Laboratory tests with discrete decision limits are often used in clinical guidelines as part of the definition of certain conditions (e.g. diabetes mellitus/gestational diabetes, acute coronary syndrome, chronic kidney disease), or serve as a trigger for initiating or changing patient management. Variation of test results around decision and reference limits may result in misclassification of conditions, including overdiagnosis. Clinical guideline teams often overlook the preanalytical, analytical and post-analytical aspects of tests which can grossly influence the clinical performance and interpretation of laboratory results especially around diagnostic cut-offs.

We reviewed the literature and carried out modelling studies to investigate the potential impact of these less known variables on diagnostic and patient management decisions. We present examples on 1/how sample collection/transport/and processing can change the diagnosis of gestational diabetes; 2/how incorrect reference intervals cited in key international guidelines for liver function tests can increase the prevalence of liver disease and generate further unnecessary testing; 3/how analytical and biological variation of glucose and HbA1c can easily shift patients across decision limits and lead to misdiagnosis or mismanagement of patients with diabetes; 4/how universal decision limits in guidelines can misinform clinical decisions when in practice multiple non-standardised laboratory methods are used for measurement; and 5/how poorly these laboratory testing-related variables are addressed and reported in large trials and guidelines.

We propose that guideline development groups include professionals from laboratory medicine especially when a laboratory test is part of a disease definition. Guideline teams should consider the uncertainty of test results due to preanalytical, analytical and postanalytical variations and define a grey zone of results around discrete decision limits for judicious use in clinical practice. Laboratory professionals should inform and educate clinical staff about the uncertainty of test results and constantly improve the performance of their laboratory methods and their pre-and postanalytical performance.