Reference intervals are an essential tool for the clinical interpretation of laboratory test results. Traditionally, these interval are determined using samples from healthy individuals, a process that is resource-intensive, time-consuming, and require the active recruitment of healthy volunteers.

In recent years, due to the increasing availability of electronic health record (EHR) databases and the growing number of laboratory tests, it is possible to determine the reference intervals indirectly. This approach relies on the analysis of routine data acquired in clinical laboratories, eliminating the need for active recruiting healthy subjects and significantly reducing costs. Moreover, the method has the potential to eliminate the selection bias of an ultra-healthy population typical of the direct methods. The indirect methods for determining reference intervals have evolved from simple strategies of isolating the healthy population using sample metadata, to sophisticated statistical models that effectively distinguish normal from pathological distributions. One of the advanced techniques, RefineR, has reached an excellent combination of accuracy, robustness, and computational efficiency, outperforming previous methods. It has been implemented as an open-source R package, facilitating its application in real-world settings.

In recognition of these advantages, the IFCC (International Federation of Clinical Chemistry and Laboratory Medicine), through its Committee on Reference Intervals and Decision Limits (C-RIDL), has promoted the adoption of indirect methods for determining reference intervals, highlighting the advantages of this strategy, including greater speed, lower costs, and the absence of a need to recruit healthy donors.

Furthermore, a recent study has highlighted age-related physiological variations in hemoglobin levels in elderly population. This underscores the need for defining age-specific reference intervals which are currently absent from most laboratory reports, potentially impacting diagnostic accuracy.

Inclusion Criteria:

• no one

Exclusion Criteria:

• no one

The study population consists of all patients undergoing routine laboratory testing during the year 2024 at the participating centers.