Summary: In 54 cardiac surgery patients with type 2 diabetes or prediabetes, a continuous glucose monitoring (CGM) based insulin titration protocol did not meet its primary endpoint of time-in-range 100-180 mg/dL (74.7% vs 71.6%, FDR-adjusted p=0.376). Several secondary glycaemic metrics, including time-in-range 70-180 mg/dL and tight-range measures, were better in the CGM arm, and a reduction in postoperative atrial fibrillation was observed as a secondary signal.
PICO Summary
| Element | Detail |
|---|---|
| Population | 54 randomized cardiac surgery patients with type 2 diabetes (n=31, 59.6%) or prediabetes (n=21, 40.4%); single-centre randomized controlled trial, Republic of Korea. 52 completed (27 CGM, 25 point-of-care). |
| Intervention | Dexcom G6 CGM with a specialised insulin titration protocol, started one day post-surgery for 7 days (CGM arm, n=27). |
| Comparison | Standard point-of-care glucose monitoring with blinded CGM, conventional titration (POC arm, n=25). |
| Outcome | Primary (not met): TIR 100-180 mg/dL 74.7% vs 71.6%, FDR-adjusted p=0.376. Secondary (favouring CGM): TIR 70-180 mg/dL 83.8% vs 75.8% (p=0.026); TITR 100-140 mg/dL 46.3% vs 36.3% (p=0.018); TITR 70-140 mg/dL 55.3% vs 40.5% (p=0.003); time below 70 mg/dL 0.03% vs 0.18% (p=0.109, ns). Postoperative atrial fibrillation 18.8% vs 55.6% (FDR-adjusted p=0.04999). No 95% CIs, ARR or NNT reported in the abstract. |
CGM-guided insulin titration after cardiac surgery
RCT · T2D/prediabetes · 7 days
The primary endpoint (TIR 100-180 mg/dL) was not met. Wider time-in-range and tight-range metrics favoured CGM, but with only 52 completers these are hypothesis-generating, not practice-changing.
Expert Commentary
This trial should be read as negative on its own terms: the pre-specified primary endpoint, time-in-range 100-180 mg/dL, did not differ between arms, so the headline question of whether a CGM-guided protocol improves the primary glucose target after cardiac surgery is answered no. The positive results sit entirely in secondary metrics (wider time-in-range 70-180 mg/dL and the tight-range measures), which are biologically coherent and consistent in direction but remain hypothesis-generating rather than confirmatory. The single most important limitation is sample size: with only 52 completers across two arms, the study is underpowered for both the primary endpoint and the clinical outcomes, and the atrial fibrillation difference, with a p-value of 0.04999, is statistically fragile and almost certainly a chance signal until replicated. The titration intervention was necessarily open-label, since clinicians acted on real-time CGM data while the comparator arm was blinded, which can bias process-of-care outcomes. Can I use this with my patients? Not yet. The data do not justify changing post-cardiac-surgery glucose protocols, though they support running a larger, adequately powered trial with atrial fibrillation as a pre-specified endpoint. I would like to see that confirmatory study before any practice change is considered.
References
Moon SJ, Kim MS, Kim YT, Lee HE, Lee YW, Lee SJ, Chung ES, Park CY. Use of an insulin titration protocol based on continuous glucose monitoring in postoperative cardiac surgery patients with type 2 diabetes and prediabetes: a randomized controlled trial. Cardiovasc Diabetol. 2025;24(1):210. doi:10.1186/s12933-025-02747-z