Biweekly January 26, 2010 / Vol.26 / No.2
invasive Streptococcal pneumoniae in Taiwan,
Reinfection of Invasive Streptococcus pneumoniae － Analysis of Notifiable
Diseases Database in Taiwan
Sheng-Yi Chuang, Che-Chieh Yen, Chi-Ching Huang
Fifth Division, Centers for Disease Control, Taiwan
This study describes the epidemiology of invasive Streptococcal pneumoniae in Taiwan,
invasive Streptococcus pneumoniae based on which may provide useful information for
Notifiable Diseases Database between October clinical practice and public health authorities.
15, 2007 and June 30, 2009. There were 1,366 Keyword：invasive Streptococcus pneumoniae
confirmed cases, with 67 % male and 9.7 % infection, notifiable diseases
mortality. Persons under five years of age and database, reinfection
over 64 years old together represented majority
of cases. According to 2008 data, the annual Introduction
incidence was 3.5 per 100,000 populations, and Streptococcus pneumoniae is one of the
mortality was 0.48 per 100,000 population. important pathogens that causes invasive
Among 27.6% of patients have underlying infections in the elderly and children by contact
diseases, the identified top seven serotypes were with respiratory secretions and droplet spread.
14, 23F, 3, 6B, 19F, 19A, 23A, and the disease The pathogen is normally found in
was more prevalent in spring and winter. nasopharyngeal cavity of human body, when
We identified seven patients, 6 were male, immunity is decreasing, it colonizes in the body
with reinfection from the notifiable diseases and causes invasive clinical symptoms including
database. These patients aged 2 to 87 years old, bacteremia, meningitis, peritonitis, septicemia,
and had one reinfection. They all had underlying and other respiratory symptoms such as
diseases which made them more vulnerable to pneumonia, and otitis media .
reinfection. The time between the two infections
in these patients ranged from 38 to 390 days.
26 Reinfection of Invasive Streptococcus
Two patients were infected twice with bacteria of pneumoniae － Analysis of Notifiable
the same serotype (23, and 6B). The seven Diseases Database in Taiwan
34 Investigation of Tuberculosis Cluster in a
patients all survived. This is the first
Keelung City Hospital
epidemiological analysis of reinfection of
27 Taiwan EB January 26, 2010
The Taiwan Epidemiology Bulletin series of same time to increase the inoculation rate.
publications is published by Centers for Disease Although there are many therapeutic
Control, Department of Health, Taiwan(R.O.C.) methods to battle against invasive Streptococcus
since Dec 15, 1984. pneumoniae infection, the morbidity and
Publisher : Hsu-Sung Kuo mortality rate were still very high , which was
Editor-in-Chief : Min-Ho Lai a serious threat for people with low immunity. In
Executive Editor : Li-Gin Wu, Hsiu-Lan Liu recent years, even thought lots prevention efforts
Telephone No : (02) 2395-9825 about the disease were made, there was an
Address : No.6,Linshen S. Road, Taipei,Taiwan increasing number of patients with reinfection
100(R.O.C.) . There were few references discussing the
Website : http://teb.cdc.gov.tw/ issue and no relevant research in the country
Suggested Citation : now.
[Author].[Article title].Taiwan Epidemiol Bull Taiwan Centers for Disease Control
2010;26:[inclusive page numbers] (Taiwan CDC) classified the Streptococcus
pneumoniae infection as a category 4 notifiable
Based on the epidemiological study in the
disease since October 15, 2007 and incorporated
USA, 40,000 people died annually due to
it in notifiable disease reporting system. The data
Streptococcus pneumoniae infection mainly in
has been collected for more than one year. This
the elderly and children less than 2 years old .
research was based on the database of the system,
According to the related research in Taiwan,
based on the onset date from October 15, 2009 to
most patients were children under 5 years old
June 30, 2009, to analyze the epidemiology of
and the elderly 60 years old or above, the highest
invasive Streptococcus pneumoniae infection
fatality rate was found in age 75 and older.
and it’s reinfection. This study could be used as a
In recent years, several medical centers in
reference for government when making
Taiwan found that drug resistance of
vaccination policy, this is also a pioneer study on
Streptococcus pneumonia in Taiwan was one of
the characteristics of reinfection in epidemiology
the worst countries in the world . As the result,
and clinical presentation in Taiwan.
the government decided to enforce the policy of
vaccination. Since October 2007, the
Material and Methods
Department of Health accepted the 23-valent
A. Notifiable disease reporting system:
pneumococcal polysaccharide vaccines donated
According to the case definition announced
by Formosa Plastics Group and the vaccines
by Taiwan CDC, all invasive disease, such as
were given in the following three years. In 2007,
septicemia, pneumonia, meningitis, arthritis,
the vaccines were first given nationwide to
osteomyelitis, pericarditis, and peritonitis,
residents in senior health-care centers and the
caused by Streptococcus pneumonia, and
elderly 75 years old or above in Yunlin county,
isolation of Streptococcus pneumoniae from
Chiayi city, and Chiayi county. In 2008-2009, the
sterile sites, including blood, cerebral spinal
elderly nationwide of 75 years old or above
fluid, synovial fluid, ascites, or pericardial fluids,
received the PPV-23 and flu vaccination at the
Vol.26 / No.2 Taiwan EB 28
should be reported within one week. The strain second time at least 30 days after the previous
should be sent to Research and Diagnostic test.
Center, Taiwan CDC for serotyping. Finally, the We selected the reinfection cases through
data is incorporated into notifiable disease checking and comparing identification from the
database. database, which should be with different report
B. Data collection: numbers and the date of a positive test identified
We collected data information from in the laboratory should be different from the
confirmed cases of invasive Streptococcus other test with at least 30 days interval. Besides,
pneumoniae infection which reported according to approve deliberation and completion of the
to the regulation by doctors from October 15, data, the local health department was notified to
2007 to June 30, 2009. The information included retrieve files of cases from medical institutions
name, identification number, date of disease that patients attended for research analysis.
onset, gender, age of disease onset, area of
infection, strain type, clinical signs, underlying Result
diseases, and survival. All collected information From October 15, 2007 to June 30, 2009,
was established into a database by using the notifiable disease reporting system received
Microsoft Excel for epidemiological analysis. reports of 1,366 confirmed cases (24% are from
C. Other underlying diseases: medical centers and 76% are from other facilities
According to the published documents include 14% from clinics, 24% from hospitals,
[2,8], the Advisory Committee on Immunization and 38% from regional hospitals). Male patients
Practices (ACIP) of U.S. suggested that the risk were about 67%. As to the age distribution and
group of the disease is patients with the the annual morbidity, the highest was for 64
following underlying diseases including human years old and above and less than 5 years old.
immunodeficiency virus (HIV) infection, The patients of 75 years old and above had the
congenital heart diseases, malignant tumors, highest mortality (4.61%). As to the case fatality,
alcohol intoxication, asthma, immune the highest rate was for 50 years old and above
suppressive or abnormal, steroid or immune which was 9.7 % on average. The definition of
suppression usage, neurological diseases, case with underlying diseases is that patient had
asplenia syndrome or splenectomy, chronic at least one of the underlying diseases suggested
obstructive pulmonary disease (COPD), and by U.S. ACIP; doctors reported the underlying
other dread diseases. Patients with underlying diseases of cases to notifiable reporting system
diseases are those who have at least one of the according to the clinical diagnosis and history of
diseases listed above. patients. Analyze the rate of underlying diseases
D. The definition of reinfection and in different age group of patients would allow
selection methods: understanding the correlation and realizing the
Definition: According to the published fatality of patients with underlying diseases.
reference  and the treatment period of the Through the study, the number of patients with
disease, patients with reinfection means patient underlying diseases was about three tenths
specimen is positive by laboratory culture for the (27.6%) of all patients and the rate increasing
29 Taiwan EB January26, 2010
with age. For example, patients with underlying 23A, each accounted for more than 50 strains,
diseases were about four tenth for those aged 50 and the 6A ranked eighth with 36 strains. As to
years and older. Overall, the case fatality rate of the high risk age groups, the top five serotypes
the patients with underlying diseases was about for children
12.5% (Table 1). less than 2 years were 19F, 6B, 14, 23F, and 19A;
The strain of Streptococcus pneumoniae the top five serotypes for the age of 65 years and
was cultured by reporting medical facilities and older were 14, 3, 23F, 6B, and 19F. The 7-valent
was sent to Taiwan CDC Research and pneumococcal polysaccharide-protein conjugate
Diagnostic Center for serotype identification. vaccine (PCV-7) has the strain coverage rate of
The test protocol followed the Manual of 72.6% and 68.8%, respectively for children less
Laboratory Testing Standards for Communicable than 2 years old and less than 5 years old. The
Diseases. Finally, the serotype data of the strain 23-valent polysaccharide vaccine (PPV-23)
were recorded into the notifiable disease contains most of the serotypes for people 65
reporting system. The majority serotypes years and older with strain coverage rate of
identified were 14, 23F, 3, 6B, 19F, 19A, and 79.6% (Table.2).
Vol.26 / No.2 Taiwan EB 30
According to the month of onset, the analyzed and found 7 reinfection patients by
infection often occurred in spring and autumn June 30, 2009, which comprised 0.51 % (7/1366)
period in 2008; the peak months were in March of all cases. The age for these patients ranged
and December (Figure.1). 2-87 years old. Six patients were male, one was
Since October 15, 2007, the disease was female. The time interval in the episodes of
classified as a category 4 notifiable disease. We reinfection was 38-390 days. All of them had
31 Taiwan EB January26, 2010
underlying diseases including alcoholism, gradually rose with age, which was similar to the
COPD, cirrhosis, diabetes, stroke, and data from other countries. [8-9]. However, the
thrombocytopenia purpura. Two patients were fatality for the age group of 50 years old and
reinfected with the same serotype, 6B and 23 A. above was higher in the country, which needs
All of them survived (Table 3). further analysis if it was correlated with other
diseases. As a whole, the case fatality rate was
Discussion 9.7 % in Taiwan, which was similar to Australia
Epidemiological analysis (9%). The mortality of the disease in the age
Streptococcus pneumoniae locates in group of 65 years old and above was the highest.
nasopharyngeal region and causes severe The annual death toll was relative high for
invasive illness when invading human body. children less than 2 years old in other countries
Infection usually occurs through close and long , but not in this country, which may due to
time contact with patients, so infection usually the data was accumulated for many years in other
happens when human immunity is weaken . countries, on the contrary, our data only collected
It was classified as a category 4 notifiable for the recent two years. We should continue to
disease since October 15, 2007 by Taiwan CDC, monitor to elucidate the result.
which requires all confirmed cases from medical The serotypes of bacterial strains was
facilities be reported to the system. In the past identified by the strain isolation of the
studies, most cases were collected from medical participating hospitals and the effort of Taiwan
centers . In this study, 76 % of cases came CDC Research and Diagnostic Center. In this
from other medical facilities, which provides a study, most serotypes identified were 14, 23F, 3,
more complete picture for the epidemiological 6B, 19F, 19A, and 23A; the orders and serotypes
analysis. were different from the previous research .
The two highest morbidity age groups of The data of the previous project was collected
invasive Streptococcus pneumoniae infections from more than 30 regional hospitals. The results
were children less than 2 years old and the of this study revealed that the data was
elderly older than 65 years old according to some influenced after vaccination policy of PCV-7 and
studies [2,8]. However, in our study, the PPV-23 has been implemented in some counties
morbidity of the disease occurred most in and cities. By analyzing the prevalent serotypes
children less than 4 years old and elderly 65 of Streptococcus pneumoniae in patients would
years old and above in the country. The infection allow understanding if the marketed vaccines
age of children in Taiwan was higher than other cover the prevalent serotypes or not, and to
countries might due to the vaccination policy prevent the disease by vaccination to the elderly
was not completely enforced in children. The at high risk. The vaccine effectiveness can be
new policy has set on July 20, 2009 to vaccinate evaluated by the strain coverage of the prevalent
the high risk group of children younger than 5 serotypes. During this study, the policy just
years old with PCV-7. With continuous started to give adults of 75 years and older a dose
monitoring, the distribution of age specific of PPV-23. After vaccination, comparing the
morbidity shall be clarified. Besides, the fatality data from January-June 2009 to the same period
Vol.26 / No.2 Taiwan EB 32
in 2008 without vaccination, the only difference the invasive disease was still increasing which
was found in the order of number of strains meant that immunity was the key factor of
detected; no obvious difference was found in the infection. Although children younger than 4
prevalent serotypes and these were covered in years old had less chance to have underlying
the vaccine. The reason possibly was that the diseases, more Streptococcus pneumoniae was
vaccination policy just started; more data was colonized in the nasopharyngeal region. Besides,
necessary for the evaluation. In the future, the immune system of children has not well
detection of drug susceptibility of the isolates developed and the colonization rate of bacteria in
would clarify which serotypes are highly drug the nasopharyngeal region of children was high.
resistant and if the marketed vaccine would be It was possible that bacterial colonization was
effective. present in children and reduced their immunity,
Climate and seasonality are the main external which caused subsequent infection. Thus, the
factors affecting the number of patients in this patient number in children group was high and
disease [12-13]. Besides, duration of daytime, the result was similar to other studies. In
temperature, and rainfall, were also correlated summary, the case fatality of patients with
with the incidence of the disease . The onset underlying diseases was very high (12.2%),
of the illness mainly occurred from November to which revealed that the high risk group of
next March, and peaks occurred in March and patients with underlying diseases should get
December in 2008. According to the report from vaccination to prevent the diseases. We hope that
Central Weather Bureau, the weather was warm the incidence rate of high risk group of patients
and rainy and the cold front affected the weather could decrease after intensive vaccination, and
for 8 times which caused obvious temperature the mortality could decrease at the same time.
variations in March 2008 and cold air masses Reinfection
moved south which caused temperature Patients with reinfection have been
fluctuations in December 2008. Climate change speculated to have high correlation with HIV
certainly has a huge impact on the susceptibility infection [5-6]. Others such as cancer, chronic
of human to the disease. In the future, to monitor pulmonary disease, even female also have
the seasonality trend of the disease with correlation with reinfection [5,14]. The
Geological Information System (GIS) platform reinfection of children usually correlates with
would be feasible. sickle cell trait . However, the real reasons
Based on the confirmed cases data, it is were not known and studies revealed that most
evident that the risk of patients with underlying reinfection patients have predisposing condition,
diseases increases with age. In general, the in other words, most of them have underlying
percentage of senior people with underlying diseases and the percentage was about 60-100%
diseases was higher than postadolescence and [16-17]. All seven patients in the study had
children. We speculate that when patients with underlying diseases (100%), which was similar
underlying diseases and weaker immune system, to other studies. Even though after vaccination,
even if Streptococcus pneumoniae at the patients were still infected again and the
nasopharyngeal region was reducing with age, percentage was not low . One of the patients
33 Taiwan EB January26, 2010
in the study was infected again within one year infectious diseases. 3rd ed., WB Saunders,
after receiving PPV-23, which indirectly Philadelphia 1994; 1117-40.
indicated that either the vaccine did not work 2. CDC. Prevention of pneumococcal disease:
effectively, or the immune system of patients recommendations of the advisory
with underlying diseases did not respond well as committee on immunization practices
expected . During the study, the percentage of (ACIP).MMWR 1997; 46:1-24.
the patients with reinfection was 0.51% (7/1366), 3. Hsueh PR, Chen HM, Lu YC, et al.
which was far less than that in other studies Antimicrobial resistance and serotype
(2.7-5.3%) [5-6, 15, 18]. This may be due to the distribution of Streptococcus pneumoniae
data collection period was short. In the future, we strains isolated in southern Taiwan. J
could retrieve cases by ICD-9 CODE from Formos Med Assoc 1996; 95:29-36.
Bureau of National Health Insurance to 4. Plouffe JF, Breiman RF, Facklam RR.
understand the reported rate regarding the Bacteremia with Streptococcus pneumoniae:
disease in the notifiable disease database, or try implications for therapy and prevention.
using Capture-Recapture  method to predict JAMA 1996; 275:194-8.
morbidity and understand the reporting 5. Turett GS, Blum S, Telzak EE. Recurrent
condition. pneumococcal bacteremia : risk factors and
Examining the bacterial serotypes from outcomes. Arch Intern Med 2001;
these seven reinfection patients, we discovered 61:2141-4.
that the onset of illness occurred most in winter 6. King MD, Whitney CG, Parekh FM, et al.
and spring seasons, so climate change was an Recurrent invasive pneumococcal disease:
important factor to the high risk group of people a population-based assessment. Clin Infect
with weaken immunity, which needs more Dis 2003; 37:1029-36.
attention. In this study, for the limitation of small 7. Hsueh PR, Luh KT. Treatment and
sample size, we did not see the high reinfection prevention of infections caused by
rate in children less than 2 years old , or high penicillin: resistant Streptococcus
mortality of patients with reinfection . More pneumoniae. J Formos Med Assoc 1991;
data are needed to understand the distribution of 1:57-65.
patients with reinfection at each age group. 8. Robinson KA, Baughman W, Rothrock G,
et al. Epidemiology of invasive
Acknowledgements Streptococcus pneumonae infections in the
We highly appreciate the efforts of United States, 1995-1998: opportunities for
colleagues from Taiwan CDC Research and prevention in the conjugate vaccine era.
Diagnostic Center and colleagues from local JAMA 2001; 285:1929-35.
health department who assisted in file retrieving, 9. Roche PW, Krause V, Cook H, et al.
so the study can proceed smoothly. Invasive pneumococcal disease in Australia,
2006. CDI 2008; 32:18-30.
References 10. Centre for Health Protection (Hong Kong).
1. Feigin RD, Cherry JD. Textbook of pediatric Commun Dis Watch. 2009; 6:21-22.
Vol.26 / No.2 Taiwan EB 34
11. Chen YY, Yao SM, Chou CY, et al. Miranda E, et al. Recurrent pneumococcal
Surveillance of invasive Streptococcus bacteremia: a warning of immunodeficiency.
pneumoniae in Taiwan, 2002–2003. J Med Arch Intern Med 1996; 156:1429-34.
Microbiol 2006; 55:1109-14.
12. Dowell SF, Whitney CG, Wright C, et al.
Seasonal patterns of invasive Investigation of Tuberculosis
pneumococcal disease. Emerg Infect Dis Cluster in a Keelung City Hospital
13. Ampofo K, Bender J, Sheng X, et al. Chin-Mei Liu1,2, Donald Dah-Shyong
Seasonal invasive pneumococcal disease in Jiang2,Ya-Jung Hu1, Hao-Shih Liu1
children: role of preceding respiratory viral Pei-Chun Chuang 3, Min-Ho Lai1
infection. Pediatrics 2008; 122: 229-37.
14. Coccia MR, Facklam RR, Saravolatz LD, et 1. First Branch, Centers for Disease Control,
al. Recurrent pneumococcal bacteremia: 34 Taiwan
episodes in 15 patients. Clin Infect Dis 1998; 2. Field Epidemiology Training Program,
26:982–5. Centers for Disease Control, Taiwan
15. Orlicek SL, Herrod HG, Leggiadro RJ, et al. 3. Research and Diagnostic Center, Centers for
Repeated invasive pneumococcal infections Disease Control, Taiwan
in young children without apparent
underlying immunodeficiency. J Pediatr Abstract
1997;130:284-8. In 2007 between October 3 and October 8,
16. Pastor P, Medley F, Murphy TV. Invasive a hospital in Keelung City reported four
pneumococcal disease in Dallas County, tuberculosis (TB) cases; three of them were
Texas: results from population-based workers of the pathology department.
surveillance in 1995. Clin Infect Dis 1998; Epidemiologic investigation of the relationship
26:590-5. between the patients, place, and time suggested
17. Harrison LH, Dwyer DM, Billmann L, et that case 1 might have transmitted the disease to
al. Invasive pneumococcal infection his spouse (case 2), who then transmitted it to
in Baltimore, MD: implications for two co-workers (cases 3 and 4). Cases 2, 3, and
immunization policy. Arch Intern Med 4 were the same cluster confirmed by restriction
2000; 160: 89-94. fragment length polymorphism (RFLP),
18. Font B, Llimiñana C, Fontanals D, et suggesting a TB cluster may have occurred in
al. Eleven-year study of recurrent the hospital. Epidemiologic evaluation and
pneumococcal bacteremia. Eur J Clin preventive measures were carried out by the
Microbiol Infect Dis 2001; 20:636-8. hospital and the Keelung City Health Bureau,
19. CDC. Completeness and timeliness of with assistance from the Tuberculosis Advisory
reporting of meningococcal disease-Maine, Team. The cluster was effectively contained and
2001-2006. MMWR 2009; 58:169-72. did not spread further in the hospital. All cases
20. Rodríguez-Créixems M, Muñoz P, completed their TB treatment on May 9, 2008.
35 Taiwan EB January26, 2010
Keyword:Tuberculosis, hospital cluster , induced coughing of hospital patients increase
restriction fragment length amount of TB droplets in air; improper personal
polymorphism(RFLP) protection facilitates transmission of the disease
[3,4]. Two sources provide most of the
Introduction TB bacterium found in a hospital:
Tuberculosis (TB) is an airborne infection already-diagnosed TB patients, and patients that
transmitted via droplets containing are later diagnosed as TB positive while in the
Mycobacterium tuberculosis from the TB hospital. The latter has a higher chance of
patient while talking, coughing, singing or causing a hospital TB infection . A hospital
laughing. The bacteria stay afloat in the air after in Taipei City had a TB outbreak in 2003, with
the droplets have dried off, and can cause an 60 cases across multiple departments . A
infection once they are inhaled and come into hospital in Yilan County also experienced a TB
contact with alveolar cells in the lung. Only outbreak in the respiratory care unit in
10-20% of the infected develop visible 2006.Twenty-two cases were reported, and nine
symptoms; fifty percent of them develop of them were of the same cluster identified by
symptoms within the first two years of initial molecular genotyping . Based on these two
infection. The rest are latent infection. nosocomial infection incidents, the difficulties
Transmission occurs most frequently between of detecting a hospital TB cluster and providing
persons with close contact, or living in the same early treatment were easily seen, emphasizing
household. the importance of TB infection control measure
Three primary factors influence in hospitals. This report described the events of
infection of Mycobacterium tuberculosis: 1) an early detection of a hospital TB cluster, and
Environmental factors: the amount of air its containment with proper infection control
flow and closed air-conditioning system. and managements.
Mycobacterium tuberculosis exists in the air as
droplets; higher concentration facilitates Background
transmission. It can also be transmitted by Hospital A in Keelung City has 1,097
air-conditioning system; longer exposure to beds, with inpatient, outpatient, and alternative
recirculated air contaminated with TB particles medication departments, to serve Keelung
leads to higher chance of infection . 2) residents. The hospital building has 13 floors,
Patient factors: patients with TB in the lungs, with centralized air circulation system. The
respiratory track, and throat, with symptoms pathology department is on the 7th floor, with 54
such as chronic coughing, positive sputum staff members, including four Environmental
diagnosis, positive chest X-ray diagnosis, Health Officials. Annual X-ray screening of all
immunosuppressive diseases such as AIDS, are staff member takes place between March and
more likely to transmit TB due to large amount May. All screening was normal this year with
of TB bacilli excreted . 3) Medical factors: the exception of four external contractors.
delayed diagnosis and incorrect medication can Investigation of the cluster
allow patients to continue spreading TB;
Vol.26 / No.2 Taiwan EB 36
The index case was a 59 year-old male Second-wave contact tracing chest X-ray was
patient with a history of minor stroke. Patient normal but all 3 sets of sputum tested positive
checked in at Hospital A on June 28, 2007 for for TB. Anti-TB treatment began on October 4,
coughing. Chest X-rays revealed pneumonia of 2007, coupled with home quarantine. The case
the lower-right lung; acid- fast stain of the was reported on October 8.
sputum tested negative for TB. Patient’s Local health bureau conducted chest X-ray
condition improved after antibiotic treatment and tuberculin test on case 3’s family (husband,
and was discharged on July 14, 2007. However, son and two grandchildren) and case 4’s family
sputum culture tested positive for TB two (husband, two sons), and all were negative.
months later. The hospital reported the case as Epidemiologic and infectious disease
TB on September 21, 2007. investigation lasted 99 days from the reporting
Case two was a 52 year-old female. She of index case to reporting of case 4. A total of
worked as a general staff member at Hospital four cases consisting of three hospital pathology
A’s pathology department, responsible for workers and one spouse were reported.
specimen collection, blood sample indexing and
organization, and cleaning of specimen Specimen collection and test results
containers. No disease history or symptoms After detection of case 2, 54 staff members
were noted. She was the spouse of the index of the pathology department (including four
case and lived together. Chest x-ray done for Environmental Health Officers) were sent for
contact tracing was negative, but patient chest X-ray, and 47 also submitted sputum
requested additional sputum tests. All three samples. Two specimens (cases 3 and 4) were
sputum sets tested positive for TB. Treatment TB positive after PCR testing, and their sputum
with anti-TB medication began on September smear and culture were also TB-positive.
27, 2007, coupled with home quarantine. The Mycobacterium tuberculosis was isolated and
case was reported on October 3. identified.
Case 3 was a 55-year old female. She was Following regulations concerning TB
also a general staff member at the pathology clustering events as outlined in the TB
department, and sit close to case 2. She had a Response Handbook established by the Taiwan
history of acute liver failure. No symptoms were Centers for Disease Control, Keelung City
observed at the time of contact tracing. Chest Health Bureau collected second sputum samples
X-ray revealed fibrosis tissues of the upper right from cases 2, 3 and 4, and sent to Taipei City
lung. However, all three sets of sputum tested Wan Fang Hospital for PCR test, sputum smear
positive for TB. Anti-TB treatment began on and culture. PCR for cases 2 and 3 were positive,
September 28, 2007, coupled with home and case 4 was not tested. Sputum smear and
quarantine. The case was reported on October 3. culture for all 3 cases were positive, with
Case 4 was a 39-year old female, mycobacterium tuberculosis isolated and
co-worker of cases 2 and 3. She had ankylosing identified. To determine if a TB cluster occurred,
spondylitis and rheumatoid arthritis and was on Keelung City Health Bureau sent samples from
long-term immunosuppressive medication. cases 2, 3 and 4 to CDC’s Center for Research
37 Taiwan EB January26, 2010
and Diagnostics for restriction fragment length
polymorphism (RFLP) test. All 3 cases were department and often had lunch together.
from the same cluster, confirming that a Cases 2 and 3 also shared office space with the
clustering event had occurred in the hospital. same air circulation system (Figure 2). Case 4
also had pre-existing autoimmune diseases.
Path of transmission These factors facilitated the transmission of TB.
Contact patterns, time and place for all 4 Results from RFLP showed cases 2, 3 and 4
cases were all epidemiologically linked (Figure were of the same cluster, confirming their
1). Record showed case 1 returned to hospital A infections were related. The suggested route of
in September for further testing, and sputum infection for this clustering event is as follows:
tests showed a high transmission level of 4+ (a case 2 contracted TB from case 1 while at home,
quantitative measure of the amount of and infected cases 3 and 4 at work. The CDC
Mycobacterium tuberculosis in the sputum, 4+ TB advisory team determined since cases 2, 3
having a high number of bacteria and highly and 4 were diagnosed early on, they were still in
transmissible). Case 2, spouse of case 1, was the incubation period and showed no external
likely to be infected by case 1, the index case. symptoms and no detectable chest X-ray
Cases 2, 3 and 4 were co-workers of the same abnormalities.
Vol.26 / No.2 Taiwan EB 38
Preventive measures team inspected laboratories, mycobacterium
Hospital A and Keelung City Health laboratory rooms, hospital environment and
Bureau initiated immediate preventive measures various staging areas on October 2, 2007, and
after the cluster. The hospital director called an found all to conform to regulations.
emergency meeting to decide on response Keelung City Health Bureau, after
actions. The following were also executed at the receiving reports of the TB clustering event
same time: hospital epidemiologic investigation, from the hospital on October 3, 2007,
sputum tests and chest X-rays for all pathology had conducted emergency epidemiologic
department staff member, medication and home investigation, contact tracing of confirmed TB
quarantine for those tested positive, reporting all cases, repeat sputum test, and submitted
TB cases to the health bureau and hosting information on all cases to the TB advisory
hospital briefing meetings, with chest and team for review in the October meeting. On
infectious diseases experts on hand to answer October 16, CDC first branch and two members
questions. Infection control personnel assisted of the TB advisory team visited the hospital to
in tracking the health of pathology department evaluate the situation and provided assistance as
staff member and those with close contact with needed. For this hospital TB cluster, the last
the staff member; chest X-rays for 2 years were case was reported on October 8, 2007, with all
also monitored. All hospital workers were surveillance completed on July 30, 2008. The
notified to exercise proper personal hygiene and hospital reported no new TB patients connected
managers were reminded to monitor workers’ to the event. The last case（case 4） completed
health status closely. The infection control treatment cycle on May 9, 2008.
39 Taiwan EB January26, 2010
comparison of the strain from cases 2, 3 and 4
Discussion with that of the index case via RFLP. Whether
Early diagnosis of TB cases greatly helps the index case’s sputum culture is the same as
the treatment of patients  and decreases the the others remains unknown. The sample from
chances of a hospital TB clustering event. The index case should have been preserved or
index case of this incident checked into hospital deliver to Taiwan CDC for testing. Also, cases 2,
A on June 28, 2007 for coughing. Two months 3 and 4 displayed no external TB symptoms;
later, the sputum culture tested TB positive. chest X-rays and CT scans revealed no
Sputum tests from the same patient taken during abnormalities in the lungs, yet sputum tests
the revisit in September tested to be 4+, a highly from all 3 cases tested strongly TB positive and
transmissible state. The hospital did not report the same cluster was isolated, suggesting cross
the case until September 21, three months after contamination may have occurred. However,
initial visit. The delay in diagnosis repeat samples were collected by 2 other health
unnecessarily increased the possibility of bureaus, using different containers from the first
transmitting TB to others . The CDC round of tests; repeat tests were also not
hospital TB prevention measures include conducted at hospital A. The same results from
3 levels: administrative management, different test facilities and newly collected
environmental control, and personal protection samples ruled out the possibility of cross
. During this TB clustering event, the contaminations during the first round of tests.
hospital called emergency meetings, decided on Research published by Pepper et al.  also
response measures (home quarantine, X-ray pointed out some patients with TB positive
screening of the pathology department staff sputum test may not show any abnormalities on
member, disinfection of the work environment, chest X-rays, especially those with autoimmune
etc.), which demonstrated competency in diseases.
handling the incidence. An outbreak often This cluster showed that hospital patients
causes panics; the hospital workers, despite have the potential to become the initial source
being medical professionals, displayed various of TB infection, and hospital workers can be
degrees of panic early in the incident, even easily infected and transmit disease to others.
leaking the information to the mass media. The Recommendations for hospital infection control
hospital hosted information meetings, measures should include implementation of
effectively decreasing the level of panic annual chest X-ray screening for all hospital
amongst workers. workers.
Because cases 2, 3 and 4 were diagnosed
chronologically close to each other, the health Conclusion
bureau was able to conduct repeat tests, obtain This incident was a TB cluster event at a
Mycobacterium tuberculosis isolates, and show Keelung City hospital. RFLP and epidemiologic
the 3 cases are epidemiologically related. studies confirmed this hospital TB cluster. One
However, the hospital does not keep staff member of the pathology department (case
Mycobacterium tuberculosis isolates, preventing 2) contracted the infection from her husband
Vol.26 / No.2 Taiwan EB 40
(case 1), then infected two of her co-workers. 5. Zhang JH, Wang FD. Pulmonary
One (case 3) was in the same office, and the tuberculosis and nosocomial infection.
other (case 4) had autoimmune disease. Notably, Infection Control 2005; 15: 286-92.
only the index case displayed external 6. Chou MY, Sun CC, Yeh PF, et
symptoms; the other 3 cases showed no al. Nosocomial transmission of
abnormalities on chest X-ray and CT scan. The Mycobacterium tuberculosis found through
hospital and local health bureau’s response screening for severe acute respiratory
protocols allowed for early detection of the syndrome --- Taipei, Taiwan, 2003.
cases, effectively containing the transmission. MMWR 2004; 53: 321-2.
Due to TB’s long incubation period, those 7. Hu YR, Wang JJ, Liu CM, et al.
infected may not develop symptoms Investigation of a suspected cluster of
immediately; therefore those in contact with the suspected cluster of in a respiratory ward
pathology department staff member should in a veteran hospital in I-Lan County.
monitor their health status closely for some Taiwan Epidemiol bull 2007; 23: 693-704.
period of time. 8. Taiwan CDC. Suspected cluster of
tuberculosis standard operation procedure；
Acknowledgement tuberculosis manual: section 12. Available
The authors would like to thank Keelung at :http://www.cdc.gov.tw/ct.asp?xItem=57
City Health Bureau for providing information 11&ctNode=1540&mp=230
and reports related to the hospital cluster 9. Okanurak K, Kitayaporn D, Akarasewi P.
incident. Factors contributing to treatment success
among tuberculosis patients: a prospective
Reference cohort study in Bangkok. Int J Tuberc
1. Taiwan CDC. Taiwan guidelines for TB Lung Dis 2008; 12: 1160-5.
diagnosis and treatment.2008; 1-6. 10. Chang RE. Patient and health system delay
2. Oeltmann JE, Varma JK, Ortega L, et al. in the diagnosis and treatment of
Multidrug-resistant tuberculosis outbreak tuberculosis in southern Taiwan. NTUR
among US-bound Hmong refugees, Unpublished Graduate Institute of Health
Thailand, 2005. EID 2008; 14: 1715-21. Care Organization Administration 2005.
3. Ohmori M. Factors for the onset of and the 11. Taiwan CDC. Nosocomial infection control
exacerbation of tuberculosis, recent guidelines of tuberculosis 2007: 14-6.
socio-medical characteristics of Available at:http://www.cdc.gov.tw/ct.asp?
tuberculosis and their perspectives in Japan. xItem=5713& ctNode=1540&mp=230.
Kekkaku 1999; 74: 759-66. 12. Pepper T, Joseph P, Mwenya C, et al.
4. Sayoki GM, Beatrice KM, Amos K, et al. Normal chest radiography in pulmonary
The magnitude and factors associated with tuberculosis: implications for obtaining
delays in management of smear positive respiratory specimen cultures. Int J Tuberc
tuberculosis in Dares Salaam, Tanzania. Lung Dis 2009;13:148-9.
BMC Health Serv Res 2008; 8: 158.