Correspondence

Vaccinating Japanese Schoolchildren against Influenza

N Engl J Med 2001; 344:1946-1948June 21, 2001

Article

To the Editor:

Reichert et al. (March 22 issue)1 compare excess mortality in Japan and the United States from 1949 through 1998. They conclude that vaccination of schoolchildren against influenza, the only independent variable assessed, reduced mortality from influenza among older persons in Japan from about 1970 through 1990. However, the authors compare populations with completely different economic and demographic characteristics. This approach most likely confounded the results. In addition, older persons living in multigenerational families also have contact with unvaccinated adults.

In the United States, the proportion of persons 65 years of age or older rose slowly from 8 percent in 1950 to 13 percent in 1998.2 In Japan, this proportion increased from 5 percent in 1950 to 10 percent in 1985 and to 16 percent in 1998.3 The decrease in mortality from 1950 through 1980 could be explained by the effect of improved hygiene and dramatically improving economic conditions, which most likely offset the increased mortality that would otherwise be expected in an aging population. The increase in excess mortality in Japan after 1990 could be explained by the increase in the number of older persons, who are more vulnerable to influenza infection.

The rate of vaccination among Japanese schoolchildren was about 80 percent. However, mortality data presented in Figure 4 of the article by Reichert et al. show a gradual decline, arguing against a consistent effect of the vaccination program. Conversely, mortality rates in the United States remained nearly constant despite an increase in vaccine coverage by a factor of more than 2.5 (among older persons) during the 1990s.

Sakae Inouye, M.D., Ph.D.
Michael H. Kramer, M.D., M.P.H., Ph.D.
National Institute of Infectious Diseases, Tokyo 162-8640, Japan

3 References

1. 1

   Reichert TA, Sugaya N, Fedson DS, Glezen WP, Simonsen L, Tashiro M. The Japanese experience with vaccinating schoolchildren against influenza. N Engl J Med 2001;344:889-896
   Full Text | Web of Science | Medline

2. 2

   National Center for Health Statistics. Health, United States, 2000: with adolescent health chartbook. Washington D.C.: Government Printing Office, 2000:123. (DHHS publication no. (PHS) 2000-1232-1.) (See http://www.cdc.gov/nchs/data/hus00.pdf.)
   

3. 3

   Trends in vital statistics of Japan for 100 years: 1899-1998: centennial of vital statistics of Japan. Tokyo, Japan: Koseisho Daijin Kanbo Tokei Johobu, Heisei 11, 1999.
   

To the Editor:

In the ecologic study by Reichert et al., the reported patterns of death may be inaccurate. The estimates of excess deaths attributed to pneumonia and influenza in the United States were higher by 3200 deaths per influenza season than the estimates in an earlier, more rigorous study.1 This very large increase appears to be related to the authors' definition of the influenza season as lasting six months. Since influenzaviruses usually do not circulate at high levels for such long periods, the patterns of deaths from pneumonia and influenza and from all causes appear to have been exaggerated by the inclusion of many deaths unrelated to influenza. Even if there is a high correlation with earlier estimates of deaths, inflating seasonal death counts by a large constant will distort comparisons between seasons.

A second concern is that several important variables, including patterns of aging, were not addressed. We question the conclusion that vaccinating children reduced mortality in the elderly, when age-specific vaccine-coverage and death rates were not shown.

Even if the patterns of death are accurate, how did Reichert et al. distinguish the effect of the factors that initiated and sustained the substantial drop in the number of deaths before 1962 (shown in Figure 3 and Figure 4 of the article) from the effect, if any, of the pediatric vaccination program? A wealth of studies supports current influenza-vaccine policies in Japan and the United States that focus on vaccinating persons at highest risk for death and serious morbidity. The evidence in this study is insufficiently persuasive to call these policies into question.

Keiji Fukuda, M.D., M.P.H.
William W. Thompson, Ph.D.
Nancy Cox, Ph.D.
Centers for Disease Control and Prevention, Atlanta, GA 30333

1 References

1. 1

   Simonsen L, Clarke MJ, Williamson GD, Stroup DF, Arden NH, Schonberger LB. The impact of influenza epidemics on mortality: introducing a severity index. Am J Public Health 1997;87:1944-1950
   CrossRef | Web of Science | Medline

To the Editor:

The report by Reichert et al. overlooks the rapid increase in the number of convalescence facilities, nursing homes, and homes and institutions for the elderly in recent years in Japan (e.g., an increase in health service facilities by a factor of approximately 10 in the past 10 years).1 This may have been an important cause of dissemination of the influenzavirus in the elderly. Living together in group facilities is an important cause of the spread of influenza among the elderly, and outbreaks of influenza and deaths in such facilities have been widely reported in Japan.2-4 This is an important cause of increased mortality from influenza among the elderly. There has been a steady decrease in three-generation households in Japan, and thus, the protective effect of the immunization of schoolchildren against influenza in regard to their grandparents is questionable.

Tsutomu Yamazaki, M.D., Ph.D.
Toru Suzuki, M.D., Ph.D.
University of Tokyo, Tokyo 113-8655, Japan

Kenji Yamamoto, M.D., Ph.D.
International Medical Center of Japan, Tokyo 162-8655, Japan

4 References

1. 1

   Annual report on health and welfare. Tokyo, Japan: Koseisho, July 2000. (See http://www.mhlw.go.jp/english/wp/wp-hw/index.html.)
   

2. 2

   Deguchi Y, Takasugi Y, Nishimura K. Vaccine effectiveness for influenza in the elderly in welfare nursing homes during an influenza A (H3N2) epidemic. Epidemiol Infect 2000;125:393-397
   CrossRef | Web of Science | Medline

3. 3

   Oshitani H, Saito R, Seki N, et al. Influenza vaccination levels and influenza-like illness in long-term-care facilities for elderly people in Niigata, Japan, during an influenza A (H3N2) epidemic. Infect Control Hosp Epidemiol 2000;21:728-730
   CrossRef | Web of Science | Medline

4. 4

   Murayama N, Suzuki H, Arakawa M, Nerome K, Mizuta K, Kameyama K. Two outbreaks of influenza A (H3N2) in a Japanese nursing home in the winter of 1996-1997, with differing vaccine efficacy. Tohoku J Exp Med 1999;188:289-298
   CrossRef | Web of Science | Medline

Author/Editor Response

The authors reply:

To the Editor: We demonstrated that the vaccination of Japanese schoolchildren against influenza was closely associated with the pattern of excess deaths during the winter season (equivalent to influenza-related deaths) and suggested that the two are related. Alternative hypotheses have been advanced by Inouye and Kramer and by Fukuda et al., who suggest that the observed pattern of excess deaths could derive from factors that altered base-line mortality rates, such as a general improvement in economic conditions and increases in the number of elderly persons. However, as Figure 1Figure 1Base-Line Incidence and Incidence of Excess Deaths from All Causes and Deaths Attributed to Pneumonia and Influenza in Japan, 1950 to 1998. clearly shows, the patterns of base-line and excess deaths are very different in terms of overall trends and especially during the period of the vaccination program.

Yamazaki et al. hypothesize that the increase in the number of excess deaths after 1990 can be explained by the movement of Japan's elderly out of three-generation households and into nursing facilities. Yet the proportion of elderly Japanese persons living in three-generation households has declined by only approximately 1 percent per year since 1980. Moreover, the 10-fold increase in nursing facilities in Japan that has occurred since 1990 has been confined to the small private sector. Overall, the proportion of elderly persons who are living in nursing facilities (<5 percent) changed little from 1975 to 1995.1 This factor cannot explain the tripling of the number of excess deaths that occurred after 1988.

Fukuda et al. are concerned that our model for estimating excess deaths is less specific than other models and that the resulting patterns appear to be exaggerated. A lower specificity, however, should decrease the likelihood of finding an influenza-specific pattern, and models that differ by only a constant do produce similar patterns and similar differences between seasons.

Fukuda et al. question our assignment of the reduction in the number of excess deaths to the elderly segment of the population and seek an alternative explanation that includes age-specific mortality and vaccination rates. In the United States, except for years in which influenza is pandemic, virtually all excess deaths attributed to influenza occur among the elderly.2 The same is most likely true for Japan. In both Japan and the United States, age-specific mortality rates have declined steadily since 1965. During the Japanese vaccination program, vaccination of the elderly was specifically contraindicated; fewer than 1 percent received the vaccine.

Inouye and Kramer cite the lack of a decrease in the number of deaths in the United States since 1990 as evidence against our conclusion that the vaccination of schoolchildren was responsible for a reduction in mortality in Japan, and Fukuda et al. suggest that our study calls into question U.S. vaccination policies. However, we cited reports in which the number of excess deaths was reduced among elderly persons who were vaccinated and noted that vaccination of elderly persons in the United States has only recently reached levels achieved in Japanese schoolchildren. Since the product of vaccine efficacy and the coverage rate is even now only approximately 50 percent, neither herd immunity nor comprehensive direct protection has been achieved, and many elderly persons remain susceptible. Others have examined the vaccination of schoolchildren and have both demonstrated a health benefit3 and projected an efficacy rate similar to that of direct vaccination of the elderly.4 Our position remains that the vaccination of schoolchildren should be viewed as a complementary strategy that can help control influenza.

The following acknowledgment was omitted from our article: We are indebted to Mizuho Murata and Mayumi Onishi, M.D., of Becton Dickinson Nippon for their assistance in obtaining and interpreting the extensive files on cause- and age-specific mortality from the Japanese Ministry of Health and Welfare.

Thomas A. Reichert, Ph.D., M.D.
Entropy Limited and Entropy Research Institute, Upper Saddle River, NJ 07458

Norio Sugaya, M.D.
Nippon Kokan Hospital, 210 Kawasaki, Japan

David S. Fedson, M.D.
Aventis Pasteur–MSD, 69367 Lyons, France

William P. Glezen, M.D.
Baylor College of Medicine, Houston, TX 77030

Lone Simonsen, Ph.D.
National Institute of Allergy and Infectious Diseases, Bethesda, MD 20892

Masato Tashiro, M.D., Ph.D.
National Institute of Infectious Diseases, Tokyo 208-0011, Japan

4 References

1. 1

   Annual report on health and welfare. Tokyo, Japan: Koseisho, July 2000. (See http://www.mhlw.go.jp/english/wp/wp-hw/index.html.)
   

2. 2

   Simonsen L, Clarke MJ, Schonberger LB, Arden NH, Cox NJ, Fukuda K. Pandemic versus epidemic influenza mortality: a pattern of changing age distribution. J Infect Dis 1998;178:53-60
   Web of Science | Medline

3. 3

   Monto AS, Davenport FM, Napier JA, Francis T Jr. Modification of an outbreak of influenza in Tecumseh, Michigan by vaccination of schoolchildren. J Infect Dis 1970;122:16-25
   CrossRef | Web of Science | Medline

4. 4

   Longini IM Jr, Ackerman E, Elveback LR. An optimization model for influenza A epidemics. Math Biosci 1978;38:141-157
   CrossRef | Web of Science

Citing Articles (3)

Citing Articles

1. 1

   Pieter LA Fraaij, Rogier Bodewes, Albert DME Osterhaus, Guus F Rimmelzwaan. (2011) The ins and outs of universal childhood influenza vaccination. Future Microbiology 6:10, 1171-1184
   CrossRef

2. 2

   Masakazu Washio, Asae Oura, Mitsuru Mori. (2008) Ecological studies on influenza infection and the effect of vaccination: Their advantages and limitations. Vaccine 26:50, 6470-6472
   CrossRef

3. 3

   Hidenori Yasuda, Nobuaki Yoshizawa, Mikio Kimura, Mika Shigematsu, Masaaki Matsumoto, Shoji Kawachi, Masamichi Oshima, Kenji Yamamoto, Kazuo Suzuki. (2008) Preparedness for the Spread of Influenza: Prohibition of Traffic, School Closure, and Vaccination of Children in the Commuter Towns of Tokyo. Journal of Urban Health 85:4, 619-635
   CrossRef