Reply to “Technical Notes on the Correct Configuration of the Alfred 60/AST Device for the Detection of Urinary Tract Infections”

Abstract

Spezzotti has identified several issues that may have resulted in an underestimation of the performance of the Alfred 60/AST device.First, Dr. Spezzotti questions the authenticity of one of the nine false-negative samples (MB12216254), based on the Alfred 60/ AST device flagging this sample as "invalid." Although this is correct, we followed the manual's recommendations (see page 66 in reference 2), which currently does not suggest excluding such a sample. Furthermore, additional samples which gave the correct result also had such issues. If the manual were to be updated, exclusion of these samples would result in a sensitivity of 88.0%.In the 3 indwelling catheter specimens or additional false negatives, Dr. Spezzotti suggests alternative Alfred 60/AST cutoffs to optimize detection. However, these recommendations are contrary to the current manual. We would argue that all indwelling catheter specimens would need to be reevaluated with the new protocol to know the true performance of the Alfred 60/AST for these samples. Nevertheless, the exclusion of all indwelling catheter samples and the use of our original cutoff results in a sensitivity of 98.0%.Standard broth, as provided by the distributor, was used in our evaluation. Although Dr. Spezzotti mentions another "specific dedicated product" to boost the detection of specific organisms, we are unaware of such a product for urine analysis, and it seems impractical to run duplicate broths for each sample to optimize detection.Although the above issues could potentially be rectified by better optimization of the instrument and culture medium inoculation, we disagree with the final point raised by Dr. Spezzotti concerning laboratory evaluations. We defined our "gold standard" prior to analysis and disagree that the results of the Alfred 60/AST should form part of the gold standard with respect to performance. Furthermore, all discordant results were investigated clinically with all patients found to have symptoms or signs consistent with urinary tract infections.