Part of the overall approach to controlling the risk of waterborne cryptosporidiosis is to improve analytical methods for detection. The microscopic technique which has been utilized and gradually improved over the past 15 years is still labor intensive, slow, expensive, and criticized for poor recovery and reproducibility (Simons et al, 2001). Furthermore, the currently used routine microscopic assay is still only capable of detecting the numbers of oocysts present. It provides no information on species, viability, or virulence of the detected oocysts. Research methods have been developed utilizing infection of cell lines to determine viability and virulence of oocysts in samples (Slifko et al, 1999; Rochelle et al, 2002). However the difficulties encountered with maintenance of sterile cell lines generally restricts use of the technique to the research setting. Two years ago the DuPont Qualicon Corporation, in association with the American Waterworks Service Company, developed and marketed an automated polymerase chain reaction (PCR) detection technique suitable for routine laboratory use. The Pittsburgh Water & Sewer Authority (PWSA) Laboratory purchased one of these units and incorporated it into routine use, alongside the microscopic method (USEPA method 1623), for routine analysis of raw and treated water samples for the presence of Cryptosporidium oocysts. This paper describes the results of a comparison of the two methods conducted over a period of more than one year. These methods include a version of the USEPA method 1623 published in April 2001 (USEPA, 2001), and a BAX PCR method. In this project the two techniques were used to analyze monthly river and drinking water samples collected at the PWSA Drinking Water Treatment Plant. The methods were also used to analyze samples of river water collected from various sites along the length of the Allegheny River as well as several samples of groundwater. The study design is discussed along with the results. Includes 12 references, tables.
