Volume 54 | September 29, 2022
|
|
A program of NIH’s National Center for Advancing Translational Sciences |
|
Sites with Signed
DUAs
326
|
|
Orgs with EHR Data Released into the Enclave
76
|
|
COVID-19
Positive
Patients
6,534,111
|
|
Rows
of Patient
Data
19.1 billion
|
|
Approved
N3C Research
Projects
389
|
|
|
What the Research Community is Saying about N3C
"We had an amazing experience in using the N3C dataset and platform for machine learning for healthcare research. The dataset is one of the largest and richest patient databases that we worked with. Their COVID patient dataset is extremely valuable and enabled our research in COVID hospitalization prediction. Thanks to N3C support, we recently published a Cell iScience paper on machine learning models for early pediatric COVID-19 hospitalization and severity prediction. Throughout the entire process, the N3C staff is also very responsive and helpful in evaluating and approving our publication request in a timely fashion. We are extremely grateful for their support.
In summary, we recommend N3C to all the health data science researchers. I hope more such open data science resources can be created and provided to researchers."
|
|
American Indian / Alaska Native Demographic Information
Now Accessible
AI/AN data will be available for research in N3C analysis that provides race and ethnicity distribution through the standard Data Use Request [covid.cd2h.org]. It is important to maintain trust with the tribal community, so we must follow important data use requirements listed below.
- Tribal affiliations and connections are NOT allowed! The N3C Data User Code of Conduct is updated so that data users will be asked to attest that they understand the N3C contains no Tribal affiliation data and that use of AI/AN data and ZIP code information to make assumptions about Tribal affiliation is not valid or appropriate.
- AI/AN data is now available in the race category in N3C.
- ZIP codes that overlap with predominantly Tribal communities is now available for research in the following manner:
- AI/AN ZIP codes containing 20,000 or fewer people are obfuscated to “000” This is standard practice for all geographic units.
- Specific ZIP codes representing rural populations predominantly with AI/AN-identifying individuals are hidden. These are now visible in both the limited dataset and de-identified data. This means that AI/AN data will be managed as others are managed with the exception that the full five-digit ZIP codes are never shown. The zip codes will contain 3 digits in both Level 2 and Level 3.
- The N3C Publication Committee will review manuscripts to ensure that tribal affiliations are not included. Violation of the Code of Conduct may result in corrective action.
|
|
A Guide for Using NLP Notes & NLP Note Table Available!
The document includes an introduction to how clinical notes are transformed into structured data and includes sample queries for using the new tables in your analysis. An additional site has been added to the NLP data for a total of six sites.
|
|
N3C Enclave Welcome Page and Phenotype Explorer Update
The N3C Enclave Welcome Page and Phenotype Explore will have a new look. Below are the details of what has and hasn’t changed.
A new version of the N3C Welcome Page will soon go live in the Enclave. This new layout arranges essential resources into four categories:
- Educational Resources
- N3C Data Overview: Summary Statistics, N3C Cohort Definition, and Phenotype Explorer
- Administrative Resources
- Data Analysis and Result Export Resources
No resources have been removed, although some links have been renamed and repositioned. Some resource links have been added to this first page.
The new Phenotype Explorer is supported by the Logic Liaison All Patients and Confirmed Covid templates and datasets. Links are provided in the Explorer so that you can import these objects into your Level 2 or 3 workspace.
-
Using the Logic Liaison Datasets, you can quickly perform more complex analysis on the large set of default variables provided.
-
Using the Logic Liaison Templates, you can generate custom variables, control parameters (such as timelines used to define Covid-associated visits and reinfections, BMI limits, etc.), and extend the pipeline by adding nodes containing custom logic.
The Phenotype Explorer supports three different views of the data:
-
All N3C Patients: Cases and controls as defined by the N3C Patient Selection Criteria.
-
Confirmed Covid Positives: N3C patients with an approximate start date for acute infection based on a Covid-19 Diagnosis (U07.1) or a Positive SARS-CoV-2 Lab Result (qualitative rt-PCR or Antigen test)
-
Possible Covid Positives: N3C patients who have no identifiable date for acute infection but have indicator(s) of possible prior infection such as a Positive Covid-19 Antibody Test, a Long Covid Clinic Visit, a Long Covid Diagnosis Code, Pneumonia due to Covid-19 or Multi-system Inflammatory Syndrome in Children
Note that confirmed patients may also have some of the features used for possible patients, but these groups are non-overlapping with the main distinction being that no Covid-19 start date is available for the possible patients. The index date in the confirmed patient group enables additional features that depend on date comparisons, such as: reinfection counts, severity at the time of first Covid infection, and the ability to split events into the before-Covid, post-Covid, and acute phases.
Feedback and Questions?
If you have questions or feedback about the new Welcome Page or Phenotype Explorer, please submit a technical ticket in the enclave (select Question or Feedback category). The N3C Logic Liaisons support generic template creation for the N3C community.
|
|
Clinician Update
Over the last few months, the Clinician Engagement Group, which consists of clinicians from diverse backgrounds, has taken a deep dive with data analysts on the following N3C-related research, offering clinical and scientific insights as well as recommendations for additional study or analyses:
● Paxlovid and Paxlovid rebound
● Analysis of historical diagnosis data using deep learning to determine risk factors for Long Covid
● Incidence and Prevalence of PASC over time
● Laboratory biomarkers, specifically comparing trajectories before and after COVID index date stratified by PASC and hospitalization status
● COVID-19 reinfections, including exploratory analysis of the frequency of reinfection and the risk of hospitalization compared with primary infection.
These meetings have provided a space for clinicians and analysts to review analytic methods results. Analysts presented data, and the group assessed the extent to which these findings dovetail with real-world clinician experiences. Overall, the working group has offered clinicians and analysts an opportunity to collaborate by engaging in robust and meaningful discussions of various analyses and analytic approaches to key clinical questions.
|
|
Publications
Published works using N3C data are listed on the N3C Cohort Exploration Dashboard. The Publications tab displays titles and links to fully published articles, articles online ahead of print, and published preprints, as well as a list of accepted conference presentations and posters.
" In JMIR Medical Informatics (September 6, 2022)
Help Us Track N3C Publications!
When you have a research product that is ready for publication or accepted for presentation, please submit it via the N3C Publication Intent Form, which will notify the Publication Committee of N3C output to be registered. (Research products include: manuscripts, posters, conference papers, blogs, press releases, podium presentations, etc.)
Per the N3C Attribution and Publication Principles, all manuscripts using N3C community resources must be reviewed by the Publication Committee. Non-manuscript products do not require review but should be submitted after they have been accepted by the conference to allow for promotion and tracking of collaborator accomplishments.
|
|
N3C In the News
Tell Bennett, MD, MS, vice chair of clinical informatics in the Department of Biomedical Informatics, recounts how CU research teams collaborated with UCHealth to develop a dynamic modeling strategy to help in case triage efforts were needed in the early
Children with sickle cell disease (SCD) are at a high risk of developing severe symptoms after infection with the SARS-CoV-2 Omicron variant, one of the most recent variants of the virus that causes COVID-19, according to a new French study.
N3C gives researchers access to a large enclave of COVID-19 data related to the entire patient journey, including symptoms and outcomes.
The implications of coronavirus disease 2019 (COVID-19) infection on outcomes after invasive therapeutic strategies among patients presenting with acute myocardial infarction (AMI) are not well studied.
In a recent study posted to the medRxiv* preprint server, researchers explored the risk factors correlated with post-acute sequelae of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (PASC).
The COVID-19 pandemic not only persists but represents a substantive ongoing threat to patients on immunosuppressive therapies. Advances in antiviral therapies and vaccines have led to substantially reduced risk, of both infection and adverse outcomes, for people who can access them, but the risk is ongoing as new variants emerge. In this continuing battle, some approaches, such as glucocorticoids, will persist in their original form and some will become irrelevant, such as older neutralising monoclonal antibodies.
The US declared monkeypox a public health emergency this month, but the decision may have come too late. Though states are now required to report cases, and commercial labs have an approved test, a testing bottleneck persists, and cases — which passed 10,000 confirmed cases this week — are likely still being underreported.
Patients previously hospitalized with COVID-19 had a 45% higher risk of heart failure than other hospitalized patients, according to the first national study of its kind, which was co-authored by a University of Arkansas for Medical Sciences (UAMS) researcher.
Vermont and Maine have the oldest populations in the U.S., which coupled with rurality, predisposes northern New Englanders to health challenges including cancer, substance use disorders, food insecurity, obesity, diabetes, Alzheimer's and cardiovascular disease.
The authors of a new review paper have concluded that CBD shows “promising anti-inflammatory effects” in the treatment of Covid-19 symptoms and have recommended that more research be carried out.
In Norway, COVID survivors did not show progressive changes in cardiac structure and function that would explain persistent symptoms months later, a cohort study showed.
The coronavirus disease 2019 (COVID-19), which is caused by infection with the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has caused over 6.4 million deaths to date.
|
|
Sign Up for the Community Forum!
Please register for your Community Forum meeting link. After registering, you will receive a confirmation email from cd2h@uw.edu containing information about joining the meeting and you will also have the ability to download the .ics file to create calendar reminders for future Community Forums.
|
|
Events
N3Community Forum
Presentations take place on select Mondays from 5–6 p.m. ET/2–3 p.m. PT. To attend these and future N3Community Forum presentations, please register here.
Missed an N3Community Forum or want to revisit a past Forum? You can find all the videos on our YouTube page.
October 3, 2022
Topic: Pregnancy Domain Team Updates; Phenotyping in Pregnancy Vaccine; Effectiveness Analysis to Date in Pregnancy
Presenters: Rena Patel, MD, Elaine Hill, PhD, Sara Jones, Crystal (Qiuyuan) Qin
University of Washington, University of Rochester, NIH, University of Rochester
October 10, 2022
Topic: Fusing Medical Knowledge and Machine Learning Models for Early Pediatric COVID-19 Hospitalization and Severity Prediction
Presenter: Jimeng Sun, PhD
University of Illinois Urbana-Champaign
October 17, 2022
Topic: Upper Airway Disease in Children With the Omicron (B.1.1.529) Variant of SARS-CoV-2-A Report From the US National COVID Cohort Collaborative
Presenter: Blake Martin, MD
University of Colorado
October 24, 2022
Topic: Extracting Patient-level Social Determinants of Health into the OMOP Common Data Model
Presenter: Jim Phuong, PhD
University of Washington
November 14, 2022
Topic: The impact of COVID-19 on clinical outcomes among acute myocardial infarction patients undergoing early invasive treatment strategy
Presenter: Sula Mazimba, MD
University of Virginia
November 28, 2022
Topic: HIV, COVID, and SDoH: Leveraging N3C resources for an R01
Presenters: Rena Patel, MD, Jing Sun, MD, Jessica Islam, PhD, Sandra Safo, PhD, Jennifer Ross, MD, Bohdan Noysk, PhD
University of Washington, Johns Hopkins University, Moffitt Cancer Center, University of Minnesota, University of Washington, Simon Fraser University
|
|
No Meeting Weeks - Mark Your Calendars!
To help our community with a positive and productive workload, CD2H-N3C will schedule several “No Meetings Weeks” throughout the year.
Most meetings will be canceled. Impromptu meetings can still occur to push through action items as needed during no meeting week. Workgroups and Domain Teams should check with their Leads to determine meeting schedules for that week.
N3C support will continue with regular operations. If you have any trouble logging on to the enclave, please contact NCATSAuthSupport@mail.nih.gov. For all other issues please use the Support Desk.
4th Quarter: November 7-11
No Community Forum during this week
|
|
N3C Domain Teams
N3C Domain Teams enable researchers with shared interests to analyze data within the N3C Data Enclave and collaborate more efficiently in a team science environment. They include multidisciplinary Clinical Domains composed of subject matter experts, statisticians, informaticists, and machine learning specialists who focus on clinical questions surrounding COVID-19's impact on health. Cross-Cutting Domains have a varied focus that applies to multiple domains. These teams provide an opportunity to collect pilot data for grant submissions, train algorithms on larger datasets, inform clinical trial design, learn how to use tools for large-scale COVID-19 data, and validate results. N3C encourages researchers of all levels to join a Domain Team that represents their interests, or to suggest new clinical areas to explore.
|
|
Offboarding
As we have recently crossed the 2-year mark for the COVID-19 pandemic, it is a good inflection point to identify colleagues who have completed their efforts with CD2H/N3C projects and have transitioned to other great opportunities. If this is you, or perhaps your colleague, we would like to ask that you complete this 2-minute form (Bit.ly/cd2h-offboarding-form) to offboard CD2H and or N3C projects. You can continue to just get the newsletter if you wish.
We thank you for your tireless efforts in CD2H/N3C projects and look forward to working with you on many other projects.
|
|
The National COVID Cohort Collaborative (N3C) is a complementary and synergistic partnership among the Clinical and Translational Science Awards (CTSA) Program hubs, the National Center for Data to Health (CD2H), distributed clinical data networks (PCORnet, OHDSI, ACT, TriNetX), and other partner organizations, with overall stewardship by NIH’s National Center for Advancing Translational Sciences (NCATS). The N3C aims to improve the efficiency and accessibility of analyses using a very large row-level (patient-level) COVID-19 clinical dataset, demonstrate a novel approach for collaborative pandemic data sharing, and speed understanding of and treatments for COVID-19.
|
|
CD2H is supported by the National Center for Advancing Translational Sciences (NCATS)
at the National Institutes of Health
(Grant U24TR002306).
|
|
|
|
|
|
|
|
