ABSTRACT
An evaluation of five commercial software systems used for immunoassay data analysis revealed numerous deficiencies. Often, the utility of statistical output was compromised by poor documentation. Several data sets were run through each system using a four-parameter calibration function, and the results were compared to those from an independent method. Comparable results between systems were obtained, but often several attempts at analysis were necessary. The evaluation process revealed that it is difficult to monitor the numerous options available on these types of programs, and that incorrect results could easily be obtained if comparison analyses were not used. Recommendations for improved software functionality and for using the four-parameter calibration model are presented.
