A review of portable quantitative and semi-quantitative devices for measurement of vitamin A in biological samples.

Background: We catalog and summarize evidence of the analytical performance of portable quantitative and semi-quantitative devices for the assessment of vitamin A status and vitamin A deficiency (VAD) in various biological samples-including whole blood, plasma, serum, and milk-in addition to VAD determination by functional indicators such as pupillary response. Methods: We searched the literature for published research articles, patents, and information from manufacturers of mobile devices, particularly those appropriate for low-resource settings. The included devices were required to be portable (lightweight and ideally not needing a power outlet) and to measure vitamin A as well as define VAD. Eligible studies compared a portable device to a reference standard of high-performance liquid chromatography for blood and milk, or a Goldmann-Weekers dark adaptometer for eyes/vision. Where available, identified devices were compared with reference methods across several performance criteria. When possible, we compared the device's performance reported in published studies against the stated performance criteria from the manufacturers' websites. Results: We catalogued 25 portable devices for measuring vitamin A and/or VAD via biological samples. We also identified 18 comparison studies (plus associated reports) assessing nine methods: the iCheck Fluoro, iCheck Carotene, CRAFTi, Tidbit with or without the HYPER filtration system, custom field-friendly immunoassays, and microfluidic assays for blood; the iCheck Fluoro and iCheck Carotene for milk; and the Scotopic Sensitivity Tester-1 for eye function. Conclusions: The iCheck Fluoro and iCheck Carotene are commercially available for use and are acceptable for measuring vitamin A in blood and milk samples, according to the available validation data. Many of the other identified devices, including other portable fluorometers, photometers, immunoassays, microfluidics-based devices, and dark adaptometers, were proofs of concept and not yet commercially available. Furthermore, none of these other devices included manufacturer-described device performance criteria to compare with descriptions from experimental studies. Several gaps remain, including studies comparing the other portable devices against a reference standard, particularly for functional indicators of vitamin A status/deficiency; available manufacturer-reported device performance criteria against which to compare future results of investigations; and more comprehensive reporting of validation metrics including sensitivity, specificity, precision, and Bland-Altman analysis.