Skip to main content
AACE Endocrinology and Diabetes logoLink to AACE Endocrinology and Diabetes
. 2026 Feb 12;13(3):452–462. doi: 10.1016/j.aed.2026.02.002

Comparison of Perioperative Glycemic Control in Patients Continuing Versus Temporarily Holding Glucagon-Like Peptide-1 Receptor Agonists or Glucose-Dependent Insulinotropic Polypeptide/Glucagon-Like Peptide-1 Receptor Agonists Preoperatively

Caia C Trykall 1, Sarah B Hill 1, Amy M James 1, Aaron J Bagnola 1, Kathleen L Wyne 2, Demond Handley 3, Kristin I Brower 1,∗
PMCID: PMC13221920  PMID: 42221394

Abstract

Objective

Glucagon-like peptide-1 (GLP-1) receptor agonists (RAs) and glucose-dependent insulinotropic polypeptide (GIP)/GLP-1 RAs are widely used medications for the treatment of type 2 diabetes (T2D). The purpose of this study is to evaluate the perioperative impact on blood glucose of temporarily holding versus continuing weekly-dosed GLP-1 or GIP/GLP-1 RAs prior to elective surgery in patients with T2D.

Methods

This was a retrospective, single-center cohort study evaluating patients before and after a change in institutional guidelines for continuing versus holding GLP-1 or GIP/GLP-1 RA medications prior to surgery (January 1-May 31, 2022 versus January 1-May 31, 2024). The primary outcome was the number of patients in the overall cohort with preoperative glycemic control, defined as blood glucose ≤ 180 mg/dL.

Results

A total of 168 encounters were included in the final analysis, 81 encounters in the pre-recommendation group and 87 encounters in the post-recommendation group. Sixty-three patients (77.8%) in the pre-recommendation group achieved glycemic control compared to seventy-two patients (82.8%) in the post-recommendation group (P = 0.443; risk difference 0.05, 95% CI −0.07 to 0.17). After matching fifty-seven encounters in both groups, there was not a significant association between the recommendation group and the primary outcome of preoperative glycemic control (80.7% vs 84.2%) (P = 0.824; risk difference 0.04, 95% CI −0.12 to 0.19).

Conclusion

In both the overall cohort and the matched cohort, there was no statistically significant difference in the rate of glycemic control whether GLP-1 or GIP/GLP-1 RA therapy was continued or temporarily held preoperatively in patients with T2D.

Key words: diabetes mellitus, glucagon-like peptide 1, glycemic control, perioperative, glucose-dependent insulinotropic polypeptide


Highlights

  • •

    In the overall cohort and matched cohort, there was no statistically significant difference in the rate of glycemic control whether glucagon-like peptide-1 (GLP-1) or glucose-dependent insulinotropic polypeptide/GLP-1 receptor agonist (RA) therapy was continued or temporarily held preoperatively in patients with type 2 diabetes

  • •

    Adverse events of interest assessed showed 1 patient in 168 total encounters having an aspiration event (though not intraoperatively) in the pre-recommendation group in which their GLP-1 RA was continued and 1 patient in 168 total encounters readmitted for surgical site infection in the post-recommendation group in which their GLP-1 RA was withheld

Clinical Relevance

This is a single-center, retrospective cohort study comparing the glycemic control of patients with type 2 diabetes who underwent an elective procedure when their glucagon-like peptide-1 (GLP-1) or glucose-dependent insulinotropic polypeptide/GLP-1 receptor agonists were either continued or temporarily held preoperatively. No previous publications have explored the effects on glucose control after this intervention.

Introduction

Glucagon-like peptide-1 (GLP-1) receptor agonists (RAs) and dual-action glucose-dependent insulinotropic polypeptide (GIP)/GLP-1 RAs are medications developed for improving glycemic control in patients with type 2 diabetes (T2D).1 These medications work by increasing glucose-dependent insulin secretion, decreasing inappropriate glucagon secretion, and slowing gastric emptying.2 Because these medications also act in the areas of the brain involved in the regulation of appetite and caloric intake, they have become a popular option for weight loss and, more recently, have been found to have positive effects in other conditions such as cardiovascular disease, chronic kidney disease, metabolic-associated steatohepatitis, and obstructive sleep apnea.1,2

Prior to June 2023, patients scheduled for surgery at The Ohio State Wexner Medical Center (OSUWMC) were instructed to continue these medications preoperatively.3,4 However, on June 29, 2023, the American Society of Anesthesiologists (ASA) released a statement recommending that patients on a weekly-dosed GLP-1 or GIP/GLP-1 RA should stop the medication 7 days prior to elective surgery due to concerns for increased risk of regurgitation and pulmonary aspiration of gastric contents while under general anesthesia.5, 6, 7, 8, 9 This recommendation stems from the pharmacokinetic properties of these medications and their ability to slow gastric emptying.5 Case reports have been cited by the ASA in support of this recommendation.10, 11, 12 The decision was made by an OSUWMC multidisciplinary team comprised of pharmacists, anesthesiologists, and endocrinologists to follow the ASA consensus statement, recommending holding weekly-dosed GLP-1 or GIP/GLP-1 RAs for 1 week prior to surgery. In January 2024, institutional guidelines were updated and published.

There is a paucity of data regarding the perioperative clinical impacts of temporarily holding weekly dosed GLP-1 or GIP/GLP-1 RAs in patients with T2D prior to surgery, nor does data exist surrounding surgical complications related to possible hyperglycemia from this intervention.

Methods

Trial Design

This was a retrospective, single-center, cohort study evaluating patients at a large academic medical center before and after the change in institutional guidelines. Patients taking a weekly-dosed GLP-1 or GIP/GLP-1 RA scheduled for elective surgery from January 1 through May 31 of 2022 (pre-recommendation group) were compared to those scheduled for elective surgery from January 1 through May 31 of 2024 (post-recommendation group). Matched analyses were conducted also to account for between-group variabilities. Patients aged 18-89 years with T2D (defined as hemoglobin A1C [HbA1C] ≥6.5% [48 mmol/mol]) were included. Patients were excluded if they were pregnant, incarcerated, or had other types of diabetes. They were also excluded if they were taking a daily-dosed GLP-1 RA, taking a GLP-1 or GIP/GLP-1 RA that is approved for the indication of weight loss, or taking a GLP-1 or GIP/GLP-1 RA approved for the treatment of T2D off-label for weight loss. Patients were screened for 5 months pre- and post-recommendation to identify surgical encounters with a GLP-1 or GIP/GLP-1 RA on the patient’s medication list prior to admission.

This study was approved by the institutional review board of The Ohio State University (IRB Study Number 2024E1080 and Approval Date 10/6/2024). The need for informed consent was waived due to the retrospective nature of the study.

Study Measures

Recommendation Groups

Pre- and post-recommendation refers to the perioperative management of GLP-1 or GIP/GLP-1 RAs prior to and after the ASA Task Force’s statement released in June 2023, respectively. Patients in the pre-recommendation group were those with a surgery or procedure between January 1, 2022, and May 31, 2022. In this group, patients were instructed to continue their GLP-1 or GIP/GLP-1 RA uninterrupted prior to surgery. Patients in the post-recommendation group were those with a surgery or procedure between January 1, 2024, and May 31, 2024. In this group, patients were instructed to hold their GLP-1 or GIP/GLP-1 RA at least 7 days prior to the procedure.

Demographic, Clinical, and Treatment Considerations

The following data points were collected via retrospective chart review. Patient demographic variables included age, sex (female, male), and race (White/Caucasian, Black/African American, Asian, and Other). Clinical characteristics included body mass index (BMI); HbA1C (%; mmol/mol); duration of diabetes diagnosis (years); diabetes therapy at time of surgery including presence of prior to admission (PTA) metformin, sodium glucose cotransporter 2 inhibitors, insulin, sulfonylureas, thiazolidinediones, or other class; GLP-1 or GIP/GLP-1 RA name, dose (mg), and duration of therapy (months); blood glucose (mg/dL; preoperative, intraoperative, and postoperative); presence of insulin administration prior to surgery to achieve blood glucose ≤ 180 mg/dL; whether insulin was administered intraoperatively; history of previous gastrointestinal (GI) surgery and gastroparesis; ASA Physical Status Classification Score; presence of documentation of perioperative GLP-1 or GIP/GLP-1 RA instructions per institutional guidance; surgery cancellation and reason (if applicable); type of surgery (non-GI surgery, GI surgery, or GI procedure); type of anesthesia (general or monitored anesthesia care); duration of surgery (hours); intraoperative pulmonary aspiration event; postoperative pneumonia; postoperative surgical site infection; length of stay; readmission within 30 days for hyperglycemia or surgical site infection; and mortality within 30 days.

Endpoints

The primary outcome was the number of patients in the overall cohort with preoperative glycemic control, defined as blood glucose ≤ 180 mg/dL. Preoperative blood glucose values were single capillary measurements taken within 2 hours of the surgery start time conducted with point-of-care StatStrip glucometers. Secondary exploratory outcomes included use of preoperative administration of insulin to achieve blood glucose ≤180 mg/dL, presence of perioperative GLP-1 or GIP/GLP-1 RA patient instructions per institutional guidance, whether intraoperative insulin was given, single highest intraoperative blood glucose value (if available), single highest postoperative blood glucose value (if available), number of surgery cancellations due to hyperglycemia, number of pulmonary complications (aspiration or pneumonia) during admission, number of postoperative surgical site infections, readmission rate for hyperglycemia or surgical site infection within 30 days, and 30-day mortality rate.

Effect Size Calculation

Effect size calculations were conducted using PASS 202 version 23.0.2. Preliminary data consisting of the expected proportion of patients who had glycemic control among those treated with GLP-1 medications before the 2023 ASA recommendation were used. Based on the anticipated proportion of 90% of patients having glycemic control in the pre-recommendation group and an anticipated sample size of 300 patients (150 patients in each group), we anticipated to have 80% power to detect a proportion difference of 0.061 and a marginal probability of 0.961 in the post-recommendation group using a two-sided, paired-sample McNemar’s test, assuming a type I error of 5%.

Statistical Analysis

Demographic factors, clinical characteristics, and treatment variables were summarized using descriptive statistics, where means and SDs, medians and IQRs, and range were used for continuous variables, and frequencies and percentages were used for categorical variables. For the unmatched cohort, differences in demographic, clinical, and treatment information were compared between encounters that occurred before the 2023 ASA recommendations and encounters that occurred after the 2023 ASA recommendation using standardized mean differences. A main effects logistic regression model was fit to predict the probability of an encounter occurring after the 2023 ASA recommendation. Covariates in the model included age, sex, HbA1C, diabetes diagnosis duration, GLP-1 or GIP/GLP-1 RA duration, and surgery type. These covariates were chosen to match patients with similar diabetes control at baseline, similar duration of the medication class in question (GLP-1 or GIP/GLP-1 RA), and similar surgical procedures (non-GI surgery, GI surgery, or GI procedure) given differences in aspiration risk. Matching on additional variables may exclude matchable encounters and further limit sample size. Using the Proc PSMatch procedure, along with a caliber of 0.25, the predicted probabilities were then used to match 1 encounter in the pre-recommendation group to 1 encounter in the post-recommendation group. For the matched cohort, differences in demographic, clinical, and treatment information were compared between encounters that occurred before the 2023 ASA recommendation and encounters that occurred after the 2023 ASA recommendation guidelines using standardized mean differences.

For the unmatched cohort, a Fisher exact test was used to assess the association between the primary outcome, glycemic control, and GLP-1 or GIP/GLP-1 RA perioperative recommendation. Fisher exact tests were used to assess the association between GLP-1 or GIP/GLP-1 perioperative recommendation and secondary outcomes of preoperative administration of insulin and intraoperative administration of insulin. A Wilcoxon rank-sum test was used to assess the association between GLP-1 or GIP/GLP-1 perioperative recommendation and secondary outcomes: intraoperative blood glucose levels and postoperative blood glucose levels. The secondary outcomes of pulmonary complications (aspiration and/or pneumonia), postoperative surgical site infection, readmission within 30 days, mortality within 30 days, presence of perioperative medication instructions, and the number of surgery cancellations due to hyperglycemia were summarized using frequencies and percentages.

For the matched cohort, a McNemar test was used to assess the association between the primary outcome, glycemic control, and GLP-1 or GIP/GLP-1 RA perioperative recommendation. A Wilcoxon signed-rank test was used to assess the association between GLP-1 or GIP/GLP-1 perioperative recommendation and secondary outcomes: preoperative blood glucose level, intraoperative blood glucose level, and postoperative blood glucose level. The secondary outcomes of preoperative administration of insulin, intraoperative administration of insulin, pulmonary complications (aspiration and/or pneumonia), postoperative surgical site infection, readmission within 30 days, mortality within 30 days, presence of perioperative medication instructions, and the number of surgery cancellations due to hyperglycemia were summarized using frequencies and percentages.

In a subset analysis among encounters where HbA1C < 9% (75 mmol/mol), a McNemar test was used to assess the association between glycemic control and GLP-1 or GIP/GLP-1 perioperative recommendation. A McNemar test was used to assess the association between glycemic control and GLP-1 or GIP/GLP-1 perioperative recommendation among encounters where insulin was not given prior to admission and among encounters where insulin was given prior to admission. For secondary outcomes with missing data, a complete case analysis was conducted. For all statistical tests, the sample size analyzed was reported. An alpha of 0.05 was used as the threshold to determine statistical significance. All analyses were conducted using SAS version 9.4 (SAS Institute). An independent statistical review was performed to ensure validity of results.

Results

Study Population

During the study period, 168 encounters met inclusion criteria (81 in the pre-recommendation cohort and 87 in the post-recommendation cohort) (Fig. 1). Prematched baseline characteristics are listed in Table 1. At baseline, the groups were well balanced based on age, race, BMI, and duration of diabetes. Differences were noted in the other reported baseline characteristics, including, but not limited to, baseline HbA1C, the other antidiabetic medication regimens, and the type of procedure or surgery (Table 1). After matching, 57 encounters in the pre-recommendation cohort were able to be matched to an encounter in the post-recommendation cohort. Matched baseline characteristics are listed in Table 2. Matched groups were well balanced with some differences in race, BMI, prior-to-admission antidiabetic medications, GLP-1 or GIP/GLP-1 RA dose, and history of gastroparesis and GI surgery (Table 2).

Figure.

Figure

Participant flowchart. This figure shows the original number of encounters screened and how many were excluded with the accompanying reasons. This figure also depicts the number of encounters that were able to be matched from each group. EGD = esophagogastroduodenoscopy; GLP-1 RA = glucagon-like peptide-1 receptor agonist; SSI = sliding scale insulin; T2D = type 2 diabetes.

Table 1.

Patient Demographic, Clinical and Treatment Information Stratified by Recommendation Guideline Status

Patient characteristic All
Recommendation guideline
SMD
Pre-recommendation
Postrecommendation
N = (168) N = (81) N = (87)
Age (years) 0.042
 Mean (SD) 61.29 (10.43) 61.06 (10.84) 61.51 (10.08)
Sex, n (%)
 Female 77 (45.8%) 41 (50.6%) 36 (41.4%) 0.186
 Male 91 (54.2%) 40 (49.4%) 51 (58.6%) 0.186
Race, n (%)
 White/Caucasian 138 (82.1%) 67 (82.7%) 71 (81.6%) 0.029
 Black/African American 21 (12.5%) 10 (12.3%) 11 (12.6%) 0.009
 Asian 1 (0.6%) 1 (1.2%) 0 (0.0%) 0.158
 Other 8 (4.8%) 3 (3.7%) 5 (5.7%) 0.096
BMI (kg/m2), n (%)
 18.5 to < 25 6 (3.6%) 5 (6.2%) 1 (1.1%) 0.270
 25 to < 30 36 (21.4%) 17 (21.0%) 19 (21.8%) 0.021
 30 to < 35 45 (26.8%) 17 (21.0%) 28 (32.2%) 0.255
 35 to < 40 40 (23.8%) 19 (23.5%) 21 (24.1%) 0.016
 ≥ 40 41 (24.4%) 23 (28.4%) 18 (20.7%) 0.180
HbA1C, n (%)
 < 7% (53 mmol/mol) 72 (42.9%) 29 (35.8%) 43 (49.4%) 0.278
 7% to < 8% (53 to < 64 mmol/mol) 55 (32.7%) 29 (35.8%) 26 (29.9%) 0.126
 8% to < 9% (64 to < 75 mmol/mol) 18 (10.7%) 11 (13.6%) 7 (8.0%) 0.179
 9% to < 10% (75 to < 86 mmol/mol) 16 (9.5%) 8 (9.9%) 8 (9.2%) 0.023
 ≥ 10% (86 mmol/mol) 7 (4.2%) 4 (4.9%) 3 (3.4%) 0.074
HbA1C (binary), n (%)
 < 8% (64 mmol/mol) 127 (75.6%) 58 (71.6%) 69 (79.3%) 0.180
 ≥ 8% (64 mmol/mol) 41 (24.4%) 23 (28.4%) 18 (20.7%) 0.180
Diabetes duration (years), n (%)
 < 5 43 (25.6%) 18 (22.2%) 25 (28.7%) 0.150
 5 to < 10 31 (18.5%) 17 (21.0%) 14 (16.1%) 0.126
 ≥ 10 94 (56.0%) 46 (56.8%) 48 (55.2%) 0.033
Diabetes duration (years, binary), n (%)
 < 10 74 (44.0%) 35 (43.2%) 39 (44.8%) 0.033
 ≥ 10 94 (56.0%) 46 (56.8%) 48 (55.2%) 0.033
eGFR (mL/min/1.73m2)
 ≤ 85.0 85 (50.6%) 37 (45.7%) 48 (55.2%) 0.191
 > 85.0 83 (49.4%) 44 (54.3%) 39 (44.8%) 0.191
Other diabetes medications PTA, n (%)
 Metformin 92 (54.8%) 46 (56.8%) 46 (52.9%) 0.079
 SGLT2i 45 (26.8%) 17 (21.0%) 28 (32.2%) 0.255
 Insulin 84 (50.0%) 49 (60.5%) 35 (40.2%) 0.414
 Sulfonylurea 35 (20.8%) 21 (25.9%) 14 (16.1%) 0.243
 TZD 12 (7.1%) 7 (8.6%) 5 (5.7%) 0.112
 Other Class 3 (1.8%) 3 (3.7%) 0 (0.0%) 0.277
GLP-1 or GIP/GLP-1 RA Name, n (%)
 Exenatide extended release 8 (4.8%) 6 (7.4%) 2 (2.3%) 0.239
 Dulaglutide 91 (54.2%) 49 (60.5%) 42 (48.3%) 0.247
 Semaglutide 59 (35.1%) 26 (32.1%) 33 (37.9%) 0.122
 Tirzepatide 10 (6.0%) 0 (0.0%) 10 (11.5%) 0.510
GLP-1 or GIP/GLP-1 RA duration (months), n (%)
 < 2 44 (26.2%) 15 (18.5%) 29 (33.3%) 0.343
 2 to < 4 27 (16.1%) 10 (12.3%) 17 (19.5%) 0.197
 4 to < 6 22 (13.1%) 11 (13.6%) 11 (12.6%) 0.028
 ≥ 6 75 (44.6%) 45 (55.6%) 30 (34.5%) 0.433
GLP-1 or GIP/GLP-1 RA duration (months, binary), n (%)
 < 4 71 (42.3%) 25 (30.9%) 46 (52.9%) 0.458
 ≥ 4 97 (57.7%) 56 (69.1%) 41 (47.1%) 0.458
History of gastroparesis, n (%) 15 (8.9%) 4 (4.9%) 11 (12.6%) 0.275
History of gastrointestinal surgery, n (%) 28 (16.7%) 16 (19.8%) 12 (13.8%) 0.160
Anesthesia, n (%)
 General 102 (60.7%) 36 (44.4%) 66 (75.9%) 0.678
 Monitored anesthesia care 66 (39.3%) 45 (55.6%) 21 (24.1%) 0.678
Surgery type (binary), n (%)
 Non-GI 127 (75.6%) 52 (64.2%) 75 (86.2%) 0.527
 GI procedure or GI surgery 41 (24.4%) 29 (35.8%) 12 (13.8%) 0.527
Surgery type, n (%)
 Non-GI 127 (75.6%) 52 (64.2%) 75 (86.2%) 0.527
 GI surgery 20 (11.9%) 10 (12.3%) 10 (11.5%) 0.026
 GI procedure 21 (12.5%) 19 (23.5%) 2 (2.3%) 0.666
Duration of surgery (hours), n (%)
 < 2 97 (57.7%) 43 (49.4%) 54 (66.7%) 0.355
 ≥ 2 71 (42.3%) 44 (50.6%) 27 (33.3%) 0.355
Length of stay (days) 0.168
 Mean (SD) 1.02 (2.89) 1.27 (3.75) 0.78 (1.75)
 Median (IQR) 0.00 (0.00, 1.00) 0.00 (0.00, 1.00) 0.00 (0.00, 1.00)
 Min, Max 0.00, 21.00 0.00, 21.00 0.00, 10.00
ASA classification score, n (%)
 II 25 (14.9%) 14 (17.3%) 11 (12.6%) 0.130
 III 128 (76.2%) 59 (72.8%) 69 (79.3%) 0.152
 IV 14 (8.3%) 7 (8.6%) 7 (8.0%) 0.022
 Unknown/Not Assessed 1 (0.6%) 1 (1.2%) 0 (0.0%) 0.158

Abbreviations: ASA = American Society of Anesthesiologists; BMI = body mass index; eGFR = estimated glomerular filtration rate; GI = gastrointestinal; GIP = glucose-dependent insulinotropic polypeptide; GLP-1 = glucagon-like peptide-1; HbA1C = hemoglobin A1C; PTA = prior to admission; RA = receptor agonist; SGLT2i = sodium glucose cotransporter 2 inhibitor; SMD = standardized mean difference; TZD = thiazolidinedione.

Table 2.

Patient Demographic, Clinical and Treatment Information Stratified by Recommendation Guideline Status After Matchinga

Patient characteristic All
Recommendation guideline
SMD
Pre-recommendation
Postrecommendation
N = (114) N = (57) N = (57)
Age (years) 0.057
 Mean (SD) 61.98 (10.49) 61.68 (11.02) 62.28 (10.03)
Sex, n (%)
 Female 58 (50.9%) 29 (50.9%) 29 (50.9%) 0.000
 Male 56 (49.1%) 28 (49.1%) 28 (49.1%) 0.000
Race, n (%)
 White/Caucasian 98 (86.0%) 48 (84.2%) 50 (87.7%) 0.101
 Black/African American 12 (10.5%) 6 (10.5%) 6 (10.5%) 0.000
 Asian 1 (0.9%) 1 (1.8%) 0 (0.0%) 0.189
 Other 3 (2.6%) 2 (3.5%) 1 (1.8%) 0.110
BMI (kg/m2), n (%)
 18.5 to < 25 4 (3.5%) 3 (5.3%) 1 (1.8%) 0.192
 25 to < 30 23 (20.2%) 13 (22.8%) 10 (17.5%) 0.131
 30 to < 35 35 (30.7%) 15 (26.3%) 20 (35.1%) 0.191
 35 to < 40 25 (21.9%) 12 (21.1%) 13 (22.8%) 0.042
 ≥ 40 27 (23.7%) 14 (24.6%) 13 (22.8%) 0.041
HbA1C, n (%)
 < 7% (53 mmol/mol) 51 (44.7%) 22 (38.6%) 29 (50.9%) 0.249
 7% to < 8% (53 to < 64 mmol/mol) 37 (32.5%) 21 (36.8%) 16 (28.1%) 0.188
 8% to < 9% (64 to < 75 mmol/mol) 12 (10.5%) 6 (10.5%) 6 (10.5%) 0.000
 9% to < 10% (75 to < 86 mmol/mol) 11 (9.6%) 7 (12.3%) 4 (7.0%) 0.179
 ≥ 10% (86 mmol/mol) 3 (2.6%) 1 (1.8%) 2 (3.5%) 0.110
HbA1C (binary), n (%)
 < 8% (64 mmol/mol) 88 (77.2%) 43 (75.4%) 45 (78.9%) 0.084
 ≥ 8% (64 mmol/mol) 26 (22.8%) 14 (24.6%) 12 (21.1%) 0.084
Diabetes duration (years), n (%)
 < 5 29 (25.4%) 13 (22.8%) 16 (28.1%) 0.121
 5 to < 10 22 (19.3%) 13 (22.8%) 9 (15.8%) 0.179
 ≥ 10 63 (55.3%) 31 (54.4%) 32 (56.1%) 0.035
Diabetes duration (years, binary), n (%)
 < 10 51 (44.7%) 26 (45.6%) 25 (43.9%) 0.035
 ≥ 10 63 (55.3%) 31 (54.4%) 32 (56.1%) 0.035
eGFR (mL/min/1.73m2)
 ≤ 85.0 55 (48.2%) 25 (43.9%) 30 (52.6%) 0.176
 > 85.0 59 (51.8%) 32 (56.1%) 27 (47.4%) 0.176
Other diabetes medications PTA, n (%)
 Metformin 63 (55.3%) 33 (57.9%) 30 (52.6%) 0.106
 SGLT2i 31 (27.2%) 13 (22.8%) 18 (31.6%) 0.198
 Insulin 52 (45.6%) 30 (52.6%) 22 (38.6%) 0.285
 Sulfonylurea 25 (21.9%) 17 (29.8%) 8 (14.0%) 0.389
 TZD 9 (7.9%) 5 (8.8%) 4 (7.0%) 0.065
 Other Class 3 (2.6%) 3 (5.3%) 0 (0.0%) 0.333
GLP-1 or GIP/GLP-1 RA name, n (%)
 Exenatide Extended Release 7 (6.1%) 5 (8.8%) 2 (3.5%) 0.221
 Dulaglutide 62 (54.4%) 33 (57.9%) 29 (50.9%) 0.141
 Semaglutide 39 (34.2%) 19 (33.3%) 20 (35.1%) 0.037
 Tirzepatide 6 (5.3%) 0 (0.0%) 6 (10.5%) 0.485
GLP-1 or GIP/GLP-1 RA duration (months), n (%)
 < 2 26 (22.8%) 13 (22.8%) 13 (22.8%) 0.000
 2 to < 4 15 (13.2%) 8 (14.0%) 7 (12.3%) 0.052
 4 to < 6 14 (12.3%) 4 (7.0%) 10 (17.5%) 0.325
 ≥ 6 59 (51.8%) 32 (56.1%) 27 (47.4%) 0.176
GLP-1 or GIP/GLP-1 RA duration (months, binary), n (%)
 < 4 41 (36.0%) 21 (36.8%) 20 (35.1%) 0.037
 ≥ 4 73 (64.0%) 36 (63.2%) 37 (64.9%) 0.037
History of gastroparesis, n (%) 11 (9.6%) 3 (5.3%) 8 (14.0%) 0.300
History of gastrointestinal surgery, n (%) 20 (17.5%) 12 (21.1%) 8 (14.0%) 0.185
Anesthesia, n (%)
 General 73 (64.0%) 29 (50.9%) 44 (77.2%) 0.570
 Monitored anesthesia care 41 (36.0%) 28 (49.1%) 13 (22.8%) 0.570
Surgery type (binary), n (%)
 Non-GI 92 (80.7%) 46 (80.7%) 46 (80.7%) 0.000
 GI procedure or GI surgery 22 (19.3%) 11 (19.3%) 11 (19.3%) 0.000
Surgery type, n (%)
 Non-GI 92 (80.7%) 46 (80.7%) 46 (80.7%) 0.000
 GI surgery 14 (12.3%) 4 (7.0%) 10 (17.5%) 0.325
 GI procedure 8 (7.0%) 7 (12.3%) 1 (1.8%) 0.421
Duration of surgery (hours), n (%)
 < 2 63 (55.3%) 36 (63.2%) 27 (47.4%) 0.322
 ≥ 2 51 (44.7%) 21 (36.8%) 30 (52.6%) 0.322
Length of stay (days) 0.183
 Mean (SD) 1.29 (3.36) 1.60 (4.38) 0.98 (1.85)
 Median (IQR) 0.00 (0.00, 1.00) 0.00 (0.00, 1.00) 0.00 (0.00, 1.00)
 Min, Max 0.00, 21.00 0.00, 21.00 0.00, 10.00
ASA classification score, n (%)
 II 19 (16.7%) 11 (19.3%) 8 (14.0%) 0.142
 III 86 (75.4%) 43 (75.4%) 43 (75.4%) 0.000
 IV 9 (7.9%) 3 (5.3%) 6 (10.5%) 0.196

Abbreviations: ASA = American Society of Anesthesiologists; BMI = body mass index; eGFR = estimated glomerular filtration rate; GI = gastrointestinal; GIP = glucose-dependent insulinotropic polypeptide; GLP-1 = glucagon-like peptide-1; HbA1C = hemoglobin A1C; PTA = prior to admission; RA = receptor agonist; SGLT2i = sodium glucose cotransporter 2 inhibitor; SMD = standardized mean differenc; TZD = thiazolidinedionee.

a

Propensity score model for matching included age, sex, HbA1C, diabetes duration, GLP-1 or GLP-1/GIP duration, and surgery type.

Endpoints

Primary Endpoint

In this retrospective, single-center study, it was found that sixty-three patients (77.8%) in the pre-recommendation group that continued prescribed GLP-1 or GIP/GLP-1 RA prior to procedure achieved glycemic control compared to seventy-two patients (82.8%) in the post-recommendation group (P = 0.443; risk difference [RD] 0.05, 95% CI −0.07 to 0.17) that temporarily held GLP-1 or GIP/GLP-1 RA prior to procedure (Table 3). Fifty-seven encounters in both groups were able to be matched for a sample size of 114. After matching, forty-six patients (80.7%) in the pre-recommendation group that continued prescribed GLP-1 or GIP/GLP-1 RA prior to procedure achieved glycemic control compared to forty-eight patients (84.2%) in the post-recommendation group that temporarily held GLP-1 or GIP/GLP-1 RA prior to procedure (Table 4). This did not result in a significant association between the recommendation group and the primary outcome of preoperative glycemic control (P = 0.824; RD 0.04, 95% CI −0.12 to 0.19). In a matched subset analysis, there was not a statistically significant association between recommendation group and glycemic control among patients with HbA1C levels <9% (75 mmol/mol) (P = 0.549; RD -0.07, 95% CI −0.21 to 0.07) (Table 5). Within the unmatched sample, numerically there were more patients in the pre-recommendation group on insulin at home compared to the post-recommendation group (60.5% vs 40.2%, respectively) (Table 1). However, when performing a matched subset analysis of patients with or without insulin on their PTA medication list, there was no association detected between recommendation group and glycemic control (P = 0.289 and P = 1.000, respectively) (RD 0.27, 95% CI −0.08 to 0.61 and RD −0.05, 95% CI −0.22 to 0.12, respectively) (Tables 6 and 7).

Table 3.

Patient Primary and Secondary Outcomes Stratified by Recommendation Guideline Status

Patient outcomes All
Recommendation guideline
Sample size analyzed P valuea Estimate (95% CI)b
Pre-recommendation
Postrecommendation
N = (168) N = (81) N = (87)
Glycemic control, n (%) 135 (80.4%) 63 (77.8%) 72 (82.8%) 168 0.443 0.05 (−0.07 to 0.17)
Preoperative blood glucose (mg/dL) 0.343 6.00 (−7.00 to 19.00)
 Median (IQR) 136.50 (106.00, 169.00) 134.00 (103.00, 172.00) 137.00 (112.00, 169.00)
 Min, Max 61.00, 374.00 71.00, 374.00 61.00, 297.00
Insulin given prior, n (%) 6 (3.6%) 4 (4.9%) 2 (2.3%) 168 0.430 −0.03 (−0.08 to 0.03)
Intraoperative blood glucose (mg/dL) 92 0.194 −14.50 (−36.00 to 7.00)
 Median (IQR) 150.00 (122.50, 192.00) 164.00 (123.00, 211.00) 142.50 (122.00, 183.00)
 Min, Max 73.00, 336.00 76.00, 336.00 73.00, 315.00
 N (Missing) 92 (76) 34 (47) 58 (29)
Intraoperative insulin given, n (%) 18 (10.7%) 9 (11.1%) 9 (10.3%) 168 1.000 −0.01 (−0.10 to 0.08)
Postoperative blood glucose (mg/dL) 120 0.302 9.50 (−10.00 to 29.00)
 Median (IQR) 157.50 (124.00, 193.00) 153.50 (114.00, 193.00) 159.00 (131.00, 199.00)
 Min, Max 72.00, 374.00 72.00, 374.00 90.00, 273.00
 N (Missing) 120 (48) 54 (27) 66 (21)
Pulmonary complication, n (%) 1 (0.6%) 1 (1.2%) 0 (0.0%) N/A N/A
Aspiration, n (%) 1 (0.6%) 1 (1.2%) 0 (0.0%) N/A N/A
Pneumonia, n (%) 1 (0.6%) 1 (1.2%) 0 (0.0%) N/A N/A
Postoperative surgical site infection, n (%) N/A N/A
 Yes 1 (0.6%) 0 (0.0%) 1 (1.1%)
 No 166 (98.8%) 81 (100.0%) 85 (97.7%)
 Unknown 1 (0.6%) 0 (0.0%) 1 (1.1%)
Readmission within 30 days, n (%) 2 (1.2%) 1 (1.2%) 1 (1.1%) N/A N/A
Mortality within 30 days, n (%) N/A N/A
 Yes 0 (0.0%) 0 (0.0%) 0 (0.0%)
 No 168 (100.0%) 81 (100.0%) 87 (100.0%)
Presence of perioperative medication instructions, n (%) N/A N/A
 Yes 147 (87.5%) 81 (100.0%) 66 (75.9%)
 No 17 (10.1%) 0 (0.0%) 17 (19.5%)
 Unknown 4 (2.4%) 0 (0.0%) 4 (4.6%)
Number of surgery cancellations due to hyperglycemia, n (%) N/A N/A
 Yes 0 (0.0%) 0 (0.0%) 0 (0.0%)
 No 168 (100.0%) 81 (100.0%) 87 (100.0%)

Abbreviation: N/A = not applicable.

a

Continuous variables were analyzed using Wilcoxon rank-sum tests, and categorical variables were analyzed using Fisher's exact tests. N/A is for variables not able to be tested.

b

Risk differences with 95% CIs are reported for categorical variables, Hodges-Lehmann pseudo-median estimates are reported for continuous variables.

Table 4.

Patient Primary and Secondary Outcomes Stratified by Recommendation Guideline Status After Matching (N = 114)

Patient outcomes All
Recommendation guideline
Sample size analyzed P valuea Estimate (95% CI)b
Pre-recommendation
Postrecommendation
N = (114) N = (57) N = (57)
Glycemic control, n (%) 94 (82.5%) 46 (80.7%) 48 (84.2%) 114 0.824 0.04 (−0.12 to 0.19)
Preoperative blood glucose (mg/dL) 114 0.411 7.50 (−12.5 to 26.5)
 Median (IQR) 137.00 (108.00, 169.00) 137.00 (96.00, 163.00) 137.00 (115.00, 169.00)
 Min, Max 61.00, 310.00 71.00, 310.00 61.00, 268.00
Insulin given prior, n (%) 2 (1.8%) 1 (1.8%) 1 (1.8%) N/A N/A
Intraoperative blood glucose (mg/dL) 34 0.863 −3.00 (−49.00 to 43.5)
 Median (IQR) 152.50 (123.00, 195.00) 164.00 (123.00, 197.00) 148.00 (123.00, 183.00)
 Min, Max 73.00, 336.00 76.00, 336.00 73.00, 315.00
 N (Missing) 66 (48) 25 (32) 41 (16)
Intraoperative insulin given, n (%) 12 (10.5%) 6 (10.5%) 6 (10.5%) N/A N/A
Postoperative blood glucose (mg/dL) 68 0.940 1.00 (−23.00 to 27.00)
 Median (IQR) 156.00 (127.50, 186.50) 148.50 (114.00, 186.00) 159.00 (133.00, 188.00)
 Min, Max 72.00, 346.00 72.00, 346.00 103.00, 273.00
 N (Missing) 88 (26) 42 (15) 46 (11)
Pulmonary complication, n (%) 1 (0.9%) 1 (1.8%) 0 (0.0%) N/A N/A
Aspiration, n (%) 1 (0.9%) 1 (1.8%) 0 (0.0%) N/A N/A
Pneumonia, n (%) 1 (0.9%) 1 (1.8%) 0 (0.0%) N/A N/A
Postoperative surgical site infection, n (%) N/A N/A
 Yes 1 (0.9%) 0 (0.0%) 1 (1.8%)
 No 112 (98.2%) 57 (100.0%) 55 (96.5%)
 Unknown 1 (0.9%) 0 (0.0%) 1 (1.8%)
Readmission within 30 days, n (%) 2 (1.8%) 1 (1.8%) 1 (1.8%) N/A N/A
Mortality within 30 days, n (%) N/A N/A
 Yes 0 (0.0%) 0 (0.0%) 0 (0.0%)
 No 114 (100.0%) 57 (100.0%) 57 (100.0%)
Presence of perioperative medication instructions, n (%) N/A
 Yes 101 (88.6%) 57 (100.0%) 44 (77.2%)
 No 11 (9.6%) 0 (0.0%) 11 (19.3%)
 Unknown 2 (1.8%) 0 (0.0%) 2 (3.5%)
Number of surgery cancellations due to hyperglycemia, n (%) N/A N/A
 Yes 0 (0.0%) 0 (0.0%) 0 (0.0%)
 No 114 (100.0%) 57 (100.0%) 57 (100.0%)

Abbreviation: N/A = not applicable.

a

McNemar's exact test was used for analysis to analyze binary variables. A Wilcoxon signed-rank test was used for Length of Stay for Surgical Procedure. N/A is for variables not able to be tested.

b

Pairwise risk differences with 95% CIs are reported for categorical variables, Hodges-Lehmann pseudo-median estimates for paired data are reported for continuous variables.

Table 5.

Analyses Showing the Association Between Recommendation Group and Glycemic Control Among Patients With HbA1C Levels <9% (75 mmol/mol)

Patient outcomes
Recommendation guideline
Sample size analyzed P valuea Estimate (95% CI)b
All
Pre-recommendation
Postrecommendation
N = (92) N = (46) N = (46)
Glycemic Control, n (%) 92 0.549 −0.07 (−0.21 to 0.07)
 Yes 81 (88.0%) 39 (84.8%) 42 (91.3%)
 No 11 (12.0%) 7 (15.2%) 4 (8.7%)

Abbreviation: HbA1C = hemoglobin A1C.

a

McNemar’s exact test was used for analysis.

b

Pairwise risk differences with 95% CIs are reported for categorical variables.

Table 6.

Analyses Showing the Association Between Recommendation Group and Glycemic Control Among Those Who Received Insulin PTA

Patient outcomes All
Recommendation guideline
Sample size analyzed P valuea Estimate (95% CI)b
Pre-recommendation
Postrecommendation
N = (30) N = (15) N = (15)
Glycemic Control, n (%) 30 0.289 0.27 (−0.08 to 0.61)
 Yes 22 (73.3%) 9 (60.0%) 13 (86.7%)
 No 8 (26.7%) 6 (40.0%) 2 (13.3%)

Abbreviation: PTA = prior to admission.

a

McNemar’s exact test was used for analysis.

b

Pairwise risk differences with 95% CIs are reported for categorical variables.

Table 7.

Analyses Showing the Association Between Recommendation Group and Glycemic Control Among Those Who Did Not Receive Insulin PTA

Patient outcomes All
Recommendation guideline
Sample size analyzed P valuea Estimate (95% CI)b
Pre-recommendation
Postrecommendation
N = (40) N = (20) N = (20)
Glycemic Control, n (%) 40 1.000 −0.05 (−0.22 to 0.12)
 Yes 37 (92.5%) 19 (95.0%) 18 (90.0%)
 No 3 (7.5%) 1 (5.0%) 2 (10.0%)

Abbreviation: PTA = prior to admission.

a

McNemar’s exact test was used for analysis.

b

Pairwise risk differences with 95% CIs are reported for categorical variables.

Secondary Endpoints

There was a difference in the documentation of preoperative instructions per institutional guidelines between groups (100% in the pre-recommendation group vs 75.9% in the post-recommendation group). There were no significant differences between pre- and post-recommendation groups for any other secondary endpoint, including number of patients given insulin prior to surgery to achieve blood glucose ≤ 180 mg/dL (4.9% vs 2.3%), whether insulin was given intraoperatively (11.1% vs 10.3%), number of pulmonary complications during admission (1.2% vs 0.0%), number of postoperative surgical site infections (0.0% vs 1.1%), 30-day readmission rate with hyperglycemia or surgical site infection (1.2% vs 1.1%), and 30-day mortality rate (0.0% vs 0.0%). No cases in either group were canceled due to hyperglycemia. There were no statistically significant differences detected between groups in the median intraoperative blood glucose (164.00 mg/dL vs 142.50 mg/dL; P = 0.194) or postoperative blood glucose (153.50 mg/dL vs 159.00 mg/dL; P = 0.302).

One patient in 168 total encounters experienced pulmonary complications during admission in the pre-recommendation group; however, this did not occur intraoperatively. One patient in 168 total encounters developed a postoperative surgical site infection in the post-recommendation group.

Discussion

This study demonstrates clinical outcome data comparing glycemic control with differences in perioperative management of GIP/GLP-1 RA medications that have not yet been published. Hulst et al and Makino et al studied the addition of liraglutide, a once-daily GLP-1 RA, to insulin and found that the combination resulted in lower insulin requirements for glycemic control during surgery, lower rates of insulin dose modification, and lower frequency of hypoglycemia.13,14 In the present study, glycemic control was defined as preoperative blood glucose ≤180 mg/dL. This definition aligned with institutional protocols based on the guidance of professional organizations and other supporting literature.15 Results of the present study suggest that in both the unmatched and matched cohorts, there were no statistically significant associations between recommendation group and glycemic control. There were numerically more patients in the post-recommendation group achieving glycemic control compared to the pre-recommendation group. There were no differences in complications of hyperglycemia including surgical site infection or readmission rates. This may have occurred because the majority of patients included in the study had well-controlled diabetes with HbA1C values < 7% (53 mmol/mol) (38.6% in the pre-recommendation group vs 50.9% in the post-recommendation group), eliminating patients who would be more sensitive to holding an antidiabetic therapy. This degree of well-controlled diabetes, which was numerically higher in the post-recommendation group, may also be a contributing factor explaining why more patients in this group achieved glycemic control prior to the procedure. Additionally, the half-life of most weekly-administered GLP-1 or GIP/GLP-1 RAs is 5-7 days.1 Drug concentration may still be present after holding these medications for only 7 days; however, pharmacokinetic and pharmacodynamic data at this concentration are unknown in relation to the degree of blood glucose reduction. Therefore, it is possible that holding these medications for 7 days may not result in clinically meaningful differences in blood glucose. At OSUWMC, the protocol is to hold GLP-1 or GIP/GLP-1 RAs indicated for treatment of T2D for 1 week, whereas GLP-1 or GIP/GLP-1 RAs used for other indications are held for 2 weeks. The protocol was adapted from the combination of ASA recommendations and small published case reports at the time of development. The results of this study may not be generalizable to other institutions with different protocols. There were numerical differences noted between groups regarding the duration of GLP-1 or GIP/GLP-1 RA therapy with several patients recently starting therapy. It is acknowledged that clinical impact on blood glucose may not be fully realized if patients started less than 2 months ago. However, these patients were still included to represent real-world practice populations. Additionally, this population was included in order to identify potential GI and anesthesia risks, as delayed gastric emptying and incidence of adverse side effects are commonly seen during GLP-1 or GIP/GLP-1 RA initiation and dose-escalation periods.16 Also, given the timeframe for inclusion criteria, only the post-recommendation group had patients taking tirzepatide, the GIP/GLP-1 RA, which has been considered clinically more effective for blood sugar control compared to GLP-1 RAs.1 This may be a contributing factor to better blood sugar control at baseline in the post-recommendation group. Given this imbalance of exposure and low number of patients on tirzepatide (n = 10), medication name was not selected as a matching parameter. However, tirzepatide sensitivity analyses were performed, and similar results were observed (see Supplementary Appendix). Overall, despite the post-recommendation cohort achieving glycemic control at a numerically higher percentage than the pre-recommendation cohort, this was not a clinically relevant difference.

In a matched subset analysis, patients with high HbA1C were compared to those with lower HbA1C to assess a potential reason the post-recommendation cohort had better glycemic control compared to the pre-recommendation group. While recognizing the selection bias, a HbA1C cut point of 9% (75 mmol/mol) was chosen because HbA1C < 8% (64 mmol/mol) and < 9% (75 mmol/mol) match the Healthcare Effectiveness Data and Information Set measures for glycemic control.17 There was not a statistically significant association between recommendation group and glycemic control among patients with HbA1C levels < 9% (75 mmol/mol). Patients with HbA1C levels ≥ 9% (75 mmol/mol) were unable to be analyzed given only 3 matched pairs met this criterion.

Concomitant antidiabetic medications were captured through data collection; however, the perioperative management of those medications was not assessed. Institutional protocol is to give patients instructions for other medications such as insulin and sodium glucose cotransporter 2 inhibitors. However, given the retrospective nature of the study, adherence to protocol could not be confirmed. It is therefore unknown if and how any adjustments to those medications may have affected the primary outcome of glycemic control. This may be more significant in patients who were on insulin therapy at home since the recommendation is to continue long-acting insulin on the day of the procedure. The proportion of patients on home insulin therapy was different between groups. When assessed as a matched subset, the presence or absence of PTA insulin did not show association with achieving glycemic control in either group. Several confounding factors may have led to this result. Providers in the preoperative assessment clinics follow institutional protocols that suggest adjusting home basal insulin doses prior to the procedure based on several factors, including type of diabetes, degree of glycemic control prior to surgery, total daily insulin dose (basal plus prandial) prior to surgery, and usual administration time of basal insulin. Dose adjustments may also be a shared decision with the managing primary care or endocrinology provider. Therefore, the differences in dose adjustments and their impact on preoperative blood glucose could not be accurately accounted for in this study.

Perioperative hyperglycemia is a contributing factor that increases the risk of surgical site infection after surgery or procedure. A 2024 systematic review cited 9 studies reporting an association between preoperative HbA1C and surgical site infections. In that review 5 studies (25 462 participants) showed no association, and 4 studies (57 771 participants) demonstrated a positive association between high preoperative HbA1C and odds of surgical site infection.18 In the present study, 1 patient in the post-recommendation group developed a surgical site infection. The patient was obese with a diabetes diagnosis of over 10 years, a HbA1C <7% (53 mmol/mol), and ASA Classification III. Preoperative blood glucose on the day of a trochanteric bursa excision was 147 mg/dL, and the patient was discharged the same day. Twenty-three days later, the patient was readmitted, found to have positive blood and intraoperative tissue cultures positive for methicillin-resistant Staphylococcus aureus, and subsequently received 4 weeks of daptomycin. Given well-controlled diabetes at baseline, it would be unlikely that holding the GLP-1 RA preoperatively was the sole risk factor precipitating hyperglycemia or surgical site infection. Although the patient was instructed to restart the GLP-1 RA upon discharge, it is unknown whether those instructions were followed, and glucose control between discharge and readmission was unknown.

Rates of pulmonary aspiration have been shown to be low, occurring in 1 of every 2-3000 operations requiring anesthesia.19 Although rates are low, the use of a GLP-1 or GIP/GLP-1 RA can cause delayed gastric emptying and increase risk for retention of gastric contents by up to four-fold.20 Delayed gastric emptying is also more likely in patients with long-standing diabetes and in patients who are in the titration phase of a GLP-1 or GIP/GLP-1 RA.16 Currently available case reports demonstrate only 1 intraoperative pulmonary aspiration event suspected from a GLP-1 RA’s delayed gastric emptying effects.10 Other published cases show surgeries were canceled due to potential risks, but no pulmonary event occurrences.11,12 In the present study, 1 patient in the pre-recommendation group experienced a pulmonary complication during their admission. At baseline, the patient was obese with long-standing uncontrolled diabetes (HbA1C greater than 9% or 75 mmol/mol), ASA Classification III, a history of previous GI surgery, and uncontrolled chronic obstructive pulmonary disease. They underwent a reverse total shoulder arthroplasty with no intraoperative complications. The patient experienced worsening hypoxia on postoperative day 1 and was transferred to the intensive care unit for bacterial pneumonia. Throughout the 21-day hospitalization, the patient suffered several aspiration events deemed to be multifactorial.

On October 29, 2024, the ASA released an updated statement recommending continuation of weekly-dosed GLP-1 or GIP/GLP-1 RA before elective surgery.21 They noted a need for balancing risks of withholding versus risks of continuation, including impacts on glycemic control and optimal use of medical resources.21 In patients at highest risk for delayed gastric emptying and subsequent pulmonary complications (such as those in the dose-escalation phase or those presenting with GI symptoms or other medical conditions that slow gastric emptying), a 24-hour liquid diet prior to surgery, assessing stomach contents using point-of-care ultrasound, or scheduling around periods of highest risk may be considered.21 The results of this study focus primarily on glycemic control but could be applied in conjunction with these new recommendations from the ASA when assessing patients’ overall risks.

Limitations

In this retrospective study, data collection was limited to documentation in the electronic medical record. The overall small sample size and the single-institution nature of the study may limit generalizability, specifically due to the primary endpoint definition of glycemic control (180 mg/dL) being based upon the institutional protocol. The small sample size limits the ability to detect statistically significant differences between groups for several of the exploratory secondary outcomes. There was also missing data for the intraoperative and postoperative blood glucose secondary outcomes, because not all procedure cases had all 3 glucose values measured (preoperative, intraoperative, and postoperative). Since the primary outcome was focused on preoperative blood sugars, the patients with missing intraoperative and postoperative blood sugars were still included in primary analyses. Analyses of the secondary outcomes (intraoperative and postoperative blood glucose) were conducted as complete case analyses using the total number and values that were available for collection. The primary endpoint, glycemic control, is a dichotomization of a continuous variable that assumes everyone above or below the cut point is equal. This may dilute true differences or not account for individual confounding variables that would impact if a patient met the primary endpoint. The study time frame of 5 months pre- and post-recommendation was short in duration for screening patients for inclusion. This timeframe also limited the number of patients on tirzepatide to the post-recommendation group and introduced a confounding variable and potential bias. There was also a nationwide backorder for GLP-1 and GIP/GLP-1 RAs making it inconclusive if patients were obtaining the medication despite it being on their home medication list. Including January 2024 was intentional given necessity for time to complete data collection; however, the researchers acknowledge the implementation of the new perioperative protocol on January 1, 2024. There may have been less adherence to holding the GLP-1 or GIP/GLP-1 RA initially, and the adherence to protocol likely improved over time. Although patients were provided perioperative instructions for their medications, adherence to these instructions is unknown. Three of 4 canceled cases in the post-recommendation group were canceled due to patients not following the hold recommendations. Adherence otherwise was not captured.

Conclusions

In the overall cohort and matched cohort, there was no statistically significant difference in the rate of glycemic control whether GLP-1 or GIP/GLP-1 RA therapy was continued or temporarily held preoperatively in patients with T2D. There was 1 patient in 168 encounters having an aspiration event (though not intraoperatively) in the pre-recommendation group in which their GLP-1 RA was continued and 1 patient in 168 encounters readmitted for surgical site infection in the post-recommendation group in which their GLP-1 RA was withheld. Larger prospective studies are needed to confirm these findings.

Ethics Approval

IRB Study Number: 2024E1080. IRB Approval Date: 10/6/2024.

Informed Consent

Due to the retrospective nature of the study, informed consent was waived.

Disclosure

Dr Kathleen Wyne has served as a consultant for New Amsterdam Pharma (9/2024) and AstraZeneca (10/2024). No other authors of this article have anything to disclose concerning possible financial or commercial entities that may have a direct or indirect interest in the subject of this article.

Supplementary materials

Supplementary data1
mmc1.docx (53.6KB, docx)

References

  • 1.American Diabetes Association Professional Practice Committee. 9 Pharmacologic approaches to glycemic treatment: standards of care in diabetes-2025. Diabetes Care. 2025;48(1 Suppl 1):S181–S206. doi: 10.2337/dc25-S009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Hinnen D. Glucagon-Like peptide 1 receptor agonists for Type 2 diabetes. Diabetes Spectr. 2017;30(3):202–210. doi: 10.2337/ds16-0026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Academy of Medical Royal Colleges. Centre for Perioperative Care . Centre for Perioperative Care (CPOC); London: 2023. Guideline for perioperative care for people with diabetes mellitus undergoing elective and emergency surgery. [Google Scholar]
  • 4.Membership of the Working Party, Barker P., Creasey P.E., et al. Peri-operative management of the surgical patient with diabetes 2015: Association of Anaesthetists of Great Britain and Ireland. Anaesthesia. 2015;70(12):1427–1440. doi: 10.1111/anae.13233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Joshi G.P., Abdelmalak B.B., Weigel W.A., et al. American Society of Anesthesiologists Consensus-Based guidance on preoperative management of patients (adults and children) on Glucagon-Like Peptide-1 (GLP-1) receptor agonists. 2023. https://www.asahq.org/about-asa/newsroom/news-releases/2023/06/american-society-of-anesthesiologists-consensus-based-guidance-on-preoperative
  • 6.Hjerpsted J.B., Flint A., Brooks A., Axelsen M.B., Kvist T., Blundell J. Semaglutide improves postprandial glucose and lipid metabolism, and delays first-hour gastric emptying in subjects with obesity. Diabetes Obes Metab. 2018;20(3):610–619. doi: 10.1111/dom.13120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Friedrichsen M., Breitschaft A., Tadayon S., Wizert A., Skovgaard D. The effect of semaglutide 2.4 mg once weekly on energy intake, appetite, control of eating, and gastric emptying in adults with obesity. Diabetes Obes Metab. 2021;23(3):754–762. doi: 10.1111/dom.14280. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Kobori T., Onishi Y., Yoshida Y., et al. Association of glucagon-like peptide-1 receptor agonist treatment with gastric residue in an esophagogastroduodenoscopy. J Diabetes Investig. 2023;14(6):767–773. doi: 10.1111/jdi.14005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Silveira S.Q., da Silva L.M., de Campos Vieira Abib A., et al. Relationship between perioperative semaglutide use and residual gastric content: a retrospective analysis of patients undergoing elective upper endoscopy. J Clin Anesth. 2023;87 doi: 10.1016/j.jclinane.2023.111091. [DOI] [PubMed] [Google Scholar]
  • 10.Klein S.R., Hobai I.A. Semaglutide, delayed gastric emptying, and intraoperative pulmonary aspiration: a case report. Sémaglutide, vidange gastrique retardée et aspiration pulmonaire peropératoire : une présentation de cas. Can J Anaesth. 2023;70(8):1394–1396. doi: 10.1007/s12630-023-02440-3. [DOI] [PubMed] [Google Scholar]
  • 11.Fujino E., Cobb K.W., Schoenherr J., Gouker L., Lund E. Anesthesia considerations for a patient on semaglutide and delayed gastric emptying. Cureus. 2023;15(7) doi: 10.7759/cureus.42153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Kittner S.L., Talbott A.L., Vishneski S.R., Narbaiza J., Shields J.S. Retained gastric contents after adequate fasting associated with GLP-1 receptor agonist use: a report of 3 cases. JBJS Case Connect. 2023;13(4) doi: 10.2106/JBJS.CC.23.00506. [DOI] [PubMed] [Google Scholar]
  • 13.Hulst A.H., Visscher M.J., Godfried M.B., et al. Liraglutide for perioperative management of hyperglycaemia in cardiac surgery patients: a multicentre randomized superiority trial. Diabetes Obes Metab. 2020;22(4):557–565. doi: 10.1111/dom.13927. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Makino H., Tanaka A., Asakura K., et al. Addition of low-dose liraglutide to insulin therapy is useful for glycaemic control during the peri-operative period: effect of glucagon-like peptide-1 receptor agonist therapy on glycaemic control in patients undergoing cardiac surgery (GLOLIA study) Diabet Med. 2019;36(12):1621–1628. doi: 10.1111/dme.14084. [DOI] [PubMed] [Google Scholar]
  • 15.Joshi G.P., Chung F., Vann M.A., et al. Society for Ambulatory Anesthesia consensus statement on perioperative blood glucose management in diabetic patients undergoing ambulatory surgery. Anesth Analg. 2010;111(6):1378–1387. doi: 10.1213/ANE.0b013e3181f9c288. [DOI] [PubMed] [Google Scholar]
  • 16.Jalleh R.J., Plummer M.P., Marathe C.S., et al. Clinical consequences of delayed gastric emptying with GLP-1 receptor agonists and tirzepatide. J Clin Endocrinol Metab. 2024;110(1):1–15. doi: 10.1210/clinem/dgae719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.NCQA HEDIS and performance Measurement. NCQA. 2023. https://www.ncqa.org/hedis/
  • 18.Yu A., Truong Q., Whitfield K., et al. Impact of preoperative haemoglobin A1c levels on postoperative outcomes in adults undergoing major noncardiac surgery: a systematic review. Diabet Med. 2024;41(10) doi: 10.1111/dme.15380. [DOI] [PubMed] [Google Scholar]
  • 19.Nason K.S. Acute intraoperative pulmonary aspiration. Thorac Surg Clin. 2015;25(3):301–307. doi: 10.1016/j.thorsurg.2015.04.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Nadeem D., Taye M., Still M.D., et al. Effects of glucagon-like peptide-1 receptor agonists on upper endoscopy in diabetic and nondiabetic patients. Gastrointest Endosc. 2024;100(4):745–749. doi: 10.1016/j.gie.2024.04.2900. [DOI] [PubMed] [Google Scholar]
  • 21.Most Patients Can Continue Diabetes Weight loss GLP-1 drugs before surgery, those at highest risk for GI problems should follow liquid diet before procedure” published 29 October 2024. https://www.asahq.org/about-asa/newsroom/news-releases/2024/10/new-multi-society-glp-1-guidance

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary data1
mmc1.docx (53.6KB, docx)

Articles from AACE Endocrinology and Diabetes are provided here courtesy of American Association of Clinical Endocrinology

RESOURCES