Citations

PURPOSEThe 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.METHODSPublished 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.RESULTSA 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.CONCLUSIONWe 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.

Objective This study aimed to examine the relationship between covert cerebrovascular disease, comprised of covert brain infarction and white matter disease, discovered incidentally in routine care, and subsequent Parkinson disease. Methods Patients were ≥50 years and received neuroimaging for non‐stroke indications in the Kaiser Permanente Southern California system from 2009 to 2019. Natural language processing identified incidentally discovered covert brain infarction and white matter disease and classified white matter disease severity. The Parkinson disease outcome was defined as 2 ICD diagnosis codes. Results 230,062 patients were included (median follow‐up 3.72 years). A total of 1,941 Parkinson disease cases were identified (median time‐to‐event 2.35 years). Natural language processing identified covert cerebrovascular disease in 70,592 (30.7%) patients, 10,622 (4.6%) with covert brain infarction and 65,814 (28.6%) with white matter disease. After adjustment for known risk factors, white matter disease was associated with Parkinson disease (hazard ratio 1.67 [95%CI, 1.44, 1.93] for patients <70 years and 1.33 [1.18, 1.50] for those ≥70 years). Greater severity of white matter disease was associated with increased incidence of Parkinson disease(/1,000 person‐years), from 1.52 (1.43, 1.61) in patients without white matter disease to 4.90 (3.86, 6.13) in those with severe disease. Findings were robust when more specific definitions of Parkinson disease were used. Covert brain infarction was not associated with Parkinson disease (adjusted hazard ratio = 1.05 [0.88, 1.24]). Interpretation Incidentally discovered white matter disease was associated with subsequent Parkinson disease, an association strengthened with younger age and increased white matter disease severity. Incidentally discovered covert brain infarction did not appear to be associated with subsequent Parkinson disease. ANN NEUROL 2022;92:620–630

Precision oncology is expected to improve selection of targeted therapies, tailored to individual patients and ultimately improve cancer patients’ outcomes. Several cancer genetics knowledge databases have been successfully developed for such purposes, including CIViC and OncoKB, with active community-based curations and scoring of genetic-treatment evidences. Although many studies were conducted based on each knowledge base respectively, the integrative analysis across both knowledge bases remains largely unexplored. Thus, there exists an urgent need for a heterogeneous precision oncology knowledge resource with computational power to support drug repurposing discovery in a timely manner, especially for life-threatening cancer. In this pilot study, we built a heterogeneous precision oncology knowledge resource (POKR) by integrating CIViC and OncoKB, in order to incorporate unique information contained in each knowledge base and make associations amongst biomedical entities (e.g., gene, drug, disease) computable and measurable via training POKR graph embeddings. All the relevant codes, database dump files, and pre-trained POKR embeddings can be accessed through the following URL: https://github.com/shenfc/POKR.

Reproducibility is an important quality criterion for the secondary use of electronic health records (EHRs). However, multiple barriers to reproducibility are embedded in the heterogeneous EHR environment. These barriers include complex processes for collecting and organizing EHR data and dynamic multi-level interactions occurring during information use (e.g., inter-personal, inter-system, and cross-institutional). To ensure reproducible use of EHRs, we investigated four information quality dimensions and examine the implications for reproducibility based on a real-world EHR study. Four types of IQ measurements suggested that barriers to reproducibility occurred for all stages of secondary use of EHR data. We discussed our recommendations and emphasized the importance of promoting transparent, high-throughput, and accessible data infrastructures and implementation best practices (e.g., data quality assessment, reporting standard).

e24070 Background: We have maintained a living interactive systematic review (LISR) on the treatment of CAT ( CAT LISR ). Recent publications of CASTA-DIVA and CANVAS trials prompted this report of updated analysis. Methods: The approach to creating LISRs has been previously published and is available on our website. ( LIvE synthesis framework ). Patient important outcomes included venous thromboembolism (VTE) recurrence, major bleeding (MB), clinically relevant non-major bleeding (CRNMB), net clinical benefit (NCB) and all-cause mortality. Mantel-Haenszel method was used to pool treatment effect estimates. A Paul-Mendel random-effects meta-analysis was conducted to compare DOACs with LMWH. Mixed-treatment comparisons were computed using fixed-effect model within a Bayesian framework owing to a sparse and open network. Effect estimates were expressed as odds ratio (OR) with 95% confidence intervals (CI). Surface under the cumulative ranking curve (SUCRA) was used to assess relative treatment rankings. Results: This systematic review currently includes six RCTs with a total of 3690 patients and four unique treatment choices. Apixaban was assessed in two trials (ADAM-VTE, CARAVAGGIO), rivaroxaban in two (SELECT-D, CASTA-DIVA) and edoxaban in one (HOKUSAI-VTE). CANVAS trial reported efficacy of DOACs as a class and hence, was only included in direct comparisons with LMWH. Direct comparisons showed that DOACs as a class significantly decreased the odds of VTE recurrence (OR 0.63, 95% CI: 0.49-0.82) without significantly increasing major bleeding (OR 1.23, 95% CI: 0.84-1.79) when compared to LMWH. However, higher odds of CRNMB (OR 1.65, 95% CI: 1.18-2.31) was observed with DOACs when compared to LMWH. Mortality rates were not different between the two groups (OR: 1.03, 95% CI: 0.87-1.21). Mixed treatment comparisons showed decreased VTE recurrences with rivaroxaban (rank 1; OR 0.49, 95% CI: 0.23-0.99) and apixaban (rank 2; OR 0.58, 95% CI: 0.36-0.90) relative to LMWH (rank 4). No significant differences were observed between edoxaban (rank 3) and LMWH in terms of VTE risk reduction. Edoxaban (rank 4) increased the risk of major bleeding (OR 1.77, 95% CI: 1.05-3.15) compared to LMWH (rank 2); there were no significant differences among rivaroxaban (rank 3) apixaban (rank 1), and LMWH. Rivaroxaban (rank 4; OR 3.16; 95% CI: 1.61-6.68) and apixaban (rank 3; OR 1.51, 95% CI: 1.01-2.28) increased the risk of CRNMB compared to LMWH (rank 1). Significant net clinical benefit was observed with apixaban (rank 1; OR 0.66; 95% CI: 0.46-0.94) when compared to LMWH (rank 4). No significant differences were observed among other mixed treatment comparisons. Conclusions: Patients treated with DOACs have 31 fewer VTE events per 1000 CAT patients compared to LMWH. Amongst DOACs, apixaban offers the greatest net clinical benefit with 39 fewer VTE and MB events per 1000 patients compared to LMWH.

Abstract Internal validation is the most popular evaluation strategy used for drug–target predictive models. The simple random shuffling in the cross-validation, however, is not always ideal to handle large, diverse and copious datasets as it could potentially introduce bias. Hence, these predictive models cannot be comprehensively evaluated to provide insight into their general performance on a variety of use-cases (e.g. permutations of different levels of connectiveness and categories in drug and target space, as well as validations based on different data sources). In this work, we introduce a benchmark, BETA, that aims to address this gap by (i) providing an extensive multipartite network consisting of 0.97 million biomedical concepts and 8.5 million associations, in addition to 62 million drug–drug and protein–protein similarities and (ii) presenting evaluation strategies that reflect seven cases (i.e. general, screening with different connectivity, target and drug screening based on categories, searching for specific drugs and targets and drug repurposing for specific diseases), a total of seven Tests (consisting of 344 Tasks in total) across multiple sampling and validation strategies. Six state-of-the-art methods covering two broad input data types (chemical structure- and gene sequence-based and network-based) were tested across all the developed Tasks. The best-worst performing cases have been analyzed to demonstrate the ability of the proposed benchmark to identify limitations of the tested methods for running over the benchmark tasks. The results highlight BETA as a benchmark in the selection of computational strategies for drug repurposing and target discovery.

Background: Concerns for side-effects makes the identification of treatment targets of hypertension in the very elderly challenging. We sought to utilize natural language processing (NLP) to evaluate the prevalence of prespecified side-effects in a cohort of very elderly (≥80 years) patients. Methods and Results: We conducted a two-year, cross-sectional study involving 5169 very elderly patients with and without hypertension seen by primary or specialist providers. Hypertension patients had to have a prior hypertension ICD-9 code in the electronic medical records problem list and ≥2 office visits in 2008 and 2009. We used NLP to analyze extracted free text reports from last clinic visits between 2008 and 2009. Of the 10 prespecified side-effects, hypotension, fatigue, and dizziness were the commonest three conditions and hypotension was significantly higher in the HTN patients. Of the 2835 patients with no prior hypertension history and 2334 hypertension patients identified, the prevalence of each prespecified item in both groups were: acute renal failure (0.3% v 0.6%, p=0.13); syncope (2.8% v 2.2%, p = 0.17), dizziness (3.1% v 4.0%, p=0.09), history of fall (1.0% v 0.9%, p=0.65), fatigue (5.4% v 4.5%, p=0.19), femoral neck fracture (0.18% v 0.04%, p=0.16), palpitations (1.2% v 1.5%, p=0.35), hypotension (4.1% v 6.6%, p<0.001), hyperkalemia (1.1% v 1.7%, p=0.04), hypokalemia (0.4% v 0.9%, p=0.01). At least one presumed side-effect was present in 17.5% of all 5169 patients. Among hypertension patients, there was no significant difference in the prevalence of the side-effects among patients with BP above (578) or below 140/90 (1756) except for acute renal failure (1.21% vs 0.40%, p = 0.03). Conclusion: Analysis of free-text outpatient notes using NLP showed that hypotension and potassium abnormalities were significantly more prevalent in very elderly patients with hypertension compared to similarly aged patients with no hypertension history.

Background Inhaled corticosteroids (ICSs) are important in asthma management, but there are concerns regarding associated risk of pneumonia. While studies in asthmatic adults have shown inconsistent results, this risk in asthmatic children is unclear. Objective Our aim was to determine the association of ICS use with pneumonia risk in asthmatic children. Methods A nested case-control study was performed in the Mayo Clinic Birth Cohort. Asthmatic children (<18 years) with a physician diagnosis of asthma were identified from electronic medical records of children born at Mayo Clinic from 1997 to 2016 and followed until 31 December 2017. Pneumonia cases defined by Infectious Disease Society of America were 1:1 matched with controls without pneumonia by age, sex and asthma index date. Exposure was defined as ICS prescription at least 90 days prior to pneumonia. Associations of ICS use, type and dose (low, medium and high) with pneumonia risk were analysed using conditional logistic regression. Results Of the 2108 asthmatic children eligible for the study (70% mild intermittent and 30% persistent asthma), 312 children developed pneumonia during the study period. ICS use overall was not associated with risk of pneumonia (adjusted OR: 0.94, 95% CI: 0.62 to 1.41). Poorly controlled asthma was significantly associated with the risk of pneumonia (OR: 2.03, 95% CI: 1.35 to 3.05; p<0.001). No ICS type or dose was associated with risk of pneumonia. Conclusion ICS use in asthmatic children was not associated with risk of pneumonia but poorly controlled asthma was. Future asthma studies may need to include pneumonia as a potential outcome of asthma management.

Abstract Objective Accurate extraction of breast cancer patients’ phenotypes is important for clinical decision support and clinical research. This study developed and evaluated cancer domain pretrained CancerBERT models for extracting breast cancer phenotypes from clinical texts. We also investigated the effect of customized cancer-related vocabulary on the performance of CancerBERT models. Materials and Methods A cancer-related corpus of breast cancer patients was extracted from the electronic health records of a local hospital. We annotated named entities in 200 pathology reports and 50 clinical notes for 8 cancer phenotypes for fine-tuning and evaluation. We kept pretraining the BlueBERT model on the cancer corpus with expanded vocabularies (using both term frequency-based and manually reviewed methods) to obtain CancerBERT models. The CancerBERT models were evaluated and compared with other baseline models on the cancer phenotype extraction task. Results All CancerBERT models outperformed all other models on the cancer phenotyping NER task. Both CancerBERT models with customized vocabularies outperformed the CancerBERT with the original BERT vocabulary. The CancerBERT model with manually reviewed customized vocabulary achieved the best performance with macro F1 scores equal to 0.876 (95% CI, 0.873–0.879) and 0.904 (95% CI, 0.902–0.906) for exact match and lenient match, respectively. Conclusions The CancerBERT models were developed to extract the cancer phenotypes in clinical notes and pathology reports. The results validated that using customized vocabulary may further improve the performances of domain specific BERT models in clinical NLP tasks. The CancerBERT models developed in the study would further help clinical decision support.

AbstractPurposeRural populations are disproportionately affected by the COVID‐19 pandemic. We characterized urban‐rural disparities in patient portal messaging utilization for COVID‐19, and, of those who used the portal during its early stage in the Midwest.MethodsWe collected over 1 million portal messages generated by midwestern Mayo Clinic patients from February to August 2020. We analyzed patient‐generated messages (PGMs) on COVID‐19 by urban‐rural locality and incorporated patients’ sociodemographic factors into the analysis.FindingsThe urban‐rural ratio of portal users, message senders, and COVID‐19 message senders was 1.18, 1.31, and 1.79, indicating greater use among urban patients. The urban‐rural ratio (1.69) of PGMs on COVID‐19 was higher than that (1.43) of general PGMs. The urban‐rural ratios of messaging were 1.72‐1.85 for COVID‐19‐related care and 1.43‐1.66 for other health care issues on COVID‐19. Compared with urban patients, rural patients sent fewer messages for COVID‐19 diagnosis and treatment but more messages for other reasons related to COVID‐19‐related health care (eg, isolation and anxiety). The frequent senders of COVID‐19‐related messages among rural patients were 40+ years old, women, married, and White.ConclusionsIn this Midwest health system, rural patients were less likely to use patient online services during a pandemic and their reasons for its use differ from urban patients. Results suggest opportunities for increasing equity in rural patient engagement in patient portals (in particular, minority populations) for COVID‐19. Public health intervention strategies could target reasons why rural patients might seek health care in a pandemic, such as social isolation and anxiety.