Genome-wide analysis provides genetic evidence that ACE2 influences COVID-19 risk and yields risk scores associated with severe disease.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) enters human host cells via angiotensin-converting enzyme 2 (ACE2) and causes coronavirus disease 2019 (COVID-19). Here, through a genome-wide association study, we identify a variant (rs190509934, minor allele frequency 0.2-2%) that downregulates ACE2 expression by 37% (P = 2.7 × 10-8) and reduces the risk of SARS-CoV-2 infection by 40% (odds ratio = 0.60, P = 4.5 × 10-13), providing human genetic evidence that ACE2 expression levels influence COVID-19 risk. We also replicate the associations of six previously reported risk variants, of which four were further associated with worse outcomes in individuals infected with the virus (in/near LZTFL1, MHC, DPP9 and IFNAR2). Lastly, we show that common variants define a risk score that is strongly associated with severe disease among cases and modestly improves the prediction of disease severity relative to demographic and clinical factors alone.

A Comparison between Human and NLP-based Annotation of Clinical Trial Eligibility Criteria Text Using The OMOP Common Data Model.

Human annotations are the established gold standard for evaluating natural language processing (NLP) methods. The goals of this study are to quantify and qualify the disagreement between human and NLP. We developed an NLP system for annotating clinical trial eligibility criteria text and constructed a manually annotated corpus, both following the OMOP Common Data Model (CDM). We analyzed the discrepancies between the human and NLP annotations and their causes (e.g., ambiguities in concept categorization and tacit decisions on inclusion of qualifiers and temporal attributes during concept annotation). This study initially reported complexities in clinical trial eligibility criteria text that complicate NLP and the limitations of the OMOP CDM. The disagreement between and human and NLP annotations may be generalizable. We discuss implications for NLP evaluation.

Pan-ancestry exome-wide association analyses of COVID-19 outcomes in 586,157 individuals.

Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) causes coronavirus disease 2019 (COVID-19), a respiratory illness that can result in hospitalization or death. We used exome sequence data to investigate associations between rare genetic variants and seven COVID-19 outcomes in 586,157 individuals, including 20,952 with COVID-19. After accounting for multiple testing, we did not identify any clear associations with rare variants either exome wide or when specifically focusing on (1) 13 interferon pathway genes in which rare deleterious variants have been reported in individuals with severe COVID-19, (2) 281 genes located in susceptibility loci identified by the COVID-19 Host Genetics Initiative, or (3) 32 additional genes of immunologic relevance and/or therapeutic potential. Our analyses indicate there are no significant associations with rare protein-coding variants with detectable effect sizes at our current sample sizes. Analyses will be updated as additional data become available, and results are publicly available through the Regeneron Genetics Center COVID-19 Results Browser.

Chia, a large annotated corpus of clinical trial eligibility criteria.

We present Chia, a novel, large annotated corpus of patient eligibility criteria extracted from 1,000 interventional, Phase IV clinical trials registered in ClinicalTrials.gov. This dataset includes 12,409 annotated eligibility criteria, represented by 41,487 distinctive entities of 15 entity types and 25,017 relationships of 12 relationship types. Each criterion is represented as a directed acyclic graph, which can be easily transformed into Boolean logic to form a database query. Chia can serve as a shared benchmark to develop and test future machine learning, rule-based, or hybrid methods for information extraction from free-text clinical trial eligibility criteria.

Ensembles of natural language processing systems for portable phenotyping solutions.

BACKGROUND: Manually curating standardized phenotypic concepts such as Human Phenotype Ontology (HPO) terms from narrative text in electronic health records (EHRs) is time consuming and error prone. Natural language processing (NLP) techniques can facilitate automated phenotype extraction and thus improve the efficiency of curating clinical phenotypes from clinical texts. While individual NLP systems can perform well for a single cohort, an ensemble-based method might shed light on increasing the portability of NLP pipelines across different cohorts. METHODS: We compared four NLP systems, MetaMapLite, MedLEE, ClinPhen and cTAKES, and four ensemble techniques, including intersection, union, majority-voting and machine learning, for extracting generic phenotypic concepts. We addressed two important research questions regarding automated phenotype recognition. First, we evaluated the performance of different approaches in identifying generic phenotypic concepts. Second, we compared the performance of different methods to identify patient-specific phenotypic concepts. To better quantify the effects caused by concept granularity differences on performance, we developed a novel evaluation metric that considered concept hierarchies and frequencies. Each of the approaches was evaluated on a gold standard set of clinical documents annotated by clinical experts. One dataset containing 1,609 concepts derived from 50 clinical notes from two different institutions was used in both evaluations, and an additional dataset of 608 concepts derived from 50 case report abstracts obtained from PubMed was used for evaluation of identifying generic phenotypic concepts only. RESULTS: For generic phenotypic concept recognition, the top three performers in the NYP/CUIMC dataset are union ensemble (F1, 0.634), training-based ensemble (F1, 0.632), and majority vote-based ensemble (F1, 0.622). In the Mayo dataset, the top three are majority vote-based ensemble (F1, 0.642), cTAKES (F1, 0.615), and MedLEE (F1, 0.559). In the PubMed dataset, the top three are majority vote-based ensemble (F1, 0.719), training-based (F1, 0.696) and MetaMapLite (F1, 0.694). For identifying patient specific phenotypes, the top three performers in the NYP/CUIMC dataset are majority vote-based ensemble (F1, 0.610), MedLEE (F1, 0.609), and training-based ensemble (F1, 0.585). In the Mayo dataset, the top three are majority vote-based ensemble (F1, 0.604), cTAKES (F1, 0.531) and MedLEE (F1, 0.527). CONCLUSIONS: Our study demonstrates that ensembles of natural language processing can improve both generic phenotypic concept recognition and patient specific phenotypic concept identification over individual systems. Among the individual NLP systems, each individual system performed best when they were applied in the dataset that they were primary designed for. However, combining multiple NLP systems to create an ensemble can generally improve the performance. Specifically, the ensemble can increase the results reproducibility across different cohorts and tasks, and thus provide a more portable phenotyping solution compared to individual NLP systems.

Doc2Hpo: a web application for efficient and accurate HPO concept curation.

We present Doc2Hpo, an interactive web application that enables interactive and efficient phenotype concept curation from clinical text with automated concept normalization using the Human Phenotype Ontology (HPO). Users can edit the HPO concepts automatically extracted by Doc2Hpo in real time, and export the extracted HPO concepts into gene prioritization tools. Our evaluation showed that Doc2Hpo significantly reduced manual effort while achieving high accuracy in HPO concept curation. Doc2Hpo is freely available at https://impact2.dbmi.columbia.edu/doc2hpo/. The source code is available at https://github.com/stormliucong/doc2hpo for local installation for protected health data.

Risk of Alzheimer's Disease Among Senior Medicare Beneficiaries Treated With Androgen Deprivation Therapy for Prostate Cancer.

Purpose To assess the relative risk of Alzheimer's disease (AD) among patients with prostate cancer who received androgen deprivation therapy (ADT), after adjustment for other cancer therapies. Methods Data from demographics, survival, diagnoses codes, procedure codes, and other information about beneficiaries age 67 years or older in the Medicare claims database was assessed to determine the unadjusted and adjusted risks of AD and of dementia from ADT. The prespecified survival analysis method was competing risk regression. Results Of the 1.2 million fee-for-service Medicare beneficiaries who developed prostate cancer in 2001 to 2014, 35% received ADT. Of these, 109,815 (8.9%) and 223,765 (18.8%) developed AD and dementia, respectively, and 26% to 33% died without either outcome. Unadjusted rates of AD and all-cause mortality per 1,000 patient-years were higher among ADT recipients; the unadjusted rates of AD were 17.0 and 15.5 per 1,000 person-years in recipients and nonrecipients, respectively, and the unadjusted rates of all-cause mortality were 73.0 and 51.6 per 1,000 person-years, respectively. The unadjusted rates for dementia in ADT recipients versus nonrecipients were 38.5 and 32.9, respectively, and the unadjusted rates of mortality were 60.2 versus 40.4, respectively. However, after analysis was adjusted for other cancer therapies and other covariates, patients with ADT treatment had no increased risk of AD (subdistribution hazard ratio [SHR], 0.98; 95% CI, 0.97 to 0.99) and had only a miniscule (1%) risk of dementia (SHR, 1.01; 95% CI, 1.01 to 1.02); patients treated with ADT were more likely to die before progression to AD (SHR, 1.24; 95% CI, 1.23 to 1.24) or dementia (SHR, 1.26; 95% CI, 1.25 to 1.26). The risks of AD and dementia were not associated with duration of ADT (ie, no dose effect). Other secondary analyses confirmed these results. Conclusion These data suggest that ADT treatment has no hazard for AD and no meaningful hazard for dementia among men age 67 years or older who are enrolled in Medicare.

Analysis of Healthcare Cost and Utilization in the First Two Years of the Medicare Shared Savings Program Using Big Data from the CMS Enclave.

The Medicare Shared Savings Program (MSSP) is the larger of the first two Accountable Care Organization (ACO) programs by the Centers for Medicare and Medicaid Services (CMS). In this study we assessed healthcare cost and utilization of 1.71 million Medicare beneficiaries assigned to the 333 MSSP ACOs in the calendar years of 2013 and 2014, in comparison to years 2010 and 2011, using the official CMS data. We employed doubly robust estimation (propensity score weighting followed by generalized linear regression) to adjust the analyses to beneficiary personal traits, history of chronic conditions, previous healthcare utilization, ACO administrative region, and ZIP code socioeconomic factors. In comparison to the care delivered to the control cohort of 17.7 million non-ACO beneficiaries, we found that the care patterns for ACO beneficiaries shifted away from some costly types of care, but at the expense of increased utilization of other types, increased imaging and testing expenditures, and increased medication use, with overall net greater increase in cost instead of smaller increase.

Identifying Repetitive Institutional Review Board Stipulations by Natural Language Processing and Network Analysis.

The corrections ("stipulations") to a proposed research study protocol produced by an institutional review board (IRB) can often be repetitive across many studies; however, there is no standard set of stipulations that could be used, for example, by researchers wishing to anticipate and correct problems in their research proposals prior to submitting to an IRB. The objective of the research was to computationally identify the most repetitive types of stipulations generated in the course of IRB deliberations. The text of each stipulation was normalized using the natural language processing techniques. An undirected weighted network was constructed in which each stipulation was represented by a node, and each link, if present, had weight corresponding to the TF-IDF Cosine Similarity of the stipulations. Network analysis software was then used to identify clusters in the network representing similar stipulations. The final results were correlated with additional data to produce further insights about the IRB workflow. From a corpus of 18,582 stipulations we identified 31 types of repetitive stipulations. Those types accounted for 3,870 stipulations (20.8% of the corpus) produced for 697 (88.7%) of all protocols in 392 (also 88.7%) of all the CNS IRB meetings with stipulations entered in our data source. A notable peroportion of the corrections produced by the IRB can be considered highly repetitive. Our shareable method relied on a minimal manual analysis and provides an intuitive exploration with theoretically unbounded granularity. Finer granularity allowed for the insight that is anticipated to prevent the need for identifying the IRB panel expertise or any human supervision.

Computationally Comparing and Analyzing All Published Scoring Systems for Diagnosis of Disseminated Intravascular Coagulation.

The clinical literature presents four different scoring systems (SS) for the diagnosis of disseminated intravascular coagulation (DIC) by four institutions: ISTH, JMHLW, JAAM and KSTH. In this study a Java program was written to retrieve medical records from the MIMIC-II database and apply the criteria of all four. The program then quantified the agreement of each DIC SS with each other and demonstrated notorious dissent. Furthermore, the average internal composition of each score was also quantified. All source code produced is available for download at https://github.com/fabkury/hedicim.

Reproducing a Prospective Clinical Study as a Computational Retrospective Study in MIMIC-II.

In this paper we sought to reproduce, as a computational retrospective study in an EHR database (MIMIC-II), a recent large prospective clinical study: the 2013 publication, by the Japanese Association for Acute Medicine (JAAM), about disseminated intravascular coagulation, in the journal Critical Care (PMID: 23787004). We designed in SQL and Java a set of electronic phenotypes that reproduced the study's data sampling, and used R to perform the same statistical inference procedures. All produced source code is available online at https://github.com/fabkury/paamia2015. Our program identified 2,257 eligible patients in MIMIC-II, and the results remarkably agreed with the prospective study. A minority of the needed data elements was not found in MIMIC-II, and statistically significant inferences were possible in the majority of the cases.

The State and Trends of Barcode, RFID, Biometric and Pharmacy Automation Technologies in US Hospitals.

The standard of safe medication practice requires strict observance of the five rights of medication administration: the right patient, drug, time, dose, and route. Despite adherence to these guidelines, medication errors remain a public health concern that has generated health policies and hospital processes that leverage automation and computerization to reduce these errors. Bar code, RFID, biometrics and pharmacy automation technologies have been demonstrated in literature to decrease the incidence of medication errors by minimizing human factors involved in the process. Despite evidence suggesting the effectivity of these technologies, adoption rates and trends vary across hospital systems. The objective of study is to examine the state and adoption trends of automatic identification and data capture (AIDC) methods and pharmacy automation technologies in U.S. hospitals. A retrospective descriptive analysis of survey data from the HIMSS Analytics® Database was done, demonstrating an optimistic growth in the adoption of these patient safety solutions.