October 2018 Cancer Epidemiology Matters E-News
Cancer Epidemiology Matters E-News
- Considerations when Developing a Data Sharing Plan
- Recordings Available from Body Composition and Cancer Outcomes Workshop
Considerations when Developing a Data Sharing Plan
NIH promotes sharing research data broadly with other researchers through several data-sharing policies and resources. De-identified data from human research participant studies are most often shared through a controlled-access model where qualified researchers may request data access to address specific research questions. An example of a controlled-access database is NIH’s database of Genotypes and Phenotypes (dbGaP). Some datasets may be truly “public” data sets, where anyone may have access to de-identified data. Available public data sets can be found at Data.gov.
For large budget research grants (those seeking $500,000 or more in subtotal direct costs), and grants that include producing large-scale genomic data, the NIH data-sharing policies expect that NIH-supported researchers will share final datasets (or data used for the main publications) through controlled-access models for de-identified data from human participants (e.g. dbGaP).
Large Budget Grants
To be compliant with NIH data sharing policies for large budget grants, the de-identified data may be made available in various ways, such as depositing the dataset to an NIH controlled-access database, securely transferring a dataset for analysis, or using an enclave model (such as a secure, access-controlled Cloud platform) for data access and analysis.
When considering how to share their data, NIH-supported investigators may choose to:
- Use an NIH data repository, and NIH will be responsible for sharing the data in accordance with the data sharing policies; or
- Store their data locally and provide access to the data in accordance with NIH policies.
If investigators choose the second option, they must consider the long-term sustainability of data storage and the resources needed to respond to data access requests. Over time, the resources required to store data and to review and respond to data requests may become more challenging, particularly once the grant supporting the original research project comes to an end.
- Regardless of the budget, projects that fall under the NIH Genomic Data Sharing (GDS) policy are expected to share data through an NIH-designated resource, such as dbGaP.
- Some Funding Opportunity Announcements have specific data-sharing expectations, such as PAR-17-233, Cohort Infrastructure and Methodological Research for Cancer Epidemiology Cohorts.
If you are considering a grant application that falls within the scientific areas of interest of the Epidemiology and Genomics Research Program (EGRP), scientific program directors in EGRP are available to advise investigators on options available to them as they develop data-sharing plans. EGRP-specific requirements related to data sharing plans for ARAs are available on EGRP’s website.
Recordings Available from Body Composition and Cancer Outcomes Workshop
In 2017, NCI’s Epidemiology and Genomics Research Program (EGRP) hosted a workshop, "Understanding the Role of Muscle and Body Composition in Studies of Cancer Risk and Prognosis in Cancer Survivors" that generated broad interest among the cancer research community.
The purpose of this workshop was to bring together scientists to discuss research examining how body composition affects cancer outcomes and to identify the research needed to inform recommendations for cancer survivors. The workshop identified key methodological challenges and approaches to optimize measurement and analysis for studying sarcopenia, cachexia, and frailty in cancer survivors.
Body composition is important for cancer survivors, as sarcopenia (low muscle mass) and obesity (excess adipose tissue) have been associated with increased chemotherapy toxicity, poorer surgical outcomes, and shorter survival. Some estimates suggest almost half of cancer survivors are sarcopenic, defined by low muscle mass and/or poor function, which is higher than the similarly-aged general population. Sarcopenia has been shown as an independent predictor of mortality after adjusting for BMI and other factors, suggesting muscle (body composition) plays an important role that can help explain cancer survival beyond measures of weight and height alone.
EGRP is pleased to share recordings of 15 presentations from this workshop for those who were unable to attend the workshop, as well as those who attended and are interested in reviewing information that was presented and discussed. Links to the recordings are embedded throughout the workshop agenda on EGRP’s website. The videos are also available on a YouTube playlist for the workshop.