No association between semaglutide and postoperative pneumonia in people with diabetes undergoing elective surgery

Abstract

Semaglutide (first marketed as Ozempic™ [Novo Nordisk]) is a glucagon-like peptide-1 receptor agonist (GLP-1RA) which has seen a 40-fold increase in use over the past 5 years. It provides not only glycaemic and cardiorenal benefits in people with type 2 diabetes (T2DM) but also produces substantial weight loss.1 Semaglutide improves glucose control by increasing insulin secretion and induces weight loss via appetite suppression and delayed gastric emptying.1, 2 In people with T2DM scheduled for surgery, initial recommendations were that patients continue their GLP-1RA treatment.3 However, case reports have described intraoperative aspiration in appropriately fasted patients on semaglutide,4-6 and small studies in people (mostly without diabetes) who used semaglutide found a three- to sevenfold increase in visible stomach contents on ultrasonography or during endoscopy after overnight fasting.2, 7 However, no prospective clinical study has formally evaluated the perioperative use of GLP-1RAs, and their safety in this setting is now unclear.8 In June 2023 the American Society of Anesthesiologists warned of a potential increased risk of intraoperative aspiration among all semaglutide users.8 We evaluated the risk of pneumonia within 14 days of surgery (which can arise from micro-aspiration during surgery9) and other specific aspiration-related outcomes among T2DM patients prescribed semaglutide at the time of elective surgery. A retrospective cohort study was conducted using multiple linked administrative databases from Ontario, Canada (ICES). In Ontario, universal healthcare and prescription medication coverage is provided for those aged ≥65 years. We identified people with T2DM aged ≥66 years, who underwent an initial elective surgery with a general or spinal anaesthetic between February 2020 and March 2023. Importantly, this period predates the published concerns regarding perioperative aspiration with GLP-1RAs, and the recommendations to hold GLP-1RA treatment preoperatively.8 We excluded people with poor data quality, those from outside the province, those without evidence of matching prespecified surgical procedure codes in two independent databases, those who had emergency surgery, and those with >1 index procedure on the same day. The primary exposure was whether these patients were using semaglutide (or not) at the time of their surgery. Pneumonia within 14 days of surgery was used as the primary outcome because pneumonia specifically linked to an aspiration event is rare. Details on the datasets and the validity of key coding algorithms are provided in the online supplement. We used the logit of a propensity score (with a calliper of 0.2) created from 29 covariates to match 1:4 or 1:3 semaglutide users to non-semaglutide users. Delirium (a potential risk factor for aspiration and pneumonia10) was measured in the postoperative period. A secondary analysis was carried out among the subgroup of patients who had a general anaesthetic (which would theoretically have the highest risk of aspiration11). To create comparable groups for this analysis, we generated a new propensity score (with a calliper of 0.2) based on the same 29 covariates to match 1:3 or 1:2 semaglutide users to non-semaglutide users. Baselines between groups were assessed using standardized differences; a standardized difference >10% is considered potentially clinically relevant. Conditional logistic regression was used to assess the odds of the outcomes between exposed and unexposed groups; a p value < 0.05 was taken to indicate statistical significance and SAS 9.4 software was used. The use of the data in this project is authorized under section 45 of Ontario's Personal Health Information Protection Act and does not require review by a Research Ethics Board. For our primary analysis, there were 48 597 patients who met our inclusion criteria (exclusions detailed in the online supplement); 3843 (7.9%) received a prescription for semaglutide preoperatively. After matching, 14 072 unexposed patients and 3833 patients exposed to semaglutide were retained, and all measured variables were similar in these groups (Table 1). The occurrence of postoperative delirium was similar in the GLP-1RA users and non-users (1.3% vs. 1.4%, respectively; p = 0.61). The odds of pneumonia were not higher in people using semaglutide (odds ratio [OR] 0.71, 95% confidence interval [CI] 0.36–1.41), and there were no significant differences among the secondary outcomes between groups (Table 2). There were 28 905/48 597 patients (59%) who had a general anaesthetic. We matched n = 2246 semaglutide to n = 6532 non-semaglutide users (baseline characteristics were similar in the two groups for this matched cohort and are shown in the online supplement). The odds of pneumonia were not significantly elevated among semaglutide users (OR 0.45, 95% CI 0.19–1.06) and secondary outcomes were also not significantly different (details in the online supplement). We did not find that people with T2DM who were using semaglutide at the time of elective surgery had an increased risk of pneumonia, or outcomes more specific to aspiration including aspiration pneumonia or need for bronchoscopy. These results add to the debate around the perioperative management of people with T2DM using a GLP-1RA; it is possible that other aspiration risk factors that overlap with semaglutide use (such as obesity or diabetes12) are responsible for the episodes of perioperative aspiration in the previously published case reports. Initial concerns regarding the risk of aspiration pneumonia with semaglutide were raised in early 2023 with the publication of case reports describing patients with full stomachs during oesophagogastroduodenoscopy.5, 6 While studies showing increased gastric contents after standard preoperative fasting are concerning, it is important to note that some of the non-semaglutide controls in these studies also had food visible on ultrasonography/gastric endoscopy.2, 7, 13 A prior study of 538 preoperative patients who were not using a GLP-1RA found that 1.7% had solid contents and 4.5% had relevant amounts of clear fluid visible on gastric ultrasonography, and the majority did not have an identifiable risk factor for poor gastric emptying.13 Other medications known to delay gastric emptying, such as opioids and antidepressants, are not held preoperatively.14 There are three relevant published studies that address aspiration risk with GLP-1RAs. The first is a large population-based study of people undergoing endoscopy between 2018 and 2020 (of whom 90% had T2DM).15 Compared to propensity-matched non-users, those using a GLP-1RA medication had a higher risk of aspiration pneumonia (absolute risk increase +0.20%, hazard ratio 1.33, 95% CI 1.02–1.74). However, on further analysis, this was only statistically significant in the group that received propofol sedation for endoscopy, and the risk was only significantly elevated in people undergoing upper endoscopy (not lower endoscopy). The results of this study may not be generalizable to surgical procedures however, as there is no airway placed during endoscopy, and the increased risk was specific to upper endoscopy, suggesting that the direct manipulation of the upper airway may play a role. A second study included a large population-based sample of people with T2DM presenting for emergency surgery; the authors assessed whether GLP-1RA users had an increased risk of a composite outcome of aspiration pneumonitis, postoperative respiratory failure, or intensive care admission compared to matched non-users.16 The adjusted OR for respiratory complications was 1.03 (95% CI 0.82–1.29). The results of this study are complementary to ours, and the authors concluded that the guidelines for withholding GLP-1RAs preoperatively could be liberalized. Finally, a large US dataset was used to identify 2256 GLP-1RAs users and 11 405 non-GLP-1RAs users with diabetes who underwent an operative procedure; a variety of GLP-1RAs were included, and patients had to be using them for ≥6 months prior to their operation. These authors also found that the risk of postoperative aspiration/pneumonitis was not significantly elevated (OR 0.78, 95% CI 0.29–2.09).17 After the American Society of Anesthesiologists communication, the American Gastroenterological Association published a clinical practice update which states that a lack of available studies precludes making a rigorous evidence-based statement on the use of GLP-1RAs at the time of endoscopy. They suggest that an individualized approach should be used in the management of GLP-1RA use at the time of endoscopy, and that low-risk patients (those who have fasted for 8 h and do not have ‘nausea, vomiting, dyspepsia or abdominal distension’) should go ahead with their procedure without an adjustment to semaglutide.18 An editorial in the British Journal of Anaesthesia makes more nuanced recommendations about the perioperative use of these medications and suggests that people with short-term use (<12 weeks) of GLP-1RAs and abdominal symptoms are likely to be at highest risk of delayed gastric emptying; importantly, they acknowledge that the negative impact of worsening glycaemic control in the preoperative period must be considered, as this can be linked to negative surgical outcomes.19 Limitations of our study are that results may not be generalizable to those without T2DM (e.g., those using semaglutide solely for weight loss who may use a higher dose) or to other GLP-1RA medications. We also could not measure intraoperative aspiration/regurgitation that did not have a clinical sequela, and we did not have information about patient's glucose control (although we controlled for duration of diabetes and diabetes-related health visits), airway management techniques (specifically the type of induction and airway and cuff pressure), or medications given during the operations (such as opioids and muscle relaxants). Importantly, the outcomes of interest in this study are generally uncommon after elective surgery; that means even with a large population-based sample, our statistical power was limited, and an OR within our study outcomes' 95% confidence is possible. Our study and others16, 17 suggest that the risk of clinically significant postoperative aspiration may not be elevated in diabetic patients using GLP-1RAs. However, there may be a risk in certain groups, such as those using higher doses of GLP-1RAs, those that recently started the medication, those with abdominal symptoms, and during certain procedures (such as upper endoscopy). The potential benefits of continuing semaglutide in the perioperative period for its glycaemic and cardiovascular benefits should be further studied.14 The results of this study support further examination of the perioperative safety of GLP-1RA medications in a prospective clinical study, which could also address factors such as airway management and perioperative medications, and could stratify patients by glycaemic control and duration of GLP-1RA use. All authors contributed to the conception and design of the work. Brooke Carter performed the statistical analysis. Blayne Welk drafted the manuscript, Andrew McClure, Luc Dubois, Colin Clarke and Kristin Clemens revised it critically for important intellectual content. All authors gave final approval and agreed to be accountable for all aspects of the work. Brooke Carter and Andrew McClure received compensation through their employment for conducting this study. Blayne Welk and Brooke Carter had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health and the Ministry of Long-Term Care. This document used data adapted from the Statistics Canada Postal CodeOM Conversion File, which is based on data licensed from Canada Post Corporation, and/or data adapted from the Ontario Ministry of Health Postal Code Conversion File, which contains data copied under licence from ©Canada Post Corporation and Statistics Canada. We thank IQVIA Solutions Canada Inc. for use of their Drug Information File. Parts of this material are based on data and information compiled and provided by Ontario Ministry of Health and the Canadian Institute for Health Information. The analyses, conclusions, opinions and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred. Blayne Welk has been a consultant for Becton, Dickinson and Company (outside of this work). Kristin Clemens has received funding from the CPD Network for delivering certified medical education (outside of this work). Andrew McClure, Brooke Carter, Colin Clarke, Luc Dubois: None. Based on the administrative nature of data used, individual patient consent was not required per section 45 of Ontario's Personal Health Information Protection Act (PHIPA). The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/dom.15711. The dataset from this study is held securely in coded form at ICES. Although legal data-sharing agreements between ICES and data providers (e.g., healthcare organizations and government) prohibit ICES from making the dataset publicly available, access may be granted to those who meet prespecified criteria for confidential access, available at www.ices.on.ca/DAS. The full dataset creation plan and underlying analytical code are available from the authors upon request, understanding that the computer programs may rely upon coding templates or macros that are unique to ICES and are therefore either inaccessible or may require modification. Data S1. Supporting information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts