Pandemic-Related School Closure Policy for Colleges and Universities:
A Report for Pandemic Planning Teams
By Steve Dalbey

Introduction

Planning for a pandemic is critical at institutions of higher learning, where the bulk of the student population (young, otherwise healthy individuals) may become a predominant at-risk population during a particularly severe pandemic. As a pandemic develops, colleges and universities may be faced with the dilemma of deciding if or when to close the classrooms to protect students and faculty members from the spread of a potentially deadly influenza virus. Health concerns are paramount but the health threats of influenza viruses vary in degrees of mortality and contagion. As such, there is no easy answer to the school closure question and there are significant consequences to the institution no matter what decision is made. School closure is only one of many pandemic planning issues, but one that can have significant impacts on students, faculty, and other stakeholders.

Influenza Virus Basics

When an influenza virus infects a human host, it replicates within the victim’s respiratory tract. The virus has its own genetic makeup that dictates its effect on the body. Two primary characteristics of an influenza virus are its mortality rate (how deadly it is) and its degree of contagion. High degrees of both characteristics are necessary for a worst-case pandemic to materialize. For example, the H5N1 (bird flu) virus that brought the subject of pandemics to the world’s attention a few years ago has proven to be very deadly, but with little or no capability for human-to-human spread. Conversely, the World Health Organization’s Phase 6 Pandemic Alert declaration in June 2009 was in response to the high degree of contagion of the H1N1 Swine Flu virus, not its fatality rate. Understanding these characteristics of influenza viruses is critical in formulating an appropriate response to the current, or next, pandemic.

As the deadly nature and degree of contagion of the viruses vary, so should the responses. In a worst-case scenario the best response to the threat of a pandemic may be to isolate oneself from the rest of society. However, it is neither necessary nor advised to treat every influenza outbreak as a worst-case scenario, at least until objective data from reputable sources warrants such a conclusion.  An organized response to the spread of a pandemic will assure that the health interests of the students and faculty are maintained as a first priority without completely sacrificing all other legitimate interests of the institution, the faculty, or the students.

School Closure Impacts

The primary mission of higher education is to educate students by offering classes and coursework.  At the same time, students' educational mission is attendance in and completion of classes and coursework. These joint missions combine to be an essential institutional function that cannot be ignored in the school-closure debate. Absent the continuous pursuit of educational goals by students, no other essential functions of the institution can be maintained.

With this in mind, if a virus has not demonstrated itself to be a significant threat to population groups other than typical at-risk groups (very young, very old, and the immunocompromised), a reasonable initial response may be to take all practical measures to prevent the spread of the virus without closing the school. While concern for the health of a mass of students is no minor matter, the response of the institution must consider the collateral, detrimental effects that result from a decision to close the school. Not only the students, but the faculty, staff, and the institution itself are stakeholders in the outcome.

As an example, school closure for part or all of a semester is likely to create a log jam of students trying to move forward with their educations after classes are resumed. The closure of colleges and universities does nothing to stop the perpetual flow of new students who graduate from high school in the spring. At the start of the next school year, if not the next term, there will be a new influx of students waiting to move into the classes that were supposed to be vacated, for example, by the sophomores at the time the freshman class arrives. The impact of clearing the backlog of students may be felt for some time after the closure is rescinded. Potential students confronting the backlog may be forced or simply elect to seek another school. A short list of other school closure consequences might include:

• Reputation issues regarding pandemic preparedness and/or ability of the institution to protect the students;
• Repetitive waves of illness that prompt multiple closures, affect student recruitment, and make the return of current students uncertain;
• Temporary and/or permanent tuition revenue losses;
• Layoffs of faculty and staff members;
• Issues of credits earned, tuition paid, etc. for partially-completed coursework;
• Inability of students to return home or leave campus for a variety of reasons;
• Return of students home to more risky environments than they left at school;
• Travel restrictions imposed on international students;
• Contractual issues with outside vendors/service providers.

These potential school closure consequences are not intended to imply that the health risks of the students, faculty, or staff should take a back seat to the financial concerns of the institution or any inconveniences closure of the school may cause. The point is that costs and benefits on both sides of the school closure arguments need to be given adequate consideration, each weighed against the others in a realistic, risk-conscious environment.

Pandemic Planning Considerations

The development of an appropriate institutional response is built on constantly monitoring and evaluating the threats to the students and faculty. A number of variables are considered in formulating the most appropriate response. The primary metrics that can be used, cumulatively, to judge the threat of the circulating influenza virus if/when it begins to affect the institution’s student population include clinical attack rates, affected populations, and case fatality ratios.

Clinical attack rates

Projections of clinical attack (illness) rates in advance of a pandemic can only be estimated. Too many variables exist that cannot be accurately predicted or controlled. Nevertheless, health officials will periodically publish estimates based on a combination of their research, historical data, and the current situation. However, quantifying infection rates over protracted periods of time may imply that everyone will become ill at roughly the same time and that the spread of the virus will be consistent across the country. Historically, pandemics manifest themselves in repeating waves, measured in weeks or months.  These waves do not correlate well with typical illness durations of 6 to 8 days when attempting to characterize the potential effects on the population. For example, a claim of illness rates of 40% does not mean that 40% of the population will be bed-ridden for 8 to 12 weeks or that all of the victims will be sick at the same time.

To illustrate, assume an illness-duration of one week during a pandemic wave of eight (8) weeks in a group of 100 people. If all illnesses were spread out evenly during that time period, five (5) people would be sick during each of the 8 weeks, totaling 40 people (40%) over the entire 8 weeks. Under those circumstances, a clinical attack rate of 40% would only impact 5% of the people in the group at any given time, statistically speaking. For a group of 100 people, 5% of them ill at once may not be significantly different than a bout of seasonal flu.

In reality, however, a wave of influenza would not affect the group with an even, across-the-board infection rate. It would start with a lower infection rate and peak somewhere in the middle, possibly at double the average rate. Given an average clinical attack rate of 5%, doubling that number would mean that, during the 8-week wave, it might peak at about 10% of the group. Infection rates would be lower than the average at the beginning and end of each wave of illness. Of course, the normal rate of seasonal flu would occur at the same time so it might be fair to estimate the total illness rate at 15% of the group for a short time when the pandemic wave peaks. So, while the clinical attack rate may be presented as 40%, it is important to understand, for planning purposes, how different the real attack rate might be.

Affected Population Characteristics

A healthy immune system will defeat the viruses that infect the body over the course of several days. Flu symptoms – fever, sore throat, cough, etc. – are evidence of the war between the virus and the immune response inside the body. In the absence of an effective immune system response, the virus spreads, killing the infected cells, and eventually takes over the respiratory system of the victim resulting in death. Very young children with underdeveloped immune systems, the elderly with failing immune systems, pregnant women, and other people with inadequately functioning immune systems or existing contributory medical conditions are all members of the at-risk group, even for seasonal influenza viruses.

Under unique circumstances with some deadly viruses, the healthy immune systems of young adults can actually create a medical crisis for them. An immune system overload known as a cytokine storm kills good cells along with infected cells in an uncontrolled manner, eventually (and relatively quickly) killing the victim. During the 1918 pandemic many of the victims were young, healthy adults who quickly succumbed to their infections due to this inappropriate immune system response. If young, college-age adults with otherwise healthy immune systems, become one of the at-risk groups during a pandemic this will be a cause for concern.

Case Fatality Ratio

The number of infected individuals who succumb to their illnesses, relative to the number of people who get infected, is referred to as the case fatality ratio. The mathematic equation involves division of the number of influenza-related fatalities by the number of influenza infections. However, those figures can be difficult to acquire. To begin with, everyone who becomes ill will not seek medical attention and not all who do will have their illnesses ‘typed’ to determine if they are infected with the H1N1 virus. Further, deaths that are facilitated by an influenza infection may be recorded as having resulted from an underlying chronic health condition, not influenza.

Practical Application of Available Data and Recommendations

The U.S. Department of Health and Human Services (DHHS), in its 2007 publication entitled Community Strategy for Pandemic Influenza Mitigation, included a Pandemic Severity Index. Using the case fatality ratio, DHHS (2007) states that the index “enable[s] estimation of the severity of a pandemic … to allow better forecasting of the impact and to enable recommendations to be made on the use of mitigation interventions” (Sec. I, para. 12). Table 1, following, identifies additional objective criteria that can be used to determine and/or confirm the pandemic severity category by comparing the excess death and illness rates as forms of checks and balances. The characteristics noted on Table 1 may not be readily available or easily accessed. Keep in mind that infection rates will vary from region to region and from community to community. Establishing a close relationship with state and/or local public health officials will pay off when seeking and interpreting the information needed to categorize the severity of the pandemic and estimate its local impact.

Table 1. Pandemic Severity Index by Epidemiological Characteristics (DHHS, 2007, Sec. XIV)

DHHS (2007) Table 2, following, displays several response options (interventions) to address the local pandemic situation through the use of the Pandemic Severity Index. Note that a very broad range of industries and institutions, personal and societal, are pigeon-holed into a limited number of ‘settings’ (home, school, and workplace/community) with recommendations for responses described as ‘generally not recommended’, ‘consider’ (some with qualifiers), or ‘recommended’. Definitions and footnotes for terms on this table are included within the reference materials for further review at http://www.flu.gov/professional/community/commitigation.html.

Table 2. Summary of the Community Mitigation Strategy by Pandemic Severity (DHHS, 2007, Sec. XIV)

The information and categories of “Interventions by Setting” on Table 2 do not specifically identify ‘colleges and universities’ as a setting for which specific interventions are recommended. The orientation of the ‘School’ setting may have some application to colleges and universities since the classroom-based learning environment is similar at all grade levels. However, school children may be considered to be at higher risk of contracting and spreading seasonal flu than college students due, in part, to immature immune systems, questionable personal hygiene habits, and the hands-on nature of their personal interactions which extend to their own families, the families of their friends, and other members of an at-large community with whom they have daily contact. It is the congregation of people of all ages, some of whom are ill whether they know it or not, that constitutes the social interaction threat that Table 2 addresses.

Given the wide range of campus activities that involve direct student interaction, the classroom may not be the most risky of them all. It is possible that cessation of a variety of other activities such as athletic contests, after-class social activities, and campus-based organizational gatherings, among others, may go a long way toward reducing the overall risk picture of the institution. Consequently, an alternative to immediate cancellation of all classes may be the phased-in elimination of all of the other non-vital opportunities for social interaction, leaving the classroom as the last target for intervention. The DHHS document from which the earlier tables were derived includes a broad range of information about non-pharmaceutical interventions (NPI) that should be considered as precursors to school closure. A detailed summary of DHHS-recommended NPI is beyond the scope of this report. However, the adoption of NPI principles is a critical element of an appropriate institutional response.

This planning process begins with an understanding and evaluation of the threats and determination of the illness and fatality rate variables. When sufficient statistical data has been acquired to quantify the current threats, a plan can be generated that outlines steps to be taken in advance of a school closure to minimize the spread of the illness. The success of non-pharmaceutical interventions may be enough to avoid closure altogether. ‘Trigger points’ for the implementation of increasingly stringent responses to increasing threats to the safety of an institution's students and staff are included in the DHHS community mitigation strategy document.

Conclusion

Making a decision on the school-closure question begins with an understanding and evaluation of the viral threats and determination of the degree of contagion and mortality rate. From that point, diligent monitoring of the illness and fatality rate variables globally, nationally, regionally, and locally will help to track the infiltration of the disease into the geographical area of the college or university. Regular examination of the most up-to-date information available from a variety of reliable sources such as the Centers for Disease Control and Prevention and the World Health Organization is a critical component of any pandemic planning strategy. Collaboration with state and local health officials will be necessary to get the most up-to-date local information on the spread of the virus.

The recommendation to use the DHHS Pandemic Severity Index is simply a way to quantify the results of the monitoring activities and to assign a degree of objectivity to the decision-making process. Adoption of non-pharmaceutical interventions may prove to be the better alternative for the stakeholders than closing the school. A number of other variables that distinguish one institution from another may play into the process as well. A one-size-fits-all approach, either for plan development or plan execution, will not necessarily meet the needs of the institution or achieve the best results for the students.

Reference

U.S. Department of Health and Human Services. (2007). Community strategy for pandemic influenza mitigation. Retrieved from http://www.flu.gov/professional/community/commitigation.html.

Additional Recommended Readings

Center for Law and the Public’s Health; Georgetown and Johns Hopkins Universities. (2008). Legal preparedness for school closures in response to pandemic influenza and other emergencies. Retrieved from http://www.flu.gov/professional/school/schoolclosures.pdf.
 
U.S. Department of Education. (2009). Preparing for the flu: Department of Education recommendations to ensure the continuity of learning for schools (K-12) during extended student absence or school dismissal. Retrieved from http://www.ed.gov/admins/lead/safety/emergencyplan/pandemic/guidance/continuity-recs.pdf.
 
World Health Organization. (2009). Assessing the severity of an influenza pandemic. Retrieved from http://www.who.int/csr/disease/swineflu/assess/disease_swineflu_assess_20090511/en/