Should quality goals be defined for multicenter laboratory testing? Lessons learned from a pilot survey on a national surveillance program for diabetes.

Quality problem Robust laboratory protocols and stringent quality control (QC) procedures are essential for meaningful collection of data from multiple sites in large-scale population-based studies. Failure to design and implement an effective QC program not only adversely affects the scientific outcome, but also affects public confidence in the acceptability of the data. Initial assessment A pilot survey was conducted to assess the analytical performance of multicenter plasma glucose measurements in a national surveillance program for diabetes in China. Choice of solution Quality goals of the imprecision in terms of coefficient of variation (CV) and total analytical error (TEa) were defined based on the Clinical Laboratory Improvement Amendments (CLIA) criteria for acceptable performance of proficiency testing (PT) for plasma glucose using commercial QC preparations. Implementation A web-based internal QC (IQC) program was established to monitor the analytical performance of the 302 centers participating in the survey. Evaluation The participation rate was 96% (289/302). Statistical analysis showed that the percentage of centers meeting the acceptable specifications of CV ≤5.0% and TEa ≤10% using the CLIA PT criteria was 91.7% while 76.4% of laboratories achieved the goals for desirable performance of CV ≤2.9% and TEa ≤6.9%, as proposed by the Laboratory Medicine Practice Guidelines for the management of diabetes mellitus based on biological criteria. Lessons learned Communications and training are important in ensuring the data integrity of multicenter population-based studies. Performance verification and IQC programs should be implemented to help identify centers that can fulfill the eligibility criteria to perform laboratory analyses.