Acute Pancreatitis in the ED: Measured Fluids, Opioids, Early Food
Chester "Chet" Shermer, MD, FACEP · November 12, 2026
About the author: Dr. Chester Shermer, MD, FACEP
What WATERFALL and the 2024-2026 guidelines change in the first hours: lactated Ringer's by weight, a six-hour recheck, opioids for severe pain, early food, no admission CT and no prophylactic antibiotics.
A 47-year-old long-haul trucker comes in at 0200 with eight hours of constant epigastric pain that bores through to his back, and he has vomited three times. He weighs 100 kg. His heart rate is 112, respiratory rate 22, temperature 37.9 °C and blood pressure 128/80. The lipase is nine times the upper limit of normal, the BUN is 24 mg/dL and the hematocrit is 47%, and your bedside ultrasound shows gallstones.
He has acute pancreatitis, and you can manage his first hours without a CT. Give him a measured volume of lactated Ringer's and recheck him early, treat his pain with an opioid now, plan to feed him as soon as he is hungry, hold the antibiotics, and admit him to a monitored bed because he meets SIRS criteria.
Make the diagnosis without the scanner
The diagnosis needs two of three criteria: typical abdominal pain, a lipase or amylase above three times the upper limit of normal, or characteristic imaging [1]. Lipase is the better test; amylase can be normal on arrival in as many as one in five patients and falls back to normal sooner [1].
The 2024 American College of Gastroenterology (ACG) guideline advises against routine CT at admission to judge severity [1]. Save CT for the patient whose diagnosis is unclear, or who has not improved 48 to 72 hours after admission and IV fluids [1]. Necrosis is usually absent on arrival and can appear after 24 to 48 hours, so an early scan can look reassuring and still be wrong [1].
Ultrasound for gallstones belongs in every workup [1]. Without gallstones or a real alcohol history, send a triglyceride level; the ACG counts it as the cause when it is above 1,000 mg/dL [1]. Once the diagnosis is made, stop trending the lipase. It has no relation to severity, prognosis, or the decision to feed or discharge [1].
Moderate fluids, not a flood
WATERFALL tested the habit of running fluids hard [2]. At 18 centers, patients with acute pancreatitis received lactated Ringer's by one of two goal-directed schemes [2]:
- Aggressive: a 20 mL/kg bolus, then 3 mL/kg per hour.
- Moderate: a 10 mL/kg bolus only if the patient was hypovolemic (no bolus if not), then 1.5 mL/kg per hour.
Both arms were reassessed and adjusted at 12, 24, 48 and 72 hours [2]. The trial stopped at its first interim analysis, after 249 of a planned 744 patients, because the safety outcome had separated [2]. Fluid overload occurred in 20.5% of the aggressive group and 6.3% of the moderate group (adjusted relative risk 2.85) [2]. Moderately severe or severe pancreatitis occurred in 22.1% and 17.3%, a difference that was not significant [2]. Median stay was 6 days versus 5 [2]. An early stop leaves the efficacy question open. The harm signal is the firm finding.
The guidelines that followed agree on direction but differ in wording. The 2024 ACG guideline suggests "moderately aggressive" resuscitation, with more boluses for hypovolemia (conditional recommendation, low-quality evidence) [1]. It reads WATERFALL as no more than 1.5 mL/kg per hour without hypovolemia and a 10 mL/kg bolus with it, yet also says most patients will likely benefit from 3 to 4 L in the first 24 hours, depending on body mass index [1]. The 2026 Ibero-Latin American guideline recommends 1.5 mL/kg per hour, preceded by 10 mL/kg over 2 hours when hypovolemia is present, repeated if it persists, and cut back or stopped if the patient becomes overloaded [3].
For the 100 kg trucker, who is tachycardic with a high BUN and hematocrit, that means 1 L of lactated Ringer's over 2 hours, then 150 mL an hour. That is a little over 4 L in the first day, just above the ACG's 3 to 4 L estimate, which it scales to body size. Write the plan as a rate per kilogram: the same scheme gives the 55 kg retiree in the next bed about 2.4 L if he is hypovolemic and gets the bolus, and about 2 L if not.
Reassess at the bedside within 6 hours: heart rate, blood pressure, urine output, and a falling BUN as the laboratory target [1]. Go slower in patients with heart or kidney disease and watch for overload [1].
Lactated Ringer's is preferred over saline, a conditional recommendation on low-quality evidence [1]. A 2025 meta-analysis of six randomized trials and four observational studies (1,500 patients) found lower odds of moderate-to-severe disease with lactated Ringer's (OR 0.48), a shorter stay and fewer ICU admissions, but no difference in mortality, organ failure or SIRS [4]. If the patient is hypotensive on arrival, you are treating shock, and the undifferentiated hypotension approach comes first.
Treat severe pain with an opioid
The 2026 European interdisciplinary guideline on pain in acute pancreatitis suggests strong opioids for patients hospitalized with the disease, because they relieved pain better than non-opioids and reduced rescue doses (low-certainty evidence, weak recommendation) [5]. The trials that favored opioids used buprenorphine and pentazocine, both mixed agonist-antagonists, against diclofenac, and the guideline names those two; applying the result to morphine, hydromorphone or fentanyl is an extrapolation [5]. It also states that current evidence suggests opioids may not increase adverse outcomes, and that safety concerns should not delay them in severe pain [5]. It urges caution with patient-controlled analgesia [5].
NSAIDs and paracetamol fit milder pain [3], with limits. A 2026 meta-analysis of 13 randomized trials found pain relief comparable to opioids and fewer local complications (RR 0.59), and its authors flag bias and heterogeneity in the trials [6]. The European guideline found no excess adverse events with NSAIDs in trials, though those trials were underpowered, and it flags renal and bleeding risk in patients prone to organ failure [5]. The trucker has a high BUN and SIRS, so his kidneys are already under strain. He gets an opioid first.
Feed early and hold the antibiotics
In mild acute pancreatitis, the ACG suggests oral feeding within 24 to 48 hours as tolerated, starting with a low-fat solid diet rather than a stepwise climb from clear liquids (conditional recommendations, low-quality evidence) [1]. The 2026 Ibero-Latin American guideline extends early oral feeding to predicted severe disease when the patient feels able to eat, and allows mild cases to eat at once [3]. A patient who wants breakfast does not need an NPO order and a lipase recheck.
The ACG suggests against prophylactic antibiotics even in severe acute pancreatitis (conditional recommendation, very low-quality evidence) [1]. Early SIRS can look exactly like sepsis. If you suspect an infection, start antibiotics while you look for the source, and stop them if the cultures return negative and no source is found [1]. Cholangitis is the exception you treat: it gets antibiotics and ERCP within 24 hours [1][3]. Biliary pancreatitis without cholangitis does not need early ERCP [1].
For mild gallstone pancreatitis, the ACG advises early cholecystectomy, preferably before discharge [1]. Put that in the admission conversation tonight.
Decide the bed by physiology
No score or early scan reliably picks out the patient who will become severe [1]. Severe acute pancreatitis means organ failure that lasts beyond 48 hours [1]. The ACG's risk factors are bedside facts: SIRS, a high or rising BUN, a high hematocrit, obesity and comorbidity, with a hematocrit of 44% or more and a BUN of 20 mg/dL or more each linked to moderately severe or severe disease [1]. SIRS in the first 24 hours is 85% sensitive for organ failure but only 41% specific for severe disease [1].
Patients with SIRS or organ failure should go to a monitored bed, and to the ICU if organ failure is already present [1]. Even a patient who looks mild needs watching for the first 48 hours [1]. The fluid rate, the time of the next BUN and the criteria for slowing down belong in the handoff.
Key Takeaways
- Run fluids by weight: lactated Ringer's at 1.5 mL/kg per hour, with a 10 mL/kg bolus only for hypovolemia; in WATERFALL, the aggressive scheme raised fluid overload from 6.3% to 20.5% without better outcomes [2][3].
- Recheck within 6 hours: heart rate, blood pressure, urine output and a falling BUN, and slow down early in heart or kidney disease [1].
- Skip the admission CT and the lipase trend: two of three criteria make the diagnosis, and CT is for doubt or no improvement at 48 to 72 hours [1].
- Give a strong opioid for severe pain without delay, feed when the patient is hungry, and give antibiotics only for cholangitis or a suspected infection [1][5].
FAQ
How much IV fluid should a patient with acute pancreatitis get in the ED?
Start lactated Ringer's at 1.5 mL/kg per hour, and give a 10 mL/kg bolus first only if the patient is hypovolemic [2][3]. Reassess within 6 hours and adjust to heart rate, blood pressure, urine output and BUN [1]. The aggressive WATERFALL scheme (20 mL/kg, then 3 mL/kg per hour) caused more fluid overload without improving outcomes [2].
Is lactated Ringer's better than normal saline for pancreatitis?
The ACG prefers lactated Ringer's, as a conditional recommendation on low-quality evidence [1]. A 2025 meta-analysis found lower odds of moderate-to-severe disease and fewer ICU admissions with lactated Ringer's, with no difference in mortality or organ failure [4].
Do you need a CT scan to diagnose acute pancreatitis?
Usually not. Typical pain plus a lipase above three times the upper limit of normal makes the diagnosis [1]. CT is for an unclear diagnosis or a patient who has not improved 48 to 72 hours after admission, because necrosis may not be visible early [1].
Dr. Chet's Take
I have practiced emergency medicine for more than 25 years, I work at a Level I academic trauma center, and as a TelEmergency medical director I also think about the small hospitals that carry this patient overnight with fewer hands. The article holds up. Its best move is to take the fluid decision away from habit and tie it to the patient's weight and physiology. Pancreatitis tempts us to do too much in the first hour and too little in the next six. The first hour feels like resuscitation, so it gets attention. The next six feel like waiting, so the follow-up gets skipped. I would rather see a modest rate with a written recheck time than a big number with no plan behind it. The evidence here is not perfect, but it points one way, and the guidelines now point the same way.
That being said, I think the article understates where this plan fails: in the hours after the order is written. A moderate rate is safe only if someone comes back to the patient. In a crowded department, the patient who is boarding is the one whose recheck slips, and an infusion pump does not page anybody when the BUN climbs. The pump keeps running at 0300 whether or not anyone has listened to the lungs or looked at the urine bag since midnight. In a small hospital with no endoscopist, the patient with cholangitis has a transfer ahead of him, and that call belongs to the emergency department. The honest answer is that I have not seen good data on how often boarding patients get the recheck on time.
If you are the attending, write three things in the order: the rate in mL/kg per hour, the time of the next BUN and vital signs, and what would make you give more or less. If the patient boards in your department, that recheck belongs to you until he physically leaves, whatever the bed board says. Order the analgesia at the same time as the fluids, and look at the creatinine before you choose the drug. Leave the NPO order off unless he is still vomiting. If you are covering a hospital without ERCP and the bilirubin and the fever tell you cholangitis, start the transfer before the admission paperwork. The fluid order is not complete until it names the time of the next look.
— Chester Shermer, MD, FACEP | Emergency Medicine, 25+ Years Clinical Experience | State Surgeon
Continue Your Training
If you're an emergency physician (or any clinician treating patients daily) trying to understand how AI will actually impact your clinical practice — not just the hype — I put together a free practical guide. You can download the AI in EM Survival Guide here.
Browse All Courses at Global MedOps Command
Relevant Reading on Global MedOps Command:
How to Avoid Becoming an AI Casualty — Dr. Shermer's guide to navigating AI tools in clinical and operational settings without compromising judgment or patient outcomes.
Emergency Department Efficiency Playbook — Practical systems for throughput, triage optimization, and operational efficiency.
Read more from Dr. Shermer on Medium
Connect with Dr. Shermer: LinkedIn — Chester "Chet" Shermer, MD, FACEP.
References
- Tenner S, Vege SS, Sheth SG, et al. American College of Gastroenterology guidelines: management of acute pancreatitis. Am J Gastroenterol. 2024;119(3):419-437. PubMed 38857482
- de-Madaria E, Buxbaum JL, Maisonneuve P, et al. Aggressive or moderate fluid resuscitation in acute pancreatitis. N Engl J Med. 2022;387(11):989-1000. PubMed 36103415
- Cárdenas-Jaén K, Guilabert L, Martínez-Moneo E, et al. AEG-AESPANC-OPGE-SIED-SPG Ibero-Latin American guidelines on acute pancreatitis (iLATAM-AP). United European Gastroenterol J. 2026;14(5):e70190. PubMed 42216617
- Zhao T, Kang Z, Zhang Q, et al. Lactated Ringer's solution versus saline fluid resuscitation for reducing progression to moderate-to-severe acute pancreatitis: a systematic review and meta-analysis. Int J Surg. 2025;111(5):3467-3480. PubMed 40085761
- Pandanaboyana S, Knoph CS, Kuhlmann L, et al. European interdisciplinary guidelines on pain management in acute pancreatitis. United European Gastroenterol J. 2026;14(5):e70230. PubMed 42299777
- Bandyopadhyay S, Khan R, Samajdar SS, et al. Early analgesia for the management of acute pancreatitis: a systematic review and meta-analysis of randomized controlled trials. Pancreas. 2026;55(9):e945-e954. PubMed 42153651
Related simulation training: Books by Dr. Shermer