Investigation of the use of serology and ultrasonography to detect hepatic cystic echinococcosis in Heilongjiang, China, using a Bayesian framework

Accepted Manuscript Title: Investigation of the use of serology and ultrasonography to detect hepatic cystic echinococcosis in Heilongjiang, China, using a Bayesian framework Author: Su Han Rui Chen Wenjuan Fang Rong Fu Jingshan Wen Xiaoli Zhang Xiao Ma PII: DOI: Reference:

S0001-706X(16)30477-6 http://dx.doi.org/doi:10.1016/j.actatropica.2016.07.002 ACTROP 3980

To appear in:

Acta Tropica

Received date: Revised date: Accepted date:

13-4-2015 7-7-2016 9-7-2016

Please cite this article as: Han, Su, Chen, Rui, Fang, Wenjuan, Fu, Rong, Wen, Jingshan, Zhang, Xiaoli, Ma, Xiao, Investigation of the use of serology and ultrasonography to detect hepatic cystic echinococcosis in Heilongjiang, China, using a Bayesian framework.Acta Tropica http://dx.doi.org/10.1016/j.actatropica.2016.07.002 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Investigation of the use of serology and ultrasonography to detect hepatic cystic echinococcosis in Heilongjiang, China, using a Bayesian framework

Short Title: Bayesian evaluation of the diagnostic methods for hepatic cystic echinococcosis Su Han

a, b☯

, Rui Chen

c, ☯

, Wenjuan Fang

d, ☯

, Rong Fu e, Jingshan Wen a, Xiaoli

Zhang a,*, Xiao Ma f,* a

Department of Parasitology, Harbin Medical University, Harbin, P. R. China

b

National Institute of Parasitic Diseases, Chinese Center for Disease Control and

Prevention; Key Laboratory of Parasite and Vector Biology, MOH, China; WHO Collaborating Center for Malaria, Schistosomiasis and Filariasis, Shanghai, PR China c

Department of Orthopedics, The Forth Affiliated Hospital of Harbin Medical

University, Harbin, P. R. China d

Department of Microbiology and Parasitology, Shanghai University of Medical ＆

Health Sciences, Shanghai, P. R. China e

Department of Epidemiology and health statistics, Public Health College, Fujian Medical University, Fuzhou, P.R. China

f

Department of Gastroenterology, The Second Affiliated Hospital of Harbin Medical

University, Harbin, P. R. China ☯These authors contributed equally to this work. * Corresponding

author:

Xiaoli Zhang, Department of Parasitology, Harbin Medical University, Harbin, 150081, China.

Tel: +86 451 86661047; Fax: +86 451 86661047.

E-mail: [email protected] Xiao Ma, The Department of Gastroenterology, The Second Affiliated Hospital of Harbin Medical University, Harbin, 150081, China E-mail: [email protected]

1

Graphical abstract

Highlights: 

Hepatic CE should not been ignored in Heilongjiang province.



This study is the first report of hepatic CE in Heilongjiang province in the last 10 years.



The Bayesian analysis was used to estimate test performance (sensitivity and specificity) and the prevalence of hepatic CE. The sensitivity and specificity of both US and ELISA were found to be high, and possibly that the predicted prevalence in outpatients was between 14 and 22%.



Serological methods could be a useful screening/surveillance tool amongst clinically suspected individuals.

2

ABSTRACT Background: Cystic echinococcosis (CE) is a public health problem in China. However, the prevalence and incidence of CE in Heilongjiang province is still poorly understood. Because there is no perfect gold standard, evaluation of ultrasound (US) and serological methods have been limited. This study evaluated the use of these two diagnostic methods for the diagnosis of CE in suspected cases. Methods: A total of 522 suspected hepatic CE patients, as well as their demographic and clinical features were collected and detected by immunoglobulin (IgG)-ELISA and US. The marginal posterior densities of sensitivity and specificity for both tests, and the prevalence of hepatic CE amongst participants, were estimated from the product of the likelihood function of observed and latent data by a Bayesian framework. Results: Most of the patients were from rural areas. The most common symptom was upper abdominal pain. The antibody-positive proportion determined by IgG-ELISA was 24.33% (127/522), significantly higher than with US examination (17.24%, 90/522). Bayesian analysis indicated that the estimated prevalence of CE amongst suspected cases was 17.70% (95% credible interval: 14.23-21.54%). The sensitivity and specificity of the ELISA test were 92.63% and 90.37%, and that of US were 93.05% and 98.44%, respectively. Among US-confirmed hepatic CE cases, the male to female ratio was 0.46 and the peak age group was 40-49 years. Conclusions: The results of the current study demonstrate that CE is present in Heilongjiang province. They also suggest that, whilst ultrasonography appears to be the detection modality of choice, serology may have a use for detection of infection in individuals suspected to be infected. This may have applications for surveillance within the province.

Keywords: Cystic echinococcosis, Diagnosis, Ultrasound, ELISA, Bayesian modeling.

3

1. Introduction Cystic echinococcosis (CE; also known as hydatidosis or hydatid disease) is one of the most geographically widespread zoonotic parasitic diseases, caused by the larval form of Echinococcus granulosus (Craig et al., 2003). Humans and intermediate hosts (commonly small ruminants) become infected by ingesting eggs which are passed in the feces of infected carnivores (Pleydell et al., 2008). The clinical manifestation of human CE ranges from asymptomatic infection to severe disease and possibly death (Li et al., 2011). CE also places an economic burden on patients, in terms of treatment costs and lost income (Benner et al., 2010; Budke et al., 2006). High prevalences of human infection have been reported in western and central Asia, China, and South America (Budke et al., 2006; Stojkovic et al., 2012). At present, CE is recognized as one of the Neglected Tropical Diseases (Junghanss et al., 2008) and represents an important public health problem worldwide, especially in China (Wang et al., 2008). Human CE is known to be widespread in 21 provinces of China, where it is especially prevalent in Qinghai, Gansu, Sichuan, Ningxia, Xinjiang, Inner Mongolia, Tibet, and Yunnan (Li et al., 2010). Additionally, CE has been the subject of major national infectious disease control programs in China since 2010. However, there have been few reports of CE in Heilongjiang province in the last 10 years. Methods of diagnosis of human CE have improved over the last few decades, including radiologic imaging, immunodiagnostics, and molecular biological techniques. However, in Heilongjiang province, molecular biological methods are expensive and impractical for point-of-care diagnosis, particularly in basic level hospitals. Therefore, radiologic imaging and immunodiagnostics are most commonly applied in this province. Imaging procedures such as ultrasound (US), computed tomography (CT), and magnetic resonance imaging (MRI) are used frequently for diagnosing CE. US has been used widely because it is simple, noninvasive, and cost-effective (Ozkol et al., 2005). However, the accuracy of US-based screening relies greatly on the skill of the ultrasonographer (Yu et al., 2008). In addition, some cyst stages may be difficult to 4

distinguish from non-parasitic cysts, which are common (Brunetti et al., 2011). The serum enzyme-linked immunosorbent assay (ELISA) is inexpensive and easy to conduct, but it has been reported to have low specificity, therefore leading to false-positive results (Carmena et al., 2006; 2007; Craig et al., 2007). Whilst the results of US examination, ELISA testing, surgical examination and tissue PCR are all potentially useful tools for the diagnosis of CE, it is necessary to quantify the limitations of these different diagnostic modalities. Bayesian methods are commonly used for validating diagnostic techniques and estimating prevalence in epidemiological studies of parasites, such as Schistosoma japonicum and Giardia duodenalis (Carabin et al., 2005; Geurden et al., 2004). In the absence of a perfect gold standard, the true prevalence and diagnostic test characteristics can be estimated successfully by using a Bayesian analysis (Joseph et al., 1995). With the aim of evaluating the accuracy of US and ELISA for diagnosis of hepatic CE and estimating the prevalence of hepatic CE amongst suspected cases in Heilongjiang province, the US and ELISA test results were evaluated in a Bayesian framework. Our results may help to improve the diagnostic accuracy, increase our knowledge of human CE in China and provide measures to facilitate the prevention and control of CE.

2. Materials and Methods 2.1 Ethics statement This study was approved by the Ethics Committee of Harbin Medical University. Written informed consent was received from all participants, including from parents/guardians of the children. The objectives, procedures and potential risks were explained to all participants. Confirmed CE cases were evaluated by a physician, and were offered suitable treatment according to the severity of the case (as assessed by cyst location, size, number, and stage of degeneration etc).

2.2 Study area and participants 5

Heilongjiang Province is situated in northeast China, with an area of approximately 454,000 km2. It has a population of approximately 38 million, comprising Han, Hui, Manchus, Koreans, and Mongols. Agriculture and stock raising are the major economic activities. All study participants were suspected of having CE upon presentation at a medical establishment, as determined by the national standardized diagnostic criteria for CE, published by the Ministry of Health of China (WS257-2006). These criteria include any individual with an appropriate epidemiological history, along with either clinical signs or suspicious lesions on abdominal imaging. An epidemiological history which would be suspicious for CE is any history of exposure to dogs, cattle, sheep and other domestic animals from endemic areas. Clinical manifestations of hepatic CE were those signs associated with hepatic space-occupying lesions or cyst rupture or leakage. A total of 522 suspected hepatic CE patients were collected from the parasite institute of Harbin Medical University and The Second Affiliated Hospital of Harbin Medical University in Heilongjiang Province from 2003 to 2012. All patients underwent US examination and serology at the time of initial assessment.

2.3 Ultrasound (US) examination A Voluson E8 (GE Healthcare, Milwaukee, WI, USA) ultrasound machine with a 3.5MHz real-time sector probe was used throughout the survey. Hepatic CE cysts were differentiated from other cystic lesions if one or more of the characteristic diagnostic criteria of hepatic CE were present, namely, a laminated membrane and/or daughter cysts (Shambesh et al., 1999, Gharbi et al., 1981). The liver, pancreas, spleen, and kidneys were examined carefully and any pathological cystic images were recorded. Cysts due to CE were differentiated from other cystic lesions if one or more of the characteristic diagnostic criteria of CE were present: specifically, a laminated membrane and/or daughter cysts. All identified hepatic CE cases were differentiated from potential AE cases based upon US morphology. The alveolar echinococcosis patients were not found. 6

2.4 Serological analysis A commercially available ELISA diagnostic kit for the diagnosis of human echinococcosis (Shenzhen Combined Biotech Co. Ltd, Guangdong, China) was used for serological analysis. This is an indirect ELISA which uses partially purified recombinant antigen B subunits derived from human hydatid cyst fluid. A total of 5 mL peripheral blood was collected from all patients and the serum stored at -20°C until tested. The absorbance at 405 nm was measured using a microplate reader. The threshold for positivity was calculated using the mean optical density (OD) value plus three standard deviations of negative control sera.

2.5 Statistical analysis Statistical analyses were performed using SPSS software (ver. 10.0; SPSS, Chicago, IL). Associations between a number of variables of interest and US-confirmed hepatic CE were investigated using univariable and multivariable methods. Age, gender, living environment, and a number of occupational risk factors were considered as variables of interest, and a chi-square test was used to identify potential associations. Unconditional multivariable analysis was used to calculate odds ratios (ORs) for the association between variables of interest and US-confirmed hepatic CE, and the corresponding 95% confidence intervals (CIs). P-values < 0.05 were considered to indicate statistical significance. All variables found to be associated with US-confirmed CE status in the univariable analysis (P < 0.05) were then added to a logistic regression model, and were removed using a stepwise process in order to develop a final model of variables associated with CE. In the absence of a perfect gold standard, a Bayesian approach was designed to evaluate diagnostic-test accuracy and the prevalence (Branscum et al., 2005). According to Bayes’ theorem, the joint posterior distribution is proportional to the product of the likelihood function and prior distribution. Using this theorem, the marginal probability density distributions of the test sensitivity and specificity parameters can be obtained from the likelihood function of the test outcomes and 7

'prior' estimates of the distribution of test sensitivity and specificity. Then, the Gibbs sampler algorithm could provide random samples from marginal posterior densities of each parameter of interest. Based on these random samples, posterior means can be constructed and used as point estimates for sensitivity and specificity, reported with 95% Bayesian credible intervals (Bayesian analogue of confidence intervals) (Joseph et al., 1995, Tarafder et al., 2010). The Bayesian model described by Johnson and others for two tests and one population (Johnson et al., 2001) was used to estimate the sensitivity and specificity of ELISA and US for diagnosis of hepatic CE, and to estimate the prevalence of CE in the study population. Since the model lacks identifiability when applied to a single population, extra attention is paid to the formulation of the priors for the test parameters. These were estimated from a review of the literature (Macpherson et al., 2003; Budke et al., 2006; Torgerson et al., 2009; Zhang W et al., 2003;Carmena D et al., 2006), and were parameterized using Beta distributions (which are bounded by 0 and 1, and are therefore suited for modeling proportions). Since the ELISA test and US examination measure different biological outcomes, the conditional independence of the tests were assumed according to infection status (Georgiadis et al., 2003). The freeware program WinBUGS 1.4 was used to run all models using Gibbs sampling. A Markov-chain Monte Carlo simulation was conducted to estimate the mean and 95% credible intervals for all parameters from the respective posterior distributions. The Markov chain was run for 155,000 iterations, with the first 5,000 discarded as burn-in.

3. Results 3.1 US and immunological examinations Among the 522 study participants, 127 (24.33%) were determined positive by ELISA and 90 (17.24%) were determined as positive by US, respectively. A total of 82 (15.71%) individuals tested positive using both methods (Table 1). 3.2 Bayesian analysis 8

The Bayesian model is shown in the Additional files 1. The WinBUGS code is available on request from the corresponding author. Table 2 also shows the properties of the two tests according to the Bayesian analysis. The mean sensitivity and specificity of the IgG-ELISA were 92.63%, 90.37%, and of US were 93.05%, 98.44%, respectively. In addition, the Bayesian analysis showed that the mean positive predictive value (PPV) and negative predictive value (NPV) of the IgG-ELISA were 67.38% and 98.27%, and of US were 92.68% and 98.49%, respectively. The prevalence of hepatic CE evaluated by the Bayesian method was 17.70% (95% CI 14.23-21.54%).

3.3 Demographic features of US-confirmed hepatic CE patients The prevalence of hepatic CE amongst those living in rural areas was higher than that amongst those living in cities, as detected by both methods (P < 0.05) (Table 3). Table 4 shows the demographic and clinical characteristics of the US-confirmed cases. The 90 patients consisted of 26 males (29%) and 64 females (71%), and the sex ratio was 0.41. The age at first detection ranged from 5 years to 83 years, with most people being between 40 and 49. In terms of the occupations of hepatic CE patients, we found that farm laborers accounted for the majority (74%), followed by workers (7%), and others, e.g., students, self-employed and teachers (19%). Most of the patients (63, 70%) came from rural areas. Univariable and multivariable logistic regression analysis found gender, age, and environment (urban/rural) to be significantly associated with US-confirmed hepatic CE (tables 5 and 6). However, worker and farmer were not significantly associated with hepatic CE. In the logistic regression, female was inversely associated with hepatic CE; every 10-year increase of age was inversely associated with hepatic CE; and rural factor was positively associated with hepatic CE in the same model. 3.4 Clinical features of hepatic CE patients According to the case definition, all patients presented with typical clinical hepatic CE symptoms. Upper abdominal pain was the most common symptom 9

(observed in 52 of the 90 individuals: 58%), and vomiting, hepatomegaly, and weakness were also described (Table 4). Six patients (7%) also presented with non-specific and subtle signs and symptoms, such as fever and the gradual development of waist/abdominal masses accompanied by lower back pain. Among all US-confirmed patients, 24 (27%) affected the hepatic left lobe, 54 (60%) affected the hepatic right lobe, and 12 (13%) affected both lobes. In addition, 4% had multiple CE lesions in extrahepatic organs, such as the kidney, and spleen. Moreover, a total of 21 (23%) patients underwent surgery for hepatic CE.

3.5 Regional distribution of hepatic CE patients The hepatic CE cases were detected by both methods in nine administrative regions in Heilongjiang province. Most patients came from SuiHua, Harbin, and DaQing, whereas no cases were found in DaXingAnLing, JiaMuSi, MuDanJiang, or JiXi (Table 7, Fig.1). Additionally, 11 patients came from outside Heilongjiang province. There was no statistically significant difference among regions (P > 0.05, χ2=5.4167). 4. Discussion CE is known to have a widespread global distribution (Giri & Parija, 2012) and China is one of the most important endemic regions of CE. However, the public health importance of CE remains underappreciated in some areas of China (Wang et al., 2008). Because early and most long-term infection with CE is often asymptomatic, many cases are identified incidentally in imaging studies performed for other reasons (Wen et al., 2010). The majority of infections remain undiagnosed and underreported. Thus, active identification of human echinococcosis cases is a necessary step for accurately evaluating the true extent of the disease (Yang et al., 2006). Diagnosis is a basic step in studies of echinococcosis. At present, many methods are used for the diagnosis of CE in humans, but a universal gold-standard is lacking (Barnes et al., 2006; Torgerson & Deplazes, 2009). Although surgery can help to confirm some cases of CE, not all patients accept operations. Bayesian approaches are commonly used to estimate the performance of tests for many parasitic diseases in the 10

absence of a gold standard (Han et al., 2012; Torgerson et al., 2009). Thus, in the present study, the Bayesian analysis was used to estimate the test performance and the prevalence of hepatic CE. In the hands of an experienced operator, US can be effective for hepatic CE diagnosis and can allow the quantification of cyst numbers, sizes, and stage of development (Nunnari et al., 2012; Yang et al., 2008). However, US examination is likely to miss very small cysts (Brunetti et al., 2011;

Shambesh et al., 1999), early

lesions and extraabdominal lesions, leading to a low reported US prevalence. Also, US may not be able to differentiate hydatid cysts from other space-occupying lesions (Nunnari et al., 2012). Although additional imaging techniques such as MRI and CT scans could be employed to improve the diagnosis, these techniques are expensive and restricted to use in hospital settings. Furthermore, inter- and intra-observer variability and differences in prevalence will affect the performance of US in different endemic settings (Macpherson & Milner, 2003). In this study, eight patients were recognized as US-positive but seronegative. This may be attributed to insufficient stimulation of the immune system due to other clinical variables related to parasite activity (Barnes et al., 2006). IgG antibody production depends on many factors, such as cyst number, size, location, and stage (Moro & Schantz, 2009). Furthermore, serological test sensitivity is inversely related to the degree of sequestration of the echinococcal antigens inside cysts (Nunnari et al., 2012). No production of antibodies has been reported when small cysts, intact cysts, and other highly calcified cysts exist (Gavidia et al., 2008). However, immunological methods, such as serum ELISA, have been used for the diagnosis of CE, because of advantages in terms of the collection, storage and transport of blood/serum (Zhang et al., 2012). This study showed that the ELISA prevalence (24.33%) was significantly higher (P < 0.05) than the US prevalence (14.75%). The specificity of the IgG-ELISA method was not as high (90.37%) as that of US (98.44%). It may be likely that there were false-positives due to cross-reactions with other parasitic infections. Some patients may be included in this category. The sensitivity and specificity of the ELISA 11

is relatively high despite coming from quite a broad prior. As both tests were applied to clinically suspected patients, the results described here only relate to this study population and cannot be extrapolated to the general population at large. It has been reported that immunodiagnosis of CE will have a higher sensitivity in a hospital setting, where CE patients are usually in an advanced condition, compared with its direct application in endemic communities (Shambesh et al., 1999). These results suggest that although US has a higher sensitivity and specificity, serological methods could provide a useful screening/surveillance tool amongst clinically suspected individuals, especially in cases where US testing is not available. Although the current study provides some useful information about the presence of CE in Heilongjiang, it should be remembered that it was based on passive surveillance data collected from hospitals, in which all participants had clinical signs suggestive of CE. Thus, there is a high likelihood of selection bias, so the results cannot be extrapolated to the general population of Heilongjiang. However, the data presented here demonstrates that further surveillance for this disease may be beneficial. The livestock population in Heilongjiang is increasing and more people are becoming involved in livestock management. Also, ownership of domestic dogs (both pets and working dogs) is common. These trends could result in an increased risk of human exposure to E. granulosus in the coming years. Whilst attention should be paid to disease prevention activities (such as education of livestock and dog owners, and praziquantel dosing of domestic dogs), the development of a provincial surveillance system would provide more information on the current situation regarding echinococcosis in the province. Provincial surveillance could mainly include: conducting epidemiological investigations to determine the epidemic scope and degree, according to local infection in hepatic hydatidosis cases; intermediate host control, such as registration of dogs and strengthening the management of dog infection (by administering praziquantel); strengthening livestock immunity; developing livestock slaughter quarantine procedures and pathological viscera disposal; providing health education to the public; and active treatment for patients.

12

Human CE is present in Heilongjiang province and should not be ignored, especially with the potential for increased circulation associated with changes in livestock and domestic dog ownership. The current study evaluated two tests for human CE in the absence of a gold standard test, which suggested that both US and ELISA had high sensitivities and specificities when applied to suspected cases of CE. Whilst US is likely to remain the diagnostic method of choice for hepatic CE, serological methods could provide a useful screening/surveillance tool in certain cases (for example, amongst clinically suspected individuals). Further work is needed to evaluate the level of echinococcosis in the province as a whole, and it is therefore suggested that a provincial surveillance scheme is developed in order to collect data from all potential hosts (humans, livestock, and domestic dogs) and better quantify the risk of human infection in Heilongjiang province.

Acknowledgements This work was supported by China Postdoctoral Science Foundation (Grant no. 2014M550786) and (Grant no. 2015T80117), National Natural Science Foundation of China (Grant no. 81401684) and Natural Science Youth Foundation of Heilongjiang Province (Grant no. QC2014C093). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References BARNES, T. S., DEPLAZES, P., GOTTSTEIN, B., JENKINS, D. J., MATHIS, A., SILES-LUCAS, M., TORGERSON, P. R., ZIADINOV, I. And HEATH, D. D. 2006. Challenges For Diagnosis And Control Of Cystic Hydatid Disease. Acta Trop, 123, 1-7. BENNER, C., CARABIN, H., SANCHEZ-SERRANO, L. P., BUDKE, C. M. And CARMENA, D. 2010. Analysis Of The Economic Impact Of Cystic Echinococcosis In Spain. Bull World Health Organ, 88, 49-57. BRANSCUM, A. J., GARDNER, I. A. And JOHNSON, W. O. 2005. Estimation Of 13

Diagnostic-Test Sensitivity And Specificity Through Bayesian Modeling. Prev Vet Med, 68, 145-163. BRUNETTI, E., GARCIA, H. H. And JUNGHANSS, T. 2011. Cystic Echinococcosis: Chronic, Complex, And Still Neglected. Plos Negl Trop Dis, 5, E1146. BUDKE, C. M., DEPLAZES, P. And TORGERSON, P. R. 2006. Global Socioeconomic Impact Of Cystic Echinococcosis. Emerg Infect Dis, 12, 296-303. CARABIN, H., BALOLONG, E., JOSEPH, L., MCGARVEY, S. T., JOHANSEN, M. V., FERNANDEZ, T., WILLINGHAM, A. L. And OLVEDA, R. 2005. Estimating Sensitivity And Specificity Of A Faecal Examination Method For Schistosoma Japonicum Infection In Cats, Dogs, Water Buffaloes, Pigs, And Rats In Western Samar And Sorsogon Provinces, The Philippines. Int J Parasitol, 35, 1517-1524. CARMENA, D., BENITO, A. And ERASO, E. 2006. Antigens For The Immunodiagnosis Of Echinococcus Granulosus Infection: An Update. Acta Trop, 98, 74-86. CARMENA, D., BENITO, A. And ERASO, E. 2007. The Immunodiagnosis Of Echinococcus Multilocularis Infection. Clin Microbiol Infect, 13, 460-475. CRAIG, P. S., MCMANUS, D. P., LIGHTOWLERS, M. W., CHABALGOITY, J. A., GARCIA, H. H., GAVIDIA, C. M., GILMAN, R. H., GONZALEZ, A. E., LORCA, M., NAQUIRA, C., NIETO, A. And SCHANTZ, P. M. 2007. Prevention And Control Of Cystic Echinococcosis. Lancet Infect Dis, 7, 385-394. CRAIG, P. S., ROGAN, M. T. And CAMPOS-PONCE, M. 2003. Echinococcosis: Disease, Detection And Transmission. Parasitology, 127 Suppl, S5-20. GAVIDIA, C. M., GONZALEZ, A. E., ZHANG, W., MCMANUS, D. P., LOPERA, L., NINAQUISPE, B., GARCIA, H. H., RODRIGUEZ, S., VERASTEGUI, M., CALDERON, C., PAN, W. K. And GILMAN, R. H. 2008. Diagnosis Of Cystic Echinococcosis, Central Peruvian Highlands. Emerg Infect Dis, 14, 260-266. 14

GEORGIADIS, M. P., JOHNSON, W. O., GARDNER, I. A. And SINGH, R., 2003. Correlation Adjusted Estimation Of Sensitivity And Specificity Of Two Diagnostic Tests. Appl. Stat. 52, 63–76. GEURDEN, T., CLAEREBOUT, E., VERCRUYSSE, J. And BERKVENS, D. 2004. Estimation Of Diagnostic Test Characteristics And Prevalence Of Giardia Duodenalis In Dairy Calves In Belgium Using A Bayesian Approach. Int J Parasitol, 34, 1121-1127. GHARBI,H.A., HASSINE,W., BRAUNER,M.W.And DUPUCH, K. 1981.Ultrasound Examination Of The Hydatic Liver.Radiology. 139,459-63. GIRI, S. And PARIJA, S. C. 2012. A Review On Diagnostic And Preventive Aspects Of Cystic Echinococcosis And Human Cysticercosis. Trop Parasitol, 2, 99-108. HAN, S., ZHANG, X., WEN, J., LI, Y., SHU, J., LING, H. And ZHANG, F. 2012. A Combination Of The Kato-Katz Methods And ELISA To Improve The Diagnosis Of Clonorchiasis In An Endemic Area, China. Plos One, 7, E46977. ITO, A. 2002. Serologic And Molecular Diagnosis Of Zoonotic Larval Cestode Infections. Parasitol Int, 51, 221-235. JOSEPH, L., GYORKOS, T. W. And COUPAL, L. 1995. Bayesian Estimation Of Disease Prevalence And The Parameters Of Diagnostic Tests In The Absence Of A Gold Standard. Am J Epidemiol, 141, 263-272. JUNGHANSS, T., DA SILVA, A. M., HORTON, J., CHIODINI, P. L. And BRUNETTI, E. 2008. Clinical Management Of Cystic Echinococcosis: State Of The Art, Problems, And Perspectives. Am J Trop Med Hyg, 79, 301-311. LI, T., CHEN, X., ZHEN, R., QIU, J., QIU, D., XIAO, N., ITO, A., WANG, H., GIRAUDOUX, P., SAKO, Y., NAKAO, M. And CRAIG, P. S. 2010. Widespread Co-Endemicity Of Human Cystic And Alveolar Echinococcosis On The Eastern Tibetan Plateau, Northwest Sichuan/Southeast Qinghai, China. Acta Trop, 113, 248-256. LI, T., ITO, A., PENGCUO, R., SAKO, Y., CHEN, X., QIU, D., XIAO, N. And CRAIG, P. S. 2011. Post-Treatment Follow-Up Study Of Abdominal Cystic 15

Echinococcosis In Tibetan Communities Of Northwest Sichuan Province, China. Plos Negl Trop Dis, 5, E1364. MACPHERSON, C. N. And MILNER, R. 2003. Performance Characteristics And Quality Control Of Community Based Ultrasound Surveys For Cystic And Alveolar Echinococcosis. Acta Trop, 85, 203-209. MCMANUS, D. P., ZHANG, W., LI, J. And BARTLEY, P. B. 2003. Echinococcosis. Lancet, 362, 1295-1304. MORO, P. And SCHANTZ, P. M. 2009. Echinococcosis: A Review. Int J Infect Dis, 13, 125-133. NUNNARI, G., PINZONE, M. R., GRUTTADAURIA, S., CELESIA, B. M., MADEDDU, G., MALAGUARNERA, G., PAVONE, P., CAPPELLANI, A. And CACOPARDO, B. 2012. Hepatic Echinococcosis: Clinical And Therapeutic Aspects. World J Gastroenterol, 18, 1448-1458. OZKOL, M., KILIMCIOGLU, A. A., GIRGINKARDESLER, N., BALCIOGLU, I. C., SAKRU, N., KORKMAZ, M. And OK, U. Z. 2005. A Discrepancy Between Cystic Echinococcosis Confirmed By Ultrasound And Seropositivity In Turkish Children. Acta Trop, 93, 213-216. PLEYDELL, D. R., YANG, Y. R., DANSON, F. M., RAOUL, F., CRAIG, P. S., MCMANUS, D. P., VUITTON, D. A., WANG, Q. And GIRAUDOUX, P. 2008. Landscape Composition And Spatial Prediction Of Alveolar Echinococcosis In Southern Ningxia, China. Plos Negl Trop Dis, 2, E287. SHAMBESH, M.A., CRAIG, P.S., MACPHERSON, C.N., ROGAN, M.T., GUSBI, A.M. And ECHTUISH, E.F.1999. An Extensive Ultrasound And Serologic Study To Investigate The Prevalence Of Human Cystic Echinococcosis In Northern Libya. Am J Trop Med Hyg. 60,462-468. STOJKOVIC, M., ROSENBERGER, K., KAUCZOR, H. U., JUNGHANSS, T. And HOSCH, W. 2012. Diagnosing And Staging Of Cystic Echinococcosis: How Do CT And MRI Perform In Comparison To Ultrasound? Plos Negl Trop Dis, 6, E1880. TAO, S., QIN, Z., HAITAO, L., LEI, Y., LANHUI, Y., QIN, X., YONGQUAN, L. 16

And HAO, W. 2012. Evaluation Of Color Doppler Ultrasonography In Diagnosing Hepatic Alveolar Echinococcosis. Ultrasound Med Biol, 38, 183-189. TARAFDER,M. R., CARABIN, H,, JOSEPH, L., BALOLONG,E J.R., OLVEDA, R. And MCGARVEY, ST. 2010. Estimating The Sensitivity And Specificity Of Kato-Katz Stool Examination Technique For Detection Of Hookworms, Ascaris Lumbricoides And Trichuris Trichiura Infections In Humans In The Absence Of A 'Gold Standard'. Int J Parasitol. 40:399-404. TORGERSON, P. R. And DEPLAZES, P. 2009. Echinococcosis: Diagnosis And Diagnostic Interpretation In Population Studies. Trends Parasitol, 25, 164-170. TORGERSON, P. R., ROSENHEIM, K., TANNER, I., ZIADINOV, I., GRIMM, F., BRUNNER, M., SHAIKEN, S., SHAIKENOV, B., RYSMUKHAMBETOVA, A.

And

DEPLAZES,

P.

2009.

Echinococcosis,

Toxocarosis

And

Toxoplasmosis Screening In A Rural Community In Eastern Kazakhstan. Trop Med Int Health, 14, 341-348. WANG, Z., WANG, X. And LIU, X. 2008. Echinococcosis In China, A Review Of The Epidemiology Of Echinococcus Spp. Ecohealth, 5, 115-126. WEN, H., AJI, T. And SHAO, Y. M. 2010. Diagnosis And Management Against The Complications Of Human Cystic Echinococcosis. Front Med China, 4, 394-398. YANG, Y. R., CRAIG, P. S., VUITTON, D. A., WILLIAMS, G. M., SUN, T., LIU, T. X., BOUFANA, B., GIRAUDOUX, P., TENG, J., LI, Y., HUANG, L., ZHANG, W., JONES, M. K. And MCMANUS, D. P. 2008. Serological Prevalence Of Echinococcosis And Risk Factors For Infection Among Children In Rural Communities Of Southern Ningxia, China. Trop Med Int Health, 13, 1086-1094. YANG, Y. R., SUN, T., Li, Z., ZHANG, J., TENG, J., LIU, X., LIU, R., ZHAO, R., JONES, M. K., WANG, Y., WEN, H., FENG, X., ZHAO, Q., ZHAO, Y., SHI, D.,

BARTHOLOMOT,

B.,

VUITTON,

D.

A.,

PLEYDELL,

D.,

GIRAUDOUX, P., ITO, A., DANSON, M. F., BOUFANA, B., CRAIG, P. S., 17

WILLIAMS, G. M. And MCMANUS, D. P. 2006. Community Surveys And Risk Factor Analysis Of Human Alveolar And Cystic Echinococcosis In Ningxia Hui Autonomous Region, China. Bull World Health Organ, 84, 714-721. YU, S. H., WANG, H., WU, X. H., MA, X., LIU, P. Y., LIU, Y. F., ZHAO, Y. M., MORISHIMA, Y. And KAWANAKA, M. 2008. Cystic And Alveolar Echinococcosis: An Epidemiological Survey In A Tibetan Population In Southeast Qinghai, China. Jpn J Infect Dis, 61, 242-246. ZHANG, W., LI, J. and MCMANUS, D. P. 2003. Concepts in immunology and diagnosis of hydatid disease. Clin Microbiol Rev 16:18-36. ZHANG, W., WEN, H., LI, J., LIN, R. and MCMANUS, D. P. 2012. Immunology and immunodiagnosis of cystic echinococcosis: an update. Clin Dev Immunol. 2012:101895. doi: 10.1155/2012/101895.

18

Figure Captions Fig.1. Heilongjiang Province map.

19

Table 1. Relationship between ELISA and US test results for a population of individuals with clinical signs of CE. Ultrasound ELISA

No.positive No.negative Total

No.positive

82

45

127

No.negative 8

387

395

Total

432

522

90

20

Table 2. Estimation of the properties of the ELISA and US tests for a population of individuals with clinical signs of CE as estimated by Bayesian analysis.

Prior Parameter

Mean

Posterior estimates %

(95% Beta (α, β) prior Source of prior Mean % (95% CI*)

probability ranges)

distribution

probabilities

95.94, 7.22

Macpherson et al., 2003; Budke et al., 2006

93.05(87.43-97.08)

Sp 96.50(0.93-1.00)

105.46, 3.83

Macpherson et al., 2003; Budke et al., 2006

98.44(96.90-99.53)

Se 63.50(0.31-0.96)

4.94, 2.84

Ultrasound Se 93.00(0.88-0.98)

Torgerson et al., 2009; Zhang W et al., 2003; ELISA

92.63(85.32-98.04) Carmena D et al., 2006 Torgerson et al., 2009; Zhang W et al., 2003;

Sp 70.00(0.40-1.00)

5.83, 2.50

90.37(87.18-93.35) Carmena D et al., 2006

Se=sensitivity. Sp=specificity. * CI, Credible intervals.

21

Table 3. Results detected by ELISA and US tests for a population of individuals with clinical signs of CE.

No. positive (%) Variable

No. examined US

ELISA

Sex Male

209

26(12.44) * 49(23.44)

Female

313

64(20.45)

Rural

349

63(18.05) * 93(26.65) *

Urban

173

27(15.61)

34(19.65)

0-19

15

8(53.33)

6(40.00)

20-29

28

8(28.57)

8(28.57)

30-39

77

24(31.17)

21(27.27)

40-49

132

27(20.45)

26(19.70)

50-59

127

15(11.81)

42(33.07)

60-69

106

6(5.66)

19(17.92)

≥70

37

2(5.41)

5(13.51)

78(24.92)

Residence

Age group

*P < 0.05

22

Table 4. Demographic and clinical characteristics of 90 US-confirmed hepatic CE in Heilongjiang Province. Variable

Subcategory

No.

Percentage (%)

Gender

Male

26

28

Female

64

71

Rural

63

70

Urban

27

30

0-9

2

2

10-19

6

7

20-29

8

9

30-39

24

27

40-49

27

30

50-59

15

17

60-69

6

7

≥70

2

2

Farmer

67

74

Worker

6

7

Other

17

19

Upper abdominal pain

52

58

Vomiting

8

9

Hepatomegaly

8

9

Weak

6

7

Other

10

11

extrahepatic

4

4

Liver

90

100

Surgery

21

23

No surgery

69

77

Residence

Age group

Occupations

Clinical symptoms¶

Lesion location

Treatment history

¶

Some patients had more than one clinical symptom.

23

Table 5. Univariable analysis of factors associated with US-confirmed hepatic CE.

Variable

β

SE

P

OR

Gender

-0.5548

0.2522

0.0278

0.574

0.350

0.941

Age

-0.6296

0.2065

0.0023

0.533

0.355

0.799

Environment

1.5938

0.2540

<0.0001

4.923

2.992

8.098

Worker

-0.5532

0.4550

0.2240

0.575

0.236

1.403

Other

0.2910

0.3820

0.5366

0.835

0.472

1.478

95% CI

Occupation

95% CI: 95% confidence interval; OR: odds ratio; SE: standard error.

24

Table 6. Multivariable logistic regression analysis of factors associated with US-confirmed hepatic CE.

Variable

β

SE

P

OR

Male

-0.5353

0.2654

0.0437

0.586

0.348

0.985

Age*

-0.6431

0.2212

0.0037

0.526

0.341

0.811

Rural

1.5529

0.2571

<0.0001

4.725

2.855

7.821

95% CI: 95% confidence interval; OR: odds ratio; SE: standard error; *every 10-year increase in age

25

95% CI

Table7. Regional distribution of US-confirmed hepatic CE patients in Heilongjiang Province. Region

No.of cases

Percentage (%)

Harbin (1)

17

19

SuiHua (2)

30

33

Jiamusi (3)

0

0

QiTaiHe (4)

1

1

MuDanJiang (5)

0

0

DaQing (6)

12

13

QiQiHar (7)

7

8

ShuangYaShan (8)

1

1

JiXi (9)

0

0

HeGang (10)

1

1

HeiHe (11)

5

6

YiChun (12)

3

3

DaXingAnLing Area (13)

0

0

Outside the province

13

14

Total

90

100

Songhua River

Nen River

Ussuri River and Heilong River

26