How the FDA improved speed & accuracy of staff forecasting with KNIME
The FDA, or Food and Drug Administration, is a crucial player in safeguarding public health in the United States. This federal agency is tasked with overseeing the safety, efficacy, and quality of a wide range of products, from foods and drugs to medical devices and cosmetics. They play an essential role in protecting people's well-being, and their regulations and monitoring have a major impact on the healthcare industry, pharmaceuticals, and beyond.
The meticulous process of getting the perfect fully loaded cost model
As a part of its regulatory functions, the FDA collects over $3 billion a year in program administration fees from over 20,000 industry partners that produce regulated products, such as drugs and medical devices, and from some other entities, such as certain accreditation and certification bodies across the world to fund its operational needs. These fees are called “user fees.” However, managing this process - with intense checks and balances in place - was a complex and time-consuming task that was handled manually using Excel spreadsheets, complex macros, and significant federal resources.
One of the key pain points for the FDA, within the various processes related to collecting user fees, was the Fully Loaded Cost (FLC) model. The FLC model is a valuable asset for the FDA when it comes to determining user fees. It helps them calculate how much funding they need to hire full-time staff to support the FDA’s critical mission. This data is also used for capacity planning and to evaluate the cost of business cases internally. The FDA keeps the model up-to-date by using the latest financial data from their system, so they can set the right fees and keep operations running smoothly while bringing in the necessary revenue to cover operational expenditures.
But the process of manually updating the model using Excel was time-consuming and error-prone, especially with data coming in from various disparate sources - the obligation data from the financial system, reports from the budget planning system, and the consumer price index data from the US Department of Labor. The FDA’s user fee team spent countless hours in Excel performing manual calculations, custom coding macros, and collaborating over multiple spreadsheets to develop the FLC model. The process often took as long as 3 months per year and spanned millions of rows to arrive at the exact obligation that the FDA incurred in that year to keep the American public health protected.
The error-prone nature of the Excel-based model and the lack of version control resulted in significant re-work for the user fee team. The ability to drill down into the millions of rows and process large datasets was limited, making it difficult to hone in on the exact cost center they were trying to project in the industry negotiations. The Excel-based macros generated FLC projections manually each time. All executive dashboards were created by the team from scratch and lacked any interactive data visualization. Additionally, they were unable to broaden their analyses with the systemic use of business data from other applications.
From 3 months to one click: The power of automating with KNIME
The Deputy Director of the Office of Financial Management (OFM) at the FDA was looking for a way to upgrade and modernize these processes to make them more efficient. After evaluating multiple options, the FDA decided to use KNIME, a powerful data analytics platform that is well-suited to automating complex business processes.
With help from KNIME’s partner Equinoxys, the FDA put KNIME in action to import data from various formats and data sources and perform current as well as historical data analysis with minimal manual intervention. KNIME's workflows allowed the FDA to automate the FLC model, and its ability to integrate with other applications like the FDA's Oracle-based budget applications made it a seamless fit for the FDA's environment.
Using KNIME, the automation of the FLC model was completed with only 3 dedicated resources within 3 months. Coincidentally, the same amount of time was spent each year in the past to accomplish this major lift. Since then, the FDA has saved hundreds of labor hours with the ability to generate cost models in minutes. By automating data processing with KNIME workflows, they are able to eliminate errors and drill-down on large datasets effortlessly regardless of the size.
Additionally, KNIME's advanced analytical dashboards provide FDA executives with the ability to make fact-based decisions using what-if analyses and scorecards with financial metrics. These analyses are further bolstered by near real-time business data from other OFM applications.
As a result of implementing KNIME, the FDA has been able to improve the speed and accuracy of staff forecasting and reduce the time and resources required to manage the user fee process. This has allowed the FDA to focus more on their core mission of protecting public health and enhancing regulatory processes.
Having witnessed the substantial benefits of automating the FLC model with KNIME, the FDA is now working on implementing a one-stop-shop to handle all user fee business processes in an automated manner using KNIME such as the five-year financial plan, annual financial reports, federal notices, business case management, undelivered orders monitoring, and much more.