Thirty-Day Major Clinical Reintervention After Gastrostomy Tube Placement: An Adjudicated Multihospital Comparison of...

Reported complication rates after gastrostomy tube (GT) placement vary widely, largely because studies count routine tube care as an adverse event. The incidence, timing, and determinants of clinically meaningful GT failure therefore remain uncertain, as does whether placement route changes the risk of catastrophic failure.

We studied patients undergoing a first or clean index GT placement at 21 hospitals and ambulatory sites within one integrated health system between April 3, 2017 and July 29, 2025. Route was adjudicated as percutaneous endoscopic gastrostomy (PEG), minimally invasive surgical gastrostomy (MIS; laparoscopic or robotic), open surgical gastrostomy, or interventional radiology (IR)-guided gastrostomy. The primary outcome, defined before analysis, was GT-specific major clinical reintervention within 30 days: any unplanned operative or invasive therapeutic procedure attributable to GT failure. Routine tube maintenance was excluded by definition. Candidate events were identified by a high-sensitivity text and CPT screen, then independently adjudicated by 2 surgeons. Associations were estimated with Firth penalized logistic regression adjusted for age category, sex, and body mass index category, with PEG and MIS combined as the reference route. Operator comparisons were prespecified within PEG cases only.

Among 6141 clean index placements [PEG 4821 (78.5%), IR 707 (11.5%), open 386 (6.3%), MIS 227 (3.7%)], 69 patients had a GT-specific major clinical reintervention within 30 days (1.12%; 95% CI: 0.88-1.42), or one event per 89 placements. Crude rates were 1.00% (PEG), 0.44% (MIS), 2.59% (open), and 1.41% (IR). Events clustered early (median: 6 d; IQR: 3-11), with 43 of 69 (62%) in the first week and 9 (13%) after day 14. After adjustment, open placement was associated with higher odds of major reintervention than PEG/MIS (OR: 2.25; 95% CI: 1.13-4.47; P=0.02), an absolute difference of 1.6 percentage points, or one extra event per 62 open placements. IR placement was not (OR: 1.28; 95% CI: 0.65-2.52), nor was any patient characteristic. Within PEG cases, gastroenterology and surgery did not differ significantly (OR: 0.58; 95% CI: 0.31-1.06). The open association persisted across 5 prespecified sensitivity analyses.

With routine tube care excluded and events adjudicated, major GT failure requiring reintervention within 30 days follows roughly 1 in 89 placements. Open placement carried about twice the odds of the endoscopic and laparoscopic routes, whereas radiologic placement did not differ despite serving more difficult anatomy. Where all routes are feasible, an open approach should not be chosen as the presumed safer one.