BACKGROUND: The beryllium lymphocyte proliferation test (BeLPT) is used to identify persons sensitized to beryllium. ATSDR convened an expert panel of physicians and scientists in April 2006 to discuss this test and to consider what BeLPT test results actually establish beryllium sensitization. The three criteria proposed by panel members were (1)one abnormal result, (2)one abnormal and one borderline result, and (3)two abnormal results. METHODS: Complete algorithms were developed for each of the three proposed criteria. Using single-test outcome probabilities developed by Stange et al. [2004. Am J Ind Med 46:453-462], we calculated and compared the sensitivity, specificity, and positive predictive values (PPVs) for each set of criteria. RESULTS: The overall sensitivity and specificity of the three criteria were similar. When the criteria required confirmation of an abnormal result the PPV was higher--whether the requirement was satisfied by a borderline result, or only by another abnormal result. Confirmation also reduced the likelihood of false positives. The differences between the three criteria decreased as the prevalence of sensitization increased. CONCLUSIONS: A single unconfirmed abnormal is usually insufficient to establish sensitization for an apparently healthy person. When the prevalence of beryllium sensitization in a group is high, however, even a single abnormal BeLPT can be a strong predictor.
BACKGROUND: The beryllium lymphocyte proliferation test (BeLPT) is used to identify persons sensitized to beryllium. ATSDR convened an expert panel of physicians and scientists in April 2006 to discuss this test and to consider what BeLPT test results actually establish beryllium sensitization. The three criteria proposed by panel members were (1)one abnormal result, (2)one abnormal and one borderline result, and (3)two abnormal results. METHODS: Complete algorithms were developed for each of the three proposed criteria. Using single-test outcome probabilities developed by Stange et al. [2004. Am J Ind Med 46:453-462], we calculated and compared the sensitivity, specificity, and positive predictive values (PPVs) for each set of criteria. RESULTS: The overall sensitivity and specificity of the three criteria were similar. When the criteria required confirmation of an abnormal result the PPV was higher--whether the requirement was satisfied by a borderline result, or only by another abnormal result. Confirmation also reduced the likelihood of false positives. The differences between the three criteria decreased as the prevalence of sensitization increased. CONCLUSIONS: A single unconfirmed abnormal is usually insufficient to establish sensitization for an apparently healthy person. When the prevalence of beryllium sensitization in a group is high, however, even a single abnormal BeLPT can be a strong predictor.
Authors: Lori J Silveira; Erin C McCanlies; Tasha E Fingerlin; Michael V Van Dyke; Margaret M Mroz; Matthew Strand; Andrew P Fontenot; Natalie Bowerman; Dana M Dabelea; Christine R Schuler; Ainsley Weston; Lisa A Maier Journal: J Immunol Date: 2012-09-12 Impact factor: 5.422