A real-time quantitative PCR assay has been developed to measure human herpesvirus 6 (HHV-6) DNA in biological specimens. The assay sensitivity was 10 copies of DNA per well, with a linear dynamic range of 10 to 10 7 copies of HHV-6 DNA. Intra-and interassay variations were, respectively, 0.88 and 0.8% for samples containing 10 2 DNA copies, 0.99 and 0.96% for samples containing 10 4 copies, and 0.76 and 0.9% for samples containing 10 6 copies. Among 34 saliva samples from healthy subjects, 26 were found to contain HHV-6 DNA (76.5%; median, 23,870 copies/ml), and following a single freeze-thaw cycle, 25 of the same samples were found to be positive for HHV-6 DNA, although at a statistically significantly lower concentration (median, 3,497 copies/ml). The assay enabled detection of HHV-6 DNA in lymph node biopsies from patients with Hodgkin's disease (

No abstract

increase the PPV to 0.59 if the result is positive, or yield an acceptable 1-NPV of 0.004 if the test is negative, and thus provides important information to the clinician attempting to make a dif-¢cult decision. Selective use in settings where the decision to biopsy may result in an adverse patient outcome may be a potential use for computerized digital dermoscopy. When assessing the clinical relevance of diagnostic technology, clinicians must consider a test's PPV and 1-NPV (not just its sensitivity and speci¢city) and the pre-test probabilities where its use may inform clinical decisions. Whereas Rubegni et al's tool appears to be a powerful diagnostic aid, whether the additional amount of clinical information the tool provides justi¢es the cost of technology requires us to consider the broad determinants of medical decision-making: physician behavior, disease prevalence, outcome, cost, and values.