Skip Navigation

Epidemiology and Burden of Rotavirus Diarrhea in Thailand: Results of Sentinel Surveillance

  1. Chuleeporn Jiraphongsa1,
  2. Joseph S. Bresee9,
  3. Yaowapa Pongsuwanna2,
  4. Pipat Kluabwang3,
  5. Urai Poonawagul4,
  6. Pramote Arporntip5,
  7. Manas Kanoksil6,
  8. Nakorn Premsri7 and
  9. Utcharee Intusoma8

    for the Rotavirus Surveillance Project Thailand Study Group

    a
  1. 1 Bureau of Epidemiology, Ministry of Public Health, Nonthaburi
  2. 2 National Institute of Health, Department of Medical Science, Nonthaburi
  3. 3 Maesod Hospital, Tak Province
  4. 4 Prapokklao Hospital, Chanthaburi Province
  5. 5 Hadyai Hospital, Songkhla Province
  6. 6 Nongkhai Hospital, Nongkhai Province
  7. 7 Sakaeo Hospital, Sakaeo Province
  8. 8 Ramathibodi Hospital, Bangkok, Thailand
  9. 9 National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia
  1. Reprints or correspondence: Dr. Chuleeporn Jiraphongsa, Bureau of Epidemiology, Ministry of Public Health, Tiwanon Rd., Muang District, Nonthaburi, Thailand 11000 (chulee{at}health.moph.go.th).
  1. Presented in part: Third Workshop of the Asian Rotavirus Surveillance Network, Manila, The Philippines, October 2003 (abstracts not published).

Abstract

Diarrhea remains an important cause of morbidity and mortality among children in Thailand, with >1 million cases reported in 2002. In anticipation of the development of vaccines against rotavirus, we evaluated the disease burden associated with rotavirus infection in Thai children and evaluated the rotavirus serotypes now circulating in Thailand. Diarrhea surveillance was conducted at 6 Thai hospitals in different geographic areas. Community-based surveillance was conducted in Huaykrajao District, Kanchanaburi Province. During the 24 months of surveillance, 4057 children were admitted to the 6 participating hospitals, and 1950 stool samples were collected. Of these stool samples, 43% (838) were positive for rotavirus. All rotavirus-positive stool samples were evaluated to identify their serotypes; 54.8% of samples were of serotype G9, which was predominant each year. Other identified rotavirus serotypes included G2, G4, G1, and G3 (17.2%, 5.3%, 0.8%, and 0.1% of isolates, respectively). Approximately one-half of the children hospitalized with rotavirus diarrhea were <1 year old. community surveillance showed the proportion of cases of rotavirus diarrhea in the community to be much lower than that in the hospitalized population (12.2% vs. 43.0%).

In recent years, morbidity and mortality associated with diarrhea have decreased worldwide, concurrently with an increase in economic development. In spite of this decrease in morbidity and mortality, the burden of severe disease (i.e., disease that requires hospitalization) remains high [14]. In 1998, the first rotavirus vaccine was licensed in the United States [5] but was withdrawn in 1999 following reports of intussusception in vaccinated infants [6]. Subsequently, in February 2002, a World Health Organization committee met to deliberate the future of rotavirus vaccine research. This committee recommended continuation of research to develop a rotavirus vaccine, along with assessment of disease burden associated with rotavirus infection in developing countries, to enable thorough evaluation of the risks and benefits of future vaccines [7].

In Thailand, diarrhea remains an important cause of morbidity and mortality among children. Even though the diarrhea-related mortality rate has decreased, from 1.11 deaths/100,000 population in 1988 to 0.23 death/100,000 population in 2002, the diarrhea-related morbidity rate has remained stable (1488 and 1687 cases/100,000 population in 1993 and 2002, respectively) [8]. In 2002, a total of 1,055,393 cases of diarrhea were reported through the Thai National Notifiable Diseases Surveillance System. Of these cases, one-third occurred in children <5 years old, and 12% of these children required hospitalization (Bureau of Epidemiology, Nonthaburi, Thailand [unpublished data]). Because this surveillance system is passive, the true burden of diarrheal disease in Thailand is likely greatly underestimated.

The importance of rotavirus infection as a cause of diarrhea in children has been recognized by Thai clinicians for almost 30 years. A long-term study conducted during 1977–1980 detected rotavirus particles in 31% of 179 stool samples, by use of electron microscopy [9]. Since 1981, ∼17%–44% of stool samples obtained from Thai children hospitalized with diarrhea have been rotavirus positive, as determined by various testing methods [919]. The Bureau of Epidemiology retrospectively reviewed the logbooks of laboratory hospitals in 3 geographically distinct areas of Thailand to assess the disease burden associated with rotavirus infection. Of 2954 stool samples screened during 1987–1996, 27% were positive for rotavirus (N.P., unpublished data). These findings are consistent with previous reports and clearly demonstrate the significance of rotavirus diarrhea in Thailand.

Before a rotavirus vaccine program can be implemented, information is needed on the magnitude and trend of rotavirus infection and the rotavirus serotypes circulating in an area. Few laboratories in Thailand are capable of serotype identification; however, data from the Thai National Institute of Health (NIH; Nonthaburi) indicate that the dominant rotavirus serotypes are changing over time. Of samples serotyped during 1983–1984, 43% were of serotype 4; during 1987–1988, 59% of samples were of serotype 1 or 2; and, during 1988–1991, 51% of samples were of serotype 1 [2022].

Progress in the development of vaccines against rotavirus encouraged us to examine the disease burden associated with rotavirus infection among Thai children and to assess the need for such vaccines. A 2-year rotavirus surveillance project was initiated in 2001, with the primary goals of determining the prevalence of rotavirus diarrhea and of noting the changes in strain circulation over time. In addition, a small, community-based surveillance project was conducted during 2001–2002 to measure the incidence of rotavirus diarrhea in the community. Finally, evaluations of the costs associated with diarrheal disease were conducted in Chanthaburi Province in 2002; those results will be presented elsewhere. The surveillance project and substudies were approved by the Ethical Review Committee for Research in Human Subjects of the Ministry of Public Health (Nonthaburi, Thailand) and by the US Centers for Disease Control and Prevention.

Methods

Hospital Sentinel Surveillance

Surveillance sites

Surveillance of rotavirus diarrhea was conducted in 6 general hospitals located throughout Thailand: Maesod General Hospital, Tak Province (Myanmar border); Nongkhai General Hospital, Nongkhai Province (near the Laos People's Democratic Republic border); Ramathibodi Hospital, Bangkok; Chanthaburi Regional Hospital, Chanthaburi Province (near the Cambodia border); Hadyai Regional Hospital, Songkhla Province; and Sakaeo Hospital, Sakaeo Province (">@@@@figure 1). These hospitals have a total catchment population of ∼10 million people, including ∼1 million children <5 years old.

Figure 1

Locations of 6 surveillance sites in the Rotavirus Surveillance Project, Thailand, 2001–2003.

All children admitted to a participating hospital as a result of acute gastroenteritis for which the onset occurred ⩽7 days earlier were eligible for inclusion in the study. Surveillance staff reviewed daily admission logs to ensure that all potentially eligible children were identified. For each child, pediatric inpatient ward nurses completed a case-reporting form within 24 h of admission. Stool samples were collected by the ward nurses and were sent to reference laboratories at the Thai NIH, where they were tested for rotavirus. A confirmed case of rotavirus diarrhea was defined as any case of acute diarrhea for which the stool sample obtained was positive for rotavirus, as determined by PAGE. Careful assessments of the fraction of the enrolled eligible population, of the population with severe diarrhea, and of deaths due to diarrhea were also performed.

Laboratory procedures

Stool samples (⩾5 mL) were obtained during a patient's hospital stay, preferably within 24 h of admission, and were placed in a screw-top container. Stool samples were refrigerated at 4°C and were sent in a box each week to the Thai NIH, together with ice packs. At the NIH, viral RNA was extracted from 10% fecal suspension with disruption solution containing SDS, 2-mercaptoethanol, and EDTA and then with phenol-chloroform. The electropherotype of each virus was determined by PAGE and electrophoresis performed on 10% acrylamide gels (2-mm thick) for 3 h at 20 mA at room temperature. RNA segments were visualized by 11-mmol/L silver staining [10, 12, 15, 23]. Stool samples positive for rotavirus were further characterized by polymerase chain reaction (PCR). PCR was performed in 2 steps. In the first amplification step, cDNA corresponding to the full-length VP7 gene was amplified with a pair of primers corresponding to the 3′ and 5′ ends of the VP7 gene. The second amplification step used a mixture of primers specific to each of the variable regions of VP7 genes of serotypes 1–4, 8, and 9, paired with a primer to the 3′ end of the VP7 gene [13, 18].

Community Surveillance

Community-based surveillance was conducted in Huaykrajao District, Kanchanaburi Province, which has a annual birth cohort of 323 infants. Every Thai child born between 1 July 2000 and 28 February 2001 (aged 6–13 months at the start of the study) was eligible. All children were followed-up for 1 year (from 1 September 2001 through 31 August 2002). Each child's parent or caretaker received a small container and a plastic spoon and was visited at home by local surveillance staff. If a child developed diarrhea, the parents or caretaker collected a stool sample, which was sent to the Huaykrajao District Health Office in an ice bag. The stool sample was refrigerated if the parents or caretaker could not send it to the office until the next day. Local staff visited the households of the children every week, inquired about diarrhea symptoms that occurred during the previous 7 days, and collected rectal swab samples. The case definition of acute diarrhea was at least 3 loose stools within 24 h or any abnormal stools (e.g., watery, with mucous, or bloody). Each month, we ensured the regularity of household visits by randomly checking 20–30 cases listed on the household visit forms, which included the dates and times of visits and were signed by health workers at every household visit.

Stool samples were stored at 4°C at the Huaykrajao District Health Office. Rectal swab samples were kept at room temperature and were sent with the stool samples to the laboratory every Tuesday and Friday for rotavirus testing. Stool samples were tested for rotavirus antigen by use of a commercial ELISA (Rotaclone; Meridian Diagnostics) [24] at Chaokhunpaibun Hospital (the community hospital for the Huaykrajao area). Rectal swabs were cultured for Escherichia coli, Salmonella species, and Vibrio species at either Kanchanaburi Provincial Hospital or the Siamed Health Care Product Laboratory (Kanchanaburi).

Results

Hospital-based surveillance

Because not all surveillance sites could begin collecting data at the same time (February 2001), we report results for the last 24-month period (July 2001–June 2003). The proportion of eligible children enrolled initially was low (35.4% in 2001) but increased after 6 months and was stable after 1 year (62.4% in 2002 and 64.5% during the last 6 months of 2003). Study children did not vary from a matched population in terms of sex and date of hospital admission; however, the mean age and duration of hospitalization differed between the 2 groups. The mean (SD) age of enrolled children and of children who were not enrolled was 1.76 (1.09) years and 2.04 (1.32) years, respectively (P < .001). the mean (SD) duration of hospitalization for groups enrolled and not enrolled was 3.63 (4.23) days and 2.98 (2.30) days, respectively (P < .001).

During the 24 months of surveillance (July 2001–June 2003), 4057 children <5 years old were hospitalized at participating surveillance sites. The overall proportion of children enrolled in the rotavirus surveillance project was 57.5%. Of these children, 94.4% had stool samples collected. Of 1950 stool samples tested for rotavirus, 43.0% were positive for rotavirus. The proportion of rotavirus-positive stool samples tended to be higher during December through March (figure 2). The proportion of rotavirus-positive stool samples varied by site, from 40% to 50% (table 1). Among all 4057 hospitalized children, 20 died. Only 6 of those deaths were related to diarrhea. One of those 6 children who died had a stool sample obtained, but the sample was not positive for rotavirus.

Figure 2

Proportion of rotavirus-positive stool samples in Thailand, by month, during July 2001–June 2003

Table 1

Proportion of Thai children enrolled in the Rotavirus Surveillance Project Thailand and proportion of stool samples positive for rotavirus, by study site, during July 2001–June 2003.

Serotypes were determined for stool samples collected during July 2001–June 2003. Of 838 stool samples positive for rotavirus by PAGE, 54.8% of the samples were compatible with serotype G9, and 17.2%, 5.3%, 0.8%, and 0.1% were compatible with serotypes G2, G4, G1, and G3, respectively. In 2001, 2002, and 2003, serotype G9 was identified in 60.3%, 61.8%, and 42.1% of the isolates tested; however, 42.4% of the samples serotyped in 2003 were of serotype G2, compared with none of those serotyped in 2001 (table 2).

Table 2

Rotavirus serotype, by study year, in the Rotavirus Surveillance Project Thailand, during July 2001–June 2003.

With regard to age group, among children 6 months to 4 years old, approximately one-half of the cases of diarrhea were due to rotavirus infection; only one-fourth of children in other age groups had rotavirus-positive stool samples. Among the children with rotavirus-positive stool samples, 58% were <1 year old. When we compared the rotavirus-positive and -negative groups, children with rotavirus-positive stool samples were more likely to have watery diarrhea and vomiting.

Community-based surveillance

The 213 children in Huaygrajua District were followed up for 212 person-years (1 child moved from the area, and 1 child died). The children had 183 cases of acute diarrhea documented during the study period (incidence, 0.86 case/person-year). The incidence decreased with age, from 1.6 cases/child/year among children 6–12 months old, to 0.7 case/child/year among children 13–18 months old, to 0.4 case/child/year among children 19–24 months old.

Stool samples were obtained from 172 (94%) of the 183 children with diarrhea; rotavirus was detected in 21 stool samples (12.2%) (95% confidence interval [CI], 7.9%–18.3%). Bacterial cultures were performed on 129 (85%) rotavirus-negative stool samples, of which 9 (7%) were positive: 2 were positive for E. coli type 1, another 2 were positive for Salmonella serogroup B, and 1 each was positive for Salmonella serogroup D, Shigella boydii (group C), Shigella sonnei, Aeromonas hydrophila, and Edwardsiella tarda. No pathogen was detected in 83% of stool samples obtained from children with acute diarrhea.

Most children (94%) visited a health care service. Five children (3%) were hospitalized for diarrhea, 1 of whom died of aspiration during intubation because of sepsis. Two of 3 stool samples obtained from hospitalized children were positive for rotavirus (no stool sample was obtained from the child who died). From this community surveillance, we estimate the rate of hospital admissions due to diarrhea as 2.4 admissions/100 person-years (95% CI, 0.77–5.42 admissions/100 person-years) and the rate of hospitalization associated with rotavirus infection as 1.6 admissions/100 person-years. With respect to sex or breast-feeding status, there were no differences between children with rotavirus-positive or -negative stool samples.

After comparison with other databases, data from routine notifiable disease surveillance, and patient registries at community, provincial, and private hospitals, we concluded that our community surveillance missed ∼10% of incident cases. Thus, the incidence of acute diarrhea in the community should be estimated as at least 0.96 case/person-year for children aged 6–24 months.

Discussion

Rotavirus infection is a common cause of hospitalization in Thailand. In the present study, the proportion of rotavirus-positive children was slightly higher than that noted in other studies performed in Thailand. Because stool samples were obtained from 94.4% of enrolled patients, our results should be reliable. The proportion of children with rotavirus-positive stool samples in 2002 was 39.0%. During the first 6 months of 2002 and 2003, at total of 50.23% and 51.15% of stool samples, respectively, were rotavirus positive. It is likely that the continuation of this surveillance will provide a good indication of trends in rotavirus infection in Thailand. Seasonal trends in rotavirus infection, especially during the cool months (December through March), were similar to previous findings in Thailand and elsewhere [25, 26]. Ramathibodi Hospital had a higher proportion of rotavirus-positive children, likely because it is a medical school and reference hospital and, therefore, is likely to care for more severely ill children. The predominance of rotavirus G9 serotypes in 2001 and 2002, which is clearly different from the results of previous studies, underscores the importance of ongoing surveillance to determine the proper vaccine candidates for Thais.

To better define the entire disease burden associated with rotavirus infection in Thailand, we determined the incidence rates of diarrhea for which hospitalization was not required and compared our findings with those of previous studies. The findings were comparable. A 1999 survey by the Thai Department of Disease Control found that, among children <5 years old, the incidence rate of acute diarrhea was 0.93 case/person-year, which was lower than the incidence rates noted during 1995 and 1991 (1.34 and 1.33 cases/person-year, respectively) [2729]. The proportion of cases of diarrhea due to rotavirus in the community (12.2%) is clearly lower than that in hospitalized populations (43.0%). From this information, we might estimate that the proportion of cases of diarrhea due to rotavirus among outpatient children is within this range. However, in the community study, the number of children who accessed health services likely was high because parents took their children along with the stool samples to the health services.

Studies that compared the incidences of both acute and rotavirus diarrhea in the community found that rates of acute diarrhea correlated inversely with the socioeconomic status of a country (i.e., the incidence was lowest in developed countries), but the proportion of cases of rotavirus diarrhea among hospitalized children increased as the economic development of a country increased. For example, a study performed in Mexico during 1987–1988 reported that the incidence of severe acute diarrhea was 3.2 cases/person-year and that the proportion of rotavirus-positive stool samples was 9% [30]. A study performed in Canada in 1976–1979 found that the incidence of severe acute diarrhea was 1.2 case/person-year and that the proportion of rotavirus-positive stool samples was 23% [31].

Rotavirus infection is the primary cause of severe diarrhea in children <5 years old. In Thailand, the percentage of children with rotavirus diarrhea did not differ by geographic area. The incidence of acute diarrhea in the community was ∼1 case/person-year and decreased with patient age. In nonhospitalized children with diarrhea, rotavirus infection was also the main identified cause of diarrhea (12%), compared with bacterial diarrhea (7%). Children with rotavirus infection were more likely to have watery diarrhea and vomiting and to require hospitalization.

To estimate the burden of rotavirus diarrhea in Thailand, we used information from our hospital-based and community-based rotavirus surveillance and from national notifiable disease surveillance. We calculated the rate of hospitalization of children with rotavirus diarrhea. The rate was 11.28 admissions/1000 population or 56,453 episodes. For all patients with diarrhea, the risks of having rotavirus diarrhea, of requiring a health care visit, and of requiring hospitalization were 1 in 8, 1 in 36, and 1 in 85, respectively (figure 3).

Figure 3

Summary of disease burden associated with rotavirus infection in Thailand

One limitation of the present study is the possibility that the study population may not be representative of all 0–5-year-old Thai children, because we used data from only 6 general hospitals in 6 large population provinces. Each site was chosen to generally represent its region. Because of the limitations of local resources, community surveillance was performed in only 1 district in a region and did not include children <6 months old, for whom the incidence of rotavirus diarrhea is low. Finally, the use of PAGE as the primary rotavirus detection method, instead of ELISA, might be less feasible to conduct in other countries.

Rotavirus Surveillance Project Thailand Study Group Members

Members of the Rotavirus Surveillance Project Thailand study group include Chuleeporn Jiraphongsa, Kumnuan Ungchusak, Wanna Hanshaoworakul, Piyanit Tharmaphornpilas, Yongjua Laosiritaworn, Jirapat Kunlayanapojporn, and Amornrat Ngowabunpat (Bureau of Epidemiology, Ministry of Public Health, Nonthaburi, Thailand); Yaowapa Pongsuwanna, Krongkaew Supawat, and Ratigorn Guntapong (National Institute of Health, Department of Medical Science, Nonthaburi, Thailand); Pipat Kluabwang, Nutthawutse Chiciri, Sorn Khunpek, Orapin Boonchamoi, and Daungporn Prasertnoi (Maesod Hospital, Tak Province, Thailand); Urai Poonawagul, Pathanee Teirsonsern, Kanya Rungratanaubol, Neeracha Leesiriwatanakul, and Lukana Salidpaisan (Prapokklao Hospital, Chanthaburi Province, Thailand); Pramote Arporntip, Bancha Nillaesonk, Sukon Kaewphibool, Sukhontha Wichian, and Rujira Minchapat (Hadyai Hospital, Songkhla Province, Thailand); Manas Kanoksil, Surasak Kowasupatha, Nipon Sakulkou, Uthaiwan Kameraniya, and Ladda Yuwanich (Nongkhai Hospital, Nongkhai Province, Thailand); Nakorn Premsri, Jakkrarin Kummuna, Supatra Ponlumyai, Pennapa Prawatwong, and Samruay Ritboon (Sakaeo Hospital, Sakaeo Province, Thailand); Utcharee Intusoma, Rattana Sritongon, Pattra Nicrotha, Vacharee Inthraphuvasak, and Rujinee Juntraruksa (Ramathibodi Hospital, Bangkok, Thailand); and Chonlada Pirunjindarat, Kasem Wetsutthanon, Visudth Pinprachanunt (Kanchanaburi Provincial Health Office, Thailand).

Acknowledgments

We thank Pathom Sawanpanyalert (National Institute of Health, Department of Medical Science, Nonthaburi, Thailand), for performing laboratory tests for the detection of rotavirus and bacteria; Wandee Varavithaya (Ramathibodi Hospital, Bangkok, Thailand); Supamit Chunsuttiwat and Sirisak Warintrawat (Department of Disease Control, Ministry of Public Health, Nonthaburi), for serving as project advisors; and Tony Nelson (Asian Rotavirus Surveillance Network), for helping with preparation of the manuscript.

Footnotes

  • Financial support: Vaccine Program at Program for Appropriate Technology in Health (PATH; Cooperative Arrangement 00-GAT.770-790-01139-COP).

  • Potential conflicts of interest: none reported.

  • The views expressed herein do not necessarily reflect the views of PATH or of the Bill and Melinda Gates Foundation.

  • a Study group members are listed after the text.

References

| Table of Contents