Systematic review and individual patient data meta-analysis on glucose- 6 - phosphate dehydrogenase activities measured by a semi-quantitative handheld biosensor.

BACKGROUND: Measurement of glucose-6-phosphate dehydrogenase (G6PD) activity guides hypnozoitocidal treatment of P vivax malaria. The G6PD Standard (SDBiosensor, Republic of Korea) here referred to as "Biosensor" is a quantitative point-of-care diagnostic that measures G6PD activity in U/gHb. The manufacturer recommends cutoffs to define G6PD deficient (≤ 4.0U/gHb), intermediate (4.1- ≤ 6.0U/gHb) and normal (> 6.0U/gHb) individuals. The aim of this individual patient data (IPD) meta-analysis was to evaluate these cutoffs (CRD42023406595). METHODS: A systematic review identified studies reporting population-level G6PD activity measured by Biosensor, published between January 2017 and May 2023. IPD were collated and standardised. The adjusted male median (AMM) was defined as 100% activity and calculated across all studies (universal AMM) and separately for each setting. The proportion of participants classified as deficient or intermediate were compared using the manufacturer-recommended cutoffs and 30% and 70% of the universal AMM and setting-specific AMM. Associations between G6PD activity and blood sampling method, malaria status, and age were assessed. RESULTS: Eleven studies with 9724 participants from eight countries were included in this analysis. The universal AMM was 7.7U/gHb and the setting-specific AMMs ranged from 6.2U/gHb to 9.9U/gHb. When using the universal AMM, 4.2% of participants were classified as deficient and 11.9% as intermediate or deficient. The corresponding values were 3.9% and 10.8% for setting-specific cutoffs, and 7.2% and 18.3% for manufacturer-recommended definitions for deficients and intermediates respectively. The manufacturer-recommended cutoff for deficient individuals fitted the distribution of G6PD activities better than definitions based on the percentage of AMM. There was no significant association between malaria status or blood sampling method and G6PD activity. Measured G6PD activity decreased in children 1-5 years and plateaued thereafter. CONCLUSION: The manufacturer's recommended cutoff is conservative but more reliable at categorising G6PD deficient individuals than those based on calculations of 30% activity using the AMM. The observed decrease in G6PD activity in children between 1 and 5 years of age warrants further investigation.