Skip to main content
Download PDF
Download ePub

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

Cardiovascular Diabetology volume 24, Article number: 210 (2025) Cite this article

Abstract

Background

Maintaining optimal glucose control is critical for postoperative care cardiac surgery patients. Continuous glucose monitoring (CGM) in this setting remains understudied. We evaluated the efficacy of CGM with a specialized titration protocol in cardiac surgery patients with type 2 diabetes (T2D) and prediabetes.

Methods

In this randomized-controlled trial, 54 cardiac surgery patients were randomized one day post-surgery, with 27 CGM and 25 point-of-care (POC) patients completing the study. The CGM group used Dexcom G6 with a CGM-specialized titration protocol, while the POC group used standard monitoring with blinded CGM. The primary outcome was time-in-range (TIR) 100–180 mg/dL for 7 days post-surgery. Secondary outcomes included various glycemic metrics and surgical outcomes. Multiple comparison adjustments were performed using false-discovery-rate (FDR).

Results

Thirty-one (59.6%) had diabetes and 21 (40.4%) had prediabetes. While TIR 100–180 mg/dL showed no difference (74.7% vs. 71.6%, FDR-adjusted p = 0.376), the CGM group demonstrated improvements in TIR 70–180 mg/dL (83.8% vs. 75.8%, FDR-adjusted p = 0.026), time-in-tight-range (TITR) 100–140 mg/dL (46.3% vs. 36.3%, FDR-adjusted p = 0.018), and TITR 70–140 mg/dL (55.3% vs. 40.5%, FDR-adjusted p = 0.003). Both groups maintained very low rates of time below range (< 70 mg/dL: 0.03% vs. 0.18%, FDR-adjusted p = 0.109). The CGM group showed lower postoperative atrial fibrillation (AF) (18.8% vs. 55.6%, FDR-adjusted p = 0.04999).

Conclusion

While the primary outcome was not achieved, CGM with a specialized titration protocol demonstrated safe glycemic control with improvements in TIR 70–180 mg/dL and TITRs in cardiac surgery patients with T2D and prediabetes. The observed reduction in postoperative AF warrants further investigation.

Trial Registration

ClinicalTrials.gov NCT06275971

Graphical Abstract

Research Insight

What is currently known about this topic?

  • Glucose dysregulation increases complications in cardiac surgery.

  • Point-of-care monitoring misses many hypo/hyperglycemic events.

  • Prior studies with standard protocols showed no time-in-range 70–180 mg/dL improvement with CGM.

What is the key research question?

  • Can CGM with specialized protocol improve outcomes in post-cardiac surgery patients with diabetes and prediabetes?

What is new?

  • First trial shows CGM improves glycemic control safely in cardiac surgery.

  • Novel protocol achieved 83.8% vs 75.8% time-in-range 70–180 mg/dL without hypoglycemia increasement.

  • CGM reduced post-op atrial fibrillation (18.8% vs. 55.6%).

How might this study influence clinical practice?

  • CGM may enable safer glucose control in cardiac surgery, but needs validation in larger trials.

Background

Cardiovascular disease (CVD) is a major complication and leading cause of death in patients with diabetes, with significant risk beginning even in the prediabetic stage. Studies show individuals with prediabetes have a 15% higher risk of CVD compared to those with normal glucose tolerance [1, 2]. Notably, patients with diabetes face higher mortality risks and adverse outcomes during cardiac surgery compared to those with normal glucose levels [3, 4]. Also, recent evidence suggests prediabetes independently increases risks of postoperative complications in cardiac surgery as well as in non-cardiac surgery [5, 6].

Perioperative hyper- or hypoglycemia leads to increased rates of surgical site infections, prolonged hospital stays, postoperative myocardial infarction, atrial fibrillation (AF), and higher mortality rates, even in patients without diagnosed diabetes [7,8,9,10]. Research has demonstrated that perioperative mean glucose levels, rather than HbA1c, more strongly influence post-surgical or procedural outcomes [8, 11, 12]. Consequently, current guidelines strongly recommend avoiding both hypo and hyperglycemia during the perioperative period [13, 14].

While achieving target glycemic levels during hospitalization requires frequent monitoring, point-of-care (POC) glucose testing has significant limitations. Studies show POC misses 31% of hyperglycemic and 86.7% of hypoglycemic episodes in hospitalized patients [15, 16]. Even in intensive care unit (ICU) settings with POC monitoring, significant hypoglycemic events below 70 mg/dL occur in 8–9% of patients, despite targeting levels ≥ 100 mg/dL [17].

Continuous glucose monitoring (CGM) offers advantages by providing real-time glucose data with numerical and graphical displays of current levels and trends. The system alerts users to actual or impending hyper- and hypoglycemia, enabling timely intervention [18]. Though initial concerns existed regarding CGM accuracy in hospitalized patients due to hypoxia, interstitial acidosis, catecholamine or vasopressin use, edema, and unstable hemodynamics, multiple studies have validated CGM's acceptable accuracy for hospitalized and perioperative use [19,20,21].

However, no interventional studies have examined CGM use specifically in cardiac surgery patients. While some randomized controlled trials (RCT) have compared CGM to POC in hospitalized patients, these were limited to stable patients in general wards and non-perioperative settings [22,23,24,25]. Additionally, these studies employed traditional POC-based protocols rather than CGM-optimized approaches, not utilizing features like trend arrows or postprandial correction injections. This limited utilization may explain their suboptimal results for time in range (TIR) of 70–180 mg/dL [22,23,24,25].

To address these knowledge gaps, we conducted an RCT evaluating real-time CGM (RT-CGM) in cardiac surgery patients with type 2 diabetes (T2D) and prediabetes, implementing a CGM-specialized in-hospital titration protocol. This study aims to provide insights into CGM's potential benefits for perioperative glycemic management in this high-risk patient population.

Methods

Trial oversight and study participants

We conducted a single-center, randomized, prospective, open-label study from November 2022 to April 2024 at Kangbuk Samsung Hospital (Institutional Review Board approval: KBSMC 2022-07-006; ClinicalTrials.gov: NCT06275971). We recruited participants scheduled for elective cardiac surgery who were > 18 years of age with T2D or prediabetes, confirmed by 75 g-oral glucose tolerance test (75 g-OGTT). Patients with high surgical severity, defined as Society of Thoracic Surgeons (STS) score ≥ 8 or European System for Cardiac Operative Risk Evaluation (EuroSCORE) ≥ 15 ", were excluded [26, 27]. Detailed in/exclusion criteria are provided in Additional file 1: Table S1.

Study design

After obtaining informed consent and baseline testing, participants wore Dexcom G6 CGM on their upper arms one or two days before surgery, linked to both smartphone and blinded receiver. All participants discontinued anti-diabetic medications and switched to insulin-only management, upon admission to the surgery ward. Prior to randomization, glycemic control relied on POC testing or arterial blood, not CGM data.

The morning after surgery, patients were randomized 1:1 to treatment or control groups, stratified by diabetes/prediabetes status. An independent statistician generated the randomization sequence using permuted blocks (size 4 or 6) using Stata version 17.0. The sequence was maintained by an uninvolved third party. In the control group, we removed the unblinded smartphone and retained only the blinded receiver, while the CGM group maintained both devices to prevent potential data loss from the smartphone. The CGM group's glycemic control was managed with alarms set at 100–180 mg/dL, with POC tests twice daily (and calibration when glucose levels differed by > 20% at ≥ 100 mg/dL, and by > 20 mg/dL at < 100 mg/dL. The calibration rule was set arbitrarily, with the cut-off value adopted from the accuracy standard of the 20/20 agreement rate [28].

CGM was removed 7 days post-surgery, with all data analyzed from the blinded receiver. Patients with < 70% available CGM data during the 7-day study period were withdrawn. We maintained standardized nursing workload with ratios of 1:2 in ICU and 1:8 in general ward, which are our hospital's routine practice, throughout the study. Endocrinologists remotely managed multiple daily insulin injections (MDI) insulin titration daily for both groups via electronic health records, with additional use of the CGM (Dexcom Clarity) interface for the CGM group, using preset algorithmic orders for correction doses. The surgical team independently determined IV-to-MDI transition timing based on post-operative recovery, regardless of glycemic control or CGM assignment. Patients received standardized hospital meals totaling 25–30 kilocalories/kilogram ideal body weight at consistent times, with no carbohydrate counting in the insulin protocol.

Titration protocol

We implemented a modified protocol incorporating CGM-specific adjustments based on existing non-CGM based protocols [29, 30]. All patients received IV insulin during fasting and continuous feeding periods, transitioning to MDI when intermittent meals resumed. In the control group, during IV insulin administration, POC testing was performed every 2–4 h, with a target range of 100–180 mg/dL. For the MDI period, POC testing was performed four times a day (every pre-meal and bedtime), and basal insulin was adjusted daily based on fasting glucose (target 100–140 mg/dL) by changing doses by 20%, while prandial insulin doses comprised a base dose plus a correction dose determined by a correction scale according to pre-prandial glucose level.

The CGM group followed similar protocols and targets but benefited from continuous monitoring and trend data. During IV insulin periods, sensor glucose was checked every 2–4 h, and trend arrows were utilized in addition to current glucose levels. For the MDI period, sensor glucose was routinely checked seven times a day (every pre-meal, 2 h post-meal, and bedtime). Basal insulin titration was guided by the nighttime CGM graph (11 PM to 5 AM in our hospital), not just by single fasting glucose level. Additional post-meal corrections were calculated and applied using a formula incorporating current glucose, trend arrows, and pre-meal insulin dose. Detailed protocols are provided in the Additional file 2.

Outcomes and clinical variables

The primary outcome was time in range (TIR) 100–180 mg/dL of sensor glucose for 7 days starting one day after surgery. The target was adopted from the current guideline for hospitalized patients without CGM [31]. Prespecified secondary glycemic outcomes included TIR 70–180 mg/dL, time in tight range (TITR) 100–140 mg/dL, time below range (TBR, < 54 mg/dL, < 70 mg/dL), time above range (TAR, > 180 mg/dL, > 250 mg/dL), mean glucose, standard deviation, and coefficient of variance (%CV). Although no CGM study has revealed TITR for cardiac surgery, since tight target (< 140 mg/dL) of cardiac surgery patients in non-CGM studies showed better outcomes and TITR 70–140 mg/dL has been regarded as a key outcome in recent CGM clinical trials, it was additionally analyzed as an exploratory secondary outcome.[32,33,34]

For surgical outcomes, we analyzed AF, pneumonia, wound infection, continuous renal replacement therapy (CRRT), symptomatic cerebrovascular accident (CVA), changes in serum C-reactive protein (CRP) level between baseline and peak within 1 week after surgery, and length of hospital stay, ICU stay, and mortality after surgery up to 30 days. As exploratory outcomes, we examined insulin dosage, TIR and TITRs across study days, and performed subgroup analyses according to diabetes and prediabetes status, and IV insulin and MDI period.

As baseline data, fasting and 2-h C-peptide levels were collected during 75 g-OGTT at the time of enrollment. The CKD-EPI equation was used to calculate estimated glomerular filtration rate. STS score and EuroSCORE were calculated based on clinical data and echocardiography findings [26, 27]. The baseline laboratory data was collected as the latest data within 2 weeks before the surgery day. CGM applied period (from one or 2 days before the surgery day to the one day before surgery day) was analyzed as baseline CGM data.

Sample size and power calculation

Based on previous research comparing POC and CGM in hospitalized patients, we calculated a required sample size of 26 per group (TIR difference 5.5%, SD 7%, 2-sided α = 0.05, power 0.8) [22]. Accounting for 30% dropout, we initially targeted 34 patients per group. However, as we experienced only two dropouts, we concluded enrollment at 54 patients, which satisfied the required sample size from our power calculation.

Statistical analysis

We conducted modified intention-to-treat (ITT) analysis as main analyses, excluding participants with < 70% CGM data or major protocol violations, with additional ITT sensitivity analyses including dropouts. We used ANCOVA for continuous outcomes and logistic regression for binary outcomes, with multiplicity adjusted using false discovery rate (FDR) based on outcome categories. FDR adjustment was not applied for the analyses across study days. For surgical outcomes, we excluded participants with pre-existing conditions at baseline. As covariates, we adjusted for diabetes and prediabetes status, baseline A1c, baseline TIR 70–180 mg/dL, and postoperative day (POD) of MDI transition due to between-group imbalance observed. For three participants with missing baseline CGM data, we performed imputation using mean values of baseline TIR 70–180 mg/dL for corresponding subgroups. We performed analyses using R version 4.1.2, with statistical significance at P < 0.05.

Results

Study population

Of 57 screened patients, 54 were randomized to CGM (n = 27) or POC (n = 27) groups. After two POC group dropouts, 52 patients were included in modified ITT analyses (Fig. 1). Mean age was 63.8 ± 10.8 years, with HbA1c of 6.3 ± 1.3% (45.1 ± 13.9 mmol/mol)(Table 1). The cohort included 31 (59.6%) diabetes and 21 (40.4%) prediabetes patients with the mean HbA1c of 6.7 ± 1.5% and 5.6 ± 0.4%, respectively (Tables S2 and S3). Newly diagnosed diabetes was identified in 11 (21.2%) patients, with 5 (20.0%) in the POC group and 6 (22.2%) in the CGM group. Baseline characteristics were similar between groups except for POD of MDI transition, with more CGM group patients transitioning after POD 4 (Table 1). Baseline characteristics by diabetes status and ITT analysis are shown in Additional file 1: Tables S2–S4.

Table 1 Baseline characteristics of total participants
Full size table
Fig. 1
figure 1

Flow diagram of study participant selection. Abbreviations: CGM, Continuous Glucose Monitoring; EURO, European System for Cardiac Operative Risk Evaluation; ITT, intention-to-treat; POC, Point-of-Care; STS, Society of Thoracic Surgeons

Full size image

Glycemic outcomes

Baseline TIR 70–180 mg/dL was 72.4 ± 26.3% and 68.8 ± 23.6% for POC and CGM groups. The primary outcome, TIR 100–180 mg/dL for the 7-day post-surgery period, showed no significant difference between POC and CGM groups (71.6 ± 17.8% vs. 74.7 ± 12.0%; adjusted difference [adj. diff.] 3.5% [95% CI − 3.4 to 10.3], P = 0.313, FDR-adjusted P = 0.376) (Table 2 and Fig. 2A).

Table 2 CGM and surgical outcomes of total participants
Full size table
Fig. 2
figure 2

Proportion of time in ranges and time in tight ranges. P values are for FDR-adjusted analyses, where diabetes and prediabetes status, baseline A1c, baseline TIR 70–180 mg/dL, and postoperative day of multiple daily injection transition were adjusted for. Abbreviations: CGM, continuous glucose monitoring; DM, diabetes; FDR, false discovery rate; POC, point-of-care; PreDM, prediabetes; TIR, time in range; TITR, time in tight range

Full size image

Several secondary glycemic outcomes improved in the CGM group. TIR 70–180 mg/dL was higher (75.8 ± 18.9% vs. 83.8 ± 10.6%; adj. diff. 8.9% [95% CI 1.9 to 15.6], FDR-adjusted P = 0.026). TITR 100–140 mg/dL and 70–140 mg/dL also improved (36.3 ± 19.0% vs. 46.3 ± 15.4%; adj. diff. 11.4% [95% CI 3.5 to 19.4], FDR-adjusted P = 0.018 and 40.5 ± 22.5% vs. 55.3 ± 15.3%; adj. diff. 16.8% [95% CI 7.5 to 26.0], FDR-adjusted P = 0.003, respectively) (Table 2 and Fig. 2B–D). Cumulative distributions and proportions of TIRs and TITRs are presented in Fig. 3 and Additional file 1: Figure S1, respectively. TIR 70–180 mg/dL and TITRs began to show improvement in the CGM group from day 2 of the study period, though not consistently across all days (Additional File 1: Figure S2).

Fig. 3
figure 3

Cumulative distributions of time in ranges and time in tight ranges among total population. Abbreviations: CGM, continuous glucose monitoring; IV, intravenous; MDI, multiple daily insulin injection; POC, point-of-care

Full size image

TAR > 180 mg/dL and mean glucose were lower in the CGM group. Both groups showed minimal TBR < 70 mg/dL (0.18 ± 0.39% vs. 0.03 ± 0.11%, FDR-adjusted P = 0.109) and < 54 mg/dL (0.00 ± 0.02% vs. 0.01 ± 0.04%, FDR-adjusted P = 0.420). CV was higher in the CGM group (22.0 ± 6.2% vs. 25.0 ± 5.9%, FDR-adjusted P = 0.029), while SD remained similar (35.3 ± 13.8 mg/dL vs. 36.2 ± 10.6 mg/dL, FDR-adjusted P = 0.733) (Table 2). Ambulatory glucose profiles according to the POC and CGM groups are presented in Additional file 1: Figure S3A. As sensitivity analyses, ITT analyses showed similar results presented in Additional file 1: Table S5.

Subgroup analyses

In patients with diabetes, TIR 100–180 mg/dL showed no difference between groups (65.2 ± 17.2% vs. 68.2 ± 10.4%; FDR-adjusted P = 0.393). However, TIR 70–180 mg/dL (68.5 ± 17.9% vs. 78.5 ± 9.1%; adj. diff. 12.4% [95% CI 2.9 to 21.8], FDR-adjusted P = 0.030) and TITR 70–140 mg/dL were improved in the CGM group (Fig. 2, Additional file 1: Table S6, Figures S1, S4, and S5). In prediabetes patients, glycemic outcomes were comparable between groups, with the CGM group showing improvement in TITRs only on several specific days (Fig. 2, Additional file 1: Table S7, Figures S1, S6, and S7).

During IV insulin periods, TIR 70–180 mg/dL showed initial improvement (64.6 ± 33.3% vs 85.0 ± 13.7%; adj. diff. 23.6% [95% CI 1.1 to 46.0]) but lost significance after FDR adjustment (Additional file 1: Table S8 and Figure S3B). In MDI periods, the CGM group showed improvements in TITR 100–140 mg/dL (37.5 ± 18.9% vs 48.4 ± 16.7%; adj. diff. 11.1% [95% CI 2.8 to 19.3], P = 0.029) and TITR 70–140 mg/dL (42.1 ± 22.7% vs 5.8.1 ± 18.1%; adj. diff. 16.3% [95% CI 6.3 to 26.3], P = 0.010) (Additional file 1: Table S9 and Figure S3C).

Surgical outcomes

Postoperative AF incidence was lower in the CGM group (55.6% vs. 18.8%; adjusted odds ratio 0.04 [95% CI 0.00 to 0.46], FDR-adjusted P = 0.04999) (Table 2). Other complications, including pneumonia, wound infection, CRRT, symptomatic CVA, and 30-day mortality, showed no differences. ICU and hospital stays were comparable between groups. Surgical outcomes in the diabetes subgroup showed no significant differences (Additional file 1: Table S10). In prediabetes patients, the CGM group showed lower CRP level increases (adj. diff. − 9.3 mg/dL [95% CI − 16.3 to − 2.2], FDR-adjusted P = 0.040) (Additional file 1: Table S11). There were no unintended adverse effects in each group regarding to the intervention. As sensitivity analyses, ITT analyses showed similar results presented in Additional file 1: Table S5.

Insulin requirements

Post-prandial positive correction insulin doses (0.000 ± 0.000 vs. 0.009 ± 0.011 U/kg/day; FDR-adjusted P < 0.001) and frequency (0.0 ± 0.0 vs. 0.3 ± 0.3 times/day, FDR-adjusted P < 0.001) were higher in the CGM group. Pre-prandial corrections, total correction doses, and total daily insulin requirements were comparable between groups (Additional file 1: Table S12). Initial differences in total correction frequency (1.1 ± 1.0 vs. 1.3 ± 1.1 times/day; adj. diff. 0.4 [95% CI 0.1 to 0.8]) lost significance after FDR adjustment. Insulin requirements for diabetes and prediabetes subgroups are presented in Additional file 1: Tables S13 and S14.

Discussion

In this randomized controlled trial of cardiac surgery patients with diabetes or prediabetes, RT-CGM with a CGM-specialized protocol showed improvements in several secondary glycemic metrics, despite no significant difference in the primary outcome of TIR 100–180 mg/dL. The CGM group achieved higher TIR 70–180 mg/dL (75.8% vs. 83.8%, FDR-adjusted P = 0.026) and TITR for both 100–140 mg/dL (36.3% vs. 46.3%, FDR-adjusted P = 0.018) and 70–140 mg/dL (40.5% vs. 55.3%, FDR-adjusted P = 0.003). Both groups maintained minimal hypoglycemia (TBR < 70 mg/dL: 0.03% vs. 0.18%, FDR-adjusted P = 0.109). Notably, AF incidence was lower in the CGM group (55.6% vs. 18.8%, FDR-adjusted P = 0.04999). These findings suggest potential benefits of CGM use in the post-operative setting, though further research is needed to confirm these observations.

Previous inpatient CGM studies have primarily focused on accuracy [19,20,21]. Moreover, there have been no intervention studies using CGM in cardiac surgery patients for perioperative use, where glucose control is critical for outcomes. Of the four existing RCTs using RT-CGM in hospitalized patients, none specifically targeted cardiac surgery patients or perioperative settings, and most were conducted in non-ICU settings [22,23,24,25]. These studies showed varying results: Fortmann et al. found no significant difference in TIR 70–180 mg/dL but improved TIR 70–250 mg/dL (64.0% vs. 72.9%, P = 0.0404) [22]; Singh et al. improved TBR < 70 mg/dL (1.88% vs. 0.40%, P = 0.002) but not TIR [23]; and neither Klarskov et al. nor Spanakis et al. found significant differences in TIR 70–180 mg/dL [24, 25]. Given that glucose control immediately after surgery is critical for cardiac patients, our study addresses an important gap.

Our inability to demonstrate significance in the primary outcome (TIR 100–180 mg/dL) may reflect our choice of target range, which was based on pre-CGM inpatient guidelines [31]. At the time of study design, there was no consensus on inpatient CGM targets. While our choice of primary outcome may have influenced the statistical findings, it is worth noting that other inpatient studies have used TIR 70–180 mg/dL as their primary outcome [24, 25], which may be better suited for capturing the full range of glycemic control in the inpatient setting. Although TIR 70–180 mg/dL was a secondary outcome in our study, the observed difference with multiplicity correction between groups suggests potential benefits. The marked improvement in TIR 70–180 mg/dL, TITRs, and extremely low hypoglycemia rate likely stemmed from our CGM-specialized titration protocol.

Previous studies used well-designed standardized protocols. Fortmann et al. adjusted basal insulin based on fasting glucose and early morning hypoglycemia, with correction insulin for high pre-meal or bedtime glucose [22]. Singh et al. applied basal dose adjustment based on fasting glucose [23], while Spanakis et al. employed a systematic approach, adjusting basal insulin based on fasting and pre-dinner glucose, with correction insulin for high pre-meal and bedtime glucose [25]. However, these protocols relied on single-point glucose measurements, which was no different from POC, and did not utilize CGM's benefits. Our study aimed to utilize CGM's advantages over POC, including graphical information, trend arrows, and ease of frequent glucose checks (Additional file 2). While total daily insulin doses remained similar between groups, additional post-prandial corrections may have contributed to improved glycemic outcomes.

Traditional approaches suggest higher glucose targets for cardiac surgery patients due to hypoglycemia risks [35, 36]; studies showed 120–180 mg/dL was preferable to 90–120 mg/dL [37], 110–140 mg/dL was better than 80–110 mg/dL [38], and 141–180 mg/dL was superior to 100–140 mg/dL [17]. However, for non-diabetic patients, there have been evidences that lower targets (100–140 mg/dL and 80–110 mg/dL) are associated with lower post-cardiac operative complications [17, 39]. Therefore, current American Diabetes Association guideline recommend 141–180 mg/dL, with 110–140 mg/dL for patients at low risk of hypoglycemia [13]. However, those studies did not utilize CGM. Our study demonstrated that RT-CGM's alarm function and trend arrows enabled safe achievement of broader, lower targets (100–180 mg/dL) while maintaining minimal hypoglycemia and high TIR/TITR.

The lower postoperative AF incidence in the CGM group, though a secondary finding, warrants attention. Postoperative AF negatively affects both short-term outcomes and long-term survival [40, 41]. Previous research showed reduced AF with better glycemic control in cardiac surgery patients (16.6% vs. 42% for targets 125–200 mg/dL vs < 250 mg/dL) [10]. Hyperglycemia can induce arrhythmia through Ca2 + /calmodulin-dependent kinase II alterations [42]. While some studies suggest links between hypoglycemia and AF [43, 44], recent evidence from a long-term study in type 1 diabetes indicates hyperglycemia, rather than hypoglycemia, may be more strongly associated with arrhythmia risk [45]. However, larger studies are needed to confirm these relationships, as our study wasn't powered for this outcome.

Our study's strengths include being the first RCT of CGM in cardiac surgery patients during immediate post-operative care, which is crucial for these patients, spanning both ICU and general ward settings. Furthermore, we presented results for a broader range of indications, including not only diabetes but also prediabetes patients. We demonstrated the feasibility of a CGM-specialized insulin titration protocol for hospitalized patients that achieved its safety objectives.

Several limitations should be noted. First, the lack of significant primary outcome difference may reflect our choice of outcome measure, though we performed multiplicity correction for secondary outcomes. Second, our single-center design and relatively small sample size, although based on proper calculations, limit generalizability. Particularly for subgroup analyses, since these weren't included in original power calculations, results should be interpreted as exploratory. Third, the open-label nature potentially introduced bias, despite following predetermined protocols. Fourth, additional healthcare provider (HCP) workload was required due to post-prandial correction insulin injections and alarms for the CGM group. While this may have been compensated by reduced POC workload, satisfaction surveys for HCPs should be included in future studies. Fifth, while CGM accuracy in inpatient settings remains a concern, we used a well-studied system (Dexcom G6) [19,20,21] with calibration protocols for significant POC differences. Sixth, since OGTTs were performed while some patients were on diuretics and beta-blockers, which can affect glucose levels, there is a possibility that diabetes or prediabetes might have been misclassified. Seventh, since there was no study comparing TIR between CGM vs. POC among prediabetes patients at the time of study design, we calculated the sample size based on studies of diabetes patients. Lastly, this study enrolled relatively well-controlled diabetes patients and prediabetes patients with relatively low TDD. However, considering the higher tendency of TDD and more frequent correction insulin in the CGM group even in this low TDD setting, it may be more effective in higher insulin-requiring participants, which may need further study.

Conclusions

In conclusion, while our study didn't achieve its primary outcome, it provides important insights into RT-CGM use with specialized protocols for cardiac surgery patients with diabetes and prediabetes. The significant improvements in secondary outcomes, including TIR 70–180 mg/dL and TITRs without increased hypoglycemia risk, along with the observed lower incidence of postoperative AF, suggest this approach warrants further evaluation in larger trials. Our findings indicate that CGM may enable safer achievement of tighter glycemic targets in this high-risk population.

Availability of data and materials

Requests for access to the study data will be considered by the investigators and may be shared at their discretion following review of the research proposal. The titration protocol is available in Additional file 2.

Abbreviations

AF:

Atrial fibrillation

CI:

Confidence interval

CGM:

Continuous glucose monitoring

CRP:

C-reactive protein

CRRT:

Continuous renal replacement therapy

CV:

Coefficient of variation

CVA:

Cerebrovascular accident

CVD:

Cardiovascular disease

DPP4:

Dipeptidyl peptidase-4

EURO:

European system for cardiac operative risk evaluation

FDR:

False discovery rate

GFR:

Glomerular filtration rate

GLP-1:

Glucagon-like peptide-1

HCP:

Healthcare provider

ICU:

Intensive care unit

IRB:

Institutional review board

ITT:

Intention-to-treat

MDI:

Multiple daily injection

NA:

Not applicable

OGTT:

Oral glucose tolerance test

POC:

Point-of-care

POD:

Post-operative day

RCT:

Randomized controlled trial

RT-CGM:

Real-time continuous glucose monitoring

SD:

Standard deviation

SGLT2:

Sodium-glucose cotransporter-2

STS:

Society of thoracic surgeons

T2D:

Type 2 diabetes

TAR:

Time above range

TBR:

Time below range

TIR:

Time in range

TITR:

Time in tight range

TZD:

Thiazolidinedione

References

  1. Rawshani A, Rawshani A, Franzén S, Sattar N, Eliasson B, Svensson AM, Zethelius B, Miftaraj M, McGuire DK, Rosengren A, et al. Risk factors, mortality, and cardiovascular outcomes in patients with type 2 diabetes. N Engl J Med. 2018;379(7):633–44.

    Article  PubMed  Google Scholar 

  2. Cai X, Zhang Y, Li M, Wu JH, Mai L, Li J, Yang Y, Hu Y, Huang Y. Association between prediabetes and risk of all cause mortality and cardiovascular disease: updated meta-analysis. BMJ. 2020;370: m2297.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Carson JL, Scholz PM, Chen AY, Peterson ED, Gold J, Schneider SH. Diabetes mellitus increases short-term mortality and morbidity in patients undergoing coronary artery bypass graft surgery. J Am Coll Cardiol. 2002;40(3):418–23.

    Article  PubMed  Google Scholar 

  4. Corazzari C, et al. Impact of preoperative glycometabolic status on outcomes in cardiac surgery: systematic review and meta-analysis. J Thorac Cardiovasc Surg. 2022;164(6):1950–19601910.

    Article  PubMed  Google Scholar 

  5. Mazhar T, Ahmed Z, Munir K, Tabassum A, Khan IH, Aslam I, Lodhi K. The impact of pre-diabetes on postoperative outcomes in cardiac surgery patients: a comprehensive analysis. J Popul Ther Clin Pharmacol. 2024;31(5):360–7.

    Google Scholar 

  6. Kwon S, Thompson R, Dellinger P, Yanez D, Farrohki E, Flum D. Importance of perioperative glycemic control in general surgery: a report from the Surgical Care and Outcomes Assessment Program. Ann Surg. 2013;257(1):8–14.

    Article  PubMed  Google Scholar 

  7. Ascione R, Rogers CA, Rajakaruna C, Angelini GD. Inadequate blood glucose control is associated with in-hospital mortality and morbidity in diabetic and nondiabetic patients undergoing cardiac surgery. Circulation. 2008;118(2):113–23.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. van den Boom W, Schroeder RA, Manning MW, Setji TL, Fiestan GO, Dunson DB. Effect of A1C and glucose on postoperative mortality in noncardiac and cardiac surgeries. Diabetes Care. 2018;41(4):782–8.

    Article  PubMed  Google Scholar 

  9. Frisch A, Chandra P, Smiley D, Peng L, Rizzo M, Gatcliffe C, Hudson M, Mendoza J, Johnson R, Lin E, et al. Prevalence and clinical outcome of hyperglycemia in the perioperative period in noncardiac surgery. Diabetes Care. 2010;33(8):1783–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Lazar HL, Chipkin SR, Fitzgerald CA, Bao Y, Cabral H, Apstein CS. Tight glycemic control in diabetic coronary artery bypass graft patients improves perioperative outcomes and decreases recurrent ischemic events. Circulation. 2004;109(12):1497–502.

    Article  CAS  PubMed  Google Scholar 

  11. Lindsay J, Sharma AK, Canos D, Nandalur M, Pinnow E, Apple S, Ruotolo G, Wijetunga M, Waksman R. Preprocedure hyperglycemia is more strongly associated with restenosis in diabetic patients after percutaneous coronary intervention than is hemoglobin A1C. Cardiovasc Revascularization Med Incl Mol Interv. 2007;8(1):15–20.

    Article  Google Scholar 

  12. Moon SJ, Ahn CH, Lee YB, Cho YM. Impact of hyperglycemia on complication and mortality after transarterial chemoembolization for hepatocellular carcinoma. Diabetes Metab J. 2024;48(2):302–11.

    Article  PubMed  PubMed Central  Google Scholar 

  13. American Diabetes Association: 16. Diabetes care in the hospital: standards of care in diabetes—2025. Diabetes Care 2024;48(Supplement_1):S321–34.

  14. Duggan EW, Carlson K, Umpierrez GE. Perioperative hyperglycemia management: an update. Anesthesiology. 2017;126(3):547–60.

    Article  PubMed  Google Scholar 

  15. Burt MG, Roberts GW, Aguilar-Loza NR, Stranks SN. Brief report: comparison of continuous glucose monitoring and finger-prick blood glucose levels in hospitalized patients administered basal-bolus insulin. Diabetes Technol Ther. 2013;15(3):241–5.

    Article  CAS  PubMed  Google Scholar 

  16. Gómez AM, Umpierrez GE, Muñoz OM, Herrera F, Rubio C, Aschner P, Buendia R. Continuous glucose monitoring versus capillary point-of-care testing for inpatient glycemic control in type 2 diabetes patients hospitalized in the general ward and treated with a basal bolus insulin regimen. J Diabetes Sci Technol. 2015;10(2):325–9.

    Article  PubMed  PubMed Central  Google Scholar 

  17. Umpierrez G, Cardona S, Pasquel F, Jacobs S, Peng L, Unigwe M, Newton CA, Smiley-Byrd D, Vellanki P, Halkos M, et al. Randomized controlled trial of intensive versus conservative glucose control in patients undergoing coronary artery bypass graft surgery: GLUCO-CABG trial. Diabetes Care. 2015;38(9):1665–72.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Cappon G, Vettoretti M, Sparacino G, Facchinetti A. Continuous glucose monitoring sensors for diabetes management: a review of technologies and applications. Diabetes Metab J. 2019;43(4):383–97.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Perez-Guzman MC, Duggan E, Gibanica S, Cardona S, Corujo-Rodriguez A, Faloye A, Halkos M, Umpierrez GE, Peng L, Davis GM, et al. Continuous glucose monitoring in the operating room and cardiac intensive care unit. Diabetes Care. 2021;44(3):e50–2.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Insler SR, Wakefield B, Debs A, Brake K, Nwosu I, Isaacs D, Bena J, Lansang MC. Continuous glucose monitoring using the Dexcom G6 in cardiac surgery during the postoperative period. Endocr Pract Off J Am Coll Endocrinol Am Assoc Clin Endocrinol. 2024;30(7):610–5.

    CAS  Google Scholar 

  21. Voglová Hagerf B, Protus M, Nemetova L, Mraz M, Kieslichova E, Uchytilova E, Indrova V, Lelito J, Girman P, Haluzík M, et al. Accuracy and feasibility of real-time continuous glucose monitoring in critically Ill patients after abdominal surgery and solid organ transplantation. Diabetes Care. 2024;47(6):956–63.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Fortmann AL, Spierling Bagsic SR, Talavera L, Garcia IM, Sandoval H, Hottinger A, Philis-Tsimikas A. Glucose as the fifth vital sign: a randomized controlled trial of continuous glucose monitoring in a Non-ICU hospital setting. Diabetes Care. 2020;43(11):2873–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Singh LG, Satyarengga M, Marcano I, Scott WH, Pinault LF, Feng Z, Sorkin JD, Umpierrez GE, Spanakis EK. Reducing inpatient hypoglycemia in the general wards using real-time continuous glucose monitoring: the glucose telemetry system, a randomized clinical trial. Diabetes Care. 2020;43(11):2736–43.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Klarskov CK, Windum NA, Olsen MT, Dungu AM, Jensen AK, Lindegaard B, Pedersen-Bjergaard U, Kristensen PL. Telemetric continuous glucose monitoring during the COVID-19 pandemic in isolated hospitalized patients in Denmark: a randomized controlled exploratory trial. Diabetes Technol Ther. 2022;24(2):102–12.

    Article  CAS  PubMed  Google Scholar 

  25. Spanakis EK, Urrutia A, Galindo RJ, Vellanki P, Migdal AL, Davis G, Fayfman M, Idrees T, Pasquel FJ, Coronado WZ, et al. Continuous glucose monitoring-guided insulin administration in hospitalized patients with diabetes: a randomized clinical trial. Diabetes Care. 2022;45(10):2369–75.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Barbash IM, Finkelstein A, Barsheshet A, Segev A, Steinvil A, Assali A, Ben Gal Y, Vaknin Assa H, Fefer P, Sagie A, et al. Outcomes of patients at estimated low, intermediate, and high risk undergoing transcatheter aortic valve implantation for aortic stenosis. Am J Cardiol. 2015;116(12):1916–22.

    Article  PubMed  Google Scholar 

  27. Andrade IN, Moraes Neto FR, Andrade TG. Use of EuroSCORE as a predictor of morbidity after cardiac surgery. Revista brasileira de cirurgia cardiovascular: orgao oficial da Sociedade Brasileira de Cirurgia Cardiovascular. 2014;29(1):9–15.

    PubMed  Google Scholar 

  28. Mader JK, Waldenmaier D, Mueller-Hoffmann W, Mueller K, Angstmann M, Vogt G, Rieger CC, Eichenlaub M, Forst T, Freckmann G. Performance of a novel continuous glucose monitoring device in people with diabetes. J Diabetes Sci Technol. 2024;18(5):1044–51.

    Article  PubMed  Google Scholar 

  29. Umpierrez GE, Smiley D, Zisman A, Prieto LM, Palacio A, Ceron M, Puig A, Mejia R. Randomized study of basal-bolus insulin therapy in the inpatient management of patients with type 2 diabetes (RABBIT 2 trial). Diabetes Care. 2007;30(9):2181–6.

    Article  CAS  PubMed  Google Scholar 

  30. Dagogo-Jack S, Alberti KGMM. Management of diabetes mellitus in surgical patients. Diabetes Spectrum. 2002;15(1):44–8.

    Article  Google Scholar 

  31. Korytkowski MT, Muniyappa R, Antinori-Lent K, Donihi AC, Drincic AT, Hirsch IB, Luger A, McDonnell ME, Murad MH, Nielsen C, et al. Management of hyperglycemia in hospitalized adult patients in non-critical care settings: an endocrine society clinical practice guideline. J Clin Endocrinol Metab. 2022;107(8):2101–28.

    Article  PubMed  PubMed Central  Google Scholar 

  32. Jin X, Wang J, Ma Y, Li X, An P, Wang J, Mao W, Mu Y, Chen Y, Chen K. Association between perioperative glycemic control strategy and mortality in patients with diabetes undergoing cardiac surgery: a systematic review and meta-analysis. Front Endocrinol. 2020;11: 513073.

    Article  Google Scholar 

  33. De Meulemeester J, Charleer S, Visser MM, De Block C, Mathieu C, Gillard P. The association of chronic complications with time in tight range and time in range in people with type 1 diabetes: a retrospective cross-sectional real-world study. Diabetologia. 2024;67(8):1527–35.

    Article  PubMed  Google Scholar 

  34. Zhang Z, Wang Y, Lu J, Zhou J. Time in tight range: a key metric for optimal glucose control in the era of advanced diabetes technologies and therapeutics. Diabetes Obes Metab. 2025;27(2):450–6.

    Article  PubMed  Google Scholar 

  35. Brunkhorst FM, Engel C, Bloos F, Meier-Hellmann A, Ragaller M, Weiler N, Moerer O, Gruendling M, Oppert M, Grond S, et al. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med. 2008;358(2):125–39.

    Article  CAS  PubMed  Google Scholar 

  36. Kansagara D, Fu R, Freeman M, Wolf F, Helfand M. Intensive insulin therapy in hospitalized patients: a systematic review. Ann Intern Med. 2011;154(4):268–82.

    Article  PubMed  Google Scholar 

  37. Lazar HL, McDonnell MM, Chipkin S, Fitzgerald C, Bliss C, Cabral H Effects of aggressive versus moderate glycemic control on clinical outcomes in diabetic coronary artery bypass graft patients. Ann Surg 2011;254(3):458–463; discussion 463–454

  38. Mulla I, Schmidt K, Cashy J, Wallia A, Andrei AC, Johnson Oakes D, Aleppo G, Li C, Grady KL, McGee E, et al. Comparison of glycemic and surgical outcomes after change in glycemic targets in cardiac surgery patients. Diabetes Care. 2014;37(11):2960–5.

    Article  CAS  PubMed  Google Scholar 

  39. Bláha J, Mráz M, Kopecký P, Stříteský M, Lipš M, Matias M, Kunstýř J, Pořízka M, Kotulák T, Kolníková I, et al. Perioperative tight glucose control reduces postoperative adverse events in nondiabetic cardiac surgery patients. J Clin Endocrinol Metab. 2015;100(8):3081–9.

    Article  PubMed  Google Scholar 

  40. LaPar DJ, Speir AM, Crosby IK, Fonner E, Jr., Brown M, Rich JB, Quader M, Kern JA, Kron IL, Ailawadi G: Postoperative atrial fibrillation significantly increases mortality, hospital readmission, and hospital costs. Ann Thorac Surg 2014;98(2):527–533; discussion 533

  41. El-Chami MF, Kilgo P, Thourani V, Lattouf OM, Delurgio DB, Guyton RA, Leon AR, Puskas JD. New-onset atrial fibrillation predicts long-term mortality after coronary artery bypass graft. J Am Coll Cardiol. 2010;55(13):1370–6.

    Article  PubMed  Google Scholar 

  42. Erickson JR, Pereira L, Wang L, Han G, Ferguson A, Dao K, Copeland RJ, Despa F, Hart GW, Ripplinger CM, et al. Diabetic hyperglycaemia activates CaMKII and arrhythmias by O-linked glycosylation. Nature. 2013;502(7471):372–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Ko SH, Park YM, Yun JS, Cha SA, Choi EK, Han K, Han E, Lee YH, Ahn YB. Severe hypoglycemia is a risk factor for atrial fibrillation in type 2 diabetes mellitus: nationwide population-based cohort study. J Diabetes Complicat. 2018;32(2):157–63.

    Article  Google Scholar 

  44. Celebi S, Celebi OO, Aydogdu S, Diker E. A peculiar medical cardioversion of atrial fibrillation with glucose infusion—a rare cause of atrial fibrillation: hypoglycemia. Am J Emerg Med. 2011;29(1):134.e131-133.

    Article  Google Scholar 

  45. Hagelqvist PG, Andersen A, Maytham K, Andreasen CR, Engberg S, Lindhardt TB, Forman JL, Pedersen-Bjergaard U, Knop FK, Vilsbøll T. Glycaemia and cardiac arrhythmias in people with type 1 diabetes: a prospective observational study. Diabetes Obes Metab. 2023;25(8):2300–9.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

We thank all members of the Division of Endocrinology and Metabolism and Department of Thoracic and Cardiovascular Surgery at Kangbuk Samsung Hospital for their valuable contributions to this study. We especially acknowledge the nursing staff in both the ICU and general ward for their dedication to protocol implementation.

Funding

This research was conducted as an investigator-led study. Although Dexcom contributed to the study by providing the CGM devices, their input was limited to technical support. The company reviewed the study protocol for funding purposes but had no role in protocol development or study execution. All aspects of data management, monitoring, and analysis were independently conducted by the coordinating center at Kangbuk Samsung Hospital in Seoul, Korea.

Author information

Author notes

  1. Sun-Joon Moon and Min-Su Kim have contributed equally as the first author.

Authors and Affiliations

  1. Division of Endocrinology and Metabolism, Department of Internal Medicine, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, 29 Saemunan-ro, Jongno-gu, Seoul, 03181, Republic of Korea

    Sun-Joon Moon, Su-Ji Lee & Cheol-Young Park

  2. Thoracic and Cardiovascular Surgery, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, 29 Saemunan-ro, Jongno-gu, Seoul, 03181, Republic of Korea

    Min-Su Kim, Ha-Eun Lee, Young-Woo Lee & Euy-Suk Chung

  3. Division of Biostatistics, Department of Academic Research, Kangbuk Samsung Hospital, Seoul, Republic of Korea

    Yun Tae Kim

Authors
  1. Sun-Joon Moon
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Min-Su Kim
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Yun Tae Kim
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Ha-Eun Lee
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Young-Woo Lee
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Su-Ji Lee
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Euy-Suk Chung
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Cheol-Young Park
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

S.J.M. was involved in the literature review and synthesis, project conception, protocol development, statistical analysis, critical discussion, and drafting of the first manuscript. M.S.K. and E.S.C. were involved in patient enrollment, project conception, protocol development, data acquisition, critical discussion, and revising the manuscript. H.E.L., Y.W.L., and S.J.L. were involved in data acquisition, device management, and conducting the protocol. C.Y.P. was involved in the project conception, critical discussion, and revising the manuscript. S.J.M. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Corresponding authors

Correspondence to Euy-Suk Chung or Cheol-Young Park.

Ethics declarations

Ethics approval and consent to participate

The study was conducted following approval from the Institutional Review Board of Kangbuk Samsung Hospital (KBSMC 2022-07-006). Written informed consent was obtained from all study participants before enrollment.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Moon, SJ., Kim, MS., Kim, Y.T. et al. 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 24, 210 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12933-025-02747-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12933-025-02747-z

Keywords

Cardiovascular Diabetology

ISSN: 1475-2840