Background: Wrong blood in tube (WBIT) is a critical pre-analytical error in laboratory medicine in which a blood sample is mislabeled with the wrong patient identity. These errors are often undetected due to the limitations of current detection strategies (e.g., delta checks). Methods: We evaluated Random Forest models for WBIT detection and conducted a detailed analyte importance analysis. In total, 799,721 samples from a German tertiary care center were analyzed and filtered for applicability. Model input features were derived by pairing consecutive same-patient samples for non-WBIT cases, simulating WBIT by pairing samples from different patients, and computing per-analyte first-order differences for each pair. We exhaustively searched all subsets of nine CBC analytes and evaluated models using F1 score, AUC, sensitivity, and PPV. Analyte importance was assessed via SHAP, permutation, and impurity decrease. Results: Models using as few as three analytes (MCV, RDW, MCH) reached F1 scores above 90%, with performance plateauing beyond six analytes. MCV and RDW were consistently top-ranked. Two-dimensional and three-dimensional visualizations revealed interpretable decision boundaries. Conclusions: Findings demonstrate that robust WBIT detection is achievable using a minimal subset of CBC analytes, offering a practical, interpretable, and broadly generalizable ML-based solution suitable for diverse clinical environments.

Background and Objectives: Clinical decision support systems (CDSS) can combine laboratory findings with additional clinical information to support diagnostic and therapeutic decision-making. This review characterizes the objectives of CDSS with relevance to laboratory medicine and examines the current evidence across six predefined domains: clinical maturity, type of evaluation, strength of evidence, integration potential, relevance to personalized medicine, and demonstrated medical benefit. Methods: PubMed was systematically searched for publications on CDSS with relevance to laboratory medicine published between 1 October 2019 and 31 August 2025. Publications without a specific laboratory-related component were excluded, and the detailed analysis was restricted to original publications. The 75 included publications were independently classified and assessed by the authors according to predefined criteria. Results were synthesized descriptively using absolute and relative frequencies across the predefined categories. Results: Publication activity showed an overall increase over the study period. The most common CDSS objectives were diagnostic support (25/75; 33%), therapy monitoring (14/75; 19%), and prognostic assessment (13/75; 17%). Only twelve publications (16%) investigated CDSS that provided specific recommendations for further diagnostic testing or therapy. Most evaluations were retrospective (31/75; 41%) or based on simulations and modelling (16/75; 21%); 13 publications (17%) used prospective non-randomized designs, and three (4%) described randomized designs, two of which were study protocols. In terms of clinical maturity, 45 CDSS were classified as experimental, six as prototypes, and 15 as clinically implemented. High integration potential was identified in 24 of the 75 publications (32%). Twenty publications (27%) met the predefined criteria for a positive effect on process or proxy parameters, and two met the criteria for clinical benefit; no publication demonstrated improvements in hard clinical endpoints. Only eight CDSS (11%) showed high relevance to personalized medicine. Conclusions: CDSS with relevance to laboratory medicine represent an active area of research. However, most systems remain at an early stage of development and have been insufficiently validated under real-world conditions. Sustainable clinical implementation will require prospective multicenter studies, patient-relevant endpoints, interoperable technical infrastructures, and stronger interdisciplinary involvement of laboratory medicine.