The screening program was undertaken from 2000 to 2002; 26 villages in 5 townships in Shiqu County, were included (Figure 1). A total of 3,199 volunteers were self-selected after the purpose of the study was explained to the communities by local village leaders; volunteers were assured free diagnosis and chemotherapeutic treatment for echinococcosis, if indicated. Study participants ranged in age from 1 to 86 years (median 32 years). Fifty-two percent (1,660) were female patients, and 48% (1,539) were male patients. Persons of Tibetan ethnicity comprised 95% of the sampled population. The other participants listed their ethnicity as Han (4.5%), Hui (0.2%), or other (0.3%). Almost half of the participants (52.9%) raised livestock, including yaks, sheep, or goats, as the primary source of their income. Other listed occupations included student (19.1%), public servant (9.8%), preschooler (3.2%), illiterate child (2.0%), semifarmer (2.5%), farmer (1.1%), employee (2.2%), or other (7.3%).

In this study, 84 (2.6%) of 3,199 participants examined by abdominal ultrasound scanning were determined to have an intrahepatic mass with a nonhomogeneous hyperechoic structure that contained scattered calcifications, and with irregular, poorly defined edges. A central necrotic cavity with a hypoechoic pseudoliquid structure and irregular borders was observed in 79 (2.5%) additional persons. In 53 of these participants the infiltrative lesions measured >10 cm in diameter and invaded or surrounded vascular structures, biliary structures, or both. In the other 26 persons, the lesions were nodular, were 3–5 cm in diameter at the longest dimension, and had calcifications. Calcified lesions, 1–3 cm in diameter, were observed in 20 persons. Thus, 163 persons were confirmed by ultrasound scanning to have AE infection, and 46 were suspected of having AE. Confirmatory serodiagnostic tests were performed in Japan and China, respectively. Serodiagnosis with the EgCF antigen in ELISA was positive in 93 of 94 persons with typical images of AE, 24 of 25 persons with nodular lesions, and 11 of 20 persons with calcified lesions. Additional serologic testing with the rEm18 antigen in ELISA and immunoblot was positive in 101 of 102 persons with typical images of AE, 16 of 25 with nodular lesions, and 8 of 14 with calcified lesions (Table 1). Therefore, positive confirmative serology in 35 study participants with a suspect AE image of a nodular lesion or calcified lesion indicated infection with AE. Another patient with a suspect AE image of a nodular lesion in the liver refused to give venous blood, so confirmative serologic tests could not be performed on him, and this case was not counted in the AE category. Thus, of 46 study participants with a suspect AE image, 35 were finally diagnosed as having AE. A total of 198 (6.2%) of 3,199 persons studied were determined to be infected with AE on the basis of abdominal ultrasound images and confirmatory serologic results; 15 (38.5%) of 39 infected persons had inactive, or abortive AE lesions. Ninety-five single AE lesions were located in the right hepatic lobe, and 31 were in the left hepatic lobe. Involvement of both right and left hepatic lobes by a single lesion was observed in 17 patients. In 55 cases, >2 distinct foci were observed.

In addition, an ultrasound image of CE in the liver was detected in 216 (6.8%) of 3,199 study participants examined. In 10 cases, ultrasound images showed unilocular, cystic lesions with uniform anechoic content, without visible cyst wall, all <5 cm; they were considered to be type CLs. Images characterized by unilocular, simple cyst with uniform anechoic content and visible cyst wall, some exhibiting a snowflake image (7 images <5 cm, 42 images ranging from 5 to 10 cm, and 26 images >10 cm) were observed in 75 patients; they were determined to be Type CE1(7 CE1s, 42 CE1m, 26 CE1l); In 54 patients, images exhibited multivesicular or multiseptate cysts with a wheel-like appearance; others displayed unilocular cysts with daughter cysts with a honeycomb appearance. Eight of these images were <5 cm, 16 images were 5–10 cm, and 30 images were >10 cm; all of these images belonged to type CE2 (8 CE2s, 16 CE2m, 30 CE2l). In 23 cases, images were characterized by anechoic content with detachment of laminated membrane from the cyst wall, visible as a water-lily design; some had a unilocular cyst containing daughter cysts, but the whole cyst form was less rounded. Five of these cysts were <5 cm, 13 cysts were 5–10 cm, and 5 cysts were >10 cm; all were confirmed to be type CE3 (5 CE3s, 13 CE3m, 5 CE3l). In 48 cases, cysts had hyperechoic degenerative contents without daughter cysts. Seventeen of these cysts were <5 cm, 19 were 5–10 cm, and 12 cysts were >10 cm; these images belonged to type CE4 (17 CE4s, 19 CE4m, 12 CE4). Cysts characterized by thick, calcified walls in an arch-shaped form with a cone-shaped shadow, were observed in 6 cases; 3 had images <5 cm, and 3 had cysts 5–10 cm in size; these were determined to be type CE5 (3 CE5s, 3 CE5m). In 18 cases, >1 cystic lesions were identified in the abdominal cavity in addition to the liver cysts. In 5 cases, additional cysts were found in the spleen; in 3 cases, additional cysts were found in the pelvic cavity; and in l case, a cyst was also found in the kidney. Serologic results in these study participants with CE at ultrasound examination are shown in Table 1. Serodiagnosis using the EgCF antigen in ELISA was negative in 16 of 161 persons with CE; 12 of 123 persons with CE were seropositive with rEm18 by ELISA and immunoblot (Table 1). No mixed infections were observed.

In this mass screening study of Tibetan communities, portable ultrasound examination combined with specific serologic tests was used for the diagnosis of both CE and AE. Survey results indicated that human echinococcosis is a serious public health problem for the inhabitants of this area, for whom a 12.9% overall prevalence was recorded. In comparison with reports on human echinococcosis in other areas, including other areas of China, the prevalence in northwest Sichuan Province was much higher for both CE and AE (1,3,12,13). The prevalence of CE was higher than in other recognized echinococcosis-endemic areas of the world, including North Africa, South America, Russia, and the Middle East (1,12,14,15). Previous ultrasound-based surveys for human AE have shown regional prevalences of <0.05% in continental Europe to 4% in Gansu Province in central northwest China (16,17). The most striking observation, however, was that both AE and CE were co-endemic in this area of Sichuan, with a prevalence of 6.8% for CE and 6.2% for AE. Only parts of Turkey, Central Asia, and Siberia have been identified as co-endemic for both human CE and AE (1,14).

In Shiqu County, China, analysis of human CE and AE indicated that prevalence of disease in female patients was significantly higher (14.7%) than in male patients (11.0%). According to traditional Tibetan custom, women are usually responsible for home chores, including feeding dogs, collecting yak dung for fuel, and milking livestock. Thus, women and girls may have more opportunity to be exposed to Echinococcus-infected dogs and the contaminated environment.

The infection prevalence for both CE and AE for persons in the age groups <20 years was markedly lower than those of other age groups. Prevalence reached a peak among the >50- to 60-year age group. The presence of CE or AE in persons as young as 4 and 8 years, respectively, indicates recent active transmission. In general, CE or AE infection increased with age. However, among persons >60 years of age prevalence of both AE and CE declined, a situation consistent with previous reports (4,15,18); this finding may be associated with early death of persons infected with forms of echinococcosis, particularly with AE. A recent analysis of the relative health impact of echinococcosis in these Tibetan communities showed that CE and AE caused an average of 0.8 disability-adjusted life years lost per person (19), which is an exceptional value.

This analysis showed that AE infection varied from 0% to 14.3% by village and that CE village prevalence ranged from 0% to 12.1%. A trend of gradual decrease in AE in villages from north to south (9.4% vs. 0.9% in the 5 townships surveyed) was observed.

Several factors may contribute to the high prevalence of human AE in this Tibetan population. High densities of small mammals are essential to maintaining the transmission cycle of E. multilocularis, and small mammal populations are also subject to ecologic changes, such as deforestation or pasture overgrazing (16,20–22). The involvement of dogs as well as foxes in transmission in eastern Tibet, together with lack of hygiene and probable contamination of the local peridomestic environment, seem to be additional major factors (23,24). For the 5 townships located in the central area of Shiqu County, the geographic conditions, apparent ecologic factors, life style, religion, livestock production, and dog ownership practices appear to be similar; however, human AE village prevalence was markedly variable. We had previously observed that local differences in small mammal abundance over time, possibly associated with overgrazing practices may contribute to variable township AE disease rates (22).

This survey disclosed that 86.8% of herdsmen were illiterate; 20% of them had either CE or AE disease. Consequently, improving the knowledge and awareness of the disease among the traditional nomadic population is imperative in any future control or prevention studies. Analysis indicated that both CE and AE risk was related to dog ownership (p<0.01), contact with dogs (p<0.01), source of drinking water, and general hygiene (p<0.01). While the role of domestic and working dogs as the major definitive host for E. granulosus is clear, such is not the case for E. multilocularis. Of particular interest therefore was the strong association between human AE risk and dog ownership or contact. Evidence from community studies in other parts of China (16), the United States (25), and Germany (26) increasingly show that the domestic dog plays a key role in the zoonotic risk for human AE.

Dogs are kept in large numbers by Tibetans and are used primarily to guard property and livestock. In this survey, 82.4% of the population owned dogs, and 21% owned >3 dogs. Buddhist practice forbids killing any animal, including dogs, and this practice leads to large numbers of stray dogs, which mainly gather around temples or townships, where they are fed by monks and herdsmen. Dogs also are predators of small mammals on adjacent pastures; these dogs are usually fed by herdsman with offal (including liver and lungs) of sheep and yaks during slaughtering season. Necropsy of intestines of stray dogs in 1995 in this region showed a 29.5% prevalence for E. granulosus and 11.5% for E. multilocularis (27,28). A recent diagnostic purgation study of dogs in this area demonstrated E. multilocularis prevalence of 12% and an E. granulosus prevalence of 8% (29). Foxes are the main sylvatic hosts of E. multilocularis, and both the Tibetan fox (Vulpes ferrilata) and the red fox (V. vulpes) are common on the Qinghai-Tibet plateau. A previous report showed a high prevalence of E. multilocularis in the Tibetan fox (59.1%) and red fox (57.1%) (28) in this area. Furthermore, Qiu et al. observed in 1995 the existence of E. strobilae in Tibetan foxes with morphologic characteristics distinct from E. multilocularis adults but considered it to be a variant of E. multilocularis. These specimens and new samples have been shown to be a new species of taeniid cestode belonging to the genus E. Rudolphi (30). However, whether the new species is involved in the transmission of a third form of human echinococcosis in this region has yet to be determined.