Data-driven alert model to guide Elon’s COVID-19 response

This fall, Elon will use a four-level system based on 12 key measures of the impact of the virus on the campus community.

Elon University’s campus is operating on a four-level COVID-19 Alert System as the fall semester begins. Data gathered from on-campus sources and health providers in the region will allow university leaders to set color-coded alert levels as follows:

Green: Level 1 – New Normal

  • Continue to check health and temperature daily, wear face coverings and practice physical distancing: Identified cases are rare and transmission of the virus is controlled

Yellow: Level 2 – Moderate Alert

  • Increase efforts to limit exposure to yourself and others and continue all precautions against the spread of COVID-19: Moderate number of cases with most from a known source

Orange: Level 3 – High Alert

  • Limit everyday activities to increase safety: Many cases, including community spread with some undetected cases

Red: Level 4 – Very High Alert

  • Take strong measures to limit all contact: Widespread, uncontrolled outbreak with many undetected cases.

The COVID-19 Alert System was developed in consultation with Elon’s Infectious Disease Committee and data scientists at Cone Health Systems. It is based on the COVID-19 Planning Guide and Self-Assessment for Higher Education published by the Center for Health Security at the Johns Hopkins University Bloomberg School of Public Health, the Council for Higher Education Accreditation and Tuscany Strategy Consulting. The design is derived from a COVID-19 alert-level system developed by Vital Strategies, a global public health organization funded by Bloomberg Philanthropies, the Bill & Melinda Gates Foundation and Gates Philanthropy Partners.

“We will base our decisions about Elon operations on extensive data and objective criteria,” said Jeff Stein, vice president for strategic initiatives who is leading Elon’s Ready & Resilient implementation team. “This risk model allows us to take into account the many different factors that will impact our community as we return to in-person classes. We will have access to daily measures of the presence of the virus at Elon and be well-prepared to make decisions about actions we should take. We understand that there will be cases in our community, like all others, and we will use this evidence-based system as a guide as we navigate this evolving situation.”

Becky Neiduski, dean of the Elon University School of Health Sciences, says the Alert System will allow the university to monitor COVID-19 and take action to protect the community. “We’re ready for the cases that will occur,” Neiduski said. “We have contact tracing, quarantine and care plans in place for our community.”

The model is based on 12 key data points that are among 45 COVID-19 metrics the university is tracking daily. Eighty percent of the model weighting is based on data from the Elon University community, which includes nearly 9,000 students, faculty and staff. Twenty percent of the model is based on data from Alamance County and Cone Health System, which serves Triad counties at Alamance Regional Medical Center, Moses Cone Hospital, Wesley Long Hospital, Annie Penn Hospital and Cone’s Green Valley Campus, which is dedicated to treating severely ill COVID-19 patients.

Elon factors in the model include the following:

  • Total number of Elon positive COVID-19 cases in the past seven days
  • Percent of students in isolation or quarantine
  • Percent of faculty in isolation or quarantine
  • Percent of staff in isolation or quarantine
  • Percent of COVID-19 surveillance tests in the most recent week with positive results
  • Percent of Elon’s designated isolation/quarantine beds in use

Regional health system factors in the model include the following:

  • Alamance County daily new cases per 100,000 population
  • Alamance County COVID-19 virus reproduction (R0) rate
  • Cone Health Systems percent of COVID-19 tests with positive results
  • Cone Health Systems daily COVID-19 hospital admissions
  • Cone Health Systems daily number of patients as a percent of capacity
  • Cone Health Systems daily intensive care unit (ICU) beds in use

Each factor in the model is assigned a weight to help determine an overall alert-level score for the university. No single factor will determine whether the university needs to move up or down in level.

In addition to these factors, Elon is monitoring a wide set of data related to the presence of COVID-19 in North Carolina; the cumulative number of cases in the campus community; the number of students, faculty and staff being tested and involved in contact tracing; the number of students, faculty and staff who are studying or working remotely; and many other factors. Elon will use this data to expand its current COVID-19 dashboard on the Ready & Resilient website with additional metrics.

Tied to Elon’s COVID-19 Alert System are a series of action steps to mitigate the spread of the disease in the campus community:

Level 1 – New Normal

  • Follow Ready & Resilient healthy habits and campus protocols
  • Classes are in-person or hybrid
  • Gatherings are allowed per Elon policies, CDC guidelines and state/local rules
  • High-risk individuals can request accommodations

Level 2 – Moderate Alert

  • Instruction is primarily in-person and hybrid
  • Some course sections may have increased reliance on remote instruction due to students and/or faculty being placed in quarantine
  • Implement limits on campus activities and events as a safeguard for spread of the disease
  • Utilize expanded surveillance testing and contact tracing
  • Consult with county public health officials and infectious disease specialists on other mitigation measures needed

Level 3 – High Alert

  • Advise campus to begin twice-daily health checks
  • Daily consultation with county public health officials and infectious disease specialists on other mitigation measures needed
  • Widespread surveillance testing and contact tracing conducted
  • Expanded remote learning and administrative functions in impacted areas
  • Significantly reduce campus activities and events
  • Limit dining services to grab-and-go
  • Suspend inter-campus visitors to residence halls and limit access to some campus facilities

Level 4 – Very High Alert

  • Activate Emergency Operations Center
  • Implement university-wide testing
  • Consider university-wide remote learning for two weeks or longer
  • Suspend campus activities for two weeks or longer
  • Limit dining hall hours and operations
  • Consider measures to reduce residence hall population
  • Transition staff to work remotely with exception of essential staff in staggered shifts
  • Building access restricted to only essential personnel and services

“We have been working for more than two months to create a model that fits our community and meets scientific criteria for controlling the presence of COVID-19,” Stein said. “Data scientists and physicians at Cone Health Systems and Duke Health Systems have provided valuable consulting advice, aligning our model with sophisticated epidemiological models they are managing.”

Daily analysis team

Cone Health data scientists will join a team of public health experts Elon is assembling for ongoing consultation and monitoring of the pandemic’s impact. The group will support the work of Elon’s Infectious Disease Committee and the administration in analyzing and assessing the evolving issues throughout the academic year. Joining these experts in daily analysis of the COVID-19 Alert System and additional metrics will be:

  • Haya Ajjan, associate professor MIS, Gordon Professor in Entrepreneurship, director of the Center for Organizational Analytics and Faculty Administrative Fellow
  • Dan Anderson, vice president for university communications
  • Steve Bailey, professor of physical therapy and member of Academic Council
  • Steven House, executive vice president and professor of biology
  • Becky Neiduski, dean of the School of Health Sciences and professor of health sciences
  • Jana Lynn Patterson, associate vice president for student life, dean of students, assistant professor
  • Jeff Stein, vice president for strategic initiatives
  • Julie White, ERP application developer, Staff Advisory Council representative