Abstract
<jats:p>Background: Continuous glucose monitoring (CGM) improves glycaemic control in people with type 1 diabetes (T1D), but high costs limit uptake in low- and middle-income (LMIC) countries. Evidence on the cost-effectiveness of CGM is limited in LMIC settings. Objective: To evaluate the short- and long-term cost-effectiveness of intermittently-scanned continuous CGM (cCGM) and intermittently-scanned periodic CGM (pCGM) (one sensor every three months), versus standard self-monitoring of blood glucose (SMBG) in the South African public-sector. Methods: A three-arm randomised controlled trial (ACCEDE) was conducted among T1D individuals with HbA1c ≥10% in South Africa. A within-trial cost-effectiveness analysis was conducted from a partial societal perspective over 9-months, using resource use data and QALYs derived from EQ-5D. A cost-utility analysis using a Markov microsimulation model was conducted for two populations: total T1D, and youth (<20 years). Results: Within-trial analysis showed no statistically significant differences in QALYs or HbA1c between arms. Costs were highest for cCGM (USD 1,504), followed by pCGM (USD 742) and SMBG (USD 467). CGM strategies were dominated in the within-trial analysis. In contrast, long-term modelling showed that CGM was more effective than SMBG and was cost-effective for youth when used periodically. cCGM delivered additional QALYs at a higher cost (ICERs USD 15,259-30,852/QALY) and was only potentially cost-effective in youth when sensor prices were reduced by >45%. Conclusions: While CGM was not cost-effective in the short term, modelling suggests pCGM use may offer value-for-money under specific assumptions and in select populations, highlighting the need for further evidence on long-term effectiveness, engagement, and pricing.</jats:p>