Novel analysis of clinically relevant diagnostic errors in point-of-care devices

Area of Science:

• Clinical Diagnostics
• Medical Device Quality Assurance
• Pharmacovigilance

Background:

• Accurate point-of-care (POC) device measurements are crucial for clinical decision-making and patient safety.
• Traditional quality assurance methods may fail to detect clinically significant diagnostic errors in POC devices.

Purpose of the Study:

• To compare the effectiveness of a novel analytic framework against traditional analyses for assessing POC International Normalized Ratio (INR) devices.
• To identify potential diagnostic errors and their clinical implications missed by conventional quality assurance.

Main Methods:

• Paired venous and fingerstick International Normalized Ratio (INR) measurements were collected from 1518 patients using Hemochron POC devices.
• Traditional quarterly quality assurance focused on correlation analysis.
• A novel analysis employed enhanced graphics and clinical decision-making assessment.

Main Results:

• While correlation between POC and laboratory INR was high (0.84-0.91), the novel analysis revealed systematic biases: inflating low values (<3) and deflating high values (>4).
• The Hemochron devices led to different clinical decisions than laboratory measures in 31% of cases.
• In 52% of cases with a low reference INR, POC devices failed to indicate appropriate dose increases, risking adverse events.

Conclusions:

• A novel, clinically relevant analysis identified significant, previously undetected deficiencies in POC INR devices.
• The proposed analytical approach offers a more sensitive method for evaluating diagnostic device performance.
• Adoption of this novel framework by industry, regulators, and institutions is recommended to improve patient safety.