Quality Control in Laboratory Medicine: Recent Advances

Overview - Internal vs. External QC

1. Quality Control in Hematology

Internal QC

• IQC methods recommended by regulatory bodies
• Frequency of running commercial controls
• Derivation of target values and action limits The ICSH is now proposing a harmonized approach to address this inconsistency.

• Levey-Jennings charts - results plotted against time with ±2SD and ±3SD warning/rejection limits
• Westgard multirules - 12S (warning), 13S (rejection), 22S, R4S, 41S, 10X rules applied in combination to optimally detect systematic and random errors
• Bull's algorithm (moving average) - a patient-based real-time QC method specific to hematology: the mean of the last N (typically 20) MCV, MCH, and MCHC results is tracked as a moving average. Because red cell indices are physiologically stable within populations, shifts in the moving average indicate instrument drift. This is particularly useful for detecting calibration shifts between scheduled QC runs.
• Average of Normals (AON) / Moving Median - patient-based real-time QC where the median of sequential results is tracked; recommended as more robust than the mean as it is less influenced by extreme values (Tietz, 7th ed.)

• All 127 surveyed centers ran regular IQC
• Materials: commercial lyophilized plasma (primary supplier), third-party, or in-house plasma pools
• Frequency: at least once per analytical batch
• Action: IQC failures trigger instrument troubleshooting, reagent review, repeat testing

• Assay lot-to-lot validation for reagents and QC materials
• Establishment of reference intervals (RIs) and therapeutic ranges
• Commutability assessment of QC materials
• Reagent lot registration in EQA schemes

External QC (EQA/PT) in Hematology

• EQA schemes for CBC: Organizations such as RCPAQAP, UK NEQAS, CAP PT distribute blood films and stabilized cell suspensions. Participants report CBC and morphology interpretations; performance is assessed against peer group medians
• UK NEQAS and EQATH (External Quality Assurance in Thrombosis and Hemostasis) are the major schemes for hemostasis, distributing lyophilized plasma panels for PT/INR, aPTT, fibrinogen, factor assays, von Willebrand disease workup, and lupus anticoagulant testing
• Commutability - a recognized challenge: stabilized or lyophilized control materials may not behave identically to fresh patient samples across different platforms, producing inter-laboratory bias not seen with real specimens. ICSH guidance recommends using commutable materials where possible
• Six Sigma metrics are increasingly applied to evaluate EQA performance: sigma = (TEa - bias) / imprecision. A 2025 evidence-based review (PMID 39964186) cautioned against uncritical application of Sigma metrics to serology/infectious disease testing (relevant to blood bank hematology screening), noting lack of standardization and risk of missing true errors even when Sigma ≥ 6

Recent Advances in Hematology QC

1. AI-Based Digital Morphology Analyzers - Systems like Mindray MC-80, Sysmex DI-60, CellaVision replace manual differential counting. AI pre-classifies leukocytes, flags atypical/immature cells, and provides reproducible morphological analysis. A study of Mindray MC-80 on 600 samples showed correlation coefficients >0.94, sensitivity of 98.8% for immature granulocytes and 97.5% for nucleated RBCs, with >80% pass rates for pathological cells. These systems serve as QC tools by reducing inter-observer variability and operator fatigue
2. Vision-Language Models (VLMs) - A 2025 review (ONMT Journal) shows VLMs integrating visual features with domain-specific prompts for cell classification, captioning, and report drafting in hematology; they enable few-shot/zero-shot generalization to rare findings and support QC through improved reproducibility and explainability
3. Patient-Based Real-Time QC (PBRTQC) - Moving average/median methods using continuous patient data streams in real time; algorithms monitor instrument stability between scheduled QC runs. The IFCC working group issued recommendations for PBRTQC in 2019, and it is now embedded in many LIS platforms
4. Optical Genome Mapping (OGM) - A 2024 framework (PMID 38164980) for clinical implementation of OGM in hematologic malignancies includes rigorous QC requirements for ultra-high molecular weight DNA, map quality scores, and software analysis pipelines, representing a new QC frontier in hematology genomics
5. Whole Bone Marrow Slide Imaging - Digital morphology extended to BM aspirates using whole-slide imaging (e.g., Metafer4 VSlide) provides quantitative results comparable to conventional microscopy with reliable detection of dysplastic features

2. Quality Control in Histopathology

Pre-Analytical QC

Analytical QC

• Standardized staining protocols with defined reagent concentrations, times, and temperatures
• Positive tissue controls run with each staining batch (e.g., tonsil for general morphology)
• Nuclear and cytoplasmic staining intensity assessed - maintained with periodic hematoxylin freshness testing
• Section thickness monitoring (ideally 3-5 μm) using micrometer checks

• Positive controls (tissue known to express the antigen) run in every batch - detects antigen retrieval failure, antibody failure, or reagent system failure
• Negative controls (primary antibody replaced by non-immune serum/buffer) - detects non-specific background staining
• External controls vs. multitissue blocks - multiblock sections containing a panel of tissues (e.g., tonsil, colon, liver, kidney, placenta) used as a single control slide covering multiple antibody targets
• Reagent lot validation - each new lot of antibody must be validated against the previous lot before clinical use
• H-score / Allred score / IHC scoring reproducibility - inter-observer concordance studies
• A 2023 review on IHC in breast diagnostics (PMID 36576522) emphasizes standardization of ER/PR/HER2/Ki-67 IHC as essential QC targets; false positives in HER2 testing lead to inappropriate trastuzumab therapy

• Rapid on-site evaluation (ROSE) adequacy assessment
• Correlation of cytological diagnoses with histology (cyto-histologic correlation)
• Detection of false-negative cases via rescreening
• 10% random rescreening of negative gynecological smears (mandatory under CLIA in the US)
• Correlation of risk of malignancy (ROM) scores with final diagnoses as a QA indicator

External QC in Histopathology

• EQA/Proficiency Testing schemes: CAP (College of American Pathologists) - distributes glass slides and digital slide sets; RCPAQAP; UK NEQAS Histochemistry; ICC/ISH modules
• Interlaboratory comparison for IHC: CAP PT programs for HER2, ER, PR, PD-L1, ALK, ROS1, MLH1/MSH2/MSH6/PMS2 (MMR proteins) - measured by % concordance with consensus diagnosis
• Peer review / second opinion: Subspecialty review of cancer diagnoses before treatment; mandatory second opinion policies for specific diagnosis categories (lymphoma, mesothelioma, soft tissue tumors, CNS tumors)
• Turnaround time (TAT) monitoring as a quality indicator
• Discordance rates: Inter-institutional discordance in sarcoma diagnosis can be >25% - reflecting a major QC challenge

Recent Advances in Histopathology QC

1. Digital Pathology and Whole Slide Imaging (WSI) - The 2024 review (PMID 37869882) on histotechnologists and digital pathology highlights that WSI introduces new QC requirements: image focus quality metrics, color calibration standards, compression artifact detection, and scanner validation (FDA-cleared scanners required for primary diagnosis in the US). Histotechnologists play a central QC role in ensuring tissue quality for digitization
2. AI/Machine Learning for Diagnostic QC - A 2025 review (PMID 40763008) on machine learning in digital pathology covers AI-based QC algorithms for tissue segmentation, automated grading (Gleason, Nottingham), mutation prediction from H&E, and detection of pre-analytical artifacts (fold detection, air bubble detection, focus failure)
3. AI in Low/Middle Income Countries - A 2025 review (PMID 41147250) discusses digital pathology and AI for molecular diagnostics in resource-limited settings, including telepathology-based EQA where remote expert review serves as the external QC mechanism
4. Proficiency Testing for Molecular Pathology - FISH, PCR, NGS panels (EGFR, BRAF, KRAS, NRAS, ALK, ROS1, NTRK) are now covered by specific PT schemes; CAP molecular pathology PT and EMQN (European Molecular Genetics Quality Network) programs
5. Tissue Biobank Quality - A 2025 systematic review (PMID 41081104) on quality assessment of biobank plasma and serum specimens extends QC principles to tissue banking; pre-analytical variables (cold ischemia time, fixative type, snap freezing protocols) significantly affect tissue biomarker integrity

3. Quality Control in Blood Bank (Transfusion Medicine)

Internal QC in the Blood Bank

• Pre-donation history questionnaire validation
• Hemoglobin/hematocrit screening (copper sulfate or automated)
• Volume of collection (450 ± 45 mL or 500 ± 50 mL depending on protocol)
• Blood bag integrity checks

• Reactive and non-reactive controls run with each batch
• Positive predictive value monitoring
• Six Sigma metrics applied to infectious disease serology - as analyzed in a 2025 review (PMID 39964186): Sigma = (TEa - bias) / CV. Studies using Sigma in blood bank serology (3 out of 4 reviewed studies) found issues with standardization of bias calculation and inappropriate use of cut-off values as the target value. The review provides evidence that even Sigma ≥ 6 assays can miss true errors

• ABO and D typing: forward (cell) typing + reverse (serum/plasma) typing must be concordant
• Antibody screening (indirect antiglobulin test) with 3-cell screening panel
• Crossmatch: electronic (computer) crossmatch (when antibody screen negative + ≥2 concordant ABO/D typings on record) or serological crossmatch
• QC of reagent red cells: grading of reaction strength (2+ to 4+) with known anti-A, anti-B, anti-D
• QC of Coombs/antiglobulin reagent: addition of check cells (IgG-sensitized cells) after negative antiglobulin tests to confirm reagent viability

• Continuous temperature data loggers for refrigerators (2-6°C for RBC), freezers (-18 to -25°C for FFP, -30°C for cryo), and platelet incubators (20-24°C with agitation)
• Alarm systems with defined response protocols
• Equipment validation and preventive maintenance records

External QC in Blood Bank

• SHOT (Serious Hazards of Transfusion) - UK hemovigilance scheme; national reporting of serious adverse events and reactions
• ISBT (International Society of Blood Transfusion) harmonization standards
• CAP Transfusion Medicine PT programs - for ABO/D typing, antibody identification, component labeling
• EFI (European Federation for Immunogenetics) - for HLA typing used in platelet refractoriness management
• Blood component EQA: National schemes test whether manufactured components meet specification; inter-laboratory comparison of QC methods
• Hemovigilance: mandatory national reporting systems (EU Blood Directive 2002/98/EC, FDA); near-miss reporting; transfusion reaction investigation protocols

• Management systems: ISO 9001 (quality management), ISO/IEC 17025 (testing labs), GMP (Good Manufacturing Practice for blood components)
• Technical operations: SOPs, batch records, deviation management, CAPA (corrective and preventive actions)
• Quality culture and sustained stewardship as the ultimate goal

Recent Advances in Blood Bank QC

1. AI-Driven Quality Control in Blood Component Manufacturing - A 2025 paper by Pereira et al. (cited in PMC12821679) focuses on AI for real-time monitoring of component manufacturing, predictive analytics, and proactive error detection to reduce clinical risks. AI-integrated blood analyzers are projected to grow from USD 2.8 billion to USD 19.3 billion over a decade (CAGR 21.3%)
2. AI Plasma Quality Control - Jenwitheesuk et al. (2025, Vox Sanguinis) describe AI revolutionizing plasma quality control "beyond the human eye" - visual inspection automation using computer vision
3. Federated Learning - Li et al. (2025) explore federated learning for transfusion demand prediction models that train across multiple hospital datasets while preserving data privacy
4. Automated Immunohematology Platforms - Column agglutination technology (gel cards, e.g., BioRad ID-System, Ortho Vision), solid-phase adherence, and microplate automation have improved objectivity and reproducibility of compatibility testing, integrating QC result archiving directly with the blood bank information system (BBIS)
5. Electronic Issue / Barcode Technology - Positive patient identification (PPID) systems using two-patient identifiers and barcode at bedside reduce ABO-incompatible transfusion (wrong blood in wrong tube errors); these form part of the pre-transfusion QC chain
6. Pathogen Inactivation Technology (PIT) - Systems like Intercept (amotosalen + UVA), Mirasol (riboflavin + UV), and Theraflex (UV-C) for platelets and plasma add a QC-validated manufacturing step that functionally reduces infectious disease transmission risk; post-PIT component quality parameters (count, pH, swirling, pH) require specific QC validation against non-inactivated component benchmarks
7. Digital Blood Bank / LIMS Integration - Computerized blood banking software enables real-time inventory, traceability from donor to recipient, automated crossmatch decision support, and immediate generation of QA reports; the LIS-BBIS integration now supports outcome-based QC monitoring tied to clinical feedback

Summary Table

• ICSH IQC review for blood cell counters (2024) - PMID 38214063
• IQC in hemostasis assays, NEQAS BC survey (2024) - PMID 37748519
• Quality in transfusion medicine: vein-to-vein chain (2025) - PMID 41197402
• Six Sigma in infectious disease serology QC (2025) - PMID 39964186
• Cytotechnologists' role in cytopathology QM (2026) - PMID 40499519