Safety of endoscopic gastrostomy tube placement compared with radiologic or surgical gastrostomy: nationwide inpatient assessment.

BACKGROUND AND AIMS

A gastrostomy tube is often required for inpatients requiring long-term nutritional access. We compared the safety and outcomes of 3 techniques for performing a gastrostomy: percutaneous endoscopic gastrostomy (PEG), fluoroscopy-guided gastrostomy by an interventional radiologist (IR-gastrostomy), and open gastrostomy performed by a surgeon (surgical gastrostomy).

METHODS

Using the Nationwide Readmissions Database, we identified hospitalized patients who underwent a gastrostomy from 2016 to 2017. They were identified using the International Classification of Diseases, 10th Revision, Procedure Coding System. The selected patients were divided into 3 cohorts: PEG (0DH64UZ), IR-gastrostomy (0DH63UZ), and open surgical gastrostomy (0DH60UZ). Adjusted odds ratios for adverse events associated with each technique were calculated using multivariable logistic regression analysis.

RESULTS

Of the 184,068 patients meeting the selection criteria, the route of gastrostomy tube placement was as follows: PEG, 16,384 (53.7 ± 29.0 years); IR-gastrostomy, 154,007 (67.2 ± 17.5 years); and surgical gastrostomy, 13,677 (57.9 ± 24.3 years). Compared with PEG, the odds for colon perforation using IR-gastrostomy and surgical gastrostomy, respectively, were 1.90 (95% confidence interval [CI], 1.26-2.86; P = .002) and 6.65 (95% CI, 4.38-10.12; P < .001), for infection of the gastrostomy 1.28 (95% CI, 1.07-1.53; P = .006) and 1.61 (95% CI, 1.29-2.01; P < .001), for hemorrhage requiring blood transfusion 1.84 (95% CI, 1.26-2.68; P = .002) and 1.09 (95% CI, .64-1.86; P = .746), for nonelective 30-day readmission 1.07 (95% CI, 1.03-1.12; P = .0023) and 1.13 (95% CI, 1.06-1.2; P = .0002), and for inpatient mortality 1.09 (95% CI, 1.02-1.17; P = .0114) and 1.55 (95% CI, 1.42-1.69; P < .0001).

CONCLUSIONS

Placement of a gastrostomy tube (PEG) endoscopically is associated with a significantly lower risk of inpatient adverse events, mortality, and readmission rates compared with IR-gastrostomy and open surgical gastrostomy.