Evaluation of an automated device for fingertip blood collection by comparing test results with venous blood

Abstract

Background

Capillary blood collection offers a less burdensome alternative to venous sampling and may improve access to laboratory testing in resource-limited settings. However, manual procedures can yield pre-analytical variation, resulting in the need for skilled personnel. To address these limitations, we developed a prototype device that automates fingertip blood collection.

Methods

The device automatically punctures a fingertip using a disposable lancet, applies intermittent compression to the fingertip using a cuff, sequentially collects blood into two microtubes, and completes hemostasis with gauze. We evaluated the device in 58 healthy volunteers aged 20–60 years by comparing capillary blood collected with the device to venous blood via standard phlebotomy. Paired samples were processed and analyzed under hospital laboratory quality control. Measurements included 8 CBC parameters, glucose and HbA1c in whole blood, 3 electrolytes, and 15 biochemical analytes in serum. Agreement was assessed using Deming regression, and bias with 95% confidence intervals at medical decision levels (MDLs) was calculated.

Results

Most analytes exhibited biases within the allowable limits defined by the Clinical Laboratory Improvement Amendments (CLIA), except for white blood cell count (WBC), potassium (K), lactate dehydrogenase (LD), and glucose. Compared with the BD MiniDraw™, which involves manual finger compression, our device showed greater biases in WBC and K, likely due to the margination effect and hemolysis, respectively—both enhanced by stronger compression.

Conclusion

The device demonstrated acceptable analytical performance for most parameters. Optimizing compression strength and duration may reduce bias. Automated capillary collection could support decentralized testing and reduce phlebotomy workload.