Household responses to school closure resulting from outbreak of influenza B, North Carolina.

Abstract

School closure is a proposed strategy for reducing influenza transmission during a pandemic. Few studies have assessed how families respond to closures, or whether other interactions during closure could reduce this strategy's effect. Questionnaires were administered to 220 households (438 adults and 355 children) with school-age children in a North Carolina county during an influenza B virus outbreak that resulted in school closure. Closure was considered appropriate by 201 (91%) households. No adults missed work to solely provide childcare, and only 22 (10%) households required special childcare arrangements; 2 households incurred additional costs. Eighty-nine percent of children visited at least 1 public location during the closure despite county recommendations to avoid large gatherings. Although behavior and attitudes might differ during a pandemic, these results suggest short-term closure did not cause substantial hardship for parents. Pandemic planning guidance should address the potential for transmission in public areas during school closure.

Full text

Page 1

School closure is a proposed strategy for reducing in-
fl uenza transmission during a pandemic. Few studies have
assessed how families respond to closures, or whether oth-
er interactions during closure could reduce this strategy’s
effect. Questionnaires were administered to 220 house-
holds (438 adults and 355 children) with school-age chil-
dren in a North Carolina county during an infl uenza B virus
outbreak that resulted in school closure. Closure was con-
sidered appropriate by 201 (91%) households. No adults
missed work to solely provide childcare, and only 22 (10%)
households required special childcare arrangements; 2
households incurred additional costs. Eighty-nine percent
of children visited at least 1 public location during the clo-
sure despite county recommendations to avoid large gath-
erings. Although behavior and attitudes might differ during a
pandemic, these results suggest short-term closure did not
cause substantial hardship for parents. Pandemic planning
guidance should address the potential for transmission in
public areas during school closure.
H
logic events (1). Preparation for the next infl uenza pandem-
ic has become a major focus of public health activities. Use
of vaccines containing antigens matched for a pandemic
infl uenza A strain is the best control measure for reducing
illness and death during a pandemic (1,2). However, spe-
cifi c vaccines will take >4–6 months to be produced once a
uman infl uenza pandemics have occurred 3 times in
the past century and are assumed to be recurring bio-
pandemic strain is identifi ed. Additionally, suffi cient doses
of antiviral drugs might not be available to supply treat-
ment and chemoprophylaxis needs (3). As a result, a vari-
ety of nonpharmaceutical interventions (NPIs) have been
proposed by US federal agencies (4) and the World Health
Organization (5) to help mitigate the impact of a pandemic
while vaccines against the pandemic strain are being pro-
duced. NPIs that have been identifi ed as potential mitiga-
tion strategies include voluntary isolation of case-patients,
voluntary quarantine of contacts of cases, and social dis-
tancing of children and adults.
School-age children have the highest attack rates dur-
ing typical seasonal infl uenza outbreaks and play a central
role in sustaining infl uenza transmission (6). Children are
being disproportionately affected by the avian infl uenza A
virus (H5N1) that is currently circulating in many coun-
tries (7). School attendance during the 1957 epidemic was
thought to amplify the transmission of virus in Japan (8).
Proposed mitigation strategies have thus focused on this age
group as a means of reducing transmission. School closure
lasting 4–12 weeks has been recommended as an option
to distance children and decrease transmission (4). Several
modeling studies have suggested that school closure might
reduce peak attack rates and overall clinical attack rates,
especially if combined with other strategies, including vol-
untary isolation and quarantine of sick persons and their
contacts (9) or household-based antiviral prophylaxis and
quarantine (10). However, few data are available to address
whether school closure can actually reduce the transmis-
sion of infl uenza viruses among susceptible children or
their family contacts (11).
Prolonged school closures might have adverse social
and economic effects (12,13). For example, some parents
Household Responses to S chool
Closure Resulting from Outbreak
of Infl uenza B, North Carolina
April J. Johnson,* Zack S. Moore,*† Paul J. Edelson,* Lynda Kinnane,‡ Megan Davies,*†
David K. Shay*, Amanda Balish,* Meg McCarron,* Lenee Blanton,* Lyn Finelli,* Francisco Averhoff,*
Joseph Bresee,* Jeffrey Engel,† and Anthony Fiore*
RESEARCH
1024 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 14, No. 7, July 2008
*Centers for Disease Control and Prevention, Atlanta, Georgia,
USA; †North Carolina Department of Health and Human Servic-
es, Raleigh, North Carolina, USA; and ‡Toe River Health District,
Burnsville, North Carolina, USA
DOI: 10.3201/eid1407.080096

Page 2

Household Responses to School Closure
will likely stay home to care for children, resulting in lost
family income, as well as adverse effects on businesses.
Children from families of lower socioeconomic status
may rely on their schools for the National School Lunch
Program, a federally assisted meal program that provides
meals and snacks to children who qualify (13). To date,
no study has evaluated parental attitudes or responses to
school closures during a seasonal infl uenza outbreak.
In late October, 2006, a rural county in North Carolina
experienced an infl uenza B virus outbreak that resulted in
a sudden increase in student and school staff absenteeism.
School offi cials closed all 9 schools in the county on No-
vember 2, and schools remained closed through November
12. These events provided an opportunity to evaluate the
response of families with schoolchildren to closing schools
and to observe the frequency of children’s excursions to
public places during the school closure.
Methods
Detection of and Response to the Outbreak
Yancey County, North Carolina, had an estimated
population in 2006 of 18,421, of whom ≈21% were <18
years of age, and an estimated 7,472 households in 2000.
This county is located in the western part of the state in
the Appalachian Highlands on the Tennessee border. From
October 26 through November 1, 2006, school offi cials
observed a marked increase in the number of student and
employee absences in the 9-school system. Many absences
were attributed to infl uenza-like illness (ILI) among chil-
dren and staff. Two elementary schools were particularly
affected, with absentee rates among students increasing
from 4% and 8.8% on October 26 to 34% and 32%, re-
spectively, on November 1. Using commercial rapid anti-
gen detection techniques, a local clinic identifi ed infl uenza
in 29 patients on October 31. The North Carolina Public
Health Laboratory subsequently confi rmed the presence of
infl uenza B virus in samples that were submitted for viral
culture from 7 of 8 children. On November 1, with 429
children (17% of schoolchildren enrolled in all 9 schools;
Figure 1, panel A) and 38 (10%) of the staff absent, school
offi cials closed all the county’s public schools because of
unmet staffi ng needs.
In an increased effort to vaccinate residents in response
to the outbreak, infl uenza vaccine clinics were established
at the county health department. A reverse 911 call was
issued to county residents by the health department and
county government on November 1. The reverse 911 sys-
tem is a notifi cation system that enables town offi cials to
deliver telephone messages during an emergency to specifi c
groups of persons on the basis of location. County residents
were given the following message by telephone: “This is
a message from Yancey County Health Department and
Yancey County Government. Due to increasing cases of
infl uenza, residents of Yancey County are being asked to
avoid large gatherings. Also, please wash hands frequently,
cover coughs, and avoid contact with sick individuals. For
more recorded information call ...”
Household Survey
A total of 1,750 households had children enrolled in the
public school system. To evaluate the response of Yancey
County residents with children in the public school system
to the infl uenza outbreak and school closings, a random
sample of these households was contacted by telephone on
November 16–18. A standard questionnaire was used. A
parent or legal guardian from each household was asked to
provide information about his or her child’s (or children’s)
activities during the school closure (November 2–12, 2006),
special arrangements that had to be made for child care, and
attitudes toward the closure. This household respondent
was also asked to provide demographic information on the
household and answer questions about how school closure
affected his or her own employment and daily routines, and
those of any other adults in the household. Respondents
were asked their perceptions of the likelihood that a child
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 14, No. 7, July 2008 1025
Figure 1. A) Percentage of schoolchildren absent from public
schools, by date, and B) total number of children surveyed with
infl uenza-like illness, by date of illness onset, Yancey County, North
Carolina, October 23–November 17, 2006.
0
9
2
4
6
8
10
12
14
16
18
Schools closed
% Children absent
No. children
A
B
0
1
2
3
4
5
6
7
8
23 24 25 26 27 28 29 30 31 12345678 9 10 11 12 13 14 15 16 17
Oct
Nov

Page 3

RESEARCH
might be infected with infl uenza virus and the likelihood
that an infected child would require hospitalization. Parents
with children who had been ill at some time since October
15, 2006, were asked about clinical manifestations of each
child’s illness. ILI was defi ned as having a fever (objective
or subjective) and either a cough or sore throat in the ab-
sence of a known cause other than infl uenza.
Four hundred telephone numbers were randomly se-
lected from a list of all households with children enrolled in
public schools, with the intent of obtaining >200 completed
surveys, which would represent >10% of the households
and >10% of children enrolled in public schools. Fami-
lies were called at various times of the day and evening
both during the week and on the weekend from November
16–18. Information was collected on each school-age child
living in the household. To control for the effect of family,
1 child was randomly selected from each household. Sta-
tistical analysis was performed by using SAS version 9.1
(SAS Institute, Cary, NC, USA).
Results
Interviewers called households over a 3-day period
until a minimum of 200 surveys were completed. Two
hundred twenty (97%) of those contacted completed the
survey; 8 (3%) households refused. The 220 households
surveyed included 438 adults (≈3% of the adult popula-
tion in the county) and 355 school-age children (≈14% of
all children enrolled in the public schools, or 9.4% of all
children <18 years of age in the county). The percentage
of children surveyed from any 1 school ranged from 9% to
21%. The percentage of children in each grade who were
surveyed ranged from 8% of third-graders to 20% of ninth-
graders. Characteristics of the household and children in
the survey are shown in Table 1. Thirty-seven (17%) of the
220 households had only 1 adult (lower than the national
rate of single-parent homes, which is 27%). Children from
87 (41%) of 212 responding households were reported as
receiving free or reduced-cost lunches through the National
School Lunch Program. This fi nding is slightly lower than
the percentage of children reported as enrolled in the coun-
ty (51%) or the state (48%) in 2005 (14), but approximately
the same as the national (37.8%) percentage of children eli-
gible for the program (15).
One hundred thirty (37%) of 355 surveyed schoolchil-
dren were ill on >1 days from October 23 through Novem-
ber 15. Among children who attended elementary schools,
50% were ill during this period, compared with 26% and
28% of surveyed middle and high school-age children, re-
spectively (Table 2). Among children who were reported
ill, 66 (51%) met the case defi nition for ILI. The reported
dates for ILI symptom onset started October 20 and peaked
on November 1, 2006, the day before schools closed (Fig-
ure 1, panel B). A total of 78 (22%) children reportedly
received infl uenza vaccine for that season as of Novem-
ber 15. Sixty-three (81%) of these children received it after
schools closed (November 2 or later).
Ninety-nine (45%) household respondents thought it
was very possible that their child could get infl uenza from
another person. However, only 15 respondents (7%) felt it
was very possible that their child would need hospitaliza-
tion if they became infected with infl uenza.
After we controlled for the effect of family, visiting
public locations during the school closure (November
2–12) was commonly reported, with 195 (89%) of 220
children visiting at least 1 public place (Table 3). Overall,
47% of children traveled outside Yancey County at least
once during the school closure. Sites children visited dur-
ing school closure differed by age group and illness status.
For example, older children were signifi cantly more likely
to go to fast food restaurants and parties (p<0.05; Figure 2).
However, they were less likely to go grocery shopping than
younger children. No differences were seen between chil-
dren who were reported as ill at any time from October 23
through November 15 and children who were not ill during
that time (Figure 3).
Among adults in surveyed households, 315 (72%)
were employed outside the home (Table 4). In 118 (54%)
of the 220 households surveyed, all adults in the household
were employed outside the home. Of 218 adults living in
those 118 homes, only 39 (18%) had occupations that per-
mitted them to work from home. Seventy-six (24%) of the
315 employed adults missed >1 day of work from Octo-
ber 23 through the date of the survey, including 36 (47%)
because of their own illness, 18 (24%) to take care of ill
1026 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 14, No. 7, July 2008
Table 1. Characteristics of children and households surveyed,
Yancey County, North Carolina, 2006*
Characteristic
Households (N = 220)
Single-adult home
Two-adult home
Three- or four-adult home
Children in home receive free/reduced-cost
lunch† (n = 212)
All adults employed outside the home
Children (N = 355)
Male
White, non-Hispanic
Median age, y (range)
*Values are no. (%) unless otherwise indicated.
†Free and reduced-cost lunches provided through the National School
Lunch Program.
Value
37 (17)
145 (66)
38 (17)
87 (41)
118 (54)
177 (50)
344 (97)
12 (5–19)
Table 2. Prevalence of child illness reported by parents by school
type, Yancey County, North Carolina, 2006
No. (%) surveyed of all
public schoolchildren
Elementary 136 (12)
Middle86 (14)
High128 (16)
School type
No. (%) ill of those
surveyed
68 (50)
22 (26)
36 (28)

Page 4

Household Responses to School Closure
family members, and 14 (18%) because of school closure.
However, all adults who reported missing work because of
school closure were school employees. Among all adults
who missed work during the period, the median number of
days missed was 3, and days missed ranged from 1 to 14.
One hundred sixty-seven (76%) households indicated
that someone was regularly available during the day to pro-
vide childcare (Table 4). Twenty-two (10%) reported that
they had to make special childcare arrangements because
of school closure, including enlisting working adult house-
hold members, grandparents or other relatives, friends, or
nonrelated adults to provide childcare; taking the child to
work; having older siblings watch younger children; or us-
ing childcare programs. Among responding households,
only 7 (3%) had to have their child spend >1 nights outside
their household for childcare purposes, and only 2 (1%) re-
ported having to spend extra money ($100 and $150) on
childcare arrangements.
A total of 201 (91%) responding households thought
the decision to close schools was appropriate (Table 5).
Eighty-two (41%) of households that provided a reason for
that opinion thought the decision was appropriate to protect
the health of the community, 71 (35%) felt there were too
many sick children for schools to remain open, 23 (11%)
thought it would help protect their child and family, and 8
(4%) thought that schools would be too understaffed if they
remained open. Of the 10 (5%) respondents who believed
school closure was inappropriate, 4 (40%) thought it could
result in lost income, 3 (30%) did not think infl uenza was
in the area, 2 (20%) did not think closure was an effective
measure, and 1 (10%) found it too diffi cult to make child-
care arrangements. Overall, 198 (90%) thought they had
enough time to prepare for school closures. A total of 180
(84%) felt well prepared and could not think of anything
that could have helped them prepare better to deal with
closure. Twenty (9%) would have appreciated more time
between notifi cation and closing. Several respondents also
mentioned that they would have liked to have been warned
that infl uenza was in the area, that children at school were
sick, and that schools might be closed as a response to ill-
nesses.
Discussion
The primary objective of this investigation was to eval-
uate the response of households to school closure caused by
an infl uenza outbreak. This study found that most adults
believed that school closure was appropriate and necessary
to slow infl uenza transmission and protect the health of the
community. Second, almost all children visited public areas
within the community while schools were closed, despite
public health recommendations to avoid large gatherings.
Lastly, the effect of school closure on work absenteeism
and childcare expenditures appeared to be minimal in this
community.
Yancey County is located in the Blue Ridge Moun-
tains, and results obtained there are likely not generalizable
to all counties. First, residents in this rural, mountainous
county are accustomed to dealing with frequent school clo-
sures resulting from adverse weather conditions, particu-
larly winter snowstorms. Families in communities where
school closures are infrequent or where extended families
are less available to provide childcare might respond differ-
ently. Additionally, only 17% of households in this survey
were single-adult homes, compared with the national aver-
age of 27%. Multiple-adult households might fi nd arrang-
ing childcare for schoolchildren during school closures to
be less challenging than single-adult homes. Also, the me-
dian age of children in this survey was 12 years. Childcare
arrangements for older children are likely easier to make
than for younger children. Lastly, no adults in this survey
reported missing work solely because of school closure,
other than those employed by the school. Only 18 (8%)
adults from the 220 households in the survey reported miss-
ing work to stay with a sick family member. This fi nding is
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 14, No. 7, July 2008 1027
Table 3. Locations visited by schoolchildren when schools were
closed, controlled for effect of family, Yancey County, North
Carolina, 2006
Location visited (N = 220)
At least 1 public location
Grocery stores
Fast food restaurants
Church services
Mall
Parties or sleepovers
Outside Yancey County
No. (%) children
195 (89)
97 (44)
77 (35)
75 (34)
42 (19)
33 (15)
103 (47)
85
39
52
24
30
16
5
90
51
40
41
35
2121
0
10
20
30
40
50
60
70
80
90
100
Any location Outside
county
Grocery
store*
Fast food* Church MallParties*
% Children
Age <12 years
Age ?12 years
Figure 2. Locations visited by schoolchildren during school closure
by age group, controlled for effect of family, Yancey County, North
Carolina, 2006. Values above bars are percentages. *p<0.05.

Page 5

RESEARCH
dissimilar to fi ndings from another study that found that at
least 1 adult in 53% of families missed work to care for an
ill child because of a winter respiratory illness (16). Other
studies have also found that epidemics of respiratory illness
can cause a substantial number of lost workdays for parents
of ill children (17,18).
Results might also have been different if schools were
closed for a longer period or if a more clinically severe
strain of infl uenza were present, causing more hospital-
izations or death. Most parents interviewed in the present
study did not think that infection would result in hospital-
ization; only 5 brief hospitalizations were reported. We did
not collect quantitative information on the frequency or du-
ration of visits to public places by schoolchildren and did
not determine whether these persons visited public places
while potentially infectious. Lastly, households with chil-
dren who attended private school or were home-schooled
were not surveyed.
The decision of the local school board to close all 9
schools in Yancey County was primarily motivated by con-
cerns about staffi ng the schools in the face of high levels of
absenteeism. Although the reduction in ILI that occurred
after schools were closed is an intriguing fi nding, results
from this investigation cannot be used to assess the effect
of school closure on the effect of illness in a community
experiencing an infl uenza outbreak. Infl uenza outbreak dy-
namics are relatively poorly understood, and the proportion
of children who were susceptible to infection might have
decreased below the number required to sustain transmis-
sion at approximately the same time schools were closed.
The fact that transmission decreased despite many school-
children in public gathering places also calls into question
the role of reduced contacts among children in ending the
outbreak.
Studies that have modeled the effects of NPIs on re-
ducing infl uenza epidemic size during a pandemic have
suggested that closing schools can be effective if imple-
mented early and if the reproductive number (R0) is low
(<1.8) (19). For example, in a network-based simulation in
which ≈50% of persons were infected, similar to the Asian
infl uenza pandemic of 1957–58, closing schools and keep-
ing those children at home reduced the calculated attack
rate by 90% (20). Studies of the effects of school closure
on respiratory disease rates in schoolchildren have shown
mixed results. An investigation in Israel showed that al-
1028 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 14, No. 7, July 2008
Table 4. Employment and childcare status of adults and households during school closure, Yancey County, North Carolina, 2006*
Characteristic
Adults employed outside home (N = 438)
Occupations of those employed outside home (N = 315)
Healthcare
Education
Industry
No. homes where all adults in home employed (N = 220)
Adults who can work from home (in homes where all adults in the house work outside the home, N = 218)
Missed worked from Oct 23 through Nov 15 (N = 315)
Median no. days missed (range)
Reason for missed work (N = 76)
Own illness
Had to take care of ill family members
School closure
Childcare (N = 220)
Someone home during the day who could provide childcare
Had to make special arrangements
Child had to spend >1 nights away from home for childcare purposes
Had to spend extra money for childcare
*Values are no. (%) unless otherwise indicated.
Value
315 (72)
35 (11)
36 (11)
27 (9)
118 (54)
39 (18)
76 (24)
3 (1–14)
36 (47)
18 (24)
14 (18)
167 (76)
22 (10)
7 (3)
2 (1)
90
45
42
34
31
18
13
87
48
49
36
37
21
18
0
10
20
30
40
50
60
70
80
90
100
Any locationOutside
county
Grocery
store
Fast food ChurchMall Parties
% Children
Healthy children
Ill children
Figure 3. Percentage of ill and healthy schoolchildren visiting
various locations during school closure, controlled for effect of
family, Yancey County, North Carolina, 2006. Values above bars are
percentages. No signifi cant differences were observed (p<0.05).

Page 6

Household Responses to School Closure
though schools were closed because of a teacher strike,
the incidence of respiratory illness diagnosed in children
who came to health clinics decreased, as did physician and
emergency department visits and purchase of medications
(11). Rates subsequently increased when schools reopened.
However, some (researchers and public health offi cials)
have proposed that school closure in urban areas might
have an opposite effect because children released from
school can more easily congregate. This effect may have
occurred in children from Chicago during the 1918–19
pandemic when rates of infl uenza among schoolchildren
increased during time off from school (21). In Connecticut,
3 cities (Bridgeport, Hartford, and New Haven) kept their
schools open during the 1918–19 pandemic (21) and ex-
perienced lower mortality rates than 2 smaller cities (New
London and Waterbury) that closed schools.
We cannot assess the effect of infl uenza vaccination
on the course of the infl uenza outbreak in Yancey County.
Large-scale vaccination programs began in the county dur-
ing late October. Although not known at the time of these
programs, the infl uenza B strain contained in the 2006–
2007 vaccine (Victoria lineage) was not a good match to
the circulating Yamagata lineage infl uenza B viruses in this
outbreak (22). However, only a few infl uenza virus isolates
were antigenically characterized at the Centers for Disease
Control and Prevention (CDC) in Atlanta, and we do not
know that the isolates tested were representative of those
circulating in the area at the time. Infl uenza A viruses were
found through CDC’s sentinel surveillance systems to be
circulating in other counties concurrently. It is possible that
other viruses were co-circulating in Yancey County and
were not detected.
The effectiveness of closing schools to reduce trans-
mission of common infectious diseases such as infl uenza is
not well studied (12). In addition, data on school closures in
response to infectious disease outbreaks in general are not
regularly collected in the United States. Additional studies
to assess these actions will be of interest to local public
health offi cials and school administrators who make deci-
sions about keeping schools open during explosive but self-
limited outbreaks with high attack rates among schoolchil-
dren, such as those commonly caused by seasonal infl uenza
or norovirus. Results from such studies will also be help-
ful in planning and implementing community mitigation
strategies for disease outbreaks whose community impact
might be severe, such as pandemic infl uenza.
This investigation provides insight into how house-
holds with school-age children in a small rural community
responded to a brief school closure precipitated by a sea-
sonal infl uenza outbreak. Overall, respondents to the sur-
vey indicated that the community was in favor of closing
schools as a way to deal with high levels of student and
staff absenteeism and potentially to reduce transmission.
Parents reported few problems in coping with the school
closure and did not miss work to provide childcare. How-
ever, many students visited public areas during school
closure. Plans for pandemic infl uenza responses should ad-
dress the potential for transmission in public areas during
school closure.
Acknowledgments
We thank Mary Alice Beaver and Darlene Tipton for assis-
tance during the investigation; the North Carolina State Labora-
tory of Public Health for laboratory support; and Kirsten Buckley,
Christina Felsen, Amanda McWhorter, John Turnbull, Daiscia
Roseberry, Emmanuel Johnson, Vivek Shinde, Nathan Almond,
Meredith Anderson, David Bergmire-Sweat, Kim Brunette, An-
drew Edmonds, Anjum Hajat, Jen Horney, Morgan Johnson, Tina
Lusk, Molly McKnight, Meagan Pressley, Mindi Russell, Lindsay
Tallon, and John Wallace for assistance with data collection and
entry.
Dr Johnson is an Epidemic Intelligence Service offi cer in the
Infl uenza Division at the Centers for Disease Control and Preven-
tion. Her research interests are the epidemiology of viral respira-
tory infections and zoonotic diseases.
References
1. Stöhr K, Esveld M. Will vaccines be available for the next in-
fl uenza pandemic? Science. 2004;306:2195–6. DOI: 10.1126/
science.1108165
2. Sellwood C, Asgari-Jirhandeh N, Salimee S. Bird fl u: if or when?
Planning for the next pandemic. Postgrad Med J. 2007;83:445–50.
DOI: 10.1136/pgmj.2007.059253
3. Germann TC, Kadau K, Longini IM Jr, Macken CA. Mitigation strat-
egies for pandemic infl uenza in the United States. Proc Natl Acad
Sci U S A. 2006;103:5935–40. DOI: 10.1073/pnas.0601266103
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 14, No. 7, July 2008 1029
Table 5. Household responses to school closure and difficulties
obtaining childcare, Yancey County, North Carolina, 2006
Response
Felt closure was appropriate (N = 220)
Reason it was appropriate (N = 201)
To protect health of the community
Too many sick children
To protect their child and family
Schools would be understaffed
Felt closure was not appropriate (N = 220)
Reason it was not appropriate (N = 10)
Could result in lost income
Did not think influenza was in the area
Did not think it was an effective measure
Too difficult to make childcare arrangements
Overall preparedness (N = 220)
Had enough time to prepare for closure
Could not think of anything that would have
made them more prepared
Could have used more time
No. (%)
201 (91)
82 (41)
71 (35)
23 (11)
8 (4)
10 (5)
4 (40)
3 (30)
2 (20)
1 (10)
198 (90)
180 (84)
20 (9)

Page 7

RESEARCH
4. Centers for Disease Control and Prevention. Community strategy
for pandemic infl uenza mitigation, 2008 [cited 2008 Apr 24]. Avail-
able from http://pandemicfl u.gov/plan/community/commitigation.
html#XV
5. World Health Organization. WHO interim protocol: rapid opera-
tions to contain the initial emergence of pandemic infl uenza, 2007
[cited 2008 Apr 24]. Available from http://www.who.int/csr/disease/
avian_infl uenza/RapidContProtMay07.pdf
6. Monto AS, Davenport FM, Napier JA, Francis T Jr. Effect of vac-
cination of a school-age population upon the course of an A2/Hong
Kong infl uenza epidemic. Bull World Health Organ. 1969;41:
537–41.
7. Chen JM, Chen JW, Dai JJ, Sun YX. A survey of human cases
of H5N1 avian infl uenza reported by the WHO before June 2006
for infection control. Am J Infect Control. 2007;35:467–9. DOI:
10.1016/j.ajic.2006.12.001
8. Reichert TA, Sugaya N, Fedson DS, Glezen P, Simonsen L, Tashi-
ro M. The Japanese experience with vaccinating schoolchildren
against infl uenza. N Engl J Med. 2001;344:889–96. DOI: 10.1056/
NEJM200103223441204
9. Haber MJ, Shay DK, Davis XM, Patel R, Jin X, Weintraub E, et al.
Effectiveness of interventions to reduce contact rates during a simu-
lated infl uenza pandemic. Emerg Infect Dis. 2007;13:581–9.
10. Ferguson NM, Cummings DA, Fraser C, Cajka JC, Cooley PC,
Burke DS. Strategies for mitigating an infl uenza pandemic. Nature.
2006;442:448–52. DOI: 10.1038/nature04795
11. Heymann A, Chodick G, Reichman B, Kokia E, Laufer J. Infl u-
ence of school closure on the incidence of viral respiratory diseases
among children and on health care utilization. Pediatr Infect Dis J.
2004;23:675–7. DOI: 10.1097/01.inf.0000128778.54105.06
12. Inglesby TV, Nuzzo JB, O’Toole T, Henderson DA. Disease mitiga-
tion measures in the control of pandemic infl uenza. Biosecur Bioter-
ror. 2006;4:366–75. DOI: 10.1089/bsp.2006.4.366
13. Pandemic infl uenza: warning, children at-risk. October 2007 report.
Washington: American Academy of Pediatrics and Trust for Ameri-
ca’s Health; 2007. p. 1–31.
14. The Annie E. Casey Foundation. CLIKS: community-level in-
formation on kids, 2006 [cited 2008 Apr 24]. Available from
http://www.kidscount.org/cgi-bin/cliks.cgi?action=profile_
results&subset=NC&areaid=101
15. Garofono A, Sable J. Characteristics of the 100 largest public el-
ementary and secondary school districts in the United States: 2004–
2005 (NCES 2008–335). Washington: National Center for Educa-
tion Statistics, Institute of Education Sciences, US Department of
Education; 2008.
16. Nettleman MD, White T, Lavoie S, Chafi n C. School absenteeism,
parental work loss, and acceptance of childhood infl uenza vaccina-
tion. Am J Med Sci. 2001;321:178–80. DOI: 10.1097/00000441-
200103000-00004
17. Li S, Leader S. Economic burden and absenteeism from infl uenza-like
illness in healthy households with children (5–17 years) in the US.
Respir Med. 2007;101:1244–50. DOI: 10.1016/j.rmed.2006.10.022
18. Neuzil KM, Hohlbein C, Zhu Y. Illness among schoolchildren dur-
ing infl uenza season; effect on school absenteeism, parental absen-
teeism from work and secondary illness in families. Arch Pediatr
Adolesc Med. 2002;156:986–91.
19. Vynnycky E, Edmunds WJ. Analyses of the 1957 (Asian) infl u-
enza pandemic in the United Kingdom and the impact of school
closures. Epidemiol Infect. 2008;136:166–79. DOI: 10.1017/
S0950268807008369
20. Glass RJ, Glass LM, Beyeler WE, Min HJ. Targeted social distanc-
ing design for pandemic infl uenza. Emerg Infect Dis. 2006;12:
1671–81.
21. Bell DM; World Health Organization Writing Group. Non-pharma-
ceutical interventions for pandemic infl uenza, national and commu-
nity measures. Emerg Infect Dis. 2006;12:88–94.
22. Centers for Disease Control and Prevention (CDC). Update: infl u-
enza activity—United States and worldwide, 2006–07 season, and
composition of the 2007–08 infl uenza vaccine. MMWR Morb Mor-
tal Wkly Rep. 2007;56:789–94.
Address for correspondence: April J. Johnson, Centers for Disease Control
and Prevention, 1600 Clifton Rd NE, Mailstop A32, Atlanta, GA 30333,
USA; email: ajjohnson@cdc.gov
1030 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 14, No. 7, July 2008
Use of trade names is for identifi cation only and does not imply
endorsement by the Public Health Service or by the U.S.
Department of Health and Human Services.
The print journal is available at no charge to public health professionals
YES, I would like to receive Emerging Infectious Diseases.
Please print your name and business address in the box and return by fax to
404-639-1954 or mail to

EID Editor
CDC/NCID/MS D61
1600 Clifton Road, NE
Atlanta, GA 30333
Moving? Please give us your new address (in the box) and print the number of
your old mailing label here_______________________________________