Biological Variation Data: Time for Development of Standards? Dr William A Bartlett, On behalf of the Biological Variation Working Group, European Federation of Clinical Chemistry and Laboratory Medicine, Blood Sciences Department, Ninewells Hospital, Dundee Scotland UK DD1 9SY. B ackground Characterization and understanding of the biological components of variation in laboratory data is a fundamental requirement for laboratory medicine specialists. The data are used in the setting of quality standards and in the determination of significance of change in two consecutive results (reference change values (RCV). The characteristics of the data define their utility to providers and users of services, and so methods of data production and other data attributes are of great importance. There are parallels with production and use of reference values as identified by the IFCC and more recently in CLSI guidance. That approach identifies need for characterization of populations studied, methods for production of data, and the statistical treatment of data. The need for this degree of definition is accepted in the context of reference values, but not so in the context of biological variation data (BV). Indiscriminate application of poorly characterized, or produced, BV data leads to delivery of erroneous quality standards and RCVs, and compromises other applications of BV data. It follows that there is a need for standards for production of BV data, standards for reporting of them, and standards for their transmission, since these all impact on their utility and commutability. A working group has been constituted by the EFCCLM to consider the issues. As a first step the group has undertaken work to deliver a checklist to help users and publishers of biological variation data to appraise existing and future studies of biological variation data in a structured way. An outline critical appraisal checklist has been developed to be further validated, and is available for wider consideration, discussion and testing ( The checklist identifies critical factors to be considered, ranging from the need to adequately describe populations studied and the analytical methods used, to the importance of detecting outlying data and appropriate use of statistics. Need for Standardisation and Checklist Biological Variation Data: Applications & factors to be considered in their Generation BV data are Reference Data Requiring Standardisation Proposed Critical Appraisal Checklist: ml Data Archetype: Transmission “ “ Our hope is that the comparability of such data might be provided by use of a common study design and analysis of data” ” Fraser & Harris 1989 Crit Rev in Clin Lab Sci. 1989;27(5) Given that biological variation (BV) data are effectively being used as reference data and that current knowledge of BV indicates a significant number of confounding factors may impact on their quality and utility, it would appear timely to consider the approach to standardising their derivation. Tools also need to be made available to enable objective assessment the veracity of the data. BV data need to be described in terms of method of production, with standards for their reporting and transmission if they are to be applied across geography, populations and time. Publications by Braga et al (Clinica Chimica Acta 2010;411: ) and Miller et al (Clin Chem 2009;55:24-38), reviewing published biological variation data pertaining to haemoglobin A 1c and urinary albumin excretion respectively, highlight the need for critical appraisal of such data. They identified limitations in experimental design used to derive the data, inappropriate study lengths, and poorly description of statistical methods. Differing methodologies with differing analytical characteristics impacted on the biological variation data as do disease states. All of these confounding factors will impact on the commutability of these important data which have an important role in not only derivation of analytical goals and quality standards, but also in assessment of significance of change through definition of reference change values. In Miller’s study 40 publications were cited. The within subject coefficients of variation (CV I ) ranged from 4 to 103%. Thus highlighting an issue. Users of biological variation data need to have an understanding of the uncertainty around these data and standards need to be set for their production, reporting and transmission. The development and validation of a critical appraisal checklist to enable assessment of suitability of data for publication and application is presented here and in more detail at The development of a data archetype that describes the attributes of biological variation will enable commutability of the data and will constitute a basic set of attributes that should be stored against it in biological variation databases. Recently Roraas et al (Clin Chem 2012;58: ) have published and an approach to delivering confidence intervals and power calculations for within-person biological variation. They looked at the effect of analytical imprecision, number of replicates, number of samples, and number of individuals on the estimates. This approach can enable objective assessment of existing published data and provide and aid to future study design. The EFCCLM Biological Variation Working Group is undertaking work to asses the veracity of this checklist and the approach.