This is the first analysis of the prevalence and medical costs of childhood rotavirus AGE for both inpatients and outpatients in Taiwan. This analysis confirmed rotavirus as a leading cause of AGE in children <5 years old who sought medical care in Taiwan. Nearly one third (32.9%) of hospitalized AGE cases were caused by rotavirus. This finding is comparable to data reported from previous studies in other countries, which ranged from 25% to 50%. Rotavirus accounted for 30–50% of diarrhoeal hospitalization in <5 years old children in England [18
], 27% in diarrhoeal consultations in El Salvador [19
], and 42 % of diarrhoeal admission in Argentina [20
]. In New York, rotavirus infection was reported in >30% of diarrhoea cases in children younger than 5 years of age [21
In a separate study of 4 hospitals in Taiwan, Chen et al found that 43% of AGE cases in <5 years old children hospitalized were attributable to rotavirus [9
]. There are several explanations for the difference in our data (32.9%) and that of Chen et al. In two of the hospitals we studied, over 33% of the young children with AGE hospitalized were attributable to rotavirus. The lower proportion (18%) of the patients with rotavirus AGE at the third hospital is likely due to the fact that eastern region of Taiwan where the hospital is located is the least populated mountainous area and one where bacterial dysentery remains an important infectious disease. Most other parts of Taiwan, including where the other two hospitals we studies are located, are highly urbanized and more densely populated. The lower rate of rotavirus infections in the hospital in the east might have underrepresented rotavirus in that region thus lowering our estimation for the whole country. Another possible explanation is that not all children with AGE in the three hospitals we studied had stool specimens tested for microbial etiology. Different testing practices might have led to an underestimation of the incidence of rotavirus infection. We therefore used the 43% reported by Chen et al [9
] as the upper range and our 32.9% as the lower range to estimate the rotavirus disease burden in children hospitalized due to AGE in Taiwan.
The true incidence of rotavirus infections in the outpatient sectors is unknown because pediatricians tend not to perform rotavirus testing in typical cases of mild watery diarrhoea. This is especially true for the outpatient sector in Taiwan as evidenced by the fact that less than 3% of the outpatients with AGE were tested for etiology in the three hospitals we studied. We feel that the use of 15–35% range reported by Glass et al [13
] provides a good estimate for outpatient rotavirus disease burden in Taiwan since rotavirus was detected in 24% of the outpatients tested in these three hospitals.
Our results indicated that the disease burden of rotavirus infections could be higher in Taiwan than many other countries. An estimated one out of every 2 to 5 children younger than 5 years of age sought medical care due to rotavirus infections in Taiwan in 2001. The risks of children in the same age group requiring an outpatient visit due to rotavirus infection were estimated to be 1 out of 7 in the United States and 1 out of 5 globally [13
]. Our estimation of the annual incidence of hospitalization due to rotavirus infections of 1,528–1,997/100,000 in the <5 years old children in Taiwan is also higher than many countries, including Finland (610/100,000 in 1985–1995) [22
], and Australia (750–870/100,000 in 1991–1996) [23
]. We do not have any biological explanation for the high hospitalization rate due to rotavirus infections in Taiwan. An earlier report revealed the incidence of rotavirus disease is similar in children in different countries, regardless of the degree of development [1
]. The high hospitalization rate in Taiwan might be due to the easy accessibility to and relatively low cost of health care. With the support of the National Health Insurance system, the medical costs are affordable to most of the population. Children with moderate to severe diarrhoea with early signs of dehydration are frequently admitted to the hospitals in Taiwan.
Of noteworthy is that although valuable information can be obtained from the extensive database maintained by the Bureau of National Health Insurance (BNHI) in Taiwan, very few AGE cases were coded as enteritis due to rotavirus (ICD9-00861) in the BNHI database. Although children with gastroenteritis who have no pathogen isolated should be correctly coded as ICD 009 (ill-defined intestinal infections), the BNHI data showed that ICD code 558.9 (other and unspecified noninfectious gastroenteritis and colitis) was often used instead. Earlier studies by Nelson et al on rotavirus disease burden also incorporated codes 558.9 and 787.91 (diarrhoea) into their analysis of disease burden of rotavirus infections [11
]. There are several explanations for the non-specific ICD coding in Taiwan. Rotavirus AGE is clinically indistinguishable from AGE caused by other pathogens. Physicians in Taiwan also do not have the incentives to look for specific etiologic agents because insurance reimbursement is not higher for more specific diagnostic ICD coding. Stool testing is seldom done by practitioners or in local hospitals. Thus laboratory diagnosis and specific ICD coding should be encouraged to increase our understanding of the roles different etiologic agents play in causing diseases in Taiwan's patients.
We also showed that although the per-capita medical cost of rotavirus infection was lower in Taiwan than that of the United States and Hong Kong, the personal economic burden was similar among the three places when normalized for gross national incomes per capita. Therefore, people in countries with different degrees of development have similar economic burdens from rotavirus infections. Other than the direct medical costs, one must also take into account associated indirect costs including an out of pocket registration fee for each outpatient visit (US $1.5–3.0), extra diapers and laundry, special diets, transportation, lost of income from work, and additional child care needs resulting from children having rotavirus infections. Thus the social and economical burden of rotavirus infections could be in actuality much higher than the above-mentioned figures.