Comparative Evaluation of Point-of-Ccare vs. Central Lab Hematology Analyses

Thesis Abstract

This study addresses the critical question of whether point-of-care (POC) hematology analyzers provide results that are sufficiently concordant with central laboratory (CL) hematology analyses to support timely clinical decision-making without compromising accuracy. The aim is to evaluate analytical agreement, diagnostic yield, and turnaround-time implications of POC versus CL hematology testing in adult patient care. Specific objectives include (i) quantifying agreement between complete blood count (CBC) parameters obtained by a validated POC analyzer and a reference CL analyzer, (ii) assessing the impact of device type on diagnostic categorization for common hematological conditions (anemia, leukocytosis, thrombocytopenia) using concordance measures, (iii) evaluating turnaround time from sample collection to result reporting, (iv) identifying systematic biases and factors influencing discordance (e.g., sample type, operator proficiency, pre-analytical variables), and (v) estimating the potential clinical and economic implications through scenario analysis. A mixed-methods design guides the investigation, integrating a cross-sectional analytical component with a prospective cohort sub-study. The population comprises adult inpatients and outpatients presenting with indications for CBC testing at a tertiary care hospital. A total of 600 paired samples will be collected over six months, with 300 analyzed by a POC hematology analyzer (manual sampling via venipuncture into EDTA tubes at the bedside) and 300 by the hospital’s standard CL platform. Sampling will employ systematic random selection of eligible encounters to ensure representation across departments (emergency, internal medicine, oncology, obstetrics). Data collection instruments include the POC device’s digital output, the CL hematology analyzer’s results, patient demographics, pre-analytical data (time to analysis, sample integrity indicators), and operator logs. Validity and reliability will be ensured through calibration protocols, daily instrument verification, and inter-operator training, with the POC device validated against the CL analyzer using Passing-Bablok regression and Bland-Altman analyses. Analytical methods encompass both agreement and impact assessments. The primary analysis will use Passing-Bablok regression to assess proportional and constant bias between CBC parameters (hemoglobin, hematocrit, mean corpuscular volume, red/white cell counts, platelets) and Bland-Altman plots to quantify limits of agreement. Intraclass correlation coefficients (ICCs) will evaluate repeatability, while Cohen’s kappa will measure diagnostic concordance for classified categories (anemia severity, leukopenia, thrombocytopenia). Subgroup analyses will test the influence of department, operator experience, and time-to-analysis on discordance. A multivariable linear regression model will identify predictors of bias magnitude, and a decision-analytic framework will compare potential clinical outcomes under POC- and CL-based decision pathways, incorporating Monte Carlo simulation to model uncertainty. The theoretical lens includes diffusion of innovations and Technology Acceptance Model (TAM) to interpret clinician adoption patterns, along with a validity framework drawing from the Clinical Laboratory Improvement context to frame analytical performance requirements. Ethical approval and patient consent procedures align with standard hospital research ethics, with de-identified data maintained in secure repositories. Expected findings anticipate high correlation for basic CBC parameters with clinically acceptable limits of agreement for most metrics, but possible systematic biases in hematocrit and platelet counts at extreme values, and modest discrepancies in red cell indices under high workload conditions. Turnaround time is anticipated to be significantly shorter for POC testing, with improved time-to-result potentially compensating minor analytical biases in routine screening scenarios. The study aims to delineate parameter-specific thresholds where POC results are clinically interchangeable with CL results and to quantify the trade-offs between expedited results and analytical accuracy. The contribution to knowledge lies in providing robust, context-specific evidence on the reliability and applicability of POC hematology in routine clinical workflows, informing laboratory governance, clinician training, and policy on when POC results can Supplement or substitute CL results without compromising patient safety. The main conclusions are expected to underscore the value of POC hematology as a rapid initial screening tool with acceptable agreement for most CBC parameters, while identifying critical exceptions requiring confirmatory CL testing. Recommendations will include standardization of pre-analytical protocols, targeted operator competency programs, clear decision thresholds for initiating CL reflex testing, and implementation strategies for integrated reporting systems to optimize patient management and cost-effectiveness in diverse care settings.