Jpn. J. Infect. Dis., 63, 422 426, 2010 Original Article Changing Epidemiology of Hand, Foot, and Mouth Disease in Hong Kong, 2001 2009 Edmond Ma*, Terence Lam, K. C. Chan, Christine Wong, and S. K. Chuang Surveillance and Epidemiology Branch, Centre for Health Protection, Department of Health, Hong Kong Special Administrative Region, Hong Kong, China (Received July 28, 2010. Accepted September 30, 2010) SUMMARY: We report the surveillance findings of hand, foot, and mouth disease (HFMD) collected from a general practitioner based sentinel surveillance system and outbreaks reported by institutions and a laboratory based enterovirus surveillance system in Hong Kong from 2001 to 2009. Aseasonal peak was detected in the warmer months (May July), along with a smaller winter peak (October December) from 2006 onwards. The number of older children (À5 years) infected increased from 25.4z in 2001 to 33.0z in 2009 (P 0.01, Mantel Haenszel chi square test). Laboratory surveillance detected a cyclical high enterovirus 71 activity every 3 to 4 years. This activity was associated with a higher average hospitalization rate for HFMD patients in the outbreaks reported in the corresponding year, although the difference was only marginally significant (P 0.09, linear regression test). The changing epidemiology of HFMD warrants continuous surveillance in order to guide preventive publichealth actions. INTRODUCTION Hand, foot, and mouth disease (HFMD), especially that associated with enterovirus 71 (EV71), has resulted in major outbreaks throughout the world, especially Asia, in the past three decades (1). The epidemic which occurred in Taiwan in 1998, included over 129,000 cases of HFMD or herpangina and 405 severe cases with complications such as encephalitis, aseptic meningitis, pulmonary edema or hemorrhage related acute flaccid paralysis, and myocarditis (2). This outbreak eventually resulted in 78 deaths, most of which were due to EV71, and was followed by epidemics in other parts of the Asia Pacific region, including Australia, Singapore, and Japan (3,4). In 2008, another wave of major HFMD epidemics which occurred in the same region, including mainland China, Taiwan, Singapore, and Hong Kong, was also associated with an increased EV71 activity (5 8). Unfortunately, there is currently no effective chemoprophylaxis or vaccination for HFMD or EV71 infection. Public health prevention and control strategies rely on good personal hygiene and social distancing measures, such as isolating the infected case from schools, to prevent further transmission of the viruses. Surveillance findings can also be used to provide important information so that the health authorities can issue an early warning to the target population. HFMD is not a statutory notifiable disease in Hong Kong, although its activity is monitored by various surveillance systems, including sentinel surveillance and voluntary *Corresponding author: Mailing address: Surveillance and Epidemiology Branch, Centre for Health Protection, Department of Health, Hong Kong Special Administrative Region, Hong Kong, China. Tel: {852 2125 2240, Fax: {852 2711 4847, E mail: smo_ss4 dh.gov.hk reporting of HFMD outbreaks and EV71 infection. Similar to many other Asian areas, Hong Kong has established a general practitioner based sentinel surveillance system to monitor the HFMD activity in the community (9,10). In addition, whenever HFMD outbreaks occurr in schools or other institutions, they are reported to the Department of Health on a voluntary basis so that the outbreaks can be controlled. Herein we analyze the surveillance data for HFMD in Hong Kong with a view to gaining a better understand of the epidemiology of this disease, with particular emphasis on the time and patient characteristics, so as to guide further preventive public health actions. METHODS Data from two independent surveillance systems, namely the sentinel surveillance and outbreak reporting systems, were analyzed. The sentinel surveillance system consists of some 40 primary care general practitioners in Hong Kong. These sentinel doctors record the number of patients diagnosed with HFMD every week according to a standardized clinical case definition. Acase is defined as a patient who has clinical symptoms of HFMD, typically vesicles on the hand or foot, and oral lesions, with or without fever. Laboratory confirmation of the pathogen causing the disease is not required. The sentinel doctors also record the age group breakdown of patients with HFMD and the total number of patients seen every week. The consultation rate for HFMD is calculated by dividing the number of patients diagnosed with HFMD by the total number of patients seen and is expressed as the number of HFMD cases per 1,000 consultations. We also analyzed the trend in the number of HFMD outbreaks reported by the institutions over the same study period. Institutions such as child care centers or kindergartens report HFMD outbreaks voluntarily to 422
the Department of Health for public health control. Pearson's correlation test was used to analyze whether the trends detected by these two independent systems correlated with each other. All calculations were performed using Statistical Package for the Social Sciences, version 14.0(SPSS, Chicago, Ill., USA). The epidemiological investigation findings, as well as the clinical and laboratory results, for each HFMD outbreak were entered into the electronic Public Health Information System. Since complete data for these outbreaks are only available for the period 2001 2008, analysis of the HFMD outbreaks was confined to this period. The geographical distribution of the HFMD outbreaks was determined on the basis of the 18 administrative districts of Hong Kong. The number of schools (including child care centers, kindergartens, and primary and secondary schools) in each administrative district in 2008 was obtained from the Education Bureau and Social Welfare Department. The incidence of HFMD outbreaks in each district was estimated by dividing the average number of HFMD outbreaks reported during 2001 2008 by the number of schools in the corresponding district in 2008. The Mantel Haenszel chi square test was used to examine whether there was any significant difference as regards the type of institution reporting HFMD over time. We determined the age and sex distribution of those people affected by HFMD outbreaks and determined whether these characteristics changed significantly with time using the Mantel Haenszel chi square test. The clinical presentation, hospitalization rate, complication rate, and case fatality of all outbreaks reported in each year of the study period were also examined. During investigation of HFMD outbreaks, clinical specimens, e.g., stool samples collected from recently affected patients, are sent to the Public Health Laboratory of the Department of Health for diagnosis of the causative agents. This laboratory is the main reference laboratory for the detection of enteroviruses in Hong Kong. Clinical specimens were tested for enterovirus upon referral from outpatient and inpatient settings in both public and private sectors. Enteroviruses were identified by viral culture following a conventional method or their genome detected using an in house real time reverse transcriptase polymerase chain reaction (RT PCR) method. The laboratory surveillance findings from 2001 to 2009 were analyzed. To assess the EV71 related morbidity, we examined whether there was any association between the yearly proportion of EV71 among all enteroviruses detected by the Public Health Laboratory and the average yearly hospitalization rate of HFMD patients reported in the outbreaks from 2001 to 2008. Linear regression was used to test for statistical significance. and troughs of the two trends match well with each other (Pearson's correlation coefficient 0.73). A seasonal peak can be seen between May and July every year except 2003, when a severe acute respiratory syndrome (SARS) epidemic occurred, and 2009, when a human swine influenza H1N1 pandemic occurred. About half (48.9z) of all HFMD outbreaks occurred during this peak season from 2001 to 2009 (Fig. 2). Apart from the traditional summer peak, a smaller winter peak between October and December, which became more obvious after 2006, was also observed. About a quarter (27.3z) of the HFMD outbreaks in this period occurred Fig. 1. Consultation rate of HFMD by sentinel surveillance system based at general practitioners and number of HFMD outbreaks reported to Department of Health, 2001 2009. Fig. 2. Number of HFMD outbreaks reported to Department of Health by month, 2001 2009. RESULTS Secular trend and seasonality: Fig. 1 shows the trends in HFMD activity detected using the sentinel surveillance data and the number of outbreaks reported. The peaks Fig. 3. Geographical distribution of HFMD outbreaks in Hong Kong, 2001 2008. 423
during the winter peak. Geospatial distribution of HFMD outbreaks: Fig. 3 shows the geographical distribution of HFMD outbreaks in Hong Kong from 2001 to 2008. The top three districts in terms of the number of HFMD outbreaks were Kowloon City, Yuen Long, and Sha Tin, where larger numbers of schools are located, thus reflecting the higher disease burden in these districts. Some districts, however, had a higher incidence rate of HFMD. Although the overall incidence for HFMD outbreaks was one in 15 schools per year, four districts, namely North, Kwun Tong, Yau Tsim Mong, and Sai Kung, had incidence rates ranging from one in 10 to one in 13 schools per year. The schools located in these districts should be targeted with special preventive measures. Demographic characteristics of people affected: The majority (75.0z) of outbreaks occurred in pre primary institutions such as child care centers and kindergartens. Since 2003, more outbreaks were reported by institutions for older age groups (i.e., primary, secondary, Fig. 4. Age distribution of HFMD detected by sentinel surveillance system based at general practitioners, 2001 2009 (n 3,512). and tertiary schools). The Mantel Haenszel chi square test showed an increasing trend in the number of HFMD outbreaks reported by these institutions ( P º 0.001). Similarly, a significant upward trend in the number of older children diagnosed with HFMD by the sentinel doctors was also observed. Children aged older than 5 years accounted for only 25.4z of cases in 2001, whereas this figure increased to 33.0z in 2009 (P 0.01, Mantel Haenszel chi square test) (Fig. 4). Nonetheless, children aged less than 5 years remained the largest group (66.4z), with adults aged 20 years or older accounting for only 2.6z of patients. More boys were affected than girls, and the male to female ratio did not change significantly with time (P 0.22, Mantel Haenszel chi square test). Clinical presentation and laboratory diagnosis: Most patients presented with a mild variant of the typical symptoms, including fever, vesicles or rash on the hands and feet, as well as oral lesions (Table 1). Only 1.3z of patients required hospital care. The hospitalization rate ranged from 0.6 to 2.8z each year. The only patient with severe complications in the past 10 years was a 5 year old boy, in good previous health, who was infected in 2008. He presented with high fever, oral lesions, and rashoverthehandsandfeetandwentontodevelop neurological symptoms, including ataxia, which required intensive care. A stool sample tested positive for EV71 by PCR. This patient attended a kindergartencum nursery where an outbreak of HFMD was reported to affect a total of 20 children and a 28 year old teacher. The outbreak was controlled after temporary class suspension to stop the transmission. There were no fatalities among the patients infected in any of the 1,121 outbreaks studied. The causative agent was identified in 312 out of 1,121 Table 1. Clinical and laboratory characteristics of HFMD outbreaks reported to Department of Health, 2001 2008 2001 2002 2003 2004 2005 2006 2007 2008Total No. of outbreaks 98234 25 155 76 222 155 156 1,121 No. of persons affected 804 1,909 178 1,034 545 1,647 1,072 893 8,082 Clinical presentation (z of persons with symptoms and signs) Fever 43.851.2 3.4 42.6 30.5 31.9 41.4 42.7 40.7 Vesicles on hand or foot 70.5 54.4 1.1 49.2 56.1 61.1 56.6 51.3 55.6 Oral lesion 81.3 88.9 5.6 76.2 93.0 91.0 89.6 88.4 85.4 Rash on face 0.0 0.5 0.0 1.5 0.9 1.5 0.7 1.3 0.9 Rash on limb 31.1 48.1 3.9 66.5 65.7 62.1 61.9 61.6 55.2 Rash on trunk 2.0 1.80.0 5.0 5.7 4.0 3.7 5.3 3.6 Clinical outcome (z of persons with adverse outcome) Hospitalization rate 2.0 1.3 1.1 0.6 0.6 0.81.2 2.8 1.3 Complication rate 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 Fatality rate 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Laboratory testing (No. and z of outbreaks with laboratory results) Positive test (no.) 46 94 10 44 21 34 21 42 312 Coxsackievirus A 80.4 94.7 100.0 65.9 76.2 88.2 71.4 50.0 79.2 Coxsackievirus B 0.0 0.0 0.0 0.0 0.0 2.9 0.0 0.0 0.3 Enterovirus 71 15.2 2.1 0.0 25.0 14.3 8.8 28.6 47.6 16.7 Other enteroviruses 2.2 0.0 0.0 4.5 9.5 0.0 0.0 0.0 1.6 Adenovirus 0.0 0.0 0.0 2.3 0.0 0.0 0.0 0.0 0.3 Mixed 2.2 3.2 0.0 2.3 0.0 0.0 0.0 2.4 1.9 Negative test (no.) 9 29 816 1 11 9 14 97 Test not done (no.) 43 111 7 95 54 177 125 100 712 424
Fig. 5. Linear regression plot of average yearly hospitalization rate of HFMD patients reported in the outbreaks against yearly proportion of EV71 detected among all enteroviruses, 2001 2008 (P 0.09). outbreaks (27.8z), with coxsackievirus A (mostly coxsackievirus A16) accounting for nearly 80z of outbreaks and EV71 being the second most common pathogen identified. Other causative agents included coxsackievirus B and various untyped enteroviruses. An apparent association was observed between the yearly proportion of EV71 detected among all enteroviruses and the average yearly hospitalization rate of HFMD patients reported in the outbreaks (Fig. 5). However, a linear regression test only showed a marginally significant result (P 0.09). Nevertheless, when EV71 became more prevalent as the circulating virus, more HFMD cases required admission to hospital for management. DISCUSSION To the best of our knowledge this is the most comprehensive review of the HFMD situation in Hong Kong, covering as it does about 1,200 outbreaks reported in the past decade affecting over 8,000 people. The two independent systems, namely the sentinel surveillance system and the outbreak reports, showed a good correlation in terms of detecting the trends of HFMD in the community. Both systems revealed a consistent seasonal trend, allowing the publichealth authorities to issue an early warning to relevant parties based on the surveillance findings. Similar sentinel surveillance data based reviews undertaken in Belgium, Taiwan, and Malaysia have also shown that the viruses were more active in the summer season (9,12). These systems are useful when large epidemic occur so that the epidemiological characteristics of the disease can be compared with those for previous outbreaks (8). The winter peak (from October to December) has become more obvious in Hong Kong since 2006, therefore it would be interesting to determine whether similar observations have been reported from the surveillance findings in other areas, including Belgium, Singapore, and Taiwan (11,12). The exact reason for this more obvious winter peak is not known, although it could be due to the warmer winters recorded in Hong Kong recently. Data obtained from the Hong Kong Observatory showed the average daily temperature for the last 12 weeks in 2006, 2007, and 2008 to be 22.3, 21.7, and 22.39C, respectively, slightly higher than that recorded for the period 2001 2003 (between 21.6 and 21.79C). Indeed, HFMD has been found to be associated with various meteorological factors, including temperature, vapor pressure, and duration of sunshine (13). Further studies regarding the effect of climatic factors on enteroviruses activity are warranted in view of emerging concerns regarding the effect of global warming on infectious diseases. In agreement with the surveillance findings from other studies (11,14), children under 5 years of age accounted for most of the HFMD cases in the outbreaks studied herein. Thus, such children have been reported to accounted for three quarters of the EV71 cases in the United States (15). However, it should be noted that more teenagers have been affected in recent years. This shift in the susceptible population suggests a need for preventive public health actions to cover this target group as well. Indeed, preventive actions targeting HFMD and EV71 infection, such as organizing health talks to provide workers with infection control knowledge and skills, have been extended to include primary and secondary schools in Hong Kong in recent years. The hospitalization and complication rates (1.3 and 0.0z, respectively) were very low for the HFMD patients studied herein, probably due to the fact that coxsackievirus A, which induces a milder illness than EV71, is responsible for the vast majority (nearly 80z) of cases. In contrast, the complication rate for EV71 infection alone is much higher. Thus, in the EV71 epidemicreported in 2008, 11.2z of 98 cases of EV71 reported to the Department of Health had complications, including meningitis or encephalitis (6.1z), pneumonia (3.1z), acute flaccid paralysis (1.0z), or shock (1.0z) (8). There was one fatal case (an 11 month old boy) in 2008, and the case fatality rate (CFR) was 1.0z. EV71 activity was found to be highly cyclical, thereby suggesting that laboratory results should be monitored closely as EV71 is known to be associated with severe complications and fatalities (17 19). EV71 activity in Japan peaks every 3 to 4 years, whereas the period in Malaysia seems to be shorter at 2 to 3 years (9). It has been suggested that the quiescent period between epidemics is a result of the development of herd immunity during the high infection rate epidemic (15). The accumulation of a susceptible population during this period, especially in younger cohorts, probably contributes to the next epidemic. Another interesting finding was the absence of a summerpeakin2003and2009whensarsandinfluenza (H1N1) pandemics occurred, respectively. During these periods, all Hong Kong citizens were advised to take particular care with personal hygiene, especially hand washing and wearing a face mask, for the whole summer. This, together with the social distancing effect arising from school closures, probably resulted in reduced contact between sick and healthy children and hence reduced viral transmission. Since coxsackievirus A and EV71 can also be transmitted via aerosol, such public health measures to deal with SARS and pandemicinfluenza also resulted in ``cross over'' protection against HFMD. Indeed, there is a good evidence to show that such community hygiene measures successfully reduced various infections spread by the respiratory route during the SARS epidemic(16). A similar observation was made in Malaysia during the 2003 EV71 outbreak. The number of EV71 cases dropped considerably by the end 425
of April, when the public health measures put into place to combat SARS also served to mitigate the transmission of enteroviruses (9).Another possible explanation is that the shift in attention from HFMD to SARS and pandemic influenza during these periods resulting in reduced reporting of HFMD cases. In summary, surveillance for HFMD must be maintained as there is no effective chemoprophylaxis or vaccine available (20 22).Social distancing, in other words avoiding contact with HFMD cases, maintaining good personal hygiene, and disinfection of the environment are probably the most effective prevention measures. These surveillance findings will assist health authorities to establish public health actions, e.g., by issuing early warnings before for both the summer and winter peaks, improving the efficient targeting of health promotion resources towards the most at risk groups. Acknowledgments We would like to thank Mr.Simon Wong for his statistical advice, and the health professionals staff in Public Health Laboratory Centre for doing the laboratory tests for HFMD; all doctors participating in the sentinel surveillance system as well as medical, nursing, and research staff in Epidemiology Section, Surveillance and Epidemiology Branch, Centre for Health, HKSAR, for contributing the epidemiology information of HFMD. Conflict of interest None to declare. REFERENCES 1. Hosoya, M., Kawasaki, Y., Sato, M., et al. (2006): Genetic diversity of enterovirus 71 associated with hand, foot and mouth disease epidemics in Japan from 1983 to 2003.Pediatr.Infect.Dis. J., 25, 691 694. 2. Ho, M., Chen, E.R., Hsu, K.H., et al. (1999): An epidemic of enterovirus 71 infection in Taiwan.N.Engl.J.Med., 341, 929 935. 3. Chan, K.P., Goh, K.T., Chong, C.Y., et al. (2003): Epidemic hand, foot and mouth disease caused by human enterovirus 71, Singapore.Emerg.Infect.Dis., 9, 78 85. 4. Fujimoto, T., Chikahira, M., Yoshida, S., et al. (2002): Outbreak of central nervous system disease associated with hand, foot, and mouth disease in Japan during the summer of 2000: detection and molecular epidemiology of enterovirus 71.Microbiol.Immunol., 46, 621 627. 5. Ding,N.Z.,Wang,X.M.,Sun,S.W.,etal. (2009):Appearance of mosaic enterovirus 71 in the 2008 outbreak of China.Virus Res., 145, 157 161. 6. Huang, S.W., Hsu, Y.W., Smith, D.J., et al. (2009): Reemergence of enterovirus 71 in 2008 in Taiwan: dynamics of genetic and antigenic evolution from 1998 to 2008.J.Clin.Microbiol., 47, 3653 3662. 7.Ministry of Health, Singapore (2008): EV71 associated epidemic hand, foot and mouth disease in Singapore.Epidemiol.News Bull., 34, 63 66. 8. Ma, E., Chan, K.C., Cheng, P., et al. (2010): The enterovirus 71 epidemic in 2008 public health implications for Hong Kong.Int. J.Infect.Dis., 14, e775 780. 9. Podin, Y., Gias, E.L., Ong, F., et al. (2006): Sentinel surveillance for human enterovirus 71 in Sarawak, Malaysia: lessons from the first 7 years.bmc Public Health, 6, 180. 10. Tseng, F.C., Huang, H.C., Chi, C.Y., et al. (2007): Epidemiological survey of enterovirus infections occurring in Taiwan between 2000 and 2005: analysis of sentinel physician surveillance data.j. Med.Virol., 79, 1850 1860. 11.Druyts Voets, E.(1997): Epidemiological features of entero nonpoliovirus isolations in Belgium 1980 94.Epidemiol.Infect., 119, 71 77. 12.Ministry of Health, Singapore (2010): Communicable Disease Surveillance in Singapore 2005 2009.Available at ºhttp://www. moh.gov.sg / mohcorp/ uploadedfiles/ Publications/ ReportsÀ. Accessed 5 March 2010. 13.Urashima, M., Shindo, N.and Okabe, N.(2003): Seasonal models of herpangina and hand foot mouth disease to simulate annual fluctuations in urban warming in Tokyo.Jpn.J.Infect. Dis., 56, 48 53. 14. Ang, L.W., Koh, B.K., Chan, K.P., et al. (2009): Epidemiology and control of hand, foot and mouth disease in Singapore, 2001 2007.Ann.Acad.Med.Singapore, 38, 106 112. 15. Khetsuriani, N., LaMonte Fowlkes, A., Oberste, S., et al. (2006): Enterovirus surveillance United States, 1970 2005.Morbid. Mortal.Wkly.Rep., 55(SS 8), 1 20. 16. Lo, J.Y., Tsang, T.H., Leung, Y.H., et al. (2005): Respiratory infections during SARS outbreak, Hong Kong, 2003.Emerg. Infect.Dis., 11, 1738 1741. 17. Lee, T.C., Guo, H.R., Su, H.J., et al. (2009): Diseases caused by enterovirus 71 infection.pediatr.infect.dis.j., 28, 904 910. 18. Huang, C.C., Liu, C.C., Chang, Y.C., et al. (1999): Neurologic complications in children with enterovirus 71 infection.n.engl. J.Med., 341, 936 942. 19. Huang, M.C., Wang, S.M., Hsu, Y.W., et al. (2006): Long term cognitive and motor deficits after enterovirus 71 brainstem encephalitis in children.pediatrics, 118, e1785 1758. 20.Wang, S.M.and Liu, C.C.(2009): Enterovirus 71: epidemiology, pathogenesis and management.expert Rev.Anti Infect.Ther., 7, 735 742. 21. Lee, M.S. and Chang, L.Y. (2010): Development of enterovirus 71 vaccines.expert Rev.Vaccines, 9, 149 156. 22. Xu, J., Qian, Y., Wang, S., et al. (2010): EV71: an emerging infectious disease vaccine target in the Far East? Vaccine, 28, 3516 3521. 426