Comparison of capillary finger stick and venous blood sampling for 34 routine chemistry analytes: potential for in the hospital and remote blood sampling.

Abstract

Objectives: This study examined the comparability of venous and capillary blood samples with regard to routine chemistry analytes.

Methods: Venous and capillary blood samples were collected from adult patients to assess comparability of alanine transaminase, albumin, alkaline phosphatase, apolipoprotein B, aspartate aminotransferase, total bilirubin, calcium, chloride, creatin kinase, creatinine, C-reactive protein, ferritin, folic acid, free T4, gamma glutamyltransferase, glucose, high density lipoprotein cholesterol, iron, lipase, lipoprotein a, magnesium, phosphate, postassium, prostate specific antigen, sodium, total cholesterol, total protein, transferrin, triglycerides, thyroid stimulating hormone, urate, urea, vitamin B12 and 25-hydroxyvitamin-D3. Furthermore, hemolysis-icterus-lipemia Index (HIL-Index) was measured for all samples. All measurements were performed using the Siemens Atellica^® CH or IH Analyzer. Deming regression analysis and mean relative differences between venous and capillary measurements of each analyte were contrasted with the desirable total allowable error (TEa) and Clinical Laboratory Improvement Amendments (CLIA) 2024 proposed acceptance limits for proficiency testing.

Results: Deming regression and mean relative differences demonstrated excellent comparability between venous and capillary samples for most measured analytes.

Conclusions: Capillary and venous samples showed comparable results for almost all studied chemistry analytes. Of the 33 studied analytes for which TEa criteria where available, 30 met TEa criteria. CLIA 2024 criteria where available for 29 of the studied analytes of which only glucose did not meet the criteria. In conclusion, capillary blood draw is a suitable alternative for venous blood sampling for measuring most of the investigated analytes. This benefits patients with fear of needles and might pave the way for remote self-sampling.