Respiratory tract viruses, including influenza, cause annual winter epidemics of respiratory disease. Annual attack rates for influenza in children are estimated to reach 20% in average years and may exceed 30–50% during pandemic years. 1–4 Viral respiratory infections are at least twice as common among preschool children cared for in child care centers than among those cared for in their own homes, 5 and children play a significant role as accelerators for the spread of respiratory illness among their household contacts. 2–4,6 In addition, febrile illness episodes and symptoms of upper respiratory tract infection in children lead to an increase in health care visits 7,8 and medication usage. 9–11
In view of the recent worldwide outbreak of a life-threatening febrile respiratory illness that has been named severe acute respiratory syndrome 12 and the likelihood of an influenza pandemic that increases proportionally with time, 13 a pandemic plan is essential to minimize infection-related illness and death. Even though vaccinating the population would be the most efficacious measure, it may not be possible within the time frame required for vaccine generation and effectiveness. Both restriction of public gatherings and school closure have been recommended to reduce the spread of viral respiratory infection. 14 Indeed the most frequently discussed and debated public health measure during the “Spanish flu” pandemic of 1918 was the closure of schools. 15,16 No quantitative evidence supporting this suggestion has, however, been presented.
In Israel, from January 16 to January 28, 2000, during an influenza outbreak that started in the last week of December 1999, there was a nationwide closure of elementary school classes because of an isolated organized labor dispute by teachers. As a consequence, children were cared for mainly at home. This presented an opportunity for a large scale natural experiment to evaluate the effect of school closure on acute respiratory tract morbidity and health care utilization among children ages 6–12 years.
The study was conducted in Maccabi Healthcare Services, which provides health care services for 1.6 million members representing every socioeconomic group. In Israel, under the National Health Insurance Law, every resident receives health care insurance and is entitled to register with 1 of 4 major health maintenance organizations for the provision of medical services. 17 Most tertiary health services are provided by the public hospital system throughout the country. Maccabi Healthcare Services has a nationwide network of >3000 independent physicians connected by a unified computer system. Among these physicians, 1435 are primary care physicians of whom 35% are pediatricians. Diagnoses are downloaded daily to a central computer, and the database is automatically updated with all hospitalizations, visits to outpatient clinics and emergency departments and prescribed medication purchases. Patients have a copayment of 15% of prescription costs.
We assembled a retrospective cohort of all 6- to 12-year-old children comprising 186,094 children. The computerized data were examined for three 2-week periods: before school closure; during closure; and after closure. The average incubation period of respiratory viruses, including influenza, is ~72 h. To account for this delay in onset of clinical symptoms, the dating for the beginning of each 2-week period was moved forward by 3 days: the before school closure period, January 4–17, 2000; the closure period, January 18–31, 2000; and the after closure period, February 1–14, 2000. We excluded Saturdays, which are nonworking days in Israel. There were no public holidays during this period.
The occurrence of respiratory tract infections was determined according to recorded diagnoses, including cough, upper respiratory tract infection, common cold, sore throat and viral infection. Utilization of health care services included visits to physicians, outpatient clinics and emergency departments, hospitalizations and purchase of prescribed medications. Medication purchases included antibiotics, antipyretics and “cold and cough” medications. Outcome rates were compared using a one-way analysis of variance, followed by pairwise Tukey post hoc test comparisons 18 between each of the 14-day periods with 1 df. Three comparisons were done, between the period during school closure versus before and after closure and between after closure period versus before. A 2-tailed P of <0.05 was considered statistically significant. Because we examined the same cohort, we did not standardize the data for age, gender or any other variable. Statistical analyses were conducted with SPSS 10.0 for Windows (SSPS, Inc., Chicago, IL). Relative risks and confidence intervals were calculated with a WHATIS 4.12 statistical calculator. 19
The daily number of physician visits and diagnoses of all respiratory tract infections for the 3 periods are shown in Figure 1. Total diagnoses decreased from 22,088 during the period before closure to 12,860 during the closure period [relative risk, 0.58; 95% confidence interval (CI), 0.57–0.59] and increased to 16,205 (relative risk, 0.79; 95% CI 0.78–0.81) after the closure ended. The combined relative risk for infectious respiratory tract disease during closure days as opposed to school days before and after the closure was 0.76 (95% CI 0.75–0.77). There was a significant (P < 0.001) drop in the rate of diagnoses of respiratory tract infections during the school closure period (Table 1). Diagnoses of upper respiratory tract infections, for example, declined from 44.0/1000 children during the 2-week period before school closure to 24.7/1000 children during the period during the closure, subsequently increasing to 30.8/1000 children during the 2-week period after the school closure.
Significant (P < 0.001) reductions occurred in the rate of visits to physicians, visits to emergency departments and medication purchases during the closure period compared with the periods before and after (Table 1). There was a 28% (95% CI 26–30%) reduction in physician visits during the closure period from 258.7 to 201.8 visits/1000 children with a subsequent increase of 13% (95% CI 11–15%) after the closure. Visits to emergency departments decreased by 28% (95% CI 11–46%) during the closure period, subsequently returning to preclosure levels. Similarly, a decrease of 35% (95% CI 32–38%) and 12% (95% CI 9–15%) occurred in medication purchases during and after the closure, respectively.
Our study demonstrates that school closure was temporally associated with decreased morbidity from respiratory tract infections, a consequent decrease in visits to physicians and emergency departments and a reduction in purchase of medications. When children become acutely ill, fever and cough are the parents' primary concerns and 32–75% of pediatric emergency departments visits are a result of these worries. 7,8 Although many parents were obliged to stay at home with their children during the school closure and thus were more available to consult with their physician, a marked decrease occurred in visits to physicians. Respiratory tract infections in school children usually do not require hospitalization; this is reflected by the fact that hospitalization rates remained unchanged. Similarly the rate of hospital outpatient visits, which are usually for complex or chronic problems, was unchanged during the closure.
Although children are seldom severely ill with influenza, there is a facilitated transmission of the virus from children to adults. Children shed greater quantities of influenza viruses than adults for a longer time. 4,6 The number of household members who became ill within 3 days of a child's absence from school was reportedly 2.2 times higher than expected during an influenza season. For every 10 children who missed school for influenza-related illness, 8 household members subsequently become ill. 11
In Israel, 33.8% of the population are children; hence our results may not be applicable to Western countries with lower percentage of children. In addition, there may be a difference in parental attitudes toward respiratory illness symptoms in other cultures that affect health care utilization. Another reason for such a difference may be the basic structure of the health system in Israel, where comprehensive health insurance is universal and provided by the law. The number of visits to physicians is one of the highest in the world, 11.2 visits per capita yearly, and there are 25 visits per 100 citizens per year to pediatric emergency departments.20 Finally there is limited availability of over-the-counter medications, and to obtain symptomatic therapeutic agents children are generally seen by a physician.
Rapid tests or cultures for viruses were not performed routinely in the community. National surveillance of influenza activity disclosed local outbreaks of influenza A (H3N2) during the last week of December 1999 with progression to the regional level by the first 2 weeks of January 2000. Most of the disease occurred among 5- to 14-year-old children (data obtained from the National Reference Center for Influenza at the Sheba Medical Center, Tel Hashomer, Israel). After the introduction of the “Asian” influenza A (H2N2) virus in 1957 and the “Hong Kong” virus (H3N2) in 1968, epidemics occurred soon after school opened. 21 Clearly the crowding that occurred in schools favored the rapid spread of the disease via aerosol transmission of the virus.
In conclusion, we noted a temporal association between school closure and decreased respiratory morbidity during an influenza outbreak. We suggest that school closure may be adopted as an effective strategy for dealing with an influenza pandemic.
1. Neuzil KM, Zhu Y, Griffin MR, et al. Burden of interpandemic influenza
younger than 5 years: a 25-year prospective study. J Infect Dis
2. Longini IM, Koopman JS, Monto AS, Fox JP. Estimating household and community transmission parameters for influenza
. Am J Epidemiol
3. Glezen WP, Couch RB. Interpandemic influenza
in the Houston area, 1974–76. N Engl J Med
4. Fox JP, Hall CE, Cooney MK, Foy HM. Influenza
virus infections in Seattle families, 1975–1979. I. Study design, methods and the occurrence of infections by time and age. Am J Epidemiol
5. Robinson J. Infectious diseases in schools and child care facilities. Pediatr Rev
6. Foy HM, Cooney MK, Hall C, Malmgren J, Fox JP. Case-to-case intervals of rhinovirus and influenza
virus infections in households. J Infect Dis
7. Wyke S, Hewison J, Russell IT. Respiratory illness in children
: what makes parents decide to consult?Br J Gen Pract
8. Kai J. What worries parents when their preschool children
are acutely ill, and why: a qualitative study. BMJ
9. Gonzales R, Malone D, Maselli J, Sande MA. Excessive antibiotic use for respiratory infections in the United States. Clin Infect Dis
10. Neuzil KM, Mellen BG, Wright PF, Mitchel EF, Griffin MR. The impact of influenza
on hospitalizations, outpatient visits and antibiotic prescriptions in children
. N Engl J Med
11. Neuzil KM, Hohlbein C, Zhu Y. Illness among schoolchildren during influenza
season: effect on school absenteeism, parental absenteeism from work, and secondary illness in families. Arch Pediatr Adolesc Med
12. Drazen JM. SARS: looking back over the first 100 days. N Engl J Med
13. Patriarca PA, Cox NJ. Influenza
pandemic preparedness plan for the United States. J Infect Dis
. 1997;176(suppl 1):S4–S7.
14. CDC National Vaccine Program Office. Pandemic influenza
: a planning guide for state and local officials. January 24, 2003. Available at: http://www.cdc.gov/od/nvpo/pandemicflu.htm
. Accessed August 12, 2003.
15. The influenza
pandemic of 1918. Available at: http://www.stanford.edu/group/virus/uda
. Accessed November 30, 2003.
16. Crosby AW. America's Forgotten Pandemic: The Influenza of 1918
. Cambridge, U.K.: Cambridge University Press; 1989.
17. Gross R, Rosen B, Shirom A. Reforming the Israeli health system: findings of a 3-year evaluation. Health Policy
18. Zar JH. Biostatistical Analysis
. Englewood Cliffs, NJ: Prentice-Hall, Inc.; 1984:185–190.
19. Abramson JH, Gahlinger PM. Computer Programs for Epidemiologic Analyses: PEPI v. 4. 0
. Salt Lake City, Utah: Sagebrush Press; 2001.
20. 20. Alhayany E, Zomra I. Utilization of emergency services in primary care clinics. Harefuah
21. Murphy BR, Webster RG. Orthomyxoviruses. In: Fields BN, Knipe DM, eds. Field's Virology.
2nd ed. New York: Raven Press; 1990:1121.