Automation and Analytics Improve CDC's COVID-19 Response

Challenge

During the COVID-19 crisis, the Centers for Disease Control and Prevention (CDC) faced an ever-changing situation in response to the ongoing and novel events regarding the safety and health of passengers and crew on board cruise ships. In April 2020, CDC formed the Maritime Unit, under the agency's Global Migration Task Force for the COVID-19 response, after CDC suspended cruise operations in March 2020. The goal of Maritime Unit was to work with the cruise industry, state, local, and territorial health departments, seaport authorities, and interagency partners to resume passenger operations in a way that prevented the introduction and transmission of COVID-19 on cruise ships operating in U.S. waters and further spread of the disease to U.S. destination communities.

Data sources used by Maritime Unit included:

• Aggregate counts of COVID-19 cases and COVID-like illnesses, testing volume, and severity indicators entered daily by ship clinicians into a REDCap platform
• Individual reports of deaths on board cruise ships entered into the Port Health Activity Reporting System (PHARS), a secure database used by CDC port health station staff
• Notifications by public health entities of individual COVID-19 case reports related to cruise travel entered into PHARS

A dedicated Maritime Unit team performed manual data extraction and analysis daily, and webpage updates weekly with ship color status (indicating onboard COVID-19 transmission risk), to inform decisions and guide communications about interventions and policies impacting cruise travel. This process required approximately 3 hours per day slowing the progress of epidemiological investigations and the development of better data to inform policy.

Solution

Guidehouse improved the Maritime Unit’s daily process by developing automated data pipelines that could evolve with the pandemic. We integrated data from multiple sources using IT platforms and servers already available in the enterprise, and we developed process-specific dashboards to speed up repetitive tasks needed for decision-making.

Guidehouse redesigned the COVID-19 data collection system used daily by cruise ships to improve data quality and reduce burden to cruise ship medical staff. Next, we developed server-based scheduled processes using application programming interfaces (APIs) to extract, clean, and process data into dynamic visualizations and reports, including risk-scoring. These included automations for internal reporting to CDC leadership for timely interventions when high-risk events occurred and external reports to communicate current risk status and disease dynamics to public health partners, cruise travelers, and industry stakeholders. As new variants emerged, we updated analytics and visualizations to show the latest risk thresholds.

Impact

By replacing manual processes with automated data processing and analytics, Guidehouse aided data collection and strategic communications development during the ever-changing pandemic landscape. Elimination of the manual data processes also allowed the Maritime Unit to use their time on epidemiological analyses and the development of evidence-based policies and guidance regarding COVID-19 on board cruise ships. The time needed to process, summarize, and review COVID-19 data was reduced from 10-30 hours to less than 50 minutes per week.

The developed dashboards enabled higher-quality, timely data visualization and centralization of information in the following ways:

Internal dashboards — Internal dashboards replaced manually written reports, simplified the daily outbreak review process to mitigate errors, reduced a three-hour process to 10 minutes, and enhanced how CDC presented data to the COVID-19 response team, CDC leadership, other US government agencies, ship representatives, and the public.

A public-facing, color-coded cruise ship status dashboard — This provided cruise ship travelers and industry stakeholders with weekly ship-by-ship COVID-19 risk status. This allowed passengers and crew to understand the potential risks of cruise travel.

Process improvement and streamlined workflows enabled additional scientific support — This included analysis of contact investigation data to understand disease spread between passengers and crew, as well as spatial aspects of disease spread such as cabin sharing and workspace proximity. These results were used to recommend interventions and improvements to COVID-19 response on cruise ships.

Reducing the Maritime Unit's efforts on repetitive data management actions enabled broader impact in evidence-based policy and epidemiologic investigations to mitigate disease spread on cruise ships.