**Background:** Flash glucose monitoring (FlashGM) is a sensor-based technology that displays current glucose levels, readings from the past 8 hours, and trend arrows, providing an ambulatory glucose profile without routine finger-pricking. Despite its rising popularity, previous systematic reviews reported inconsistent findings and largely included observational studies without control groups, introducing bias. HbA1c has been the traditional gold standard for glycemic assessment, but it does not capture intra- and interday glucose variability. Time in range and hypoglycemic episodes are increasingly recognized as important complementary metrics linked to microvascular complications and life-threatening outcomes. This meta-analysis aimed to assess the effect of FlashGM versus SMBG on HbA1c, time in range, time in hypo/hyperglycemia, and frequency of hypoglycemic episodes using only RCT data.
**Methods:** A systematic search of MEDLINE, EMBASE, and CENTRAL was conducted for articles published between January 1, 2014 and September 13, 2021. The Cochrane Highly Sensitive Search Strategy was used to restrict to RCTs. Inclusion criteria were: participants aged ≥18 years with type 1 or type 2 diabetes, FlashGM as intervention, SMBG as control, reporting of HbA1c and at least one other glycemic outcome, intervention duration ≥10 weeks, and RCT design. Studies that blinded participants to sensor glucose data were excluded. Two independent reviewers screened studies using Covidence and extracted data using a piloted form. Risk of bias was assessed using the Cochrane RoB 2 tool. Meta-analyses used a random-effects model; heterogeneity was quantified using I² statistics. Standard errors were estimated from confidence intervals when needed. Five RCTs involving 719 participants (400 intervention, 319 control) were included, published between 2016 and 2020, with study durations of 10–24 weeks. Participants were from Europe, Australia, Japan, and Israel. One study focused on type 1 diabetes, three on type 2 diabetes, and one included both types. Mean patient age ranged from 40.2 to 67.6 years; mean baseline HbA1c ranged from 50.6 mmol/mol (6.78%) to 72.1 mmol/mol (8.75%).
**Key Results:** FlashGM did not lead to a statistically significant reduction in HbA1c compared to SMBG (mean difference: -0.17%, 95% CI -0.41 to 0.07, p = 0.164; I² = 87.2%, p < 0.001). FlashGM was associated with a statistically significant increase in time in range (3.9–10.0 mmol/L) by 1.16 hours per day (95% CI 0.13 to 2.19, p = 0.027; I² = 71.7%, p = 0.029). There was a statistically significant reduction in frequency of hypoglycemic episodes per 24 hours (mean difference: -0.28 episodes, 95% CI -0.53 to -0.04, p = 0.022; I² = 71.4%, p = 0.030). No statistically significant differences were found for time in hypoglycemia (<3.9 mmol/L) (mean difference: -0.51 hr, 95% CI -1.17 to 0.16, p = 0.137; I² = 91.4%, p < 0.001) or time in hyperglycemia (>10 mmol/L) (mean difference: -0.69 hr, 95% CI -2.45 to 1.07, p = 0.440; I² = 89.1%, p < 0.001). Risk of bias was low in all studies except one with randomization concerns; lack of blinding was noted but considered impractical for real-world FlashGM use.
**Clinical Implications:** FlashGM improves day-to-day glycemic management by extending time in target range and reducing hypoglycemic episode frequency, even without lowering HbA1c. This is clinically significant because time in range is associated with microvascular complication risk, and hypoglycemia is linked to neurocognitive dysfunction and cardiovascular events including myocardial infarction and cardiac arrhythmias. The instant access to glucose data with a painless scan promotes self-efficacy and timely behavioral adjustments regarding food, exercise, and insulin dosing. Previous meta-analyses that included observational studies reported larger HbA1c reductions (e.g., -0.55% to -0.51%), likely due to selection and performance bias. The exclusive use of RCTs in this analysis provides higher-quality evidence. Limitations include the small number of studies (precluding subgroup analysis by diabetes type), open-label design, and substantial heterogeneity across outcomes. The longest study was 24 weeks, so long-term sustainability of benefits remains unclear. Future research should include more RCTs, longer follow-up, and structured educational programs to optimize FlashGM use. The recent FDA approval of Freestyle Libre 2 with alert capabilities may further improve glycemic outcomes.