The Kato-Katz assay has long been the backbone of schistosomiasis diagnosis in endemic areas. However, even when multiple samples are tested, the Kato-Katz method has inadequate sensitivity, especially in areas with lower rates of transmission 
. This results in the challenging task of trying to evaluate other tests in comparison to an imperfect ‘gold standard’. When using the Kato-Katz as the ‘gold standard’ to determine the sensitivities and specificities of the CCA urine assays, the sensitivities are rather high but the specificities are quite low. The low specificities obtained when using Kato-Katz as the ‘gold standard’ may be explained by the low sensitivity of the Kato-Katz method. Thus, to more accurately assess the sensitivities and specificities of these assays, we analyzed the results using LCA. We chose the Bayesian approach to Latent class analysis, considering the two CCA diagnostic assays correlated because they test for the same antigen in the same sample (urine). Sensitivities and specificities of each of the four tests were calculated based on the latent variable ‘true disease status’.
As expected, the Kato-Katz method had the lowest sensitivity when analyzed by LCA. The SWAP ELISA and the cassette urine CCA assay were the most sensitive assays, albeit also the least specific. It is possible that false positive antibody responses could be present in children who were exposed to schistosome antigens but not infected (e.g., in utero, during the pregnancy of an infected mother) or in older children who had been infected then cleared the infection due to natural worm death. Schistosomes that typically infect wild mammals, including some new and hybrid species, have also been described in the Lake Vitoria Basin near Kisumu 
. However, it is not known how human exposure to these other schistosomes may affect performance of the ELISA.
False positive CCA results may occur if the individual being tested has haematuria or pio-uria due to urinary tract infection, as stated on the CCA assay technical brochure. However due to the age of these children we feel that this would not likely make a large impact on our study. Similarly, we do not know what the effect of having frozen the urine had on the assay performance as we did not test any urine that had not been frozen. The product insert states that urine can be frozen for up to a year but we are unable to comment on whether or not the frozen urine was any less sensitive than urine collected and tested immediately.
Of the 53 individuals who were negative for eggs in stool but positive by both of the CCA assays and the SWAP ELISA it is likely that they had light infections that were missed by the Kato-Katz even though stools were collected on three consecutive days. Of the four children who were positive by stool examination and not by any other test, three children had lighter infections (96, 102, and 18 EPG) that may have been missed by the other diagnostic tests. However, one of the children had a high intensity infection (1,382 EPG), which was unlikely to be missed by the other assays. It is possible that the stool samples were mislabeled and not from the same individual that provided the other two samples types. This highlights a limitation of stool and urine data in that sample collection is typically not directly observed and either urine or stool may be substituted with samples from other individuals. In contrast, this is usually not a concern for blood samples collected by study personnel.
Other factors that may have contributed to incongruous results between the various assays include a poor understanding of whether CCA levels demonstrate diurnal variations. In addition, volume of fluid intake or number of previous urinations in a given day may affect urine CCA concentration. Also, while our assumption was that most of the children in this study had not previously been treated for schistosomiasis as no mass treatments had recently been performed in this area, it is possible that some families had migrated into the area and that children may have been treated elsewhere and could thus be antibody positive but egg and antigen negative. Finally, although HIV-1 prevalence levels are high in this area of western Kenya and could theoretically affect antibody responsiveness, we do not expect that HIV-1 coinfection would significantly affect our observed results as the age of the majority of participants was such that previous exposure to HIV-1 was unlikely.
In addition to the accuracy of a screening test, it is important to also consider practicality when comparing different tests 
. Although the Kato-Katz test can be done at a relatively low cost, there are other factors that make this method less than ideal. For example, the Kato-Katz method requires one or more individuals trained in stool sample preparation and microscopy. There are also significant expenses associated with transport costs and the field staff time necessary for multiple field trips, to obtain consecutive stool samples and to provide treatment. In contrast, the cassette CCA assay is designed for point of contact use. Thus, health workers could combine screening and treatment during a single trip which would save on costs and possibly reduce the number of missed treatments of individuals who cannot be found on a follow-up visit. Currently, the individual test cost of the CCA assays is relatively high ($1.98 US per test) but future comparisons of expenses should include personnel time, transport and equipment expenses to determine which approach is more cost effective. In addition, cost of the CCA test may be reduced with greater use and increased scale of production.
Although the SWAP ELISA had similar sensitivity as the cassette CCA assay, this assay is less practical in field based control programs because of the requirements for reagents and equipment, as well as the need for specialized training of laboratory personnel. Also, the inability of the total IgG antibody test to distinguish between past and current infections makes this a poor test for assessing prevalence in populations that have already received treatment.
Our results are consistent with those of a recently published study that found good agreement between urine CCA assay results and a single stool analyzed by the Kato-Katz method 
. This study evaluated a total 171 children, ages 6 to 17, in 11 schools along the Kenyan and Tanzanian shore of Lake Victoria, including 14 children from Usoma, the school that serves the children in our study. Both studies found a strong correlation between stool egg concentration and intensity of the CCA test band (
Like the authors of this previous study 
, we conclude that the CCA urine assays are an effective screening tool for S. mansoni
infections in areas of high prevalence. The CCA urine assays were more sensitive than examination of three stools by Kato-Katz and were as sensitive as the adult worm-specific antibody tests. The urine CCA assays are also easy to use and less time consuming than the other methods currently employed for S. mansoni
screening. CCA assays also have the potential to asses cure, as CCA should not be present after the resolution of infection 
. However, in this study, we were not able to assess cure rates or the ability of the tests to detect resolution of infections as we did not collect additional samples after treatment. Future studies including urine CCA assays should address this question as well as evaluating the performance of the tests in areas with lower intensities of infection. summarizes the strengths and weaknesses of the assays used in this study for population screening purposes. How data from these different assays can be used to inform treatment decisions at the local level has been discussed by Standley et al. in a recent publication 
Comparison of characteristics important for effective screening test.
Results from this study indicate that the urine CCA assays are at least as sensitive as the Kato-Katz method of testing for schistosome eggs in stool. The company that produced both CCA tests has now halted production of the carbon CCA assay. Results from the discontinued test are included here because results from that assay contributed to the study analysis. Production of the cassette CCA assay continues and its ease of use and relatively simple sample collection make it an attractive tool for screening for S. mansoni infections in control programs.