Related Experiment Video
Updated: Oct 13, 2026
Enumeration of Major Peripheral Blood Leukocyte Populations for Multicenter Clinical Trials Using a Whole Blood Phenotyping Assay
Published on: September 16, 2012
A three-parameter complete blood count triage tool for monocytosis to prioritize hematology evaluation: development
1Department of Hematology, Fuyang People's Hospital (The Affiliated Fuyang People's Hospital of Anhui Medical University), Fuyang, Anhui, China.
Introduction:
Monocytosis is a common complete blood count (CBC) abnormality, but in non-hematology settings it may be difficult to distinguish reactive causes from patterns warranting urgent evaluation for acute myeloid leukemia (AML) or myelodysplastic syndromes (MDS). We aimed to develop and externally validate a simple, report-level CBC triage model to prioritize hematology assessment in patients with monocytosis.
Methods:
MonoxGB was developed using admission-level data from MIMIC-IV v2.2 and three routinely reported CBC variables: absolute monocyte count, monocyte percentage, and platelet count. No proprietary analyzer signals, scattergram-derived parameters, morphology flags, or manufacturer-specific cell-population data were used. External validation included a retrospective Fuyang cohort of 91 eligible patients with monocytosis, all of whom underwent bone marrow examination, and an eICU Collaborative Research Database cohort of 73,650 intensive care admissions with monocytosis.
Results:
In fivefold cross-validation within MIMIC-IV, the mean area under the receiver operating characteristic curve (AUC) was 0.931 ± 0.022. In the Fuyang cohort, 29 of 91 patients had AML/MDS; the AUC was 0.840 (95% confidence interval, 0.753-0.915), and average precision was 0.682. In eICU-CRD, the AUC was 0.886 (95% confidence interval, 0.839-0.928). At an exploratory local high-sensitivity threshold of 0.0028, sensitivity was 0.966, specificity was 0.500, and negative predictive value was 0.969. Across cohorts, AML/MDS was associated with a higher monocyte burden and lower platelet counts; platelet count monocyte percentage, and absolute monocyte count were the dominant model features.
Discussion:
MonoxGB is a three-parameter, CBC report-based triage tool that may help non-hematologists prioritize patients with monocytosis for earlier smear review or hematology assessment. It is not a diagnostic substitute for bone marrow evaluation or specialist judgment. Prospective workflow-based validation using a prespecified deployment threshold is required before routine implementation.
