Abstract
Hematological and clinical chemistry measurements are an integral part of vaccine safety monitoring. While adopting a conservative approach is important to exclude potential risks for patients, the rationale and methodology underlying the assessment of given adverse events have to be well grounded to avoid raising unfounded concerns. Using asymptomatic transient neutropenia as an example, this paper aims to address the complexity of interpreting abnormal hematological values in vaccine clinical trials and to evaluate the validity of using neutrophil count cut-off points to assess neutropenia in the context of safety monitoring. The validity of the neutrophil count cut-off point methodology was assessed in terms of content validity (i.e., the extent to which a single neutrophil count below the cut-off point corresponds to a clinically significant adverse event), criterion validity (i.e., the extent to which a neutrophil count below a given cut-off point correlates with another manifestation of neutropenia, namely bacteremia), and construct validity (i.e., the exactness of the assumption that a neutrophil count below a given cut-off point corresponds to a reactogenic event caused by the vaccination). We argue that, because of within-individual physiological fluctuations, variations according to population demographics, and poor predictive potential with regard to neutropenia-associated infection, the application of the cut-off point methodology to neutropenia safety monitoring presents major limitations. Based on this assessment, we conclude that hematological laboratory values must be evaluated on a case-by-case basis by investigators to determine their clinical significance.