Outbreaks of infectious diseases—such as novel coronavirus and pandemic flu—have raised concerns about how federal agencies use modeling to predict a disease’s course. Models can help decision makers set disease control policies and allocate resources. If models are unsound, they may not produce the reliable predictions needed to make good decisions.
The Government Accountability Office examined how Health and Human Services, which includes the Centers for Disease Control and Prevention, uses and assesses models. GAO recommended that HHS improve coordination of modeling across agencies and ensure models are reproducible, which helps build confidence in their results.
Within the Department of Health and Human Services (HHS), the Centers for Disease Control and Prevention (CDC) and the Office of the Assistant Secretary for Preparedness and Response (ASPR) used models to inform decision-making during and after outbreaks of Ebola, Zika, and pandemic influenza. These agencies’ modeling efforts informed public health planning, outbreak response, and, to a limited extent, resource allocation. Four CDC centers perform modeling.
HHS agencies reported using multiple mechanisms to coordinate modeling efforts across agencies, but they do not routinely monitor, evaluate, or report on the extent and success of coordination. Consequently, they risk missing opportunities to identify and address modeling challenges—such as communicating clearly, and obtaining adequate data and resources—before and during an outbreak. As a result, agencies may be limiting their ability to identify improvements in those and other areas. Further, there is potential for overlap and duplication of cross-agency modeling efforts, which could lead to inefficiencies.
CDC and ASPR generally developed and assessed their models in accordance with four steps GAO identified as commonly-recognized modeling practices: (1) communication between modeler and decision maker, (2) model description, (3) verification, and (4) validation. However, for four of the 10 models reviewed, CDC did not provide all details needed to reproduce model results, a key step that lets other scientists confirm those results. GAO found that CDC’s guidelines and policy do not address reproducibility of models or their code. This is inconsistent with HHS guidelines and may jeopardize the reliability of CDC’s research.
GAO recommends that HHS (1) develop a way to routinely monitor, evaluate, and report on modeling coordination efforts across multiple agencies and (2) direct CDC to establish guidelines to ensure full reproducibility of its models. HHS agreed with GAO’s recommendations.