P1474 - Effect of Hospital Teaching Status on Percutaneous Endoscopic Gastrostomy Tube Placement in Ascites Patients: A Population-Based Study

Amrita Chawla, MD, Faiz Afridi, MD, Reza Hashemipour, MD, Sushil Ahlawat, MD
Rutgers New Jersey Medical School, Newark, NJ

Introduction: Ascites is viewed as a relative and often absolute contraindication to the insertion of a percutaneous endoscopic gastrostomy (PEG) tube. Fluid accumulation in the peritoneal cavity may prevent proper positioning of the gastric body along the abdominal wall hindering healing of a PEG tract. Moreover, fluid may drain through the PEG tract increasing the risk of infection, bleeding, and aspiration pneumonia. Nonetheless, PEG tube placement may be required in certain circumstances to ensure proper nutrition. This study aims to assess teaching versus nonteaching hospital inpatient outcomes in PEG tube placement in ascites patients.



Methods: Cases with ascites and associated procedure of PEG tube placement were identified from the 2001-2014 National Inpatient Sample. Baseline characteristics were analyzed with Rao Scott chi squared and Mann-Whitney tests. Teaching hospital admissions were propensity score matched 1:1 against controls across covariates. Primary outcomes of complications rates of pneumonia, respiratory failure, shock, peritonitis, and blood transfusion were assessed with adjusted multivariable logistic regression. Secondary outcomes of mortality, total charges, and length of stay (LOS) were examined via multivariable regression.

Results: A total 15,251 cases of PEG tube placement in ascites were identified for teaching hospitals versus 9,305 for nonteaching. Among all cases of ascites, teaching hospitals had a higher rate of PEG tube placement than nonteaching hospitals (0.94% vs 0.73%, OR: 1.28, 95% CI 1.18 - 1.4, P< 0.001). Post propensity match, teaching hospitals had lower complication rates of pneumonia (aOR: 0.78, 95% CI 0.65 - 0.93, P=0.006), respiratory failure (aOR: 0.83, 95% CI 0.7 - 0.98, P=0.03), blood transfusion (aOR: 0.78, 95% CI 0.65 - 0.93, P=0.007), and shock (aOR: 0.83, 95% CI 0.7 - 1, P=0.046) [Table 1]. There was no significant difference in rates of peritonitis, overall mortality, disposition, or total charges. Teaching hospitals had a higher median LOS by one day (23 vs 22 days, aIRR: 1.13, 95% CI 1.04-1.21, P=0.002).

Discussion: PEG tube placement in ascites patients is associated with fewer severe complications at teaching hospitals compared to nonteaching hospitals. Given the high risk nature of PEG tube insertion in ascites, further review is needed to understand the drivers of worse outcomes in nonteaching hospitals in order to ensure consistent care and adherence to best practice.

	Teaching	Nonteaching	Adj Odds Ratio1	95% Conf Interval	Pvalue
	N = 7183	N = 6990
Complications
Peritonitis/Intestinal Abscess	11.4%	11.4%	0.987	(0.78 - 1.26)	0.912
Pneumonia	29.0%	34.5%	0.779	(0.65 - 0.93)	0.006*
Shock	23.9%	27.8%	0.833	(0.7 - 1)	0.046*
Respiratory Failure	48.5%	53.8%	0.827	(0.7 - 0.98)	0.027*
Blood Transfusion	39.1%	45.2%	0.777	(0.65 - 0.93)	0.007*
In-Hospital Outcomes
Total Charges2,3	$179,379 ($80,536 - $397,328)	$188,905 ($83,878 - $381,537)	1.039	(0.95 - 1.14)	0.429
Mortality	17.5%	17.8%	0.981	(0.8 - 1.2)	0.852
Disposition	9.0%	6.8%	1.275	(0.92 - 1.76)	0.143
Length of Stay2,4	23 (13 - 38)	22 (13 - 35)	1.126	(1.04 - 1.21)	0.002*

Table: Title: Complication Rates and In-hospital Outcomes of the Propensity Matched Cohorts

Footnotes:
1. Adjusted for age, sex, race, liver disease, comorbidities, payer, income, hospital size, hospital region, admission type
2. Median (Interquartile Range)
3. Gamma GLM regression coefficient
4. Incident Rate Ratio from Poisson regression
Counts weighted with National Inpatient Sample trend weights post propensity match
*P < 0.05


Disclosures:
Amrita Chawla indicated no relevant financial relationships.
Faiz Afridi indicated no relevant financial relationships.
Reza Hashemipour indicated no relevant financial relationships.
Sushil Ahlawat indicated no relevant financial relationships.


Amrita Chawla, MD, Faiz Afridi, MD, Reza Hashemipour, MD, Sushil Ahlawat, MD. P1474 - Effect of Hospital Teaching Status on Percutaneous Endoscopic Gastrostomy Tube Placement in Ascites Patients: A Population-Based Study, ACG 2021 Annual Scientific Meeting Abstracts. Las Vegas, Nevada: American College of Gastroenterology.