US Emergency Department Visits Clinicians Attributed to Semaglutide Adverse Events, 2022–2023

Introduction

Semaglutide is a glucagon-like peptide-1 (GLP-1) receptor agonist approved for treatment of type 2 diabetes and obesity. In randomized controlled trials, semaglutide was associated with increased rates of nausea, diarrhea, vomiting and constipation, and post-marketing studies of GLP-1 agonists prior to US Food and Drug Administration (FDA) approval of semaglutide for weight loss found an increased risk of gastroparesis, pancreatitis, and bowel obstruction.(1–3) The clinical manifestations that most commonly lead to emergency department (ED) visits for semaglutide adverse events in routine clinical practice, and the national burden of those visits, is unknown. We used ED-based nationally-representative surveillance data to characterize and quantify US ED visits clinicians attributed to semaglutide adverse events.

Methods

The National Electronic Injury Surveillance System–Cooperative Adverse Drug Event Surveillance (NEISS-CADES) project, an active public health surveillance system based on a nationally-representative size-stratified probability sample of US EDs, is a collaborative effort of the US Centers for Disease Control and Prevention, US Consumer Product Safety Commission (CPSC), and FDA designed to detect ED visits attributable to medication harms. NEISS-CADES data were used to identify cases of ED visits clinicians attributed to semaglutide adverse events by adults aged ≥18 years from January 1, 2022 through December 31, 2023, the 2 years following FDA approval for weight loss.

As described previously, trained abstractors at 82 hospitals that voluntarily participate in NEISS-CADES reviewed all ED records to identify adverse events that treating clinicians attributed to use of drugs.(4) To ensure the national representativeness of the data, each case is weighted based on the inverse probability of selection for hospitals in each stratum, and weights are adjusted for nonresponse and hospital nonparticipation and poststratified to adjust for the total number of hospital ED visits annually. National estimates and 95% CIs were calculated using the PROC SURVEYMEANS procedure in SAS version 9.4 to account for the weights and complex sample design. This activity was reviewed by CDC, deemed not research, and was conducted consistent with applicable federal law and CDC policy.¹

Results

Based on 551 cases, there were an estimated 24,499 ED visits (95% CI, 17,022–31,976) clinicians attributed to semaglutide adverse events in 2022–2023 (Table 1); 82.6% (20,226 visits; 95% CI, 13,607–26,846) were in 2023. Of ED visits clinicians attributed to semaglutide adverse events, mean patient age was 50.9 years, and 73.2% of patients were female. Semaglutide was often the only medication implicated (81.7%), and was nearly always in injectable form (94.0%); 9.0% of visits involved medication errors. Overall, ED visits most commonly involved gastrointestinal effects (69.3%), hypoglycemia (16.5%), and allergic reactions (5.5%)(Table 2). Of the gastrointestinal effects, nausea/vomiting(83.2%), abdominal pain(36.2%), and diarrhea (17.7%) were most commonly reported. Hospitalization was required in 37.6%(95% CI, 23.7%-51.4%) of ED visits for hypoglycemia and 15.0%(95% CI, 9.4%-20.6%) for gastrointestinal effects.

Table 1.

Characteristics of Emergency Department Visits that Clinicians Attributed to Semaglutide Adverse Events -- US, 2022–2023^a

Patient and Case Characteristics	Cases (Unweighted)	Nationally-Weighted Estimateb
	n	n	% (95% CI)
Age, y
18–24	26	--	--
25–34	75	3,472	14.2 (10.9 – 17.4)
35–44	98	3,847	15.7 (12.0 – 19.4)
45–54	113	5,513	22.5 (18.0 – 27.0)
55–64	128	5,939	24.2 (20.1 – 28.4)
65+	111	4,719	19.3 (14.5 – 24.0)
Sex
Female patient	405	17,937	73.2 (68.8 – 77.6)
Male patient	146	6,562	26.8 (22.4 – 31.2)
Race
White	317	14,870	60.7 (46.1 – 75.3)
Black	113	3,831c	15.6 (7.2 – 24.1)
Other	18	--	--
Unspecified	103	5,164c	21.1 (5.8 – 36.4)c
Ethnicity
Hispanic	73	2,071c	8.5 (4.0 – 12.9)
Non-Hispanic	374	16,748	68.4 (53.8 – 82.9)
Unspecified	104	5,679c	23.2 (7.2 – 39.2)c
Semaglutide dosage form
Injectable	523	23,029	94.0 (90.9 – 97.1)
Oral	28	1,470	6.0 (2.9 – 9.1)
No. of implicated medications d
1	440	20,005	81.7 (77.0 – 86.4)
>1	111	4,493	18.3 (13.6 – 23.0)
Error Documented
Yes	55	2,216	9.0 (6.1 – 12.0)
No	496	22,283	91.0 (88.0 – 93.9)
Total	551	24,499	100.0

^aNational estimates and 2-sided 95% CIs were calculated from statistical weighting of 551 cases from 82 nationally representative hospitals participating in the National Electronic Injury Surveillance System–Cooperative Adverse Drug Event Surveillance Project for 2022–2023, Centers for Disease Control and Prevention.

^bNational estimates may vary for similar numbers of cases because of statistical weighting. Estimates based on <20 cases or total estimates of <1200 visits are considered statistically unstable and are not shown (--). Estimates with a coefficient of variation >30% may be statistically unstable and are noted.

^cCoefficient of variation >30%.

^dIncludes prescription or over-the-counter (OTC) medications, herbal/dietary supplements, homeopathic products, or vaccines.

Table 2.

Clinical Manifestations that Emergency Department Clinicians Attributed to Semaglutide Adverse Events -- US, 2022–2023^a

Clinical Manifestation and Disposition	Cases (Unweighted)	Nationally-Weighted Estimatesb
	n	n	% (95% CI)
Adverse Event Manifestation c
Hypoglycemia	111	4,043	16.5 (10.6 – 22.4)
with other diabetes agentsd	90	3,205	13.1 (7.8 – 18.3)
without other diabetes agents	21	--	--
Gastrointestinal effectse	363	16,979	69.3 (63.2 – 75.4)
Nausea and/or vomiting	301	14,123	57.6 (52.3 – 63.0)
Abdominal pain and/or discomfort	133	6,139	25.1 (18.9 – 31.2)
Diarrhea	67	3,001	12.2 (9.2 – 15.3)
Constipation	21	--	--
Allergic reaction	29	1,338	5.5 (3.7 – 7.2)
Dehydration	15	--	--
Other effects	28	--	--
Disposition
Not hospitalized	453	20,124	82.1 (77.4 – 86.9)
Hospitalizedf	98	4,375	17.9 (13.1 – 22.6)
Total	551	24,499	100.0

Abbreviations: ED, emergency department

^aNational estimates and 2-sided 95% CIs were calculated from statistical weighting of 551 cases from 82 nationally representative hospitals participating in the National Electronic Injury Surveillance System–Cooperative Adverse Drug Event Surveillance Project for 2022–2023, Centers for Disease Control and Prevention.

^bNational estimates may vary for similar numbers of cases because of statistical weighting. Estimates based on <20 cases or total estimates of <1200 visits are considered statistically unstable and are not shown (--).

^cAdverse event manifestation categories are mutually exclusive and were assigned hierarchically as listed (from top to bottom). For example, an ED visit in which a patient experienced both hypoglycemia and gastrointestinal effects would be categorized only under hypoglycemia; an ED visit in which a patient experienced an allergic reaction and dehydration would be categorized under allergic reaction. Five cases in which the patient experienced a semaglutide medication error without any documented clinical manifestations are not shown.

^dOther diabetes agents included: insulins (70 cases), biguanides [metformin] (25 cases), sodium-glucose cotransporter-2 [SGLT-2] inhibitors (15 cases), sulfonylureas (11 cases), thiazolidinediones (2 cases), oral combination products (2 cases), injectable combination products (1 case), and dipeptidyl peptidase 4 [DPP-4)] inhibitors (1 case).

^eNational estimates of ED visits involving specified gastrointestinal effects are not mutually exclusive (i.e., ED visits involving nausea and constipation would be include in the estimates for both nausea and constipation).

^fDefined as hospital admission, observation status admission, or transfer to another facility.

Of 551 cases, 6 patients(1.1%) were diagnosed with pancreatitis and 4 patients(0.7%) were diagnosed with semaglutide-attributed biliary disease. Semaglutide was implicated alone, without other documented antidiabetic medications, in 21 of 111 cases(18.9%) involving hypoglycemia.

Discussion

With more than 5 million patients dispensed semaglutide in 2023, the estimated 20,226 ED visits in 2023 suggests a rate of less than 4 ED visits for adverse events per 1,000 patients dispensed semaglutide. The majority of ED visits were due to adverse effects previously observed during randomized controlled trials of semaglutide, including gastrointestinal adverse effects and hypoglycemia when combined with other diabetes agents.(1,2)

This study has limitations: implicated drugs are identified based on patient report; comorbid conditions and indication for semaglutide use are not systematically collected; and attribution of adverse events was based solely on ED clinicians’ documentation. These limitations might explain the unexpected finding that semaglutide alone contributed to serious hypoglycemic episodes, even without concurrent insulin or sulfonylureas. Severe hypoglycemia is unexpected given semaglutide’s mechanism of action and lack of hypoglycemic adverse events in randomized controlled trials of semaglutide for the treatment of obesity. However, a recent Poison Center study also raises the possibility that semaglutide alone might cause severe hypoglycemia.(5)

Given these findings, clinicians could counsel patients when initiating semaglutide about the potential of severe gastrointestinal adverse effects and adjust co-prescribed antidiabetic medications to decrease hypoglycemia risk. Additionally, when evaluating patients with new onset gastrointestinal symptoms or hypoglycemia, clinicians could consider recent semaglutide use.