Potential Community and Campus Covid-19 Outcomes Under University and College Reopening Scenarios

Abstract

Background: Significant uncertainty exists about the safety of, and best strategies for, reopening colleges and universities while the Covid-19 pandemic is not well-controlled. Little also is known about the effects that on-campus outbreaks may have on local non-student and/or higher-risk communities. Model-based analysis can help inform decision and policy making across a wide range of assumptions and uncertainties. Objective: To evaluate the potential range of campus and community Covid-19 exposures, infections, and mortality due to various university and college reopening plans and precautions. Methods: We developed and calibrated campus-only, community-only, and campus-x-community epidemic models using standard susceptible-exposed-infected-recovered differential equation and agent-based modeling methods. Input parameters for campus and surrounding communities were estimated via published and grey literature, scenario development, expert opinion, Monte Carlo simulation, and accuracy optimization algorithms; models were cross-validated against each other using February-June 2020 county, state, and country data. Campus opening plans (spanning various fully open, hybrid, and fully virtual approaches) were identified from websites, publications, communications, and surveys. All scenarios were simulated assuming 16-week semesters and best/worst case ranges for disease prevalence among community residents and arriving students, precaution compliance, contact frequency, virus attack rates, and tracing and isolation effectiveness. Day-to-day student and community differences in exposures, infections, and mortality were estimated under each scenario as compared to regular and no re-opening; 10% trimmed medians, standard deviations, and probability intervals were computed to omit extreme outlier scenarios. Factorial analyses were conducted to identify inputs with largest and smallest impacts on outcomes. Results: As a base case, predicted 16-week student infections and mortality under normal operations with no precautions (or no compliance) ranged from 472 to 9,484 (4.7% to 94.8%) and 2 to 61 (0.02% to 0.61%) per 10,000 student population, respectively. In terms of contact tracing and isolation resources, as many as 17 to 1,488 total exposures per 10,000 students could occur on a given day throughout the semester needing to be located, tested, and if warranted quarantined. Attributable total additional predicted community exposures, infections, and mortality ranged from 1 to 187, 13 to 820, and 1 to 21, respectively, assuming the university takes no additional precautions to limit exposure risk. The mean (SD) number of days until 1% and 5% of on-campus students are infected was 11 (3) and 76 (17) days, respectively; 34.8% of replications resulted in more than 10% students infected by semester end. The diffusion first inflection point occurred on average on day 84 (+/- 20 days, 95% interval). Common re-opening precaution strategies reduced the above consequences by 24% to 26% fewer infections (now 360 to 6,976 per 10,000 students) and 36% to 50% fewer deaths (now 1 to 39 per 10,000 students). Perfect testing and immediate quarantining of all students on arrival to campus at semester start further reduced infections by 58% to 95% (now 200 to 468 per 10,000 students) and deaths by 95% to 100% (now 0 to 3 per 10,000 students). Uncertainties in many factors, however, produced tremendous variability in all median estimates, ranging by -67% to +370%. Conclusions: Consequences of reopening college and university physical campuses on student and community Covid-19 exposures, infections, and mortality are very highly unpredictable, depending on a combination of random chance, controllable (e.g. physical layouts), and uncontrollable (e.g. human behavior) factors. Important implications at government and academic institution levels include clear needs for specific criteria to adapt campus operations mid-semester, methods to detect when this is necessary, and well-executed contingency plans for doing so.

Predictors of compliance with COVID-19 related non-pharmaceutical interventions among university students in the United States

“…A study looking at Fall 2020 classes has shown that counties with large universities with in-person instruction had a large increase in COVID-19 cases in the three weeks after school started, while counties with universities with remote instruction had a decrease in number of cases during this similar time frame [22]. While college students are typically at a low risk for severe infection and death, spread of COVID-19 on campuses has the potential to spill over into more vulnerable communities adjacent to campuses [23,24]. Additionally, long-term consequences of COVID-19 infections are not well understood [25], thus widespread infections amongst college students could plausibly lead to complications in the future.…”

Empirical networks for localized COVID-19 interventions using WiFi infrastructure at university campuses

“…Additionally, depending on campus priorities and resource limitations, different campuses can use this same data to model policies differently. The effectiveness of reopening policies is expected to be sensitive to a campus’ specific context that includes physical infrastructure, overarching guidelines, and human compliance [5]. For certain campuses policies might not need to be constrained by exposure risk as testing might be frequent, ubiquitous, and voluminous.…”

“…In the main paper, LC uses PageRank [42] as an illustrative algorithm to identify important location nodes. For robustness, we apply various additional algorithms to identify highly authoritative nodes in H — betweenness centrality [15], eigenvector centrality [5], and load centrality [29]. In the SI Appendix, we distinguish these different policies as LC PRank , LC BCen , LC ECen , LC LCen .…”

“…In the wake of the Coronavirus Disease (COVID-19) [41], the U.S. alone witnessed more than half a million cases at universities [54], and colleges are still left with decisions for operations in Fall 2021 [34, 44]. Controlling the disease at universities can be pivotal to securing the surrounding environment [5]. To reduce on-campus infections and the likelihood of superspreading events, a recommended form of non-pharmaceutical intervention (NPI) is partial closure of the campus [21].…”

Using structural equation modeling to model compliance with COVID-19 related non-pharmaceutical interventions amongst university students in the United States

“…As of November 19, 2020, there have been over 321,000 cases at over 1,700 American colleges and universities (21). While college students are typically at a low risk for severe infection and death, spread of COVID-19 on campuses has the potential to spill over into more vulnerable communities adjacent to campuses (22,23). Additionally, long-term consequences of COVID-19 infections are not well understood (24), thus widespread infections amongst college students could plausibly lead to complications in the future.…”