The Food and Drug Administration oversees everything from food safety to the efficacy of pharmaceutical drugs and medical devices. Despite an expansive mission, it wants to stand up a data management system that serves as a common foundation for all its work.
As part of its Data Modernization Action Plan, FDA intends to create a master data management system that will help it “understand what we’re doing in terms of common efforts to provide success on the mission of public health,” said Thomas Beach, the agency’s associate director for data governance, in an interview with Federal News Network.
FDA consists of nine center-level organizations, each with its own IT systems and own way of storing and cataloging data. Getting all of FDA working on a common enterprise data system will help the agency better handle food inspections, drug testing, laboratory work and all other aspects of its mission, Beach said.
“Can we all say, at the same time, we know exactly how many firms are doing how many things, in how many different places — and where and when? To be able to answer those questions quickly across an organization is just the testament of this,” he said.
Beach said the data modernization effort will build on three pillars: developing use cases that provide business value to the agency, creating repeatable data practices across FDA and building up a data-savvy workforce.
FDA data management efforts comparable to others across government
The majority of CDOs surveyed, 65%, told the foundation that “developing and implementing data strategy or governance was ‘very’ important to achieving the CDO mission,” and 55% pointed to improving data infrastructure as key.
The FDA’s Data Modernization Action Plan ties back to several other agency policy documents, including modernization action plans for technology, cybersecurity and the enterprise as a whole.
Beach said the agency is particularly busy developing the next generation of its data workforce, and that the agency is working with the second cohort of its data talent team, which is disbursed across the agency’s nine centers.
That’s another common CDO effort: 70% told the Data Foundation that they see influencing data culture as a critical focus of their role. “CDOs are change agents, engaging with stakeholders across their organization to improve the quality and use of data,” the report noted.
At FDA, its centers specialize in issues that range from pharmaceutical drugs to tobacco products to food safety, but Beach said they face some common data challenges.
“What they look at is how do we formulate the data science team in small groups and tackle maybe a specific issue or challenge, but also what are the enterprise lessons learned — like natural language processing, data wrangling and understanding how to convert data into decision-making. Those are some of the activities we have going on,” he said.
By way of example, Beach pointed to the data team’s portfolio of work tackling challenges related to FDA approval of tobacco products.
“Organizations will send in submissions for tobacco products, and they come in by the thousands. It’s just an unwieldy amount of data that has to be sifted through, and it’s a classic example of what you do with a data typhoon of information, to get to what is really relevant to the examiner or the person on the other side,” he said.
Future cohorts will tackle data challenges related to artificial intelligence and natural language processing. FDA Chief Data Officer Ram Iyer is considering scaling up the data cohort and other ways to build up the agency’s data workforce.
“Do we train them up in data science and continue to provide that [training], or do we go to another approach, which is to bring in new data scientists and have them work on the challenges that we have? It’s arguably a combination of both,” Beach said.
Engaging data stakeholders across FDA
Meanwhile, FDA stood up a Data Modernization Steering Committee to serve as the agency’s data governance board. The board addresses data challenges common across the agency, including data catalogs and inventories, as well as data governance.
“The focal point here is that we have all the representation from the centers involved. That’s really key to all of this,” Beach said.
FDA is further putting its data modernization efforts into action by reorganizing the Office of the Chief Information Officer into its commissioner’s office.
Beach said this reorganization helps puts the agency’s data expertise “in the spotlight.”
“If you have data acumen at the highest level, it’s connected to an organization that can go forth and do. It’s not in a vacuum, just sort of a policy with no action,” he said. “This reorganization is really crucial, and it’s a huge undertaking … in an organization as complex and broad and detailed as FDA.”
As part of this reorganization, FDA’s CDO can also help further the CIO’s mission and ensure that data sets purchased by one office within FDA are generally accessible to the rest of the agency, Beach said.
“Where in that whole requirements process is, ‘Have you looked and understood if this data is already available in the organization? Is there already another system that’s already creating this? Are you creating another system that’s a copy of that or a partial copy of it?’ That’s the net result, in my mind — that this partnership will lead to really create thoughtful requirements from the beginning,” Beach said.
FDA has made significant progress putting data at the forefront of day-to-day decision-making, as outlined under the Foundations for Evidence-Based Policymaking Act and the Federal Data Strategy.
“They’re really living and breathing the Evidence Act by positioning the CDO in the organization in a way that has the profile necessary to get things done. It’s our job to be this stakeholder-in-chief. I feel like that’s part of the role … taking in multiple points of view and building it out on the front end,” Beach said.
But putting policy into practice is no small feat. Beach said he’s been able to see both the data policy and implementation sides at work, having previously served on detail at the Office of Management and Budget and led the data governance team behind the Federal Data Strategy.
“It’s interesting to be on one side of writing playbooks and suggestions, and now living on the other side — how we do it and make it happen,” Beach said.
Data Foundation shares 4 ways to help CDOs better meet the goals of the Evidence Act
The government’s chief data officers are generally experienced technologists with federal pedigrees. And they’re making steady progress against the mandates of the Foundations for Evidence-Based Policymaking Act. Those were two findings that the Data Foundation reported after surveying the government’s CDOs in 2022.
Even so, the foundation offered four recommendations, based on its survey of federal CDOs, that could improve the efforts of CDO organizations to successfully improve the government use and management of data to inform and improve decision-making:
Recommendation 1: Congress should increase CDO funding flexibility and provide more direct resources.
Recommendation 2: The Office of Management and Budget should clarify the required CDO responsibilities, which would make full implementation of the OPEN Government Data Act possible.
Recommendation 3: Congress should create a federal CDO at OMB.
Recommendation 4: Congress should remove the statutory sunset for the CDO Council.
“It is imperative for the maturing CDO community that Congress, leaders in the executive branch, and external partners continue to find ways to support and encourage federal CDOs,” the foundation concluded in its report.