Field assessment of the operating procedures of a semi-quantitative G6PD Biosensor to improve repeatability of routine testing

Abstract

In remote communities, diagnosis of G6PD deficiency is challenging. We assessed the impact of modified test procedures and delayed testing for the point-of-care diagnostic STANDARD G6PD (SDBiosensor, RoK), and evaluated recommended cut-offs. We tested capillary blood from fingerpricks (Standard Method) and a microtainer (BD, USA; Method 1), venous blood from a vacutainer (BD, USA; Method 2), varied sample application methods (Methods 3), and used micropipettes rather than the test's single-use pipette (Method 4). Repeatability was assessed by comparing median differences between paired measurements. All methods were tested 20 times under laboratory conditions on three volunteers. The Standard Method and the method with best repeatability were tested in Indonesia and Nepal. In Indonesia 60 participants were tested in duplicate by both methods, in Nepal 120 participants were tested in duplicate by either method. The adjusted male median (AMM) of the Biosensor Standard Method readings was defined as 100% activity. In Indonesia, the difference between paired readings of the Standard and modified methods was compared to assess the impact of delayed testing. In the pilot study repeatability didn't differ significantly (p = 0.381); Method 3 showed lowest variability. One Nepalese participant had <30% activity, one Indonesian and 10 Nepalese participants had intermediate activity (≥30% to <70% activity). Repeatability didn't differ significantly in Indonesia (Standard: 0.2U/gHb [IQR: 0.1-0.4]; Method 3: 0.3U/gHb [IQR: 0.1-0.5]; p = 0.425) or Nepal (Standard: 0.4U/gHb [IQR: 0.2-0.6]; Method 3: 0.3U/gHb [IQR: 0.1-0.6]; p = 0.330). Median G6PD measurements by Method 3 were 0.4U/gHb (IQR: -0.2 to 0.7, p = 0.005) higher after a 5-hour delay compared to the Standard Method. The definition of 100% activity by the Standard Method matched the manufacturer-recommended cut-off for 70% activity. We couldn't improve repeatability. Delays of up to 5 hours didn't result in a clinically relevant difference in measured G6PD activity. The manufacturer's recommended cut-off for intermediate deficiency is conservative.

Fig 1. Schematic workflow of the pilot study in Australia, and the field studies in Indonesia and Nepal.

Fig 2. Illustrations of components and step-by-step guide of the STANDARD G6PD Test (SD Biosensor, RoK) from Adhikari et al (2022) [29].

Fig 3. The Ezi tube, a single use plastic pipette included in the Biosensor test kit.

When the tip is inserted into a blood sample, 10 μL are collected through capillary action. The black line indicates 10μL.

Fig 4. Absolute differences between paired measurements in U/g Hb for each Biosensor method used in the pilot study.

Fig 5. Histogram of G6PD activity as measured by the Biosensor of female and male participants from Indonesia (Standard Method, n = 60) and Nepal (Standard Method n = 60, and Method 3 n = 60).

Vertical lines denote (from left to right) 30%, 70%, and 100% of AMM.

Fig 6

Boxplot of absolute differences between paired Biosensor G6PD activity readings per method (in U/g Hb) from field studies in Indonesia (left), Nepal (middle), and both sites (right).

Fig 7

Scatterplot of G6PD activity as measured by the Biosensor using the Standard Method vs spectrophotometry (left); and by the Biosensor using the Standard Method vs Method 3 (right). The diagonal line denotes the line of equality. The red horizontal and vertical lines, from the point of origin outwards, mark the 30%, 70%, and 100% of the AMM (normal G6PD activity). Individuals whose G6PD categorization changed when measured by the Biosensor and spectrophotometry were marked as red dots.