Scarlet Fever Outbreak, Hong Kong, 2011

To the Editor: Scarlet fever is a notifiable disease in Hong Kong, Guangdong Province, and Macau in the People’s Republic of China. All 3 areas reported substantial increases in cases during 2011 (Figure, panel A). In Hong Kong, individual data, including age, geographic location, date of notification, and travel history within the incubation period, were collected from all locally notified scarlet fever case-patients. As of December 31, 2011, a total of 1,535 cases (21.7 cases/100,000 population) were reported, which was ≈10× higher than the average number of annual cases reported during the preceding 10 years (1). Of those, 730 cases were laboratory confirmed; 46 cases were imported; and 2 cases, 1 each in a 7-year-old girl and a 5-year-old boy co-infected with chickenpox, resulted in death (2). 
 
 
 
Figure 
 
Trends in scarlet fever during outbreak in Hong Kong, Guangdong, and Macau, People’s Republic of China, 2011. A) Monthly scarlet fever notifications in Hong Kong, Guangdong (data obtained from Department of Health Guangdong Province, www.gdwst.gov.cn/a/yiqingxx ... 
 
 
 
Group A Streptococcus (GAS), the bacterium that causes scarlet fever, is mainly transmitted by direct contact with saliva and nasal fluids from infected persons (3). Many children can also carry GAS or be asymptomatically infected (4). A recent study in China showed that GAS is commonly resistant to macrolides and tetracycline but sensitive to penicillin, chloramphenicol, cefradine, and ofloxacin (5). In Hong Kong, GAS emm type 12 dominated among the isolates cultured during 2011 (6). Most of the cases reported were in children <10 years of age (range 1 month–51 years; median 6 years [interquartile range 4–7 years]). The age distribution is similar to that reported during previous years (data not shown). 
 
In the United Kingdom during the mid-19th century, scarlet fever epidemics were found to follow a 5- to 6-year cycle, but this pattern disappeared as incidence decreased (7). Annual scarlet fever notifications in Hong Kong remained low during 2001–2010 (<4 cases/100,000 population) and did not demonstrate any apparent long-term pattern. The recent increase in scarlet fever notifications might be attributable to antigenic drift, increase in virulence of GAS (8), or increased circulation of GAS. However, other than mandatory notification of medically attended case-patients, systematic laboratory testing of GAS isolates was not conducted in Hong Kong, and these possibilities could not be further investigated. 
 
Notifications of scarlet fever usually peak during December–March in Hong Kong, but the outbreak in 2011 peaked in June (Figure, panel B). The rise in scarlet fever cases in Guangdong Province and Macau slightly preceded that in Hong Kong; cases in Guangdong peaked in April (Figure, panel A). Maximum cross-correlations between spline-interpolated weekly scarlet fever notifications in Guangdong and Macau and those in Hong Kong were found at 1- and 2-week lags, respectively (ρ = 0.45 and 0.58) (Technical Appendix). 
 
In 2011, scarlet fever notification rates were elevated in all 4 regions of Hong Kong: New Territories East, New Territories West, Kowloon, and Hong Kong Island at 27.2, 21.7, 18.9, and 19.6 cases per 100,000 population, respectively. However, a distinctly higher proportion of imported cases before July 2011 (12 of 14, p value for exact binomial test = 0.01) were notified in New Territories East and New Territories West, where the main border crossings to mainland China are located. This finding suggests a link to the outbreak in Guangdong in these regions during the early phase of the local outbreak. 
 
We estimated the instantaneous reproduction number (Rt), which measures the time-dependent frequency of transmission per single primary case (Technical Appendix) (9). An Rt consistently >1 would indicate sustained local transmission. We estimated Rt on the basis of the daily scarlet fever notification data in different periods, adjusted for imported cases. For 19 cases (1.2% of all cases), we could not determine whether infection was local or imported. We estimated Rt in 2 different ways: either by assuming that all of these cases were local or by assuming that they all were imported, to represent possible extreme values of Rt. Rt fluctuated between 0.6 and 2.0 and was consistently >1 from mid-May through the end of June. Rt fell quickly to <1 beginning in early July after 2 fatal scarlet fever cases were reported on May 29 and June 21, which raised widespread concern in the community (Figure, panel C). Heightened surveillance, publicity, health education to the public (Technical Appendix) were implemented by the Centre for Health Protection in early June and could have contributed to the reduction in transmissibility. The health education measures included guidance on prevention and control measures, such as updates of antimicrobial drug resistance profile of GAS issued to all doctors and strengthening reporting of scarlet fever cases by child care centers and schools for prompt epidemiologic investigations. 
 
In summary, we analyzed the notification data of scarlet fever and investigated spatiotemporal spreading patterns of the disease with certain time lags in Hong Kong, Macau, and Guangdong. The estimated Rt in 2011 indicated the potential for local transmission and persistence. Such a borderless spread indicates a critical need to enhance cross-border communication and timely sharing of epidemic information so that future disease control efforts can be made at multiple geographic levels. 
 
Technical Appendix: 
Statistical methods and discussion of impact of public notification of disease transmission, scarlet fever outbreak, Hong Kong, 2011. 
 
Click here to view.(80K, pdf)

is mainly transmitted by direct contact with saliva and nasal fl uids from infected persons (3). Many children can also carry GAS or be asymptomatically infected (4). A recent study in China showed that GAS is commonly resistant to macrolides and tetracycline but sensitive to penicillin, chloramphenicol, cefradine, and ofl oxacin (5). In Hong Kong, GAS emm type 12 dominated among the isolates cultured during 2011 (6). Most of the cases reported were in children <10 years of age (range 1 month-51 years; median 6 years [interquartile range 4-7 years]). The age distribution is similar to that reported during previous years (data not shown).
In the United Kingdom during the mid-19th century, scarlet fever epidemics were found to follow a 5to 6-year cycle, but this pattern disappeared as incidence decreased (7). Annual scarlet fever notifi cations in Hong Kong remained low during 2001-2010 (<4 cases/100,000 population) and did not demonstrate any apparent long-term pattern. The recent increase in scarlet fever notifi cations might be attributable to antigenic drift, increase in virulence of GAS (8), or increased circulation of GAS. However, other than mandatory notifi cation of medically attended case-patients, systematic laboratory testing of GAS isolates was not conducted in Hong Kong, and these possibilities could not be further investigated.
Notifi cations of scarlet fever usually peak during December-March in Hong Kong, but the outbreak in 2011 peaked in June (Figure, panel B). The rise in scarlet fever cases in Guangdong Province and Macau slightly preceded that in Hong Kong; cases in Guangdong peaked in April ( Figure, panel A). Maximum cross-correlations between spline-interpolated weekly scarlet fever notifi cations in Guangdong and Macau and those in Hong Kong were found at 1-and 2-week lags, respectively (ρ = 0.45 and 0.58) (online Technical Appendix, wwwnc. cdc.gov/EID/pdfs/12-0062-Techapp. pdf).
In 2011, scarlet fever notifi cation rates were elevated in all 4 regions of Hong Kong: New Territories East, New Territories West, Kowloon, and Hong Kong Island at 27.2, 21.7, 18.9, and 19.6 cases per 100,000 population, respectively. However, a distinctly higher proportion of imported cases before July 2011 (12 of 14, p value for exact binomial test = 0.01) were notifi ed in New Territories East and New Territories West, where the main border crossings to mainland China are located. This fi nding suggests a link to the outbreak in Guangdong in these regions during the early phase of the local outbreak.
We estimated the instantaneous reproduction number (R t ), which measures the time-dependent frequency of transmission per single primary case (online Technical Appendix) (9). An R t consistently >1 would indicate sustained local transmission. We estimated R t on the basis of the daily scarlet fever notifi cation data in different periods, adjusted for imported cases. For 19 cases (1.2% of all cases), we could not determine whether infection was local or imported. We estimated R t in 2 different ways: either by assuming that all of these cases were local or by assuming that they all were imported, to represent possible extreme values of R t . R t fl uctuated between 0.6 and 2.0 and was consistently >1 from mid-May through the end of June. R t fell quickly to <1 beginning in early July after 2 fatal scarlet fever cases were reported on May 29 and June 21, which raised widespread concern in the community (Figure, panel C). Heightened surveillance, publicity, health education to the public (online Technical Appendix) were implemented by the Centre for Health Protection in early June and could have contributed to the reduction in transmissibility. The health education measures included guidance on pre-vention and control measures, such as updates of antimicrobial drug resistance profi le of GAS issued to all doctors and strengthening reporting of scarlet fever cases by child care centers and schools for prompt epidemiologic investigations, In summary, we analyzed the notifi cation data of scarlet fever and investigated spatiotemporal spreading patterns of the disease with certain time lags in Hong Kong, Macau, and Guangdong. The estimated R t in 2011 indicated the potential for local transmission and persistence. Such a borderless spread indicates a critical need to enhance cross-border communication and timely sharing of epidemic information so that future disease control efforts can be made at multiple geographic levels.

Hand, Foot, and Mouth Disease Caused by Coxsackievirus A6
To the Editor: Coxsackievirus A6 (CVA6) is a human enterovirus associated with herpangina in infants. In the winter of 2012, we evaluated a cluster of 8 patients, 4 months-3 years of age, who were brought for treatment at Boston Children's Hospital (Boston, MA, USA) with a variant of hand, foot, and mouth disease (HFMD) that has now been linked to CVA6 (Table, Appendix, wwwnc.cdc.gov/ EID/article/18/10/12-0813-T1.htm). During this same period, the Boston Public Health Commission's syndromic surveillance system detected a 3.3-fold increase in emergency department discharge diagnoses of HFMD. In the United States, HFMD typically occurs in the summer and early autumn and is characterized by a febrile enanthem of oral ulcers and macular or vesicular lesions on the palms and soles; the etiologic agents are most often CVA16 and enterovirus 71.
In contrast to the typical manifestation, the patients in the Boston cluster exhibited symptoms in late winter (Table, Appendix Patients experienced a prodrome lasting 1-3 days, consisting of fever (8 patients), upper respiratory tract symptoms (4 patients), and irritability (7 patients). This prodrome was followed by the development of a perioral papular rash (8 patients), which was often impetiginized with secondary crusting; a prominent papulovesicular rash on the dorsum of the hands and feet (6 patients); and a perirectal eruption (7 patients). Half of the patients had intraoral lesions. Fever abated in most of the patients within a day after onset of the exanthem. The rash resolved over 7-14 days with no residual scarring. Samples from the oropharynx, rectum, and vesicles from these patients were sent to the Centers for Disease Control and Prevention (Atlanta, GA, USA) for analysis. Reverse transcription PCR and sequencing by using primers specifi c for a portion of the viral protein 1 coding region identifi ed CVA6 (1) (Table, Appendix).
Outbreaks of HFMD caused by CVA6 have been described in Singapore, Finland, Taiwan, and most recently in Japan; most cases have occurred in the warmer months (2-6). Cases in the cluster described here are likely related to an emerging outbreak of CVA6-associated HFMD in the United States (7). The atypical seasonality of the outbreak, during the winter in Boston, could be related to the unusually mild temperatures in the winter of 2012.
Recent CVA6 outbreaks have been characterized by a febrile illness associated with an oral enanthem and lesions on the palms, soles, and buttocks. CVA6 infections in Taiwan during 2004-2009 were associated with HFMD in 13% of cases, with disease defi ned as oral ulcers on the tongue or buccal mucosa and vesicular rashes on the palms, soles, knees, or buttocks (2). In Singapore, where CVA6 accounted for 24% of HFMD cases, patients had oral lesions and <5 peripheral papules, placing them on a spectrum closer to the herpangina more typically observed in CVA6 infection (8).
The patients we report in this cluster most typically had perioral and perirectal papules in addition to vesicles on the dorsum of their hands. Two reports of CVA6-associated HFMD outbreaks describe cases that more closely resemble patients in the Boston outbreak. In a series from Finland in 2008, representative patients had both perioral lesions and vesicles on the dorsum of their hands (6). In a large series of patients with