Fluid Resuscitation in Acute Pancreatitis: Aggressive Bolusing or a Moderate, Goal-Directed Protocol?
A patient with newly diagnosed gallstone pancreatitis and chronic kidney disease sits at the center of a strategy the field itself changed recently — the debate isn't whether to give fluids, but how a trial that stopped early for harm actually applies to a kidney with less room to give.
Walter H. spent thirty-four years driving long-haul routes across the Midwest, retiring eighteen months ago to a small apartment near his daughter's family, where he has slowly adjusted to eating on a fixed schedule instead of whatever a truck stop had ready. He has carried a hypertension diagnosis for over fifteen years, managed on lisinopril and amlodipine, and a routine screening ultrasound two years ago incidentally found gallstones he was told to watch rather than treat — advice he followed until last night, when a large dinner with his daughter's family was followed within hours by epigastric pain boring straight through to his back, then vomiting that brought no relief. He arrived in the emergency department dehydrated and uncomfortable, unable to find a position that eased the pain.
His lipase returned at four times the upper limit of normal, and bedside ultrasound confirmed the gallstones already on record, with no evidence yet of a stone lodged in the common bile duct. He meets two systemic inflammatory response criteria — heart rate 108, temperature 38.1°C — and his labs tell a real, specific story about his volume status: hematocrit 48%, up from a baseline of 41% drawn at a primary-care visit eighteen months ago, and BUN 28, up from his own recent baseline of 22. That is genuine hemoconcentration, not admission noise, and it means real plasma volume has already moved out of his vasculature and into his inflamed peritoneum before the first bag of fluid has even been hung. Layered on top of that honest hypovolemia is his chronic kidney disease, stage 3a, eGFR hovering around 52 for at least the past three years on his outpatient labs — a baseline that makes him more vulnerable to acute injury from under-resuscitation and less able to tolerate the fluid overload that aggressive resuscitation can cause. The question dividing the team isn't whether he needs intravenous fluids — nobody disputes that — it's how fast and how much, in a patient whose kidneys make both wrong answers more costly than they'd be in someone else.
Admission orders, before the first bag is hung
Go with the moderate, goal-directed protocol from the start — that's the strategy WATERFALL (de Madaria et al., NEJM 2022) actually validated, and the trial was stopped early specifically because the aggressive arm caused real harm: fluid overload in 20.5% of patients versus 6.3% on the moderate protocol, with no difference in whether patients actually progressed to moderate or severe disease.
Walter's stage 3a CKD only sharpens that argument. His kidneys have less diuretic reserve to correct an overload than the average trial participant's did, so if anything he's a worse candidate for the aggressive arm than the patients the trial stopped enrolling for.
You're right that the overload signal is real and that it's exactly why the trial stopped early — I'm not arguing for the aggressive protocol. But "moderate" isn't the same as "slow," and his own numbers are already telling us he's genuinely volume-down: a hematocrit that's risen seven points and a BUN that's climbed six points off his own recent baselines, not population norms. If we start him at a flat maintenance-range infusion because "moderate" sounds conservative, we can just as easily undertreat him into acute kidney injury on a baseline that has less room to give than most of WATERFALL's own cohort.
Both of you are actually describing the same protocol correctly, and the trial doesn't force a choice between them the way it's being framed here. WATERFALL's moderate arm was never a fixed drip rate — it's goal-directed: a 10 mL/kg lactated Ringer's bolus specifically for a patient who's hypovolemic on exam, exactly what Walter's exam and labs already show, followed by 1.5 mL/kg/hour with reassessment at defined intervals — against the aggressive arm's 20 mL/kg bolus and 3 mL/kg/hour regardless of exam findings.
He should get the bolus the protocol calls for, not a flat lower rate. What the trial genuinely doesn't answer is how tightly to space reassessment in a patient with his renal history, since WATERFALL wasn't stratified by baseline CKD — that piece we set locally, not something the trial hands us.
Agreed: the moderate, goal-directed protocol, with the 10 mL/kg lactated Ringer's bolus his exam and labs actually earn, then 1.5 mL/kg/hour maintenance. Renal function and urine output reassessed every four hours rather than the trial's own less frequent windows, given his baseline vulnerability — a local addition on top of, not a departure from, WATERFALL's protocol.
Not agreed: the exact threshold for adding a second bolus if his creatinine or urine output hasn't improved by the four-hour mark. The nephrologist wants a low bar to intervene again; the hospitalist is wary of redefining "moderate" case by case until the strategy that actually produced the trial's safety signal quietly erodes. Both orders are written pending that four-hour recheck.