Responses to Participant Questions

Dr. Bevc: Yesterday, seminar registrants should have received two PDF files. The PDF file titled Data Access and Availability Guide (PDF, 1 page) is going to be your starting point to learn how to access RECOVER data and understand which types of RECOVER data are available for study through the National Heart, Lung, and Blood Institute (NHLBI) BioData Catalyst® (BDC). You can visit the web app for BDC Powered by PIC-SURE (BDC-PIC-SURE) to start exploring the data. 

If you submit an application in response to the Research Opportunity Announcement (ROA) and are selected for funding, your research plan is approved through the ROA application process, and you’ll receive instructions for accessing the data through BDC. You could also submit a study plan and data request through RECOVER Ancillary Studies.

The second PDF document you received, Research Opportunities and Data Access Flowchart (PDF, 1 page), will help guide you through how to submit your research plan to request access to RECOVER data. Please look at both documents and submit any other questions through the resources our panelists have provided—they are continuing to drop links into the chat. A recording will be posted to the RECOVER website, so if you miss something, it’s going to be available to you. Additional resources on RECOVER data can be found on the RECOVER Data page. 

Ms. Hughes: If you go to the Discover page on the BDC-PIC-SURE website (accessed by clicking the “Discover” tab), you can search for different terms of interest—for example, “olfaction tests.” A tip for searching in BDC-PIC-SURE is to try a few search terms to see what’s returned. Someone asked about loss of smell, which might be related to olfaction tests, so try a couple of different searches to see what comes back and what data are available.

Dr. Horwitz: The study protocols are all posted publicly on the RECOVER Studies website, so you’ll always be able to see exactly what data we’ve been collecting in the study.

Dr. Gross: I also want to highlight that each of RECOVER’s observational cohorts has a study protocol manuscript. That publication gives an overview of the study design and all the assessments included. Read the adult and pediatric study protocol manuscripts on the RECOVER website. The most up-to-date information will be contained in the study protocols posted online that Dr. Horwitz mentioned.

Dr. Fisher: Yes—one of the strengths of BDC is that you can set up secure workspaces and invite other researchers. You can set role-based permissions that give different levels of access, such as copying files or executing workflows. That’s all handled through a registration process; you need an eRA Commons login to use the service. This makes it easy for researchers to collaborate and share information, data, and scripts without resorting to disallowed practices like emailing Excel files.

Dr. Fisher: I won’t sugarcoat it—it’s not necessarily easy, but we know you’ll need support to assess costs. All computational costs are direct charges from Amazon Web Services or Google Cloud and vary depending on the size of the computer resource that is utilized (more RAM or CPUs cost more). Generally, for working with RECOVER data, you’ll likely use a resource that costs about $0.50 per hour. You have to figure out how long setup will take and how long the actual analyses will run. Costs are stable at that per-hour charge and prorated by the minute. You can get into the weeds here, and there is lots of support to help you figure out expected spending, like this guide on estimating costs when using the BDC Powered by Seven Bridges (BDC-Seven Bridges) platform.

Mr. Chan: I would recommend the RECOVER Network of Biostatisticians website. It has lots of materials including the data dictionaries or codebooks and many resources to help you understand them.

Dr. Fisher: Absolutely. There are multiple ways to do this, and you can co-analyze RECOVER data with other pertinent data you may have.

Dr. Bevc: The current ROA will be accepting applications on a rolling basis through April 2027, depending on the availability of funds. Be sure to check the FAQs on the ROA webpage.

Dr. Horwitz: Our comorbidity/new condition form asks whether the participant is aware of a prior infection with EBV, cytomegalovirus, influenza, or Lyme disease. In addition, specimens collected for storage can be used post hoc by investigators to explore evidence of prior infection by such pathogens if so desired.

Dr. Horwitz: In the adult cohort, we use the University of Pennsylvania Smell Identification Test, which is a scratch-and-sniff test of 40 odors.

Dr. Horwitz: We do not do an oral examination; however, we do ask every 3 months about problems with teeth, rhinitis/sinusitis (runny nose), and changes in taste or smell.

Dr. Bevc: Autopsy data will be presented and discussed in the next R3 Seminar on January 13, 2026.

Dr. Horwitz: “Uninfected” participants are tested for active infection on enrollment (nasal swab for antigen polymerase chain reaction [PCR]) and for evidence of prior infection by nucleocapsid antibody (which is not affected by immunization) using a clinically certified blood test. “Uninfected” participants testing positive on enrollment are reclassified as infected. In addition, we have recently received data from a research laboratory that has run more sensitive multiplex antibody tests and has identified some more “uninfected” participants who actually have evidence of prior infection. These updated data will be indicated in future releases of the dataset. We do not do t-cell proliferation tests.

Dr. Horwitz: Timepoint 0 (“index” date) is set as the date of first infection, per participant, for infected participants. Participants may enroll on or after their index date and pick up the study schedule from the time point since infection at which they enroll.

Dr. Horwitz: Clinical trial results are not included in the observational study datasets.

Dr. Horwitz: The surveys are built in REDCap, an electronic database, and participants complete them online.

Dr. Horwitz: Yes. In the adult cohort, we have saliva samples (one time only, on enrollment), as long as the participant agreed to the possibility of future genetic testing. We also have nasal swabs on enrollment.

Ms. Hughes: In BDC-PIC-SURE, the variables are not currently linked to diagnosis or International Classification of Diseases (ICD) codes. I encourage you to do a couple of searches with different terms related to your research question.

Dr. Gross: The RECOVER Studies website has a page with instructions for how samples were collected.

Mr. Chan: The enrolled parent in the pregnancy cohort goes through the full adult protocol. So, saliva is collected at baseline, which can be any time since the actual pregnancy.

Dr. Horwitz: A subset of participants agreed to link their EHR data (to the extent that the data were available to them through the patient portal); those data are included in the dataset.

Mr. Chan: The Adolescent Brain Cognitive Development (ABCD) Study worked with us to deliver the RECOVER surveys to pediatric observational study participants in the ABCD cohort.

Dr. Horwitz: All symptoms include a follow-up question about how burdensome the symptom is to the participant (although these questions were not included early in the study). For many symptoms, we also ask about frequency and severity, if those data are required to calculate other disease scores, such as the Composite Autonomic Symptom Score-31 (COMPASS-31) for dysautonomia, the DePaul Symptom Questionnaire for Myalgic Encephalomyelitis/Chronic Fatigue Syndrome, or the Mast Cell Activation Syndrome (MCAS) instrument for MCAS.

Dr. Bevc: RECOVER’s Data Access and Availability Guide (PDF, 1 page) describes the types of RECOVER data available in open access versus controlled access.

Ms. Hughes: Researchers can use BDC-PIC-SURE to select participants who had specific tests performed. These include cardiac MRI and chest CT scans. However, I do not have information on whether these imaging data are available for research.

Ms. Hughes: Yes. BDC-PIC-SURE makes the data available based on the “freezes” or versions of the data shared by RECOVER. To check which version is currently available, you can go to the Data Dashboard page and refer to the accession number. This identification number corresponds to the specific version of the data currently available in BDC-PIC-SURE.

Dr. Horwitz: -Omics data for RECOVER are not housed in the All of Us database. We do not have COVID viral genetic data.

Dr. Horwitz: We do not provide clinical guidance on diagnosis or management of Long COVID.