Source Electronic Health Record Research Articles

Vaccine Effectiveness Against Long COVID in Children.

Vaccination reduces the risk of acute coronavirus disease 2019 (COVID-19) in children, but it is less clear whether it protects against long COVID. We estimated vaccine effectiveness (VE) against long COVID in children aged 5 to 17 years. This retrospective cohort study used data from 17 health systems in the RECOVER PCORnet electronic health record program for visits after vaccine availability. We examined both probable (symptom-based) and diagnosed long COVID after vaccination. The vaccination rate was 67% in the cohort of 1 037 936 children. The incidence of probable long COVID was 4.5% among patients with COVID-19, whereas diagnosed long COVID was 0.8%. Adjusted vaccine effectiveness within 12 months was 35.4% (95 CI 24.5-44.7) against probable long COVID and 41.7% (15.0-60.0) against diagnosed long COVID. VE was higher for adolescents (50.3% [36.6-61.0]) than children aged 5 to 11 (23.8% [4.9-39.0]). VE was higher at 6 months (61.4% [51.0-69.6]) but decreased to 10.6% (-26.8% to 37.0%) at 18-months. This large retrospective study shows moderate protective effect of severe acute respiratory coronavirus 2 vaccination against long COVID. The effect is stronger in adolescents, who have higher risk of long COVID, and wanes over time. Understanding VE mechanism against long COVID requires more study, including electronic health record sources and prospective data.

Sexual orientation and gender identity measures and viral suppression for people living with HIV: a protocol for a population-based cohort study

IntroductionThe measure of sexual orientation and gender identity (SOGI) data in electronic health records (EHR) has been critical for addressing health disparities and inequalities, especially for HIV care. Given that...

Methods for Identification of Blood Product Transfusions in a Cohort of Myelofibrosis Patients: An Assessment of Data Capture Using Insurance Claims and Electronic Health Record Abstraction

Introduction Transfusion data completeness is critical in assessing real-world response and progression in myeloid malignancies. In the real-world setting transfusion data can be collected via insurance claims or by Electronic Health Record (EHR) abstraction. Transfusion events captured from claims have previously been shown to be 71% sensitive and 92% specific compared to EHR abstraction in the community hospital setting, however this has not been evaluated in outpatient community oncology setting (Howard DH, et al. Transfusion Medicine. 2016;26:457-459). Our objective was to assess the completeness of transfusion data from claims in the The Komodo Healthcare Map™(KHM) to EHR abstraction in the Flatiron Health Research Database (FHRD) in a group of patients with myelofibrosis (MF). Methods This study used the nationwide Flatiron Health EHR-derived de-identified database and The Komodo Healthcare Map™, a database of de-identified claim-based healthcare encounters from insured patients in the US (2015-2022).This is a cross-sectional study of a cohort of patients with MF that underwent EHR abstraction (EHRC) (n = 507) from FHRD, capturing transfusion dates, and number of units from date of diagnosis up to data cutoff on 30 November 2022 (observation window). Those patients were linked to KHM to create a Linked Claims Cohort (LCC), and a sub-cohort of patients with any closed claims during the observation window to make a Closed Claims Cohort (CCC), where transfusions were assessed by reviewing transfusion-related HCPCS and ICD codes. Analyses were performed on the matched patients in the LCC and CCC, respectively. To assess the level of concordance between claims and EHR data sources on the number of transfusions per patient the intraclass correlation coefficient (ICC) was used. Also, percent overlap and sensitivity, specificity, and positive and negative predictive values were calculated. Results Out of a total of 507 in the EHRC, 498 were found in KHM and included in the LCC, and 245 had closed claims and comprised the CCC. Demographics are listed below (Table 1) and are similar between groups, however the CCC had a higher percentage of patients with more recent diagnosis. A total of 319 (64%) vs 168 (34%) in the LCC, and 143 (58%) vs 95 (39%) patients in the CCC had at least one transfusion identified by EHR and claims data sources, respectively (Table 2). In the LCC, there was 64.1% agreement of having a transfusion across data sources, with a kappa of 0.34 (p = 0), and in the CCC, there was 72.24% agreement, with a kappa of 0.47 (p = 0). The unweighted kappa for the number of transfusions recorded during the observation period was 0.11 (p = 0) in the LCC and 0.18 (p = 0) in the CCC. Treating the EHR source as the source of truth, the claims data source in LCC had a sensitivity of 48.28%, specificity of 92.18%, PPV of 91.67%, and NPV of 50% to identify transfusion events, while the claims in CCC had a sensitivity of 59.44%, specificity of 90.20%, PPV of 89.47%, and NPV of 61.33% Conclusions There was an observed low agreement between transfusions identified via EHR abstraction and Claims data sources, with the EHR having more transfusions identified, but a moderate agreement when selecting for patients with closed claims. As a source for transfusion data EHR provided more data points during the observation window for this group of patients. Future investigation of transfusions in patients with continuous closed claims during the observation window is warranted to determine if it improves data capture. This study demonstrates that EHR abstraction may have higher utility for identifying transfusions with lower attrition in a cohort of MF patients with long observation periods, and with variable insurance coverage in the outpatient oncology setting.

EHR-QC: A streamlined pipeline for automated electronic health records standardisation and preprocessing to predict clinical outcomes

The adoption of electronic health records (EHRs) has created opportunities to analyse historical data for predicting clinical outcomes and improving patient care. However, non-standardised data representations and anomalies pose major challenges to the use of EHRs in digital health research. To address these challenges, we have developed EHR-QC, a tool comprising two modules: the data standardisation module and the preprocessing module. The data standardisation module migrates source EHR data to a standard format using advanced concept mapping techniques, surpassing expert curation in benchmarking analysis. The preprocessing module includes several functions designed specifically to handle healthcare data subtleties. We provide automated detection of data anomalies and solutions to handle those anomalies. We believe that the development and adoption of tools like EHR-QC is critical for advancing digital health. Our ultimate goal is to accelerate clinical research by enabling rapid experimentation with data-driven observational research to generate robust, generalisable biomedical knowledge.

Advancing Toward a Common Data Model in Ophthalmology: Gap Analysis of General Eye Examination Concepts to Standard Observational Medical Outcomes Partnership Concepts

Evaluate the degree of concept coverage of the general eye examination in one widely used electronic health record (EHR) system using the Observational Health Data Sciences and Informatics Observational Medical Outcomes Partnership (OMOP) common data model (CDM). Study of data elements. Not applicable. Data elements (field names and predefined entry values) from the general eye examination in the Epic foundation system were mapped to OMOP concepts and analyzed. Each mapping was given a Health Level 7 equivalence designation-equal when the OMOP concept had the same meaning as the source EHR concept, wider when it was missing information, narrower when it was overly specific, and unmatched when there was no match. Initial mappings were reviewed by 2 graders. Intergrader agreement for equivalence designation was calculated using Cohen's kappa. Agreement on the mapped OMOP concept was calculated as a percentage of total mappable concepts. Discrepancies were discussed and a final consensus created. Quantitative analysis was performed on wider and unmatched concepts. Gaps in OMOP concept coverage of EHR elements and intergrader agreement of mapped OMOP concepts. A total of 698 data elements (210 fields, 488 values) from the EHR were analyzed. The intergrader kappa on the equivalence designation was 0.88 (standard error 0.03, P<0.001). There was a 96% agreement on the mapped OMOP concept. In the final consensus mapping, 25% (1% fields, 31% values) of the EHR to OMOP concept mappings were considered equal, 50% (27% fields, 60% values) wider, 4% (8% fields, 2% values) narrower, and 21% (52% fields, 8% values) unmatched. Of the wider mapped elements, 46% were missing the laterality specification, 24% had other missing attributes, and 30% had both issues. Wider and unmatched EHR elements could be found in all areas of the general eye examination. Most data elements in the general eye examination could not be represented precisely using the OMOP CDM. Our work suggests multiple ways to improve the incorporation of important ophthalmology concepts in OMOP, including adding laterality to existing concepts. There exists a strong need to improve the coverage of ophthalmic concepts in source vocabularies so that the OMOP CDM can better accommodate vision research. Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Advancing heart failure research using machine learning

Advancing heart failure research using machine learning

Identifying subtypes of heart failure from three electronic health record sources with machine learning: an external, prognostic, and genetic validation study

Machine learning has been used to analyse heart failure subtypes, but not across large, distinct, population-based datasets, across the whole spectrum of causes and presentations, or with clinical and non-clinical validation by different machine learning methods. Using our published framework, we aimed to discover heart failure subtypes and validate them upon population representative data. In this external, prognostic, and genetic validation study we analysed individuals aged 30 years or older with incident heart failure from two population-based databases in the UK (Clinical Practice Research Datalink [CPRD] and The Health Improvement Network [THIN]) from 1998 to 2018. Pre-heart failure and post-heart failure factors (n=645) included demographic information, history, examination, blood laboratory values, and medications. We identified subtypes using four unsupervised machine learning methods (K-means, hierarchical, K-Medoids, and mixture model clustering) with 87 of 645 factors in each dataset. We evaluated subtypes for (1) external validity (across datasets); (2) prognostic validity (predictive accuracy for 1-year mortality); and (3) genetic validity (UK Biobank), association with polygenic risk score (PRS) for heart failure-related traits (n=11), and single nucleotide polymorphisms (n=12). We included 188 800, 124 262, and 9573 individuals with incident heart failure from CPRD, THIN, and UK Biobank, respectively, between Jan 1, 1998, and Jan 1, 2018. After identifying five clusters, we labelled heart failure subtypes as (1) early onset, (2) late onset, (3) atrial fibrillation related, (4) metabolic, and (5) cardiometabolic. In the external validity analysis, subtypes were similar across datasets (c-statistics: THIN model in CPRD ranged from 0·79 [subtype 3] to 0·94 [subtype 1], and CPRD model in THIN ranged from 0·79 [subtype 1] to 0·92 [subtypes 2 and 5]). In the prognostic validity analysis, 1-year all-cause mortality after heart failure diagnosis (subtype 1 0·20 [95% CI 0·14-0·25], subtype 2 0·46 [0·43-0·49], subtype 3 0·61 [0·57-0·64], subtype 4 0·11 [0·07-0·16], and subtype 5 0·37 [0·32-0·41]) differed across subtypes in CPRD and THIN data, as did risk of non-fatal cardiovascular diseases and all-cause hospitalisation. In the genetic validity analysis the atrial fibrillation-related subtype showed associations with the related PRS. Late onset and cardiometabolic subtypes were the most similar and strongly associated with PRS for hypertension, myocardial infarction, and obesity (p<0·0009). We developed a prototype app for routine clinical use, which could enable evaluation of effectiveness and cost-effectiveness. Across four methods and three datasets, including genetic data, in the largest study of incident heart failure to date, we identified five machine learning-informed subtypes, which might inform aetiological research, clinical risk prediction, and the design of heart failure trials. European Union Innovative Medicines Initiative-2.

Replication of Real-World Evidence in Oncology Using Electronic Health Record Data Extracted by Machine Learning.

Meaningful real-world evidence (RWE) generation requires unstructured data found in electronic health records (EHRs) which are often missing from administrative claims; however, obtaining relevant data from unstructured EHR sources is resource-intensive. In response, researchers are using natural language processing (NLP) with machine learning (ML) techniques (i.e., ML extraction) to extract real-world data (RWD) at scale. This study assessed the quality and fitness-for-use of EHR-derived oncology data curated using NLP with ML as compared to the reference standard of expert abstraction. Using a sample of 186,313 patients with lung cancer from a nationwide EHR-derived de-identified database, we performed a series of replication analyses demonstrating some common analyses conducted in retrospective observational research with complex EHR-derived data to generate evidence. Eligible patients were selected into biomarker- and treatment-defined cohorts, first with expert-abstracted then with ML-extracted data. We utilized the biomarker- and treatment-defined cohorts to perform analyses related to biomarker-associated survival and treatment comparative effectiveness, respectively. Across all analyses, the results differed by less than 8% between the data curation methods, and similar conclusions were reached. These results highlight that high-performance ML-extracted variables trained on expert-abstracted data can achieve similar results as when using abstracted data, unlocking the ability to perform oncology research at scale.

A Comprehensive and Improved Definition for Hospital-Acquired Pressure Injury Classification Based on Electronic Health Records: Comparative Study

Patients develop pressure injuries (PIs) in the hospital owing to low mobility, exposure to localized pressure, circulatory conditions, and other predisposing factors. Over 2.5 million Americans develop PIs annually. The Center for Medicare and Medicaid considers hospital-acquired PIs (HAPIs) as the most frequent preventable event, and they are the second most common claim in lawsuits. With the growing use of electronic health records (EHRs) in hospitals, an opportunity exists to build machine learning models to identify and predict HAPI rather than relying on occasional manual assessments by human experts. However, accurate computational models rely on high-quality HAPI data labels. Unfortunately, the different data sources within EHRs can provide conflicting information on HAPI occurrence in the same patient. Furthermore, the existing definitions of HAPI disagree with each other, even within the same patient population. The inconsistent criteria make it impossible to benchmark machine learning methods to predict HAPI. The objective of this project was threefold. We aimed to identify discrepancies in HAPI sources within EHRs, to develop a comprehensive definition for HAPI classification using data from all EHR sources, and to illustrate the importance of an improved HAPI definition. We assessed the congruence among HAPI occurrences documented in clinical notes, diagnosis codes, procedure codes, and chart events from the Medical Information Mart for Intensive Care III database. We analyzed the criteria used for the 3 existing HAPI definitions and their adherence to the regulatory guidelines. We proposed the Emory HAPI (EHAPI), which is an improved and more comprehensive HAPI definition. We then evaluated the importance of the labels in training a HAPI classification model using tree-based and sequential neural network classifiers. We illustrate the complexity of defining HAPI, with <13% of hospital stays having at least 3 PI indications documented across 4 data sources. Although chart events were the most common indicator, it was the only PI documentation for >49% of the stays. We demonstrate a lack of congruence across existing HAPI definitions and EHAPI, with only 219 stays having a consensus positive label. Our analysis highlights the importance of our improved HAPI definition, with classifiers trained using our labels outperforming others on a small manually labeled set from nurse annotators and a consensus set in which all definitions agreed on the label. Standardized HAPI definitions are important for accurately assessing HAPI nursing quality metric and determining HAPI incidence for preventive measures. We demonstrate the complexity of defining an occurrence of HAPI, given the conflicting and incomplete EHR data. Our EHAPI definition has favorable properties, making it a suitable candidate for HAPI classification tasks.

Mapping progress in intravascular catheter quality surveillance: An Australian case study of electronic medical record data linkage.

Background and significanceIntravascular (IV) catheters are the most invasive medical device in healthcare. Localized priority-setting related to IV catheter quality surveillance is a key objective of recent healthcare reform in Australia. We sought to determine the plausibility of using electronic health record (EHR) data for catheter surveillance by mapping currently available data across state-wide platforms. This work has identified barriers and facilitators to a state-wide EHR surveillance initiative.Materials and methodsData variables were generated and mapped from routinely used EHR sources across Queensland, Australia through a systematic search of gray literature and expert consultation with clinical information specialists. EHR systems were eligible for inclusion if they collected data related to IV catheter insertion, care, or outcomes of hospitalized patients. Generated variables were mapped against international recommendations for IV catheter surveillance, with data linkage and data export capacity narratively summarized.ResultsWe identified five EHR systems, namely, iEMR, MetaVision ICU®, Multiprac, RiskMan, and the Nephrology Registry. Systems were used across jurisdictions and hospital wards. Data linkage was not evident across systems. Extraction processes for catheter data were not standardized, lacking clear and reliable extraction techniques. In combination, EHR systems collected 43/50 international variables recommended for catheter surveillance, however, individual systems collected a median of 24/50 (IQR 22, 30) variables. We did not identify integrated clinical analytic systems (incorporating machine learning) to support clinical decision making or for risk stratification (e.g., catheter-related infection).ConclusionCurrent data linkage across EHR systems limits the development of an IV catheter quality surveillance system to provide timely data related to catheter complications and harm. To facilitate reliable and timely surveillance of catheter outcomes using clinical informatics, substantial work is needed to overcome existing barriers and transform health surveillance.

Fitness of real-world data for clinical trial data collection: Results and lessons from a HARMONY Outcomes ancillary study.

Despite the extensive use of real-world data for retrospective, observational clinical research, our understanding of how real-world data might increase the efficiency of data collection in patient-level randomized clinical trials is largely unknown. The structure of real-world data is inherently heterogeneous, with each source electronic health record and claims database different from the next. Their fitness-for-use as data sources for multisite trials in the United States has not been established. For a subset of participants in the HARMONY Outcomes Trial, we obtained electronic health record data from recruiting sites or Medicare claims data from the Centers for Medicare & Medicaid Services. For baseline characteristics and follow-up events, we assessed the level of agreement between these real-world data and data documented in the trial database. Real-world data-derived demographic information tended to agree with trial-reported demographic information, although real-world data were less accurate in identifying medical history. The ability of real-world data to identify baseline medication usage differed by real-world data source, with claims data demonstrating substantially better performance than electronic health record data. The limited number of lab results in the collected electronic health record data matched closely with values in the trial database. There were not enough follow-up events in the ancillary study population to draw meaningful conclusions about the performance of real-world data for identification of events. Based on the conduct of this ancillary study, the challenges and opportunities of using real-world data within clinical trials are discussed. Based on a subset of participants from the HARMONY Outcomes Trial, our results suggest that electronic health record or claims data, as currently available, are unlikely to be a complete substitute for trial data collection of medical history or baseline lab results, but that Medicare claims were able to identify most medications. The limited size of the study population prevents us from drawing strong conclusions based on these results, and other studies are clearly needed to confirm or refute these findings.

NeuroBlu, an electronic health record (EHR) trusted research environment (TRE) to support mental healthcare analytics with real-world data

PurposeNeuroBlu is a real-world data (RWD) repository that contains deidentified electronic health record (EHR) data from US mental healthcare providers operating the MindLinc EHR system. NeuroBlu enables users to perform...

Optimizing Data on Race and Ethnicity for Veterans Affairs Patients.

Maintaining accurate race and ethnicity data among patients of the Veterans Affairs (VA) healthcare system has historically been a challenge. This work expands on previous efforts to optimize race and ethnicity values by combining multiple VA data sources and exploring race- and ethnicity-specific collation algorithms. We linked VA patient data from 2000 to 2018 with race and ethnicity data from four administrative and electronic health record sources: VA Medical SAS files (MedSAS), Corporate Data Warehouse (CDW), VA Centers for Medicare extracts (CMS), and VA Defense Identity Repository Data (VADIR). To assess the accuracy of each data source, we compared race and ethnicity values to self-reported data from the Survey of Health Experiences of Patients (SHEP). We used Cohen's Kappa to assess overall (holistic) source agreement and positive predictive values (PPV) to determine the accuracy of sources for each race and ethnicity separately. Holistic agreement with SHEP data was excellent (K > 0.80 for all sources), while race- and ethnicity-specific agreement varied. All sources were best at identifying White and Black users (average PPV = 0.94, 0.93, respectively). When applied to the full VA user population, both holistic and race-specific algorithms substantially reduced unknown values, as compared to single-source methods. Combining multiple sources to generate race and ethnicity values improves data accuracy among VA patients. Based on the overall agreement with self-reported data, we recommend using non-missing values from sources in the following order to fill in race values-SHEP, CMS, CDW, MedSAS, and VADIR-and in the following order to fill in ethnicity values-SHEP, CDW, MedSAS, VADIR, and CMS.

Identifying Subtypes of Heart Failure with Machine Learning: External, Prognostic and Genetic Validation in Three Electronic Health Record Sources with 320,863 Individuals

Identifying Subtypes of Heart Failure with Machine Learning: External, Prognostic and Genetic Validation in Three Electronic Health Record Sources with 320,863 Individuals

Enhancing PCORnet Clinical Research Network data completeness by integrating multistate insurance claims with electronic health records in a cloud environment aligned with CMS security and privacy requirements.

The Greater Plains Collaborative (GPC) and other PCORnet Clinical Data Research Networks capture healthcare utilization within their health systems. Here, we describe a reusable environment (GPC Reusable Observable Unified Study Environment [GROUSE]) that integrates hospital and electronic health records (EHRs) data with state-wide Medicare and Medicaid claims and assess how claims and clinical data complement each other to identify obesity and related comorbidities in a patient sample. EHR, billing, and tumor registry data from 7 healthcare systems were integrated with Center for Medicare (2011-2016) and Medicaid (2011-2012) services insurance claims to create deidentified databases in Informatics for Integrating Biology & the Bedside and PCORnet Common Data Model formats. We describe technical details of how this federally compliant, cloud-based data environment was built. As a use case, trends in obesity rates for different age groups are reported, along with the relative contribution of claims and EHR data-to-data completeness and detecting common comorbidities. GROUSE contained 73 billion observations from 24 million unique patients (12.9 million Medicare; 13.9 million Medicaid; 6.6 million GPC patients) with 1674134 patients crosswalked and 983450 patients with body mass index (BMI) linked to claims. Diagnosis codes from EHR and claims sources underreport obesity by 2.56 times compared with body mass index measures. However, common comorbidities such as diabetes and sleep apnea diagnoses were more often available from claims diagnoses codes (1.6 and 1.4 times, respectively). GROUSE provides a unified EHR-claims environment to address health system and federal privacy concerns, which enables investigators to generalize analyses across health systems integrated with multistate insurance claims.

FACILITATING REAL WORLD CARDIOLOGY STUDIES THROUGH VALIDATION OF THE APOLLO ELECTRONIC DATA CAPTURE SYSTEM IN A COMMUNITY CARDIOLOGY ELECTRONIC HEALTH RECORD DATABASE

<h3>BACKGROUND</h3> Real world data (RWD) can provide valuable insights to complement evidence from randomized control trials (RCTs). While the proliferation of electronic health records (EHRs) has led to an abundance of RWD, much of the valuable information from these sources are unstructured and unanalyzable. Emerging technologies promise to address this limitation using natural language processing and other computing techniques. Many have not been validated, raising questions about their suitability for research. We performed a preliminary validation study of one such solution, the Apollo Electronic Data Capture (EDC) system by Ensho Health, in a community cardiology setting. <h3>METHODS AND RESULTS</h3> The Apollo EDC system was used to extract the most recent estimated LVEF and heart failure medications [ace inhibitors (ACEi), angiotensin receptor-neprilysin inhibitors (ARNi), angiotensin II receptor blockers (ARB), beta blockers (BB) and mineralocorticoid receptor antagonists (MRA)] from the de-identified electronic health records of 25 consecutive patients that visited an outpatient cardiology practice in November of 2020 with ICD-9 codes for congestive heart failure. The extracts included the name and date of the originating document, the method for estimating LVEF, the name and daily dose of heart failure medications, and whether the medications were titrated, discontinued or held at last mention. Apollo extracted LVEF measurements in 100% of patients (25/25), BB data for 100% of patients (25/25), ACEi data for 60% of patients (15/25), MRA data for 60% of patients (15/25), ARB data for 28% of patients (7/25) and ARNi data for 25% of patients (6/25). We compared the data extracted from Apollo to the information contained in the source EHR to assess the fidelity of the extraction. The accuracy of LVEF and LVEF estimation methods was 100% (25/25 for both value and method). Prescribed medications for heart failure were reported correctly 98.5% (67/68) of the time. The lone error involved a BB that was changed during a hospitalization. We did not find evidence that an ACEi, ARNi, ARB, BB or MRA was prescribed in the 57 instances Apollo did not report one. <h3>CONCLUSION</h3> The Apollo EDC system by Ensho Health extracted complex data from the EHR of a cardiology practice with high fidelity. As such, Apollo EDC is a promising tool to facilitate real world cardiology studies. Its ability to extract challenging data elements like LVEF and current medications from EHRs could allow researchers to ask more complex questions of RWD. Further validation across other data fields, EMR systems and therapeutic areas are warranted.

Assessing HIPAA Compliance of Open Source Electronic Health Record Applications

Electronic health record (EHR) applications are digital versions of paper-based patient health information. EHR applications are increasingly being adopted in many countries. They have resulted in improved quality in healthcare, convenient access to histories of patient medication and clinic visits, easier follow up of patient treatment plans, and precise medical decision-making process. The goal of this paper is to identify HIPAA technical requirements, evaluate two open source EHR applications (OpenEMR and OpenClinic) for security vulnerabilities using two open-source scanner tools (RIPS and PHP VulnHunter), and map the identified vulnerabilities to HIPAA technical requirements.

Transforming and evaluating electronic health record disease phenotyping algorithms using the OMOP common data model: a case study in heart failure.

ObjectiveThe aim of the study was to transform a resource of linked electronic health records (EHR) to the OMOP common data model (CDM) and evaluate the process in terms of syntactic and semantic consistency and quality when implementing disease and risk factor phenotyping algorithms.Materials and MethodsUsing heart failure (HF) as an exemplar, we represented three national EHR sources (Clinical Practice Research Datalink, Hospital Episode Statistics Admitted Patient Care, Office for National Statistics) into the OMOP CDM 5.2. We compared the original and CDM HF patient population by calculating and presenting descriptive statistics of demographics, related comorbidities, and relevant clinical biomarkers.ResultsWe identified a cohort of 502 536 patients with the incident and prevalent HF and converted 1 099 195 384 rows of data from 216 581 914 encounters across three EHR sources to the OMOP CDM. The largest percentage (65%) of unmapped events was related to medication prescriptions in primary care. The average coverage of source vocabularies was >98% with the exception of laboratory tests recorded in primary care. The raw and transformed data were similar in terms of demographics and comorbidities with the largest difference observed being 3.78% in the prevalence of chronic obstructive pulmonary disease (COPD).ConclusionOur study demonstrated that the OMOP CDM can successfully be applied to convert EHR linked across multiple healthcare settings and represent phenotyping algorithms spanning multiple sources. Similar to previous research, challenges mapping primary care prescriptions and laboratory measurements still persist and require further work. The use of OMOP CDM in national UK EHR is a valuable research tool that can enable large-scale reproducible observational research.

Completeness and accuracy of national cancer and death registration for outcome ascertainment in trials\u2014an ovarian cancer exemplar

BackgroundThere is a trend to increasing use of routinely collected health data to ascertain outcome measures in trials. We report on the completeness and accuracy of national ovarian cancer and death registration in the United Kingdom Collaborative Trial of Ovarian Cancer Screening (UKCTOCS).MethodsOf the 202,638 participants, 202,632 were successfully linked and followed through national cancer and death registries of Northern Ireland, Wales and England. Women with registrations of any of 19 pre-defined ICD-10 codes suggestive of tubo-ovarian cancer or notification of ovarian/tubal/peritoneal cancer from hospital episode statistics or trial sites were identified. Copies of hospital and primary care notes were retrieved and reviewed by an independent outcomes review committee. National registration of site and cause of death as ovarian/tubal/peritoneal cancer (C56/C57/C48) obtained up to 3 months after trial censorship was compared to that assigned by outcomes review (reference standard).ResultsOutcome review was undertaken in 3110 women on whom notification was received between 2001 and 2014. Ovarian cancer was confirmed in 1324 of whom 1125 had a relevant cancer registration. Sensitivity and specificity of ovarian/tubal/peritoneal cancer registration were 85.0% (1125/1324; 95% CI 83.7–86.2%) and 94.0% (1679/1786; 95% CI 93.2–94.8%), respectively. Of 2041 death registrations reviewed, 681 were confirmed to have a tubo-ovarian cancer of whom 605 had relevant death registration. Sensitivity and specificity were 88.8% (605/681; 95% CI 86.4–91.2%) and 96.7% (1482/1533, 95% CI 95.8–97.6%), respectively. When multiple electronic health record sources were considered, sensitivity for cancer site increased to 91.1% (1206/1324, 95% CI 89.4–92.5%) and for cause of death 94.0% (640/681, 95% CI 91.9–95.5%).Of 1232 with cancer registration, 8.7% (107/1232) were wrongly designated as ovarian/tubal/peritoneal cancers by the registry and 4.0% (47/1172) of confirmed tubo-ovarian cancers were mis-registered. In 656 with death registrations, 7.8% (51/656) were wrongly assigned as due to ovarian/tubal/peritoneal cancers while 6.2% (40/645) of confirmed tubo-ovarian cancer deaths were mis-registered.ConclusionFollow-up of trial participants for tubo-ovarian cancer using national registry data will result in incomplete ascertainment, particularly of the site due in part to the latency of registration. This can be reduced by using other routinely collected data such as hospital episode statistics. Central adjudication by experts though resource intensive adds value by improving the accuracy of diagnoses.Trial registrationISRCTN: ISRCTN22488978. Registered on 6 April 2000

Learning Health System in Crisis: Lessons From the COVID-19 Pandemic.

The novel coronavirus disease (COVID-19) pandemic is the gravest public-health crisis that the United States has seen in more than a century. Healthcare delivery systems are the focal point for interfacing with COVID-19; however, many were and remain unprepared for this or similar outbreaks. In this paper, we describe the Learning Health System (LHS) as an ideal organizing principle to inform an evidence-based response to public-health emergencies like COVID-19. We further describe barriers and challenges to the realization of the LHS and propose a call to action for a substantial investment in the LHS, with a focus on public health. Specifically, we advocate for a Learning Health Network that promotes collaboration between health systems, community-based organizations, and government agencies, especially during public health emergencies. We have approached this commentary through the unique lens of researchers embedded within a large, integrated healthcare delivery system, with direct experience working with clinical and operational units in response to the COVID-19 pandemic.