MIMIC II cohort). In conclusion, context-dependent clinical reference ranges are critical to clinical decision making; furthermore, important findings can be discovered through EMR-driven phenome association studies.