GLP-1 Agonists in the Hospital: What Every Intern Should Know

GLP-1 Agonists in the Hospital: What Every Intern Should Know

The patient on Ozempic is in bed 4, and now it's your problem

You're an intern. It's 7 a.m. Your new admission is a 58-year-old woman with type 2 diabetes who's been on semaglutide for a year. She's here for a pneumonia. As you do medication reconciliation, you realize the weekly shot she took on Saturday is still very much in her system, her glucose is drifting, and anesthesia is going to want to talk to you before her bronchoscopy tomorrow.

Five years ago this was a rare curiosity. Today it's a Tuesday. GLP-1 receptor agonists have gone from a specialty endocrinology drug to one of the most common things you'll see on a medication list. Roughly one in eight U.S. adults is now on one of these drugs, for diabetes, for weight, or for their heart. And because the big trials mostly excluded the sick and the surgical, there's no neat textbook chapter telling you what to do with them once they're in your hallway.

So here's the practical version, the stuff that actually changes what you do at the bedside.

First: you probably shouldn't be starting one

The single most common mistake is trying to be helpful and initiating a GLP-1 during an admission. Don't. Acute illness is a terrible time to start a drug whose main side effects are nausea, vomiting, and early satiety. You want gradual dose escalation and a clinician who will follow the patient longitudinally, not an intern who will be gone in three weeks.

The right move is to defer initiation and hand it to a weight-management or endocrine specialist, ideally right before or after discharge. If a patient is admitted for something non-urgent and genuinely needs to start, that's the one narrow exception, and it comes with clear outpatient follow-up. Otherwise, let the outpatient team do it.

Hold it or keep going? (And the restart trap)

For most medical admissions, holding the GLP-1 is the reasonable default. The patient is often sick, possibly NPO, and may need a procedure. You're not committing any crime by pausing it, and there's no need to taper these drugs off. Just hold it and manage glucose with insulin in the meantime.

Here's the part that trips people up on the way out. If a patient has missed about two doses, do not just restart them at their old, higher dose. The GI side effects will come roaring back. Re-titrate from the lowest dose and climb back up. It feels bureaucratic, but it's the difference between a smooth discharge and a patient vomiting in the car on the way home.

That said, not everyone needs a hold. A stable patient who's eating well, has no GI symptoms, and isn't heading to a procedure can often just keep taking their medication. The decision comes down to oral intake, GI symptoms, hemodynamic stability, and whether a procedure is on the horizon. When in doubt, hold it, document that you held it on purpose, and write down when you'll reassess.

Hypoglycemia: the quiet one

On their own, GLP-1s carry a low risk of hypoglycemia. Their insulin-stimulating effect is glucose-dependent, so they tend to dial down when the sugar is already low. The danger shows up in combination. The moment you add or adjust insulin, or the patient is also on a sulfonylurea, the risk becomes additive. If you're starting an insulin drip or a basal-bolus regimen for a patient who's on a GLP-1, be deliberate about not over-correcting.

There's a subtler trap with diabetic ketoacidosis. A GLP-1 doesn't directly cause DKA, but it can mask a rising glucose in an insulin-deficient patient, so the numbers look "fine" while the ketones climb. In type 1 diabetics and insulin-deficient type 2s, check the ketones, not just the glucose. This is the kind of thing that looks innocent on the first check and ugly on the second.

The stomach is the whole story

If you remember one mechanism, make it this: GLP-1s slow gastric emptying. That's not a side effect you can ignore, it's a core pharmacologic effect, and it's the reason every perioperative conversation circles back to aspiration.

The numbers are modest but real. Liraglutide cut one-hour gastric emptying by roughly 13 to 23 percent in studies. Semaglutide at 2.4 mg slowed paracetamol absorption by about 8 percent, which is a neat way to see the same effect on drug delivery. And semaglutide's half-life is about seven days, meaning it takes on the order of three weeks for levels to fall below 10 percent. A short pre-op hold does not reset the stomach to baseline.

There's a case in the literature of a woman on weekly semaglutide who showed up for an endoscopy after an 18-hour fast and still had a stomach full enough that they had to suction it before intubation. That's the scenario anesthesia is worried about.

There's a second, less discussed effect: slowed emptying can blunt the absorption of oral medications that matter, like the antiplatelet drugs clopidogrel and ticagrelor. If a patient's anticoagulation or antiplatelet plan depends on a pill being absorbed on a tight timeline, that's worth a thought.

Perioperative management: the guidelines moved

This is where the guidance has actually evolved, and it's worth knowing the current position rather than the old reflex of "hold everything."

The older American Society of Anesthesiologists position was to hold the daily dose on the day of the procedure and hold weekly formulations about a week before. That's still a reasonable institutional default in a lot of places.

The 2024 multi-society guidance (ASA, AGA, ASMBS, ISPO, and SAGES) pushed in a different direction: most patients can continue their GLP-1 before elective surgery. The emphasis shifted to risk stratification. Patients at the highest risk for delayed emptying, meaning those in the early dose-escalation phase (the first four to eight weeks), those with active GI symptoms like nausea or vomiting, and those on higher doses, should follow a liquid-only diet for 24 hours before the procedure. Point-of-care ultrasound to look at stomach contents is the tool of choice for the highest-risk patients right before the case.

And if the risk is high or the ultrasound is inconclusive, you fall back on full-stomach precautions: rapid-sequence intubation, nasogastric aspiration, and a cuffed endotracheal tube. The gastroenterology society's take is similar, noting there's no hard evidence of increased aspiration or pneumonia risk and favoring an individualized approach with a liquid diet the day before.

The practical summary for your service: ask about GI symptoms, consider a gastric ultrasound if you have one, and push a liquid diet for 24 hours before the procedure. Then follow whatever your institution's anesthesia team actually does, because that's what will govern the case.

The pancreatitis question

Every intern asks about this, so let's settle it. Back in the exenatide era there were early reports linking GLP-1s to acute pancreatitis, and the labeling carried a warning. But the larger observational studies and meta-analyses on the newer agents (semaglutide, liraglutide, dulaglutide) have not confirmed a meaningful increase in risk. The absolute risk is low. Keep a reasonable threshold to evaluate new or worsening abdominal pain, but don't reflexively blame the GLP-1 for every pancreatitis you see.

Use the admission to fix the medication list

Here's an underused opportunity. Hospitalizations are when you get to look at the whole list, and a lot of it is quietly promoting weight gain. If your patient's obesity is part of the picture, this is the moment to make swaps:

  • Beta-blockers cause weight gain; carvedilol is usually the most weight-neutral of the class, and ACE inhibitors, ARBs, and calcium channel blockers are alternatives when appropriate.
  • Corticosteroids wreck glucose tolerance; reach for NSAIDs when they'll do the job.
  • Insulin is anabolic; a basal-only regimen tends to add less weight than combo or pre-mixed regimens.
  • SSRIs, TCAs, and atypical antipsychotics increase appetite and cravings; bupropion is often a better fit when it suits the patient.
  • Metformin can help offset the weight gain from a drug you genuinely need to keep.

It's a small thing, but it's the kind of thinking that separates managing a number from managing a person.

Why they're on these drugs in the first place

It helps to understand why the patient is on a GLP-1, because the answer is increasingly "not just diabetes." The evidence has spread well beyond glucose:

  • Kidney: the FLOW trial showed semaglutide slowed the decline in kidney function and cut cardiovascular death in diabetic kidney disease.
  • Heart, without diabetes: the SELECT trial showed semaglutide reduced the composite of cardiovascular death, heart attack, or stroke by about 20 percent in people who didn't even have diabetes.
  • Heart failure: the SUMMIT trial showed tirzepatide reduced worsening heart failure events and cardiovascular death in preserved-ejection-fraction heart failure with obesity.
  • Liver: the ESSENCE study showed semaglutide reversed steatohepatitis in a majority of patients with MASH and fibrosis.

None of this changes what you do at the bedside today, but it explains why these patients are on the drugs, why they're anxious about stopping them, and why "just hold it" deserves a follow-up plan rather than a silent discontinuation.

The bedside checklist

When the patient on a GLP-1 lands in your hallway, here's the short list:

  1. Don't start one in the hospital. Refer initiation to the outpatient team.
  2. Holding is reasonable for most admissions. No taper needed.
  3. If about two doses were missed, re-titrate from the lowest dose on restart.
  4. Watch for additive hypoglycemia with insulin or sulfonylureas; check ketones in insulin-deficient patients.
  5. Treat them as potentially "full stomach" for procedures: ask about GI symptoms, consider gastric ultrasound, and use a liquid diet for 24 hours prior.
  6. Defer elective surgery during the dose-escalation phase or active GI symptoms.
  7. Remember slowed emptying can blunt oral drug absorption.
  8. Use the admission to swap weight-promoting drugs for weight-neutral ones.
  9. Keep a low threshold to evaluate abdominal pain, but don't over-attribute it to the GLP-1.

These drugs are here to stay, and the inpatient team is going to be the one making the real-time calls. You don't need to be the prescriber. You just need to know what the drug is doing to the stomach, the glucose, and the procedure, and to document your decisions so the next team isn't guessing.