Social distancing refers to actions to reduce the number and duration of contacts and increase the physical distance between individuals to slow the spread of a communicable disease.1 In an influenza pandemic or other infectious disease outbreak, public health officials may recommend preemptive school closures to disrupt transmission before many students and staff members become ill and thereby decelerate community-wide spread of the disease. They may also recommend other social distancing measures in other community congregate settings (eg, workplaces, mass gatherings) to slow the spread of disease and thereby relieve pressure on overburdened health care and public health systems.2–4
Schools are important settings for social distancing. Because public schools are socially dense environments where more than 50 million students congregate across the United States each day, schools can fuel community-wide disease transmission.5–7 In addition to promoting respiratory etiquette and hand hygiene and engaging in frequent environmental cleaning, schools can implement practices that promote social distancing to potentially protect large numbers of vulnerable children, as well as limit secondary transmission to adults within their households and communities.
Over the past decade, research and guidance on social distancing in US schools have mainly focused on school closure as the most impactful, albeit disruptive, social distancing practice.1 While a substantive evidence base documents that school closure can mitigate influenza pandemics, there may be potentially less disruptive opportunities to increase social distance among students who remain in school.1–3,8 Nonetheless, feasibility, acceptability, and effectiveness of the full range of school social distancing practices have not been explored.2–4
To address this gap, the US Centers for Disease Control and Prevention and RAND Health conducted a large qualitative field study to examine the feasibility of social distancing practices other than school closures in K-12 schools. Through focus groups with senior education administrators across the United States, we aimed to identify potential social distancing practices beyond school closure, describe barriers to and facilitators of implementing these practices for at least 3 weeks to decrease the spread of influenza before many students become ill, and rank practices by feasibility.
In the summer of 2017, we conducted 36 focus groups with education and public health officials via webinar. Participants were selected from all 10 US Department of Health & Human Services (HHS) regions to explore perspectives that might differ by region of the country. We also recruited participants representing both primary and secondary schools to compare how barriers and facilitators might vary by grade level. We recruited superintendents, principals, teachers, school nurses, state school nurse consultants, district transportation directors, state health pandemic planners, and school safety representatives.
For each HHS region, we assembled a purposive sample of participants by searching professional association Web sites and LexisNexis and conducting snowball sampling with focus group attendees. Within professional association Web sites, we searched for lists of members, conference attendees, and association leaders. In LexisNexis, we looked for names of school leaders who were quoted in the media about emergency preparedness.
A total of 158 participants, representing all 10 regions, participated in a total of 36 focus groups. The number of participants per focus group ranged from 2 to 7, with a mean of 4. Each participant participated in a single focus group.
A team of 6 moderators trained in qualitative research conducted the focus groups via webinar. Most of the groups consisted of participants from a single HHS region, but 3 groups included a mix of several HHS regions to accommodate scheduling preferences. Participants were contacted via e-mail and offered a $50 gift card as an incentive. All focus group discussions were recorded and then transcribed.
Each group followed a semistructured protocol, and participants were asked to identify, mention experience with, and discuss the feasibility of practices within 1 of 2 categories: practices that could be implemented within a normal school schedule (“within-school practices”) and those that would require an altered school schedule (“reduced-schedule practices”). To limit focus group duration to 90 minutes, 23 groups discussed within-school practices (9 focused on elementary schools, 11 focused on secondary schools, and 3 discussed both elementary and secondary schools) and 13 discussed reduced-schedule practices (both elementary and secondary schools). In each group, participants were first presented with a list of practices assembled by the study team and asked to brainstorm any additional practices that could be implemented in K-12 schools. Second, they discussed any direct experiences with listed practices as well as implementation barriers and facilitators of each individual practice. Finally, they selected the most and least feasible practices from among the full list of practices. Feasibility was defined as “ease of implementation” in this context. This study was approved by RAND's Institutional Review Board, and all participants gave oral consent to participate at the start of each focus group.
Standard qualitative analysis techniques were used to identify and characterize instances of themes arising from the various topics covered in focus group protocols (eg, each identified practice) as well any unanticipated themes that emerged. Two of the authors read each transcript and independently coded. To ensure that different coders were interpreting the data as similarly as possible, we (1) developed descriptive and precise codebooks that gave each code a clear definition and meaning; (2) performed intercoder agreement checks prior to analyses where all analysts read the same text, coded independently, and discussed areas of disagreement; and (3) performed supervisory reviews of the analysis at regular intervals. We compared themes by HHS region as well as by secondary versus elementary school settings, and we reported any differences identified. Dedoose qualitative research software was used to facilitate data handling, coding, and thematic analyses.
For ranking, participants in within-school practices focus groups could vote for 3 practices they perceived as the most feasible and 3 perceived as the least feasible. For the focus groups on reduced-schedule practices, participants voted for the single most feasible and single least feasible practice. Individual votes were summed.
We first present the full menu of social distancing practices (other than school closure) discussed by focus group participants as well as high-level summaries of perceived barriers, facilitators, and variation by region and/or grade level. We then present details on the barriers and facilitators for the within-school and reduced-schedule practices that (1) were deemed most feasible by participants and (2) could be implemented continuously over at least 3 weeks.
Participants discussed a total of 29 school practices (25 within-school practices and 4 reduced-schedule practices). While 23 practices were identified by the study team in advance through policy and literature reviews, 6 additional within-school practices were identified by focus group participants. Additional practices included limiting group work, limiting congregation at arrival and dismissal, encouraging solo physical activity, canceling cross-school transfer for special programs such as dual enrollment, reducing congestion in the school health office, and educating students and family members to maintain their distance.
Practices with which participants reported prior experience as part of routine operations or in response to the 2009 H1N1 pandemic are shown in Table 1. For elementary schools, the more common within-school practices that some schools use as part of routine operations included homeroom stay, restricting hall movement (walking single file, a foot apart), segregating recess area by class, staggering recess times, segregating the cafeteria by class, and staggering lunch periods. Common within-school practices implemented for the first time in response to the 2009 H1N1 influenza pandemic included canceling field trips, canceling assemblies, limiting visitors, and reducing congestion in the health office. Several participants also reported experience with reduced-schedule practices as part of routine operations and in response to emergencies, including operating on a 4-day week because of budget constraints, and shortening the school day (delayed start or early dismissal for severe weather).
The perceived barriers and facilitators for each practice, as well as variability by grade level, are shown in Table 2. Within-school practices were generally perceived to be less feasible for secondary schools than elementary schools for a variety of reasons (eg, lack of homeroom; individualized and complex class schedules, including off-campus education activities; classes have students from multiple grades; need to use lockers to retrieve textbooks and other belongings). Reduced-schedule practices, such as shortening the school week or the school day, were perceived to be less feasible than within-school practices in both elementary and secondary schools because of complexities related to scheduling, transportation, staff work hours, communication to families, food preparation and provision, and a variety of regulatory issues (eg, required in-person instructional hours, union rules requiring duty-free periods for teachers) listed in Table 3. The need to arrange childcare was especially challenging for the parents of elementary school students.
As shown in Tables 4 and 5, the within-school practices most frequently perceived as feasible in both primary and secondary settings included canceling field trips (46 votes), canceling assemblies (41 votes), rearranging desks to increase space between students in the classroom (22 votes), restricting hall movement (elementary only) (20 votes), and limiting nonessential staff and visitors (19 votes). The practices most frequently identified as least feasible included moving class outdoors (41 votes), staggering class start and dismissal times (30 votes), separating classes into smaller groups (24 votes), and shortening and staggering lunch times (19 votes). The reduced-schedule practice considered the most feasible was a shortened school week affecting the entire school (25 votes), and the least feasible was selective dismissal of one class or one grade in a school (23 votes). We did not identify any substantive variation in perceived feasibility by US region; however, moving class outdoors was considered more feasible in locations with mild climates.
Feasible within-school practice: rearranging desks to increase space
Moving desks at least 3-ft apart to increase the physical distance between students is a social distancing measure that could be implemented in some classrooms. One participant discussed experience with this practice during the 2009 H1N1 pandemic, noting that he configured students' seats so that they were no longer facing each other in a pod formation and indicated that this required some substantial “geometry.” Barriers to this practice did not vary by location in the United States; however, there was variation by grade level because different types of seating arrangements are used for different ages. Barriers to implementing this practice included insufficient classroom space to spread desks out, inflexible seating arrangements and furniture, and negative impacts on students with special needs (eg, those who need to be located in the front row during instruction).
Participants indicated that flexible seating arrangements, such as desks and chairs that can be moved separately, would facilitate reconfiguring a classroom to create additional space between students during a pandemic. Also, a region 8 participant discussed how schools with limited space could rearrange students (eg, have all students face front, limiting face-to-face contact) without necessarily rearranging furniture. Finally, a region 10 participant discussed encouraging elementary-age students to fully utilize all the spaces in the classroom (eg, rug on the floor, rocking chair) rather than be limited to using desks and chairs at all times. He explained,
They like gathering in a carpet area or ... like a rocker or pillows. So, when they're doing their work, they can move around the room and find a different spot that's not at their table, and so that's a little bit more of a common practice at the elementary level.
Feasible reduced-schedule practice: shortened school week for the entire school
With a shortened school week, students do not attend 5 days in a row. This practice can apply to the entire school (eg, all students attend Monday through Wednesday only), or the school can alternate so that some students, for example, attend Monday and Tuesday and the others attend Thursday and Friday. Several focus group participants noted that their districts were already operating on a 4-day week because of budget constraints. The leading barriers to implementation of this practice include burden on parents to find childcare, impact on students who rely on schools for meals, need to make up missed instructional hours, challenging communication with parents about complex school schedules especially with alternating days, lower educational quality, and burden on staff. These barriers did not vary significantly by region of the country or grade level, with the exception of childcare needs being more pronounced for elementary-aged children.
Participants identified several facilitators to this practice. Many felt that a shortened school week would be less of a burden than shortened days on parents and on the bus transportation system because the disruption would be limited to 1 to 2 days per week rather than all 5 school days. Participants recommended having the whole district or at least all schools within the same feeder cluster operate on the same schedule to reduce some of the complexities for families with multiple children. Factors that could increase the feasibility of this practice include giving parents plenty of advance notice, engaging key stakeholders to explain why and how the new schedule will work, obtaining a waiver from the state so that district funding is not reduced, putting a distance learning plan in place to continue instruction when students are at home, and offering meal programs off the campus for students not attending school on a given day.
Through 36 focus groups in all 10 HHS regions, we identified and reviewed 29 social distancing practices that schools could implement in an influenza pandemic while continuing to operate. Participants reported prior experience with several within-school practices in elementary schools as a part of routine (nonemergency) operations. These practices included homeroom stay, restriction of hall movement, segregation of recess area by class, staggering of recess times, segregation of cafeteria by class, and staggering of lunch periods. Within-school practices were generally perceived to be more feasible for elementary schools than secondary schools. Reduced-schedule practices, such as shortening the school week or the school day, were perceived to be less feasible than within-school practices in both settings.
The practices considered the most feasible were those that involved isolated events like field trips and all-school assemblies or furniture rearrangement that had little to no staffing, curricular, or cost implications. In general, the practices considered the least feasible were those that required smaller classes, more staff, more square footage, or staggered start and end times. These were considered less feasible because they require hiring more staff on short notice, rewriting lesson plans, altering parent work schedules on a daily and prolonged basis, or finding more space in already crowded buildings. Within-school practices that do not change the academic calendar in any way are easier to implement than reduced-schedule practices. In addition, reduced-schedule practices that affect the entire school rather than subgroups of students are easier to implement because an alternating schedule presents additional challenges (eg, need for additional bus routes, extra burden on parents with students on different schedules, and need to explain complex scheduling with clear, consistent messages). State and district policies on required instructional time, required physical activity hours, duty-free periods for teachers, professional development hours for teachers, and/or teacher qualifications were frequently mentioned as key barriers to making significant changes to school schedules and supervision of students. Schools can surmount some of these regulatory legal and regulatory barriers by seeking waivers and building flexibility into staff contracts.
Our findings indicate that several within-school practices could be implemented in elementary schools to reduce the transmission of influenza (eg, keeping students in their homerooms for the entire school day, restricting hall movement, segregating recess area by class, staggering recess times, segregating the cafeteria by class, staggering lunch periods). These practices might be more effective in reducing disease transmission across classrooms than within a classroom. Practices such as canceling assemblies or field trips might be feasible but might not have a sustained impact on disease transmission because they are one-time events rather than practices that alter the nature of social interactions at school. Implementation of some practices, such as rearranging desks to increase space, might be feasible but might not keep students from mixing, as it would be difficult to limit students' movement in class (eg, require students to stay seated at all times). Students are likely to continue moving within the room and interacting with their peers. Among the reduced-schedule practices, shortening the school week for the entire school was perceived as the most feasible. Shortening the school week might be an alternative to prolonged school closure.
To our knowledge, our study represents the first one to comprehensively assess the feasibility of social distancing practices in schools. This study builds on our previous review of the literature, which found that neither the peer-reviewed and gray literature nor the pandemic influenza guidance and plans included details on the range of potential practices or on the barriers schools would likely encounter in implementing social distancing practices. Therefore, public health and school leaders had limited evidence to inform decisions about social distancing in schools.9 Our previous literature review identified 1 epidemiological and 5 modeling studies that assessed the effect of selected school practices on reducing influenza transmission.9 In addition to school closures, these studies considered limiting use of congregation spaces such as the cafeteria as well as a number of social distancing practices that our study identified as infeasible (eg, class and grade dismissal, classroom movement restrictions, and staggered classroom schedules). An additional modeling study reported that shortening school week may be effective in reducing virus transmission.10 Hence, further epidemiologic and modeling efforts are needed to explore the range of effects of social distancing practices that our study identified as feasible according to educators.
This study has a number of limitations. First, we did not include parents because we chose to focus on educators' implementation challenges in school settings. While parents are crucial to ensuring compliance, they do not make the decision to implement social distancing measures and are not tasked with enforcing them in schools. Nonetheless, many focus group participants are also parents or report to them, and they commented on how practices would be received by parents. Second, although we engaged numerous stakeholders and achieved thematic saturation, findings represent the perspectives of focus group participants and might not be generalizable. Third, we focused on the United States and did not engage school leaders from other countries, nor assess the implications of this work for an international audience. Future research should explore the feasibility of school social distancing practices in a wide range of developed and developing countries. Fourth, due to time constraints, some focus groups discussed within-school practices and others discussed reduced-schedule practices. Fifth, we explored perceptions of feasibility. More research is needed on the effectiveness of identified practices on reducing disease transmission. Before such evaluations are completed, infectious disease transmission experts can consider our data to determine which of the top-rated practices may be helpful in reducing school-based transmission of the disease(s) of interest. The study's strengths included the following: the study was based on a large number of focus groups, included participants from all HHS regions, and assessed feasibility by primary versus secondary school.
Implications for Policy & Practice
- Schools and public health officials can jointly consider multiple practices to reduce influenza transmission during a pandemic as an alternative to closing.
- These practices can also be considered in other outbreaks of infectious disease that affect school settings.
- Practices vary with respect to feasibility, with practices that can be implemented as part of the school day (vs those that affect school hours) considered the most feasible.
In summary, our findings suggest that schools have options to increase social distance during an influenza pandemic or other infectious disease outbreak as alternatives to closing. Given that influenza pandemics occur in waves and can last for months, it is critical to identify and consider alternatives to extended school closure, which is burdensome to students, parents, and employers.
Future research should evaluate the effects and optimal timing and duration of a set of seemingly feasible practices on influenza transmission in schools, given that selecting effective social distancing practices is the ultimate goal of policy makers and practitioners. In addition, feasibility and acceptability of the most promising practices will eventually need to be evaluated among other audiences, including parents and students.
1. Qualls N, Levitt A, Kanade N, et al Community mitigation guidelines to prevent pandemic influenza
—United States, 2017. MMWR Recomm Rep. 2017;66(1):1–34.
2. Jackson C, Mangtani P, Hawker J, Olowokure B, Vynnycky E. The effects of school closures on influenza outbreaks and pandemics: systematic review of simulation studies. PLoS One. 2014;9(5):e97297.
3. Jackson C, Vynnycky E, Hawker J, Olowokure B, Mangtani P. School closures and influenza: systematic review of epidemiological studies. BMJ Open. 2013;3(2):pii: e002149.
4. Rashid H, Ridda I, King C, et al Evidence compendium and advice on social distancing
and other related measures for response to an influenza pandemic. Paediatr Respir Rev. 2015;16(2):119–126.
5. Chao DL, Halloran ME, Longini IM Jr. School opening dates predict pandemic influenza
A(H1N1) outbreaks in the United States. J Infect Dis. 2010;202(6):877–880.
6. Gog JR, Ballesteros S, Viboud C, et al Spatial transmission of 2009 pandemic influenza
in the US. PLoS Comput Biol. 2014;10(6):e1003635.
7. National Center for Education Statistics. Back to school statistics. https://nces.ed.gov/fastfacts/display.asp?id=372
. Published 2017. Accessed November 12, 2017.
8. Community Preventive Services Task Force. Emergency preparedness and response: school dismissals to reduce transmission of pandemic influenza
. Published 2015. Accessed December 10, 2017.
9. Uscher-Pines L, Schwartz HL, Ahmed F, et al School practices to promote social distancing
in K-12 schools: review of influenza pandemic policies and practices. BMC Public Health. 2018;18(1):406.
10. Cooley P, Bartsch S, Brown S, Wheaton W, Wagener D, Lee B. Weekends as social distancing
and their effect on the spread of influenza. Comput Math Organ Theory. 2016;22(1):71–87.