The Observational Health Data Sciences and Informatics (or OHDSI, pronounced "Odyssey") program is a multi-stakeholder, interdisciplinary collaborative to bring out the value of health data through large-scale analytics. All our solutions are open-source. OHDSI has established an international network of researchers and observational health databases with a central coordinating center housed at Columbia University. Our Mission To improve health by empowering a community to collaboratively generate the evidence that promotes better health decisions and better care. Our Vision A world in which observational research produces a comprehensive understanding of health and disease. Our Objectives Innovation: Observational research is a field that will benefit greatly from disruptive thinking. We actively seek and encourage fresh methodological approaches in our work. Reproducibility: Accurate, reproducible, and well-calibrated evidence is necessary for health improvement. Community: Everyone is welcome to actively participate in OHDSI, whether you are a patient, a health professional, a researcher, or someone who simply believes in our cause. Collaboration: We work collectively to prioritize and address the real-world needs of our community’s participants. Openness: We strive to make all our community’s proceeds open and publicly accessible, including the methods, tools and the evidence that we generate. Beneficence: We seek to protect the rights of individuals and organizations within our community at all times.
External link for OHDSI
622 West 168th Street, PH-20
New York, NY 10032, US
OHDSI reposted this
Head of the Data Integration Center at the University Hospital Carl Gustav Carus Dresden
I am thrilled- Michele Zoch and I are kickikg off our first German OHDSI (Observational Health Data Sciences & Informatics) Studyathon in Dresden. With 40 passionate professionals and researchers by our side, we‘ll explore and learn how to run #OHDSI studies for the next two days. Stay tuned for outcomes. 🚀🚀🚀
Aggregating and harmonizing registry databases for comparative analyses – lessons learnt #OHDSISocialShowcase #JoinTheJourney Lead: Eva-Maria Didden Team: James Weaver, Dmitry Dymshyts, Amélie Beaudet, Audrey Muller, Andrius Kavaliunas, PhD Background: Pulmonary Arterial Hypertension [PAH] is a rare group of diseases characterized by progressive increase of pulmonary vascular resistance leading to right ventricular failure and premature death. Several risk assessment algorithms have been published for stratification of PAH patients into 1-year mortality risk (e.g., low, intermediate-low, intermediate-high or high risk), based on their disease status and severity. PAH is incurable, but treatable with the goal of achieving low risk status. Endothelial receptor antagonists [ERA] targeting the endothelin pathway, phosphodiesterase type 5 inhibitors [PDE5i] and soluble guanylate cyclase stimulators and [sGCS] targeting the nitride oxide pathway, and prostacyclin analogue and prostacyclin receptor agonist [PA and PRA] targeting the prostacyclin pathway are used as monotherapy or in combination to treat patients with PAH. According to the 2022 ERS/ECS guidelines, patients should receive PRA as add-on therapy if they are at intermediate low risk at follow-up. While the benefits of these agents and new treatment modalities have been intensely studied in the randomized controlled clinical trials, evidence from real-world clinical practice is lacking. The objectives of this study were: 1) describe the demographic and clinical characteristics of patients treated for PAH in real-world settings and 2) assess the feasibility of a comparative effectiveness analysis between patients treated with PRA add-on combination (target cohort) and patients treated with other PAH-specific treatment regimens (PDE5i ERA or sGCS ERA combination therapy without PRA (comparator cohort)). The study intended to assess whether target cohort therapy resulted in greater benefit (e.g., time to hospitalization, death, parenteral therapy, disease worsening) than comparator cohort therapy in real-world settings. Such evidence would add knowledge and understanding of PAH therapy use in routine care where patients have multiple comorbidities and variable individual treatment regimens. https://lnkd.in/eMAnbRjq
Predicting the risk of new onset of type 2 diabetes following exposure of Statin within patient with coronary artery disease #OHDSISocialShowcase Lead: Septi Melisa Team: Christianus Heru Setiawan, Muhammad Solihuddi Muhtar, Phuc Phan Thanh, Nguyen Phung-Anh, Jason Hsu, Ph.D. Background: Statins are widely used to prevent the progression of atherosclerosis, particularly in patients with coronary artery disease. However, statin use has been associated with an increased risk of new incident type 2 diabetes mellitus compared to non-statin use. Statin use can impair insulin signaling, affect downstream metabolites, and disrupt adipocyte maturation and differentiation, potentially leading to the development of diabetes. This presents a trade-off, as the benefits of statins may greatly outweigh the risks, yet understanding their potential long-term side effects is crucial. Predicting the risk of type 2 diabetes can help the physician to do preventive treatment. Therefore, this study aims to develop a prediction model for the risk of new-onset type 2 diabetes in patients with coronary artery disease using statins. https://lnkd.in/eT6V9Uvq #JoinTheJourney
Estimation of Causal Effects under Treatment Misclassification: A Semi-Parametric Bias Correction Framework with Application to Vaccine Effectiveness Study #OHDSISocialShowcase #JoinTheJourney Lead: Qiong Wu Team: Huiyuan Wang, Yong Chen Background: The effectiveness of COVID-19 vaccines in preventing symptomatic and severe COVID-19 has been assessed through randomized controlled trials (RCTs) and subsequent observational vaccine effectiveness studies. However, our understanding of how a COVID-19 vaccine administered prior to infection impacts long COVID outcomes is still unclear. Research conducted to date has been mainly centered on adults and has produced inconsistent findings. Some studies suggest a significant protective effect, e.g., a reduced risk of the diagnosis of post-acute sequalae of COVID-19 (PASC) or experiencing certain PASC symptoms. Meanwhile, other studies indicate mixed effects revealing considerable variations across different age groups, various dominant virus strains, and distinct PASC symptoms, or even suggesting counter-protective effects. One of the primary challenges in extracting reliable evidence from real-world data (e.g., electronic health records) in the United States is the under-reporting of vaccine status in EHR data due to the lack of immunization registry linkage. In the U.S., immunization records are often disconnected due to logistic, technical, and political barriers. Such incomplete vaccine information can lead to biased estimates in the comparative effectiveness research of COVID-19 vaccines. Motivated by this important challenge, we developed a novel semi-parametric bias correction framework for estimating average treatment effects in the presence of treatment misclassification. We introduce the identifying formula and the efficient influence function where such partial validation of vaccination status is available. This is a reasonable assumption given a substantial number of patients in the PEDSnet network were able to conduct linkage with their local immunization registry data. Subsequently, we constructed an efficient estimator to correct the potential bias of the incomplete vaccine data while achieves full statistical efficiency. Our method allows the misclassification rate of vaccine status captured by the EHR system to be dependent on the characteristics of the patients (i.e., differential misclassification) in a flexible form. We demonstrated the validity of our method using simulated synthesis data, mimicking the characteristics of the patients in the PEDSnet dataset, and studied the effectiveness of BNT162b2 on long COVID risks during the Omicron period. https://lnkd.in/e6q9JpRp
Evaluating Synthea: Comprehensive Analysis of a Leading Synthesized Medical Record Generator #OHDSISocialShowcase #JoinTheJourney Lead: Zach Wagner Co-Author: Clair Blacketer, MPH Background: Observational healthcare research has grown significantly due to the availability of electronic healthcare record (EHR) data, aiding in policy, healthcare delivery, procedural advancements, and outbreak responses [1]. However, accessing EHR data is challenging due to privacy regulations like the Health Insurance Portability and Accountability Act (HIPAA) in the US and General Data Protection Regulation (GDPR) in the EU. Synthetic data—simulating real populations using AI and ML models—emerges as a promising solution to bypass patient privacy concerns and expedite research. Developed by MITRE in 2017, Synthea is an open-source generator of synthesized EHRs for research and innovation without legal or privacy restrictions. According to the developers, it provides state-level patient information, mimicking real populations in geographic distributions, disease rates, doctor visits, hospitalizations, drug usages, and social demographics. Unlike other methods, Synthea avoids re-identification risks and does not rely on real patient data. Despite its promise, further research is needed to fully understand and improve Synthea's capabilities. Despite its potential, Synthea faces limitations. Hodges et al. found Synthea's data unreliable without external modification, like the "Medication Diversification Tool" (MDT) for realistic medication distributions. The lack of diversity in Synthea’s modeling capacities hinders its fidelity. Further analysis is required to validate Synthea’s capabilities, especially in tracking chronic diseases, which are significant healthcare concerns. This study aims to expand on prior research by comparing Synthea data generated to emulate the population of California (CA) to real-world data reported by the state. https://lnkd.in/eK5uD7Yd
Please join us Tuesday, Nov. 5 (11 am ET) to meet our 2024 Titan Award winners. Nominated and awarded by fellow community members, these collaborators did exceptional work to further OHDSI's mission this year. You will learn about their journeys and work this Tuesday! If you do not already have a call invite, the meeting link is on the community calls page below! #JoinTheJourney The 2024 Titan Award winners are: Data Standards: Alexander Davydov, Vlad Korsik, Anna Ostropolets, Aleh Zhuk Methodological Research: Linying Zhang Open-Source Development: Martin Lavallee Clinical Applications: Cindy Cai, Jung Ho Kim, Jack Janetzki Community Collaboration: Natthawut Max A. Community Support: Montse Camprubí, Elisse Katzman Community Leadership: Clair Blacketer, MPH https://lnkd.in/eMg4DKbe
Dynamic Mapping Tools: Keeping Up to Date with Vocabulary Changes #OHDSISocialShowcase #JoinTheJourney Co-Leads: Melanie Philofsky, Hanan S. Team: Izzie (Margaret) Clinton, Jue Wang, Krista Miller, Michael Kahn, Michelle Edelmann, Ian Brooks Background: Health Data Compass, the clinical data warehouse at the University of Colorado Anschutz, uses the #OMOP CDM for delivery of research datasets. A weekly ETL process transforms the raw EHR data into the CDM and updates the OHDSI vocabulary tables with the latest codes and mappings. However, when vocabulary tables are updated concepts can change to “nonstandard” or change field / table location, which can cause data to appear to be missing when in fact they have moved to a different field and/or table from where they were originally located. To alert users, we created two tools: the concept mapping stored procedure and the concept mapping table. These tools inform users of the current location of the concept and allow them to find and utilize these and their associated data. The stored procedure also identifies errors in the codes or concept IDs and corrects for common variations in vocabulary input. These tools will allow the OHDSI community to produce reliable evidence more swiftly and accurately. ➡️ https://lnkd.in/e8G8NbDV
How is it possible that #OHDSI2024 happened a week ago already? It was amazing to see so many collaborators from around the world join us for three days of learning, sharing, networking and driving OHDSI's mission forward together! Be on the lookout for all the video presentations coming soon, and check out the slides and showcase presentations on our event homepage! #JoinTheJourney https://lnkd.in/eiwwdwGg
Cloud-based, Automated Solution for Transforming Clinical Datasets to the OMOP CDM #OHDSISocialShowcase #JoinTheJourney Lead: Simon Thompson Co-Author: Abigail Carter https://loom.ly/vJ1-VHo
dsOMOP: Federated Analysis of Harmonized Clinical Data Combining OMOP CDM and DataSHIELD in a DATOS-CAT Cohort Use Case #OHDSISocialShowcase #JoinTheJourney Lead: David Sarrat González Team: Judith Martinez Gonzalez, Xavier Escribà Montagut, Aikaterini Lymperidou, Ramon Mateo Navarro, Rafael de Cid, Juan R Gonzalez, Alberto Labarga https://loom.ly/xbNvJo8