Part I. Community Guide Update and Rationale for Intervention
| Author & Year | Intervention Setting, Description, and Comparison Group(s) | Study Population Description and Sample Size | Effect Measure (Variables) | Results including Test Statistics and Significance | Follow-up Time |
| Kawano & Kakehashi, 2015 | Reactive school closures in Oita City, Japan during the 2009 H1N1 Pandemic.
Children were not allowed to come to school when symptomatic, and until 2 days after resolution of fever. Classes closed when 10% of students were infected. Grades closed when multiple classes within that grade were closed and schools were closed when multiple grades within the school were closed. They used data from the actual infection to create a model to compare schools with and without closure. |
This study used data from the Oita City Board of Education which had daily reports from all schools of number of infected students and subsequent closures. Data included all 134 schools in Oita City, with a total of 51,872 students from kindergarten through high school. | 1. Which schools closed – when, which classes/grades, for how long. 2. Daily reports of infected students |
At the peak of the pandemic, school closure decreased the influenza prevalence by 24%. The total number of flu cases was decreased by 8.0% by school closure. |
August 2009-March 2010
(8 months) |
| Egger et al., 2012 | Active surveillance by the NYC Department of Health and Mental Hygiene (DOHMH), and subsequent selective school closure in New York City public schools during the 2009 H1N1 outbreak.
About 1/3 of the schools that met dismissal criteria actually dismissed students (for a median of 6 days). The existing surveillance data was used to create a model to predict epi curves for ILI for schools that met the dismissal criteria. Data for schools that did not dismiss were compared to projections for schools that did (surveillance data was not collected after dismissal). |
Out of 434 total public and private schools, 64 NYC public schools met criteria for dismissal. Criteria were based on presentation for influenza-like-illness (ILI) to the school nurse. Dismissal criteria included a school-wide single-day ILI prevalence of ≥2% or ≥1% for 2 consecutive days. | 1. Whether or not a school that met dismissal criteria closed 2. Percentage of students presenting to the school nurse for ILI (daily surveillance) |
Compared to similar schools that did not close, schools that closed were predicted to have 7.1% decreased ILI prevalence over the entire outbreak. | May 11 – June 26, 2009 |
| Davis et al., 2015 | There were 567 school closures in the state of Michigan in response to the 2009 H1N1 influenza outbreak. School districts were compared based on when they closed (in relation to peak # of ILI cases in the district), and what proportion of schools closed. | 559 public and private schools closed during the fall semester of 2009. Some schools closed multiple times, resulting in 567 total closures. |
Level 1: 0% Level 2: 1-50% Level 3: 50-100% |
65% of schools closures were reactive.
Districts with the highest proportion of schools closed had the highest ILI incidence. Districts with no closed schools had higher ILI incidence than districts with up to 50% of schools closed. |
September 1 – December 31, 2009 |
The Community Guide last made recommendations about school closures to limit transmission of pandemic influenza in August of 2012, and used literature published through February of 2011 to make its recommendations. After conducting further literature search of literature published since this time, I agree with the most recent recommendations. Whether or not school closure is more beneficial than costly depends greatly on the severity of disease of the pandemic influenza and the timing, coordination, and scale of school closures. My updated recommendation is aligned with the 2012 Community Guide recommendation. I recommend preemptive coordinated school closures during pandemics of severe influenza. There is still insufficient evidence for school closures during moderate or mild influenza pandemics, as well as for less coordinated and reactionary closures.
My literature search was mostly investigating the 2009 H1N1 pandemic influenza and the 2012 recommendations also included similar literature that was published before February of 2011. This particular strain of influenza was novel, but did not produce very serious disease in most cases. In the 2015 paper by Davis et al., the authors even stated that many schools that met dismissal criteria but did not close down were not experiencing very serious disease, even though they had an increased prevalence of influenza-like-illness (ILI). Parents, teachers, and administrators are less likely to perceive the benefits of a potential school closure if the disease is mild. The recommendation for school closures by the community guide is for severe pandemic influenza – the 1918 influenza pandemic is used as a reference. The recommendation takes into account studies that use data from the 1918 pandemic when school closures occurred in a few major cities in the United States. These studies concluded that areas that utilized school closures during this pandemic benefited. The benefits for a more moderate or mild influenza are more marginal, and may not outweigh the costs such as lost classroom time, stress and cost to households where both parents work and scramble to find child care during school closures.
In addition to the H1N1 influenza pandemic being mostly moderate disease, the school closures in 2009 were reactionary (after a high incidence of ILI and sometimes even after peak incidence). There is a much greater benefit (reduced influenza transmission) as a result of preemptive school closures when compared with reactive closures, especially during severe disease. School closures also provide the most benefit when they are coordinated within states and districts. Schools/school districts should set thresholds of ILI incidence and triggers for when to execute school closures, then disseminate and follow these plans.
Part II. Theoretical Framework/Model
Since school dismissals are not a decision made at the individual level, the Social Ecological Model is the best model to apply to this intervention strategy. This model takes factors at five different levels into account and is depicted as concentric circles that represent the five levels: individual/intrapersonal, interpersonal, institutional, community, and policy. The Social Ecological Model allows us to explore how factors at multiple levels interact and influence each other within the scope of school closures during an influenza pandemic. All five levels influence each other. For example, if school closure is decided at the level of a principal (that is, if no state or school-district level policy exists), then that individual’s perception of the effectiveness of school closures affects the culture of the institution. This example demonstrates how policy, individual factors, and institutional culture are intertwined.
Individual
There are many individuals involved in pandemic influenza school closures, including students, teachers, parents, and school administrators. I also consider the characteristics of the virus “individual” factors, as the virulence of the virus, transmissibility, infectiousness, and severity of disease play a major role in the efficacy of school closures to prevent transmission. Since this is an infectious disease, individual student, parent, and teacher immunity (including vaccination behaviors) plays a role in whether closures are necessary. Parents’, teachers’, and administrators’ perception of the risks of the disease influence motivation for school closures. The perception of the effectiveness of school closures at reducing transmission is another key determinant of whether a school closes, when, and based on what criteria.
Interpersonal
One major interpersonal factor in school closures is the ability of parents or guardians to provide child care in the event of a closure. This is a reflection of the support network within the family. Support systems within schools affect teachers, students, and administrators. School support systems are especially important after school closures, and help to reduce negative effects of lost class time.
Institutional
The core institutions involved in school closures due to pandemic influenza are schools, school districts, and local/state health departments. The presence and preparedness of school nurses affects detection of increased incidence of influenza-like illness (ILI) in students. Many school closure policies use increased ILI incidence as triggers for closures. Local health departments also can use surveillance data about ILI in the general population to determine if there is an ongoing outbreak. The diagnostic capabilities and clinical skills of area physicians also plays a role in ILI surveillance.
Community
Community factors include communication and relationships between schools in the same area, between schools and districts, and between schools, districts, physicians, and health departments. Having efficient case reporting systems in place greatly affect quality of surveillance. Financial support of institutions plays a huge role in their ability to respond and act in case of pandemic influenza. Funding influences resources for surveillance and having competent staff at every institution. This is where some health disparities come in, especially since most school districts are funded through local property taxes. This means that schools in low-income areas have fewer resources, including school nurses.
Policy
The highest level includes policies/plans at the state and school district level. Many states/school districts have plans in place for triggers of school closures during influenza outbreaks. These may also be influenced by the national response to the pandemic. The CDC tracks influenza trends, so the dissemination of information about disease severity and national trends can inform decision making at lower levels.
Part III. Logic Model, Causal and Intervention Hypotheses, and Intervention Strategies
The target population for this intervention is the Horry County School System. The Horry County School System is a district comprised of 51 schools and serving over 42,000 students. At this time the district does not have a school closure policy in place for severe influenza pandemics. The SC Department of Health and Environmental Control does currently conduct active influenza surveillance with SC ILINet providers. Horry County is in the northeast corner of the state of South Carolina. This semi-urban county is highly diverse with a variety of races and income levels represented. Despite being diverse, there is a sizeable proportion of low-income families serviced by Horry County Schools. Low-income families often disproportionately feel the negative effects of school closures due to decreased child care flexibility when both parents are working. According to Egger et al., during the 2009 H1N1 pandemic in New York City, school-aged children (5-18 years old) had 3.5 times more emergency room visits due to ILI than all other ages (Egger et al., 2012). Schools and school buses have high social densities, which makes them sources of increased transmission and epidemic amplification (US Centers for Disease Control and Prevention, 2007). Social distancing is one of the most common non-pharmaceutical interventions (NPIs) for influenza, and school closures are one form of social distancing that targets this population that has greater susceptibility and transmission potential for pandemic influenza.
| Intervention Method | Alignment with Theory | Intervention Strategy |
| Social distancing during influenza pandemics | In the Social Ecological Theory, all levels are connected and influence each other. This intervention method has strategy developed at the institutional level but also has direct impact on the community, interpersonal, and individual levels. In theory, school closures decrease ILI incidence in the entire community, not just within a school. School closures require individuals to rely on each other for alternative child care. Individual students, teachers, and parents are affected when schools close as well (in both positive and negative ways). Making plans apply to an entire district helps even out some of the health disparities associated with influenza school closures. | Develop district-wide policies for pre-emptive, coordinated school closures in the event of a severe influenza pandemic. Policies should establish threshold levels of ILI incidence that would trigger a school closure. |
| Increase individual knowledge about influenza and efficacy of school closures | Increased individual knowledge about influenza and school closures helps make closures smoother if/when they are needed. This fits in at the community, institutional, interpersonal, and individual levels of the Social Ecological Model. . By increasing awareness about the efficacy of school closures at reducing transmission of serious pandemic influenza, you are also encouraging other social distancing measures and NPIs, which increases the likelihood of decreased influenza transmission throughout the Community. | Schools hold informational sessions for parents/guardians about influenza, school closures, and other non-pharmaceutical influenza interventions. |
| Increase social support | Social support is most pertinent to the interpersonal level. Families will feel support from schools and other individuals that they make child care plans with. This intervention method also improves Interpersonal relationships via familial support – by helping parents develop plans for child care in the event of a school closure ahead of time. School administrators will be more comfortable when implementing school closures knowing that parents are prepared to some extent. | Parent/guardian info sessions will also include guidance about how to plan for alternative childcare in the case of school closures. |
| Optimize district ILI surveillance | Surveillance is most relevant to the Community level, with multiple institutions (physicians, local health departments, schools/school nurses) all collaborating with each other for the best result. Having reliable surveillance information also ensures that school closure policies will be implemented at the right time and on the right scale. Although national surveillance data (at the Policy level) is crucial in determining the severity of the pandemic, having reliable local surveillance is essential for appropriate timing of school closures, as pre-emptive closures are most effective. | The district will hold continuing education for school nurses about influenza and surveillance.
School nurses will be required to report all ILI cases to a designated individual at the Horry County Department of Health and Environmental Control (DHEC) Office. Relationships will be formed between Horry Co DHEC and Horry Co Schools. |
| Inputs/Resources | Activities | Outputs | Short-term Outcomes | Intermediate Outcomes | Long-term Outcomes |
| Personnel: school nurses, local health department employees (epidemiologist), at least one person who works for the school district to take charge of developing and implementing the program, including education
Funding: personnel salaries, not much needed to diagnose ILI (thermometer, stethoscope), access to journals to find literature on best practices for school closures Materials: educational materials (such as pamphlets, newsletters, presentations/posters) Relationships: relationships are needed between schools, districts, physicians, and local health departments |
1. Develop a specific policy for school closures in the event of pandemic influenza based on best practices. This should include what level of ILI incidence will trigger a closure. Policies should be made available online for the public to read.
2. Inform parents through presentations at parent organization meetings, flyers/pamphlets sent home, etc about the new policy and how it will benefit their children. 4. Encourage families to come up with plans for child care in the event of school closures. 4. Require school nurses to report ILI cases to DHEC. |
1. Each district will develop a school closure policy which will be published on their website.
2. Each school should have at least one informational session for teachers and parents on the benefits of school closures and how a school closure would work. 3. Every school nurse will be educated on ILI symptoms, diagnosis, and reporting 4. Local health departments will conduct ILI surveillance and communicate findings with school districts. |
1. Increased individual knowledge about influenza and how school closures can reduce transmission through social distancing 2. Improved social support from schools to families and between schools, districts, and local health departments 3. District-wide influenza school closure policy put in place. 4. Accurate and reliable school ILI surveillance data |
In the event of a pandemic of severe influenza, surveillance will indicate when peak incidence will occur and there will be coordinated school closures before peak incidence. | Decreased morbidity and mortality associated with future pandemic influenza events |
Intervention Hypotheses:
- Development of school closure policies will lead to improved support between schools, districts, and individuals.
- Holding informational sessions for parents will lead to increased individual knowledge about influenza and how school closures can reduce transmission.
- Encouraging families to plan for alternative child-care during school closures will lead to improved social support from schools to families and within families and communities.
- Mandatory reporting of ILI cases by school nurses will lead to accurate and reliable school ILI surveillance data.
Causal Hypotheses:
- Increased individual knowledge about influenza will increase compliance of individuals during school closures.
- Increased support from schools to families and improved individual knowledge about ILI will increase compliance and decrease strain on families during school closures
- Having school closures in place will decrease transmission of pandemic influenza through social distancing.
- Having accurate and reliable school ILI surveillance data will ensure that school closures are pre-emptive, leading to decreased influenza transmission.
All four of these intermediate outcomes will lead to decreased morbidity and mortality associated with pandemic influenza.
SMART Objectives
- Immediately following parent informational sessions (held within 6 months of policy finalization), average scores on an influenza/school closure knowledge test will be at least 80%.
- At the same time as the knowledge tests, a questionnaire about parent satisfaction with school support (related to influenza policy) will be distributed. 50% of parents will report being at least “satisfied” on a likert scale from “very dissatisfied” to “very satisfied.”
- There will be a statistically significant (p<0.05) difference between actual and projected influenza incidence rates in the event of a severe influenza pandemic and subsequent school closures.
- Within the first year, the school ILI surveillance system will have a sensitivity of at least 75%.
Part IV. Evaluation Design and Measures
| Stakeholder | Role in Intervention | Evaluation Questions from Stakeholder | Effect on Stakeholder of a Successful Program | Effect on Stakeholder of an Unsuccessful Program |
| Students | Participants | What will I miss in school?
Will school closures decrease my risk of getting influenza? |
Reduced likelihood of pandemic influenza transmission at school.
Minimal loss of class time. |
The same (high) risk of pandemic influenza transmission at school.
Loss of class time. |
| Parents/guardians/family members | Participant | How will I have to adapt in the case of a school dismissal?
Will school closures decrease the risk of my child getting influenza? Will they decrease my risk of getting influenza? |
Minimal disruption to work/life balance. Decreased transmission of influenza to their child. Increased awareness and knowledge of influenza. | Disruption to work schedule, strain on resources, stress among family members.
No change to child’s influenza risk. |
| School Administrators | Program implementation | How smoothly will communication of surveillance data to me be?
Will school closures affect academic performance of students? |
Minimal disruption to academic calendar. Reduced absenteeism due to influenza before and after closure.
Adequate notification of increased ILI incidence from local health department. |
High transmission of influenza among students, high absenteeism before and after closure. |
| School District Personnel | Program development, implementation, and funding | How much money and personnel will be needed to implement the program?
Will school closures decrease influenza transmission among students and/or negatively impact academic performance of students? |
A program that does not cost an excessive amount of money.
Decreased influenza transmission. No disruption to academic calendar. |
An expensive program that does not decrease influenza transmission. |
| Local Health Department | Program implementation and monitoring | Will school nurses and local physicians report ILI cases in a timely manner? | Timely and accurate reporting of ILI cases. Compilation of data and notification to school administrators of threshold incidence being reached. | Incomplete reporting of ILI cases, inaccurate surveillance data. |
| Community Members | Provide support for schools and families. | How will school closures affect influenza transmission in the community? | Decreased influenza prevalence and transmission from students, teachers, and staff to community members. | No change in influenza transmission and prevalence in the community. |
The outcome of parent/guardian education about influenza and school closures can be measured after development of the school closure policy. Parents will be given a pre-test at the beginning of a parent organization meeting, and then later in the meeting an informational session about the influenza school closure policy. This is a One group pre-test/post-test study design, represented below:
O1 X O2
The biggest threat to internal validity of this measure is selection bias – not all parents go to parent organization meetings. This is especially true to lower-income parents who may work non-business hours. In order to reach more parents, informational pamphlets should be sent home with students. To test parents’ knowledge, tests could be sent home as well, or given at parent-teacher conferences or other events where parents and teachers/administrators come into contact with each other. Information can also be presented in a video that can be distributed through email and the school district’s social media sites with online pre and post-tests. By conducting the pre and post-tests close together, it reduces the risk of maturation, history, and attrition threats.
Another possible method to measure efficacy of education would be to randomize which schools hold informational settings and do a randomized control group design, depicted below:
R O1 X O2
R O1 O2
However, with something as serious as pandemic influenza it is not ethical to withhold information on how to respond if a pandemic occurred.
Although education and increased self-efficacy of parents are important, the primary outcome of interest for school closures is reduced influenza transmission. The efficacy of school closures to reduce influenza transmission can only be measured in the event of a pandemic of severe influenza (which has not occurred since 1918). Since school closures are hypothesized to decrease influenza transmission in the event of a pandemic with the same severity of the 1918 Spanish Flu, it would be unethical to have control schools in this district that did not dismiss schools after reaching threshold influenza incidence. For this and other reasons, this would have to be a strictly observational study. Surveillance data will be used to create a model for projected influenza incidence without closure and will be compared to incidence with closure. In this case, the model is a stand-in for a control group.
An observational study inherently has many threats to internal validity. Influenza transmission is extremely complex with multiple variables to consider and these variables will be practically impossible to control in the event of a severe influenza pandemic. Historical threats include variations in individual immunity that affect whether or not a person will get influenza. In addition, school is just one of countless places children could come into contact with an infected person. The incidence of influenza may have reduced without school closures for unrelated reasons (maturation effect).
| Short-term or Intermediate Outcome Variable | Scale, Questionnaire, Assessment Method | Brief Description of Instrument | Example item (for surveys, scales, or questionnaires) | Reliability and/or Validity Description |
| Increased awareness of pandemic influenza and school closures | Adapted questionnaire from (Koep et al., 2014) distributed | Koep et al described a questionnaire that they developed for school-aged children. This survey is 20 questions and asks about pharmaceutical and non-pharmaceutical interventions for influenza, as well as risk perception (Koep et al., 2014). This questionnaire should be adapted for parents of school-aged children. | Risk perception: I am likely to get the flu (Likert scale from strongly disagree to strongly agree).
Non-pharmaceutical interventions: Staying at home when I am sick will keep others from getting sick (Likert scale). |
Out of 20 questions, 16 met researcher’s criteria for high inter-class reliability after 1 year (>70% question repeatability between both administrations). |
| Decreased influenza incidence | Direct observation (retrospective observational study after severe influenza pandemic) | Structure similar to (Kawano & Kakehashi, 2015) study, where they used daily case numbers to create a model and compare projected vs actual incidence. | Daily reports of ILI cases from school nurses to local health department personnel. | Reliability of model depends on accuracy of surveillance data. High probability of reporting bias, especially in underfunded schools (reduced resources = increased likelihood of underreporting).
This has good construct validity since we are directly measuring incidence (as a reflection of transmission). |
Part V. Process Evaluation and Data Collection Form
RECRUITMENT
Administrator “Recruitment”/Informational Letter
Dear [school administrator],
This year, Horry County Schools is taking an important step in preparation for Flu Season. In the unlikely event of a pandemic of severe influenza, evidence has shown that school dismissals can reduce the number of cases of influenza in school-aged children. The district will be developing a policy in the coming months for the best way to respond (in coordination with the local health department) to an influenza pandemic. This policy will be distributed to all schools in the district after it is developed, as well as informational materials. Schools will be expected to hold informational sessions (with the aid of district personnel) for parents within 3 months of policy development. It is in the best interest of our teachers, students, and staff to be prepared for severe pandemics.
Thank you for your cooperation, please do not hesitate to contact ____________ at _____@hcs.net with any relevant questions or concerns.
Sincerely,
Horry Country Schools
Enrollment Form
Schools/administrators that respond with concerns:
1.
2.
3.
4.
5.
Parent Recruitment Letter
Dear [parent name],
This year, Horry County Schools is taking an important step in preparation for Flu Season. In the unlikely event of a pandemic of severe influenza, evidence has shown that closing schools can reduce the number of cases of influenza in school-aged children. In order to protect your child, your family, and you, Horry County Schools has developed a policy for the best way to respond to a possible influenza pandemic. This policy has been developed with Horry County DHEC.
In order to fully prepare for the event of a school closure, your child’s school will be holding an informational session for parents and guardians on Wednesday, October 5th, 2016 at 7:00pm. We strongly encourage your attendance at the informational session in order for all parents to be informed about school closures and how their families will be affected.
Please contact _______________ at _____@hcs.net with any questions or concerns. You may also contact _____________ if you cannot attend the informational session and we will email you an online version of the presentation.
Thank you for your participation,
[principal name]
Principal, Your child’s school
ATTRITION
Parent Influenza Knowledge Tests:
| Quizzes Handed Out | Quizzes Returned | |
| Pre-test | ||
| Post-test |
Names of missing post-tests:
Family Childcare Plan Email:
Dear [parent],
Thank you for attending our influenza school closure informational session on October 5, 2016. We appreciate your participation in this program and hope you found the session to be helpful. As you remember, a significant portion of the program was dedicated to guidance on planning for alternative child care in the case of an influenza pandemic. In order to ensure and encourage families to make these plans, we would like for you to fill out this survey [survey link] to tell us how you planned!
Again, thank you for your participation. Having informed and prepared parents is in the best interest of our students, faculty, staff, and community!
Sincerely,
[principal name]
Childcare Plan Survey:
Your name: __________________
Your student’s name: __________________
Did you attend the influenza school closure informational session? Y/N
Since attending the session, has your family developed an alternative child care plan in the event of a school closure? Y/N
School Nurse Education:
# of School Nurses in District: _________
# of School Nurses attending educational sessions: _________
% attendance: _____%
School Closure Verification:
I verify that [school name] will be closed from [closure dates], due to increased incidence of influenza cases among students. There will be no students, teachers, staff, or administrators present in the school during the dates mentioned above.
____________________________ [signature of principle] __________________________ [date]
____________________________ [signature of school nurse] __________________________ [date]
____________________________ [signature of district personnel] _________________________ [date]
FIDELITY
Info Session Tracking (district level):
| School Name | Info Session Scheduled? | Date Info Session Completed | Contact Person |
| St. James High | Yes | 10/5/16 | Jim Stewart |
| Socastee High | No | Sue Platford | |
| Seaside Elementary | Yes | Tammy Smith |
Info Session Tracking (school level):
Name of Person Leading Session: ______________________________
Date session completed: ___________________________
Time session started: ______________
Time session ended: _______________
Number in attendance: _____________
Info Session Checklist:
____ Flu background info
____ School closures: what works and why it works
____ Alternative child care
____ Administer knowledge tests
Parent Info Session Satisfaction Survey:
Thank you for attending tonight’s informational session about influenza and school closures. Please fill out this form and return it to school with your child in the next week. We will be sending reminder emails in one week.
Please indicate your agreement with the following statements below:
“The informational session was effective at informing be about influenza.”
| Strongly Disagree | Disagree | Neutral | Agree | Strongly Agree |
“School closures are effective at reducing influenza cases.”
| Strongly Disagree | Disagree | Neutral | Agree | Strongly Agree |
“I feel that the school district had my best interests in mind while developing the school closure policy.”
| Strongly Disagree | Disagree | Neutral | Agree | Strongly Agree |
“The informational session was well-organized.”
| Strongly Disagree | Disagree | Neutral | Agree | Strongly Agree |
“I feel supported by the school and school district having a school closure policy.”
| Strongly Disagree | Disagree | Neutral | Agree | Strongly Agree |
Thank you again for your participation, please contact ___________________ at ___@hcs.net with any questions or concerns.
References:
Davis, B. M., Markel, H., Navarro, A., Wells, E., Monto, A. S., & Aiello, A. E. (2015). The Effect of Reactive School Closure on Community Influenza-Like Illness Counts in the State of Michigan During the 2009 H1N1 Pandemic. Clinical Infectious Diseases, 60(12), e90-e97. doi:10.1093/cid/civ182
Egger, J. R., Konty, K. J., Wilson, E., Karpati, A., Matte, T., Weiss, D., & Barbot, O. (2012). The Effect of School Dismissal on Rates of Influenza-Like Illness in New York City Schools During the Spring 2009 Novel H1N1 Outbreak. Journal of School Health, 82(3), 123-130 128p. doi:10.1111/j.1746-1561.2011.00675.x
Kawano, S., & Kakehashi, M. (2015). Substantial Impact of School Closure on the Transmission Dynamics during the Pandemic Flu H1N1-2009 in Oita, Japan. PLoS ONE, 10(12), 1-15. doi:10.1371/journal.pone.0144839
US Centers for Disease Control and Prevention. (2007). Interim Pre-pandemic Planning Guidance: Community Strategy for Pandemic Influenza Mitigation in the United States – Early, Targeted, Layered Use of Nonpharmaceutical Interventions. Retrieved from http://www.flu.gov/planning-preparedness/community/community_mitigation.pdf