Quantifying sampling method-induced imprecision in user-friendly lateral upper arm blood collection: Introducing σ3-methodology (S3) for the verification of alternative sampling methods with TAP® II as a use case

Clinica Chimica Acta, vol. 569, pp. 120107

Abstract

• Sampling method-induced imprecision is the largest source of imprecision and must be addressed in assessing upper-arm blood collection devices. • We introduce σ3-methodology to quantify sampling method-induced imprecision in alternative sampling like lateral upper-arm blood collection. • 5 out of 14 analytes were validated for clinical use with TAP II. Alternative means of blood sampling continue to grow due to the scarcity of phlebotomists and the need for person-centered care. It is crucial to consistently support these alternative blood sampling innovations with scientific evidence to guarantee the quality of care, especially when implementing for instance Lateral Upper-arm Blood Collection (LUBC) for use at home. Knowledge gaps remain in how to quantify imprecision introduced by the collection method and its impact on clinical use. We developed the Six Sigma analysis for alternative Sampling methods, including sampling method-induced imprecision, accuracy and precision, called σ 3 -methodology (S3). We performed a two-step verification using σ 3 -methodology for LUBC TAP® II, which included fourteen routine clinical analytes. We used venipuncture as the gold standard. The biggest source of imprecision for all analytes was sampling method-induced imprecision, which, for the first time, was quantified, resulting in a varying effect on the clinical usability. TAP® II showed acceptable analytical performance for ALP, bilirubin, HDL-cholesterol, CRP and sodium. Unacceptable analytical performance was found for ALT, AST, cholesterol, creatinine, GGT, HbA1c, potassium, LDH and triglycerides. Alternative blood sampling innovations hold promise for advancing diagnostic care, aiming to deliver accessible decentralised sample collection at home that does not require phlebotomist involvement. However, sampling method-induced imprecision should not be overlooked in the performance assessment to guarantee responsible development that will contribute to the success and desirable societal impact of alternative sampling technologies.

Authors 6

1. Vincent J Molenaar Aachen

   Leiden University · Leiden University Medical Center

   Affiliation as printed

   Department of Public Health and Primary Care, National eHealth Living Lab (NeLL), Leiden University Medical Centre, Leiden, the Netherlands; Greiner Bio-One, Kremsmünster, Austria

2. María José Villalobos-Quesada Aachen Department of Public Health and Primary Care (PHEG) National eHealth Living Lab

   Leiden University · Leiden University Medical Center

   Affiliation as printed

   Department of Public Health and Primary Care, National eHealth Living Lab (NeLL), Leiden University Medical Centre, Leiden, the Netherlands

3. Tjin L. Njo

   Unilabs (Switzerland)

   Affiliation as printed

   Unilabs Group, Geneva, Switzerland

4. Niels Henrik Chavannes Aachen Department of Public Health and Primary Care (PHEG) National eHealth Living Lab

   Leiden University · Leiden University Medical Center

   Affiliation as printed

   Department of Public Health and Primary Care, National eHealth Living Lab (NeLL), Leiden University Medical Centre, Leiden, the Netherlands

5. Esther P. W. A. Talboom-Kamp corresponding Aachen

   Leiden University · Leiden University Medical Center

   Affiliation as printed

   Department of Public Health and Primary Care, National eHealth Living Lab (NeLL), Leiden University Medical Centre, Leiden, the Netherlands; Zuyderland, Sittard-Geleen-Heerlen, the Netherlands. Electronic address: esther@talboom.info

6. Maarten B. Kok corresponding

   Unilabs (Switzerland)

   Affiliation as printed

   Unilabs Group, Geneva, Switzerland. Electronic address: maarten.b.kok@gmail.com