Patient-Centered Outcomes Research Network Research Articles

Identification and assessment of classification criteria attributes for systemic lupus erythematosus in a regional medical record data network

ObjectiveTo assess the application and utility of algorithms designed to detect features of SLE in electronic health record (EHR) data in a multisite, urban data network.MethodsUsing the Chicago Area Patient-Centered...

Predictive Modeling of Lapses in Care for People Living with HIV in Chicago: Algorithm Development and Interpretation.

Reducing care lapses for people living with HIV is critical to ending the HIV epidemic and beneficial for their health. Predictive modeling can identify clinical factors associated with HIV care lapses. Previous studies have identified these factors within a single clinic or using a national network of clinics, but public health strategies to improve retention in care in the United States often occur within a regional jurisdiction (eg, a city or county). We sought to build predictive models of HIV care lapses using a large, multisite, noncurated database of electronic health records (EHRs) in Chicago, Illinois. We used 2011-2019 data from the Chicago Area Patient-Centered Outcomes Research Network (CAPriCORN), a database including multiple health systems, covering the majority of 23,580 people with an HIV diagnosis living in Chicago. CAPriCORN uses a hash-based data deduplication method to follow people across multiple Chicago health care systems with different EHRs, providing a unique citywide view of retention in HIV care. From the database, we used diagnosis codes, medications, laboratory tests, demographics, and encounter information to build predictive models. Our primary outcome was lapses in HIV care, defined as having more than 12 months between subsequent HIV care encounters. We built logistic regression, random forest, elastic net logistic regression, and XGBoost models using all variables and compared their performance to a baseline logistic regression model containing only demographics and retention history. We included people living with HIV with at least 2 HIV care encounters in the database, yielding 16,930 people living with HIV with 191,492 encounters. All models outperformed the baseline logistic regression model, with the most improvement from the XGBoost model (area under the receiver operating characteristic curve 0.776, 95% CI 0.768-0.784 vs 0.674, 95% CI 0.664-0.683; P<.001). Top predictors included the history of care lapses, being seen by an infectious disease provider (vs a primary care provider), site of care, Hispanic ethnicity, and previous HIV laboratory testing. The random forest model (area under the receiver operating characteristic curve 0.751, 95% CI 0.742-0.759) revealed age, insurance type, and chronic comorbidities (eg, hypertension), as important variables in predicting a care lapse. We used a real-world approach to leverage the full scope of data available in modern EHRs to predict HIV care lapses. Our findings reinforce previously known factors, such as the history of prior care lapses, while also showing the importance of laboratory testing, chronic comorbidities, sociodemographic characteristics, and clinic-specific factors for predicting care lapses for people living with HIV in Chicago. We provide a framework for others to use data from multiple different health care systems within a single city to examine lapses in care using EHR data, which will aid in jurisdictional efforts to improve retention in HIV care.

Cardiovascular health trajectories from age 2–12: a pediatric electronic health record study

PurposeMany children have non-ideal cardiovascular health (CVH), but little is known about the course of CVH in early childhood. We identified CVH trajectories in children and assess the generalizability of these trajectories in an external sample. MethodsWe used data spanning 2010–2018 from children aged 2–12 years within the Chicago Area Patient-Centered Outcomes Research Network—an electronic health record network. Four clinical systems comprised the derivation sample and a fifth the validation sample. Body mass index, blood pressure, cholesterol, and blood glucose were categorized as ideal, intermediate, and poor using clinical measurements, laboratory readings, and International Classification of Diseases diagnosis codes and summed for an overall CVH score. Group-based trajectory modeling was used to create CVH score trajectories which were assessed for classification accuracy in the validation sample. ResultsUsing data from 122,363 children (47% female, 47% non-Hispanic White) three trajectories were identified: 59.5% maintained high levels of clinical CVH, 23.4% had high levels of CVH that declined, and 17.1% had intermediate levels of CVH that further declined with age. A similar classification emerged when the trajectories were fitted in the validation sample. ConclusionsStratification of CVH was present by age 2, implicating the need for early life and preconception prevention strategies.

Using pooled electronic health records data to conduct pharmacoepidemiology safety studies: Challenges and lessons learned.

We assessed the suitability of pooled electronic health record (EHR) data from clinical research networks (CRNs) of the patient-centered outcomes research network to conduct studies of the association between tumor necrosis factor inhibitors (TNFi) and infections. EHR data from patients with one of seven autoimmune diseases were obtained from three CRNs and pooled. Person-level linkage of CRN data and Centers for Medicare and Medicaid Services (CMS) fee-for-service claims data was performed where possible. Using filled prescriptions from CMS claims data as the gold standard, we assessed the misclassification of EHR-based new (incident) user definitions. Among new users of TNFi, we assessed subsequent rates of hospitalized infection in EHR and CMS data. The study included 45 483 new users of TNFi, of whom 1416 were successfully linked to their CMS claims. Overall, 44% of new EHR TNFi prescriptions were not associated with medication claims. Our most specific new user definition had a misclassification rate of 3.5%-16.4% for prevalent use, depending on the medication. Greater than 80% of CRN prescriptions had either zero refills or missing refill data. Compared to using EHR data alone, there was a 2- to 8-fold increase in hospitalized infection rates when CMS claims data were added to the analysis. EHR data substantially misclassified TNFi exposure and underestimated the incidence of hospitalized infections compared to claims data. EHR-based new user definitions were reasonably accurate. Overall, using CRN data for pharmacoepidemiology studies is challenging, especially for biologics, and would benefit from supplementation by other sources.

Rapid identification of chronic kidney disease in electronic health record database using computable phenotype combining a common data model.

Rapid identification of chronic kidney disease in electronic health record database using computable phenotype combining a common data model.

Evaluation of weight change and cardiometabolic risk factors in a real-world population of US adults with overweight or obesity

Whether individuals in real-world settings are able to lose weight and improve cardiometabolic risk factors over time is unclear. We aimed to determine the management of and degree of body weight change over 2 years among individuals with overweight or obesity, and to assess associated changes in cardiometabolic risk factors and clinical outcomes. Using data from 11 large health systems within the Patient-Centered Outcomes Research Network in the U.S., we collected the following data on adults with a recorded BMI ≥25 kg/m2 between January 1, 2016 and December 31, 2016: body-mass index (BMI), systolic blood pressure (SBP), diastolic blood pressure (DBP), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDLC), triglycerides and glycated hemoglobin (HbA1c). We found that among 882,712 individuals with BMI ≥25 kg/m2 (median age 59 years; 56% female), 52% maintained stable weight over 2 years and 1.3% utilized weight loss pharmacotherapy. Weight loss of 10% was associated with small but significant lowering of mean SBP (−2.69 mmHg [95% CI -2.88, −2.50]), DBP (−1.26 mmHg [95% CI -1.35, −1.18]), LDL-C (−2.60 mg/dL [95% CI -3.14, −2.05]), and HbA1c (−0.27% [95% CI -0.35, −0.19]) in the same 12 months. However, these changes were not sustained over the following year. In this study of adults with BMI ≥25 kg/m2, the majority had stable weight over 2 years, pharmacotherapies for weight loss were under-used, and small changes in cardiometabolic risk factors with weight loss were not sustained, possibly due to failure to maintain weight loss.

Commentary on Verdejo-Garcia et al.: Cognitive training and remediation: identifying priorities for intervention research.

The perspectives of youth, adults and family with lived experience and key knowledge users such as health-care providers and decision-makers are critical to further inform these priorities for cognitive intervention research. Verdejo-Garcia and colleagues [1] outline a large-scale and collaborative effort to identify research priorities to address cognitive deficits in substance use disorders. This commentary outlines additional considerations to ensure that strategic and coordinated research priorities have a meaningful impact upon health systems. Put most clearly, the input of youth, adults and family with lived experience and key knowledge users such as health-care providers and decision-makers is critical for cognitive intervention research with impact. As long noted in health science research, the substantial gaps between research and practice require greater connection between academic and knowledge user groups. Increased connectedness between researchers and knowledge users facilitates knowledge mobilization [2] and research impact [3]. It is increasingly considered essential to engage knowledge users in all stages of research focusing upon issues relevant to them, to improve research quality and relevance [4]. Frameworks have been developed to support and advocate for this approach, including Canada's Strategy for Patient-Oriented Research (SPOR) Framework [5], the United Kingdom's National Institute for Health Research [6] and the American Patient-Centered Outcomes Research Network [7]. Substance use research is commonly conducted with limited to no involvement of people who use substances; this practice has the potential not only to fail to address key issues but also to even intensify stigma and marginalization [8, 9]. Verdejo-Garcia and colleagues highlight the potential for the identified consensus to provide a roadmap to direct future participatory research; however, the incorporation of knowledge users to inform the roadmap itself holds considerable value. Indeed, frameworks for engagement such as those noted above recognize the skills and expertise that stakeholders can bring to research and decision-making processes [10]. Research guidelines and frameworks developed by and for people who use substances are available to promote more effective research partnerships at this early stage and may be particularly helpful, as other guidance materials can emphasize health-care providers and decision-makers rather than those with lived experience or the broader community [11]. Recently published key issues in the field further underscore the key importance of engagement from the very beginning of the research process in the substance use field [12]. The input of broad knowledge user groups is particularly important to extend this novel consensus to include harmonized and equitable (a) outcomes and (b) implementation approaches. Regarding outcomes, Verdejo-Garcia and colleagues carefully derive consensus on cognitive targets in cognitive intervention research, and it is likely that primary outcomes would include measures of these targets. However, cognitive intervention research in other domains highlights the importance of functional recovery outcome measures within this research [13, 14]. Indeed, these functional recovery outcomes have greater ecological validity than many cognitive outcomes and often align more closely with patient treatment goals. Such goals are also aligned with long-standing calls in substance use research to go beyond substance use frequency and abstinence as primary outcomes [15]. Nevertheless, those with lived experience as well as health-care providers and decision-makers are critical to determining harmonized outcomes that are relevant and meaningful to health systems and those that access them. Regarding implementation, it is increasingly recognized that nuanced efficacy research must be followed by equally rigorous implementation science research to support the integration of interventions into health systems. Verdejo-Garcia and colleagues do not explicitly endorse a specific approach to this essential phase of research, which would similarly benefit from broad consensus and harmonization. Frameworks such as the Damschroder et al. [16] Consolidated Framework for Implementation Research may support an approach to implementing cognitive interventions globally in a standardized manner and harmonizing implementation outcomes in the future. Consensus upon implementation outcome measures such as acceptability, adoption, appropriateness, cost, feasibility, penetration, sustainability and fidelity is key to a relevant, common and comprehensive list of key outcomes, which may more effectively advance the field and mobilize this knowledge. In summary, health research has long been characterized by a ‘know–do’ gap wherein the results of research frequently fail to be realized in practice and policy [17]. Investigations of cognitive interventions can meaningfully advance both theory and practice. Restructuring health research to prioritize partnerships between researchers and knowledge users, even at the earliest stage of identifying research priorities, is integral to conduct meaningful and impactful research for all [18]. None. There are no financial or other relevant links to companies with an interest in the topic of this article. Lena C. Quilty: Conceptualization (lead); writing—original draft (lead); writing—review and editing (lead).

Comparison of algorithms for identifying people with HIV from electronic medical records in a large, multi-site database.

ObjectiveAs electronic medical record (EMR) data are increasingly used in HIV clinical and epidemiologic research, accurately identifying people with HIV (PWH) from EMR data is paramount. We sought to evaluate EMR data types and compare EMR algorithms for identifying PWH in a multicenter EMR database.Materials and MethodsWe collected EMR data from 7 healthcare systems in the Chicago Area Patient-Centered Outcomes Research Network (CAPriCORN) including diagnosis codes, anti-retroviral therapy (ART), and laboratory test results.ResultsIn total, 13 935 patients had a positive laboratory test for HIV; 33 412 patients had a diagnosis code for HIV; and 17 725 patients were on ART. Only 8576 patients had evidence of HIV-positive status for all 3 data types (laboratory results, diagnosis code, and ART). A previously validated combination algorithm identified 22 411 patients as PWH.ConclusionEMR algorithms that combine laboratory results, administrative data, and ART can be applied to multicenter EMR data to identify PWH.

Multijurisdictional Analyses of Birth Defects: Considering the Common Data Model Approach.

Multijurisdictional Analyses of Birth Defects: Considering the Common Data Model Approach.

Effect of Two Interventional Strategies on Improving Continuous Positive Airway Pressure Adherence in Existing COPD and Obstructive Sleep Apnea Patients: The O2VERLAP Study.

Obstructive sleep apnea (OSA) is a sleep disorder prevalent in >10% of individuals diagnosed with chronic obstructive pulmonary disease (COPD). Continuous positive airway pressure (CPAP) is the first-line therapy for OSA, but many do not use it enough during sleep to effectively manage OSA. The O2VERLAP study compared proactive care (PC)-structured web-based peer-coaching education and support intervention versus reactive care (RC)-education and support based on limited scheduled interactions and patient-initiated contacts. Participants were primarily recruited from patient communities (COPD, OSA, and the National Patient-Centered Outcomes Research Network [PCORnet]) through electronic methods. Inclusion criteria: ≥40 years old, diagnosis of both COPD and OSA, and currently using CPAP. Participants were then randomly assigned to either the PC or RC group, with outcomes assessed at baseline and 6 and 12 weeks. The primary study outcome was CPAP adherence (hours of use/night) and secondary outcomes were daytime functioning, sleep quality, and daytime sleepiness. Changes in outcomes over time were examined using random effects models. The study enrolled 332 participants of which 294 were randomized. While groups differed significantly in CPAP adherence at baseline (PC: 6.1±3.1, RC: 7.3±2.4 hours/night; P<0.001), there were no significant differences in change of primary and secondary outcomes at either 6 or 12 weeks. In this group of patients with both COPD and OSA on CPAP therapy, no difference was found between the provision of PC and RC. The study did find unexpectedly high baseline CPAP adherence levels, which suggests that any improvement from the intervention would have been very small and difficult to detect.

Structure, Process and Outcomes in Patient Engagement

Although patient engagement in research is gaining acceptance by researchers and funding bodies, descriptions of implementation options and associated outcomes are limited. In this appreciative inquiry of the 12 institutions involved in the Great Plains Collaborative of the Patient-Centered Outcomes Research Network (PCORnet), we interviewed patient engagement officers and patient partners to enhance understanding of approaches to organizational structure, research engagement processes, and associated outcomes of the engaged research. Multiple structures have been identified including operational affiliations in both hospital and multiple university departments. Professional affiliations of patient engagement officers vary widely, including nurses, social workers, and public health professionals, among others. Patient engagement processes also vary, but with the majority using forms of advisory boards. All sites reported outcomes of their work including completed and/or ongoing research and co-authored publications.

Automated production of research data marts from a canonical fast healthcare interoperability resource data repository: applications to COVID-19 research

ObjectiveThe rapidly evolving COVID-19 pandemic has created a need for timely data from the healthcare systems for research. To meet this need, several large new data consortia have been developed that require frequent updating and sharing of electronic health record (EHR) data in different common data models (CDMs) to create multi-institutional databases for research. Traditionally, each CDM has had a custom pipeline for extract, transform, and load operations for production and incremental updates of data feeds to the networks from raw EHR data. However, the demands of COVID-19 research for timely data are far higher, and the requirements for updating faster than previous collaborative research using national data networks have increased. New approaches need to be developed to address these demands.MethodsIn this article, we describe the use of the Fast Healthcare Interoperability Resource (FHIR) data model as a canonical data model and the automated transformation of clinical data to the Patient-Centered Outcomes Research Network (PCORnet) and Observational Medical Outcomes Partnership (OMOP) CDMs for data sharing and research collaboration on COVID-19.ResultsFHIR data resources could be transformed to operational PCORnet and OMOP CDMs with minimal production delays through a combination of real-time and postprocessing steps, leveraging the FHIR data subscription feature.ConclusionsThe approach leverages evolving standards for the availability of EHR data developed to facilitate data exchange under the 21st Century Cures Act and could greatly enhance the availability of standardized datasets for research.

PROVIDE-HF primary results: Patient-Reported Outcomes inVestigation following Initiation of Drug therapy with Entresto (sacubitril/valsartan) in heart failure

In PARADIGM-HF, sacubitril/valsartan improved quality of life (QOL) versus enalapril in heart failure with reduced ejection fraction (HFrEF), yet limited data are available regarding QOL changes after sacubitril/valsartan initiation in routine practice. PROVIDE-HF was a prospective study within a national research network (Patient-Centered Outcomes Research Network) of HFrEF outpatients recently initiated on sacubitril/valsartan versus controls with recent angiotensin-converting enzyme inhibitor/angiotensin receptor blocker initiation/dose change. The primary end point was mean Kansas City Cardiomyopathy Questionnaire (KCCQ) change through 12 weeks. Other end points included responder analyses: ≥5-point and ≥20-point KCCQ increase. Adjusted QOL change was estimated after propensity score weighting. Overall, 270 patients had both baseline and 12-week KCCQ data (151 sacubitril/valsartan; 119 control). The groups had similar demographics and HF details: median EF 28% and N-terminal pro-brain natriuretic peptide 1083 pg/mL. Sacubitril/valsartan patients had larger improvements in KCCQ (mean difference +4.76; P = .027) and were more likely to have a ≥5-point and ≥20-point response (all P < .05). Adjusted comparisons demonstrated similar numerical improvements in the change in KCCQ (+4.55; 95% CI -0.89 to 9.99; P = .101) and likelihood of ≥5-point increase (odds ratio 1.55; 95% CI: 0.84-2.86; P = .16); ≥20-point increase remained statistically significant (odds ratio 3.79; 95% CI 1.47-9.73; P = .006). In this prospective HFrEF study of sacubitril/valsartan initiation compared with recent angiotensin-converting enzyme inhibitor/angiotensin receptor blocker initiation/dose change, the between-group difference in the primary end point, mean KCCQ change at 12 weeks was not statistically significant following adjustment, but sacubitril/valsartan initiation was associated with early improvements in QOL and a higher likelihood of ≥20-point improvement in KCCQ at 12 weeks. These data add additional real-world evidence related to patient-reported outcomes following the initiation of sacubitril/valsartan in routine clinical practice.

Electronic Measurement of a Clinical Quality Measure for Inpatient Hypoglycemic Events

Hypoglycemia related to antidiabetic drugs (ADDs) is important iatrogenic harm in hospitalized patients. Electronic identification of ADD-related hypoglycemia may be an efficient, reliable method to inform quality improvement. Develop electronic queries of electronic health records for facility-wide and unit-specific inpatient hypoglycemia event rates and validate query findings with manual chart review. Electronic queries were created to associate blood glucose (BG) values with ADD administration and inpatient location in 3 tertiary care hospitals with Patient-Centered Outcomes Research Network (PCORnet) databases. Queries were based on National Quality Forum criteria with hypoglycemia thresholds <40 and <54 mg/dL, and validated using a stratified random sample of 321 BG events. Sensitivity and specificity were calculated with manual chart review as the reference standard. The sensitivity and specificity of queries for hypoglycemia events were 97.3% [95% confidence interval (CI), 90.5%-99.7%] and 100.0% (95% CI, 92.6%-100.0%), respectively for BG <40 mg/dL, and 97.7% (95% CI, 93.3%-99.5%) and 100.0% (95% CI, 95.3%-100.0%), respectively for <54 mg/dL. The sensitivity and specificity of the query for identifying ADD days were 91.8% (95% CI, 89.2%-94.0%) and 99.0% (95% CI, 97.5%-99.7%). Of 48 events missed by the queries, 37 (77.1%) were due to incomplete identification of insulin administered by infusion. Facility-wide hypoglycemia rates were 0.4%-0.8% (BG <40 mg/dL) and 1.9%-3.0% (BG <54 mg/dL); rates varied by patient care unit. Electronic queries can accurately identify inpatient hypoglycemia. Implementation in non-PCORnet-participating facilities should be assessed, with particular attention to patient location and insulin infusions.

Establishment of a Patient-Centered Outcomes Research Network for Individuals with TBI and Neuropsychiatric Symptoms

ABSTRACTAims: The overarching goal of this project was to establish a group comprised of a variety of TBI stakeholders for the purpose of: (1) determining facilitators and barriers in management of neuropsychiatric symptoms after TBI; (2) identifying strategies for maintaining a TBI PCOR network; (3) enumerating research topics related to TBI neuropsychiatry; and (4) highlighting policy changes related to TBI neuropsychiatry.Methods: Twenty-nine TBI stakeholders participated in focus group discussions. Qualitative analyses were conducted both manually and using Dedoose software.Results: Participant-identified barriers included stigma associated with experiencing neuropsychiatric symptoms and poor insurance coverage. Facilitators included treatment focused on education of neuropsychiatric symptoms after TBI and having a comprehensive caregiver plan. Best strategies for maintaining TBI PCOR network included having a well-defined project, continued regular meetings, and on-going education of network members. Pertinent research topics included TBI and aging, factors influencing outcomes after TBI, substance use disorders related to TBI, and effectiveness of telemental health services. Needed policy changes included making TBI neuropsychiatry education accessible to stakeholders and improving accessibility of TBI neuropsychiatric care.Conclusion: TBI stakeholders identified several facilitators of care for neuropsychiatric symptoms after TBI and suggested research topics and best practices for conducting PCOR in this area.

The Patient-Centered Outcomes Research Network Antibiotics and Childhood Growth Study: Implementing Patient Data Linkage.

PCORnet, the Patient-Centered Outcomes Research Network, is comprised of health systems and health plans that transform electronic health records (EHRs) and claims data to a common data model (CDM) to facilitate real-world clinical research. Because patients receive health care in multiple care delivery settings, linking health records across systems and health plan claims would provide a more comprehensive and accurate picture of health care for patients. The current study expanded on a PCORnet Antibiotics and Childhood Growth (ABX) study to (1) identify and implement a privacy-preserving patient linkage solution among a clinical data research network and a health plan network within the ABX Study, and (2) assess overlap in prescribed and dispensed antibiotics and additional data gained from claims among the linked patients. This manuscript describes the linkage process and resulting overlap analysis. The authors identified 549 patients from the EHR record study cohort who had claims records with the health plan. Sixty percent (n = 329) of patients had consistent antibiotic exposure data across the 2 sources, indicating antibiotic exposure (44.3%) or nonexposure (15.7%). Among total antibiotic prescribing records, 43.1% had a matched claims record for dispensing within 60 days. Among antibiotic dispense records 26.5% were not associated with a prescribing record in the linked health systems. These findings showcase the feasibility of linking health plan claims data to PCORnet CDM in a privacy-preserving manner while also demonstrating continued gaps in data that may occur. The study highlights the importance of combining multiple health data sources for comprehensive clinical research.

Fast Healthcare Interoperability Resources (FHIR) as a Meta Model to Integrate Common Data Models: Development of a Tool and Quantitative Validation Study.

BackgroundIn a multisite clinical research collaboration, institutions may or may not use the same common data model (CDM) to store clinical data. To overcome this challenge, we proposed to use Health Level 7’s Fast Healthcare Interoperability Resources (FHIR) as a meta-CDM—a single standard to represent clinical data.ObjectiveIn this study, we aimed to create an open-source application termed the Clinical Asset Mapping Program for FHIR (CAMP FHIR) to efficiently transform clinical data to FHIR for supporting source-agnostic CDM-to-FHIR mapping.MethodsMapping with CAMP FHIR involves (1) mapping each source variable to its corresponding FHIR element and (2) mapping each item in the source data’s value sets to the corresponding FHIR value set item for variables with strict value sets. To date, CAMP FHIR has been used to transform 108 variables from the Informatics for Integrating Biology & the Bedside (i2b2) and Patient-Centered Outcomes Research Network data models to fields across 7 FHIR resources. It is designed to allow input from any source data model and will support additional FHIR resources in the future.ResultsWe have used CAMP FHIR to transform data on approximately 23,000 patients with asthma from our institution’s i2b2 database. Data quality and integrity were validated against the origin point of the data, our enterprise clinical data warehouse.ConclusionsWe believe that CAMP FHIR can serve as an alternative to implementing new CDMs on a project-by-project basis. Moreover, the use of FHIR as a CDM could support rare data sharing opportunities, such as collaborations between academic medical centers and community hospitals. We anticipate adoption and use of CAMP FHIR to foster sharing of clinical data across institutions for downstream applications in translational research.

Abstract P2050: The Pcornet Blood Pressure Control Laboratory

Background: New technology and healthcare delivery approaches may help improve blood pressure (BP) control. Randomized controlled trials (RCTs) can demonstrate effectiveness of these emerging approaches, both overall and especially within underserved populations, but RCTs are expensive. Our objective is to establish a national platform to support BP control surveillance and efficient RCTs. Methods: The PCORnet BP Control Laboratory, designed and launched in partnership with the American Heart Association (AHA) and the American Medical Association (AMA), leverages electronic health record (EHR) systems throughout the US, the National Patient-Centered Outcomes Research Network (PCORnet), and the NIH-funded Eureka Research Platform, which enables direct-to-participant research including collection of patient-reported outcomes and mHealth data from wearable devices and smartphones. We will use these resources to 1) Conduct national surveillance through quarterly EHR data queries to calculate and track institution- and clinic-level BP control metrics across 23 PCORnet institutions, overall and within key subgroups, and support participation in the AHA/AMA’s Target:BP recognition program (BP Track); 2) Compare effectiveness of a Full Support versus Self-Guided version of the AMA’s Measure Accurately, Act Rapidly, and Partner with Patients (MAP) program at improving BP control in a cluster randomized trial of 24 safety net clinics (BP MAP); and 3) Compare effectiveness of Smartphone-linked versus standard home BP monitoring at reducing systolic BP in an individual-level, direct-to-participant RCT of 2000 patients with uncontrolled BP (BP Home). Results: Preliminary BP Track queries demonstrate an overall BP control rate of 56%, lower control in Black patients (51%), substantial heterogeneity in clinic-level metrics (e.g., 33% vs. 64% BP control), and low rates of medication intensification (9% prescribed a new medication class when BP was high). BP Track will launch PCORnet-wide queries in June; BP MAP in May; and BP Home in August 2019. Conclusions: Feasibility testing is underway to determine if the PCORnet BP Control Laboratory will be useful for investigators interested in conducting efficient RCTs of BP control interventions.

Data model harmonization for the All Of Us Research Program: Transforming i2b2 data into the OMOP common data model.

BackgroundThe All Of Us Research Program (AOU) is building a nationwide cohort of one million patients’ EHR and genomic data. Data interoperability is paramount to the program’s success. AOU is standardizing its EHR data around the Observational Medical Outcomes Partnership (OMOP) data model. OMOP is one of several standard data models presently used in national-scale initiatives. Each model is unique enough to make interoperability difficult. The i2b2 data warehousing and analytics platform is used at over 200 sites worldwide, which uses a flexible ontology-driven approach for data storage. We previously demonstrated this ontology system can drive data reconfiguration, to transform data into new formats without site-specific programming. We previously implemented this on our 12-site Accessible Research Commons for Health (ARCH) network to transform i2b2 into the Patient Centered Outcomes Research Network model.Methods and resultsHere, we leverage our investment in i2b2 high-performance transformations to support the AOU OMOP data pipeline. Because the ARCH ontology has gained widespread national interest (through the Accrual to Clinical Trials network, other PCORnet networks, and the Nebraska Lexicon), we leveraged sites’ existing investments into this standard ontology. We developed an i2b2-to-OMOP transformation, driven by the ARCH-OMOP ontology and the OMOP concept mapping dictionary. We demonstrated and validated our approach in the AOU New England HPO (NEHPO). First, we transformed into OMOP a fake patient dataset in i2b2 and verified through AOU tools that the data was structurally compliant with OMOP. We then transformed a subset of data in the Partners Healthcare data warehouse into OMOP. We developed a checklist of assessments to ensure the transformed data had self-integrity (e.g., the distributions have an expected shape and required fields are populated), using OMOP’s visual Achilles data quality tool. This i2b2-to-OMOP transformation is being used to send NEHPO production data to AOU. It is open-source and ready for use by other research projects.

Identifying heart failure using EMR-based algorithms

BackgroundHeart failure (HF) is a major clinical and public health problem, the management of which will benefit from large-scale pragmatic research that leverages electronic medical records (EMR). Requisite to using EMRs for HF research is the development of reliable algorithms to identify HF patients. We aimed to develop and validate computable phenotype algorithms to identify patients with HF using standardized data elements defined by the Patient Centered Outcomes Research Network (PCORnet) Common Data Model (CDM). MethodsWe built HF computable phenotypes utilizing the data domains of HF diagnosis codes, prescribed HF-related medications and N-terminal B-type natriuretic peptide (NT-proBNP). Algorithms were validated in a cohort (n = 76,254) drawn from Olmsted County, MN between 2010–2012 a sample of whose records were manually reviewed to confirm HF according to Framingham criteria. ResultsThe different algorithms we tested provided different tradeoffs between sensitivity and positive predictive value (PPV). The highest sensitivity (78.7%) algorithm utilized one HF diagnosis code and had the lowest PPV (68.5%). The addition of more algorithm components, such as additional HF diagnosis codes, HF medications or elevated NT-proBNP, improved the PPV while reducing sensitivity. When added to a diagnostic code, the addition of NT-proBNP (>450 pg/mL) had a similar impact compared to additional HF medication criteria, increasing PPV by ∼3–4% and decreasing sensitivity by ∼7–10%. ConclusionsAlgorithms derived from PCORnet CDM elements can be used to identify patients with HF without manual adjudication with reasonable sensitivity and PPV. Algorithm choice should be driven by the goal of the research.