Back to 2014 Annual Meeting Abstracts

---

Prevalence, Impact and Predictors of Hospital Acquired Conditions After Major Surgical Resection for Cancer: a NSQIP Analysis
Daniela Molena^*, Benedetto Mungo, Miloslawa Stem, Anne O. Lidor
Surgery, Johns Hopkins University, Baltimore, MD

Background: The Centers for Medicare and Medicaid Services (CMS) initiated a nonpayment policy for certain hospital acquired conditions (HAC) in 2008. As of 2013, 11 HAC have been identified; however, since their occurrence is linked - at least in part - to preoperative comorbidities, the preventability of HAC in these patients is questionable. This study aimed to determine the rate of the 3 most common HAC in patients undergoing major surgical resections for cancer: surgical site infection (SSI), urinary tract infection (UTI), and deep vein thrombosis (DVT). Additionally, the association of HAC with patients' characteristics and their effect on post-operative outcomes were investigated.
Methods: Patients 18 years of age and older with a diagnosis of esophageal, gastric, hepatic, gallbladder, biliary, pancreatic, colic, anal and lung cancers, who underwent surgical resection were identified using the American College of Surgeons' National Surgical Quality Improvement Program (ACS-NSQIP) database (2005-2012). Patients were grouped into two categories for comparison: HAC versus non-HAC patients. Outcomes, including 30-day mortality, mean length of stay (LOS), return to operating room, readmission and discharge destination were compared. Multiple pre-operative patient variables were considered and multivariate regression analysis was performed to identify risk factors for developing HAC.
Results: 74,381 patients were identified, of whom 9,478 (12.74%) developed at least one of the three HAC. SSI was the most common (7.52%), followed by UTI (2.93%) and VTE (2.30%). The rate of HAC decreased from 15% to 11% over the study period. Pancreatic surgery was associated with the highest rates of SSI (10.88%) and UTI (3.83%), while the highest rates of VTE were observed in esophagectomy (5.92%). HAC patients had significantly higher rates of 30-day mortality (3.65% vs. 2.18%, p<0.001), return to operating room (12.34% vs. 4.61%, p<0.001), 30-day readmission (25.88% vs. 9.36%, p<0.001), and had longer LOS (10 vs. 6 days, p<0.001). Moreover, HAC patients were significantly less likely to be discharged home and more likely to be directed towards rehabilitation, acute care and skilled care. Multivariate analysis revealed that several peri-operative patients' factors, including dyspnea, steroid use and emergent surgery, were significantly associated with HAC (Table).
Conclusion: Our data demonstrate that the development of HAC is strongly associated with pre-operative patient characteristics. These data suggest that the nonpayment policy might be excessively penalizing healthcare providers, since inherent patient factors are not modifiable and seem to play an important role in the development of HAC in this population. These findings are important to help inform health care policy decisions regarding access to care for patients undergoing cancer surgery.
TABLE. Multivariable logistic regression analysis of variables associated with hospital acquired condition (HAC)

Variable	OR	p	95% CI
Age≥80	1.08	0.032	(1.01-1.17)
Race
White	Reference	Reference	Reference
Black	0.98	0.643	(0.89-1.07)
Other	0.94	0.298	(0.83-1.06)
ASA classification
No Disturb/Mild Disturb	Reference	Reference	Reference
Severe Disturb	1.30	<0.001	(1.23-1.38)
Life Threat/Moribund	1.57	<0.001	(1.40-1.76)
BMI≥30	1.36	<0.001	(1.29-1.44)
Dyspnea	1.27	<0.001	(1.17-1.37)
Steroid use	1.32	0.001	(1.12-1.54)
Radiation	1.52	<0.001	(1.40-1.65)
Emergency surgery	1.38	<0.001	(1.17-1.64)
Procedure type
Lung surgery	Reference	Reference	Reference
Esophagectomy	3.91	<0.001	(3.28-4.65)
Hepatectomy	2.28	<0.001	(1.95-2.67)
Colectomy	2.65	<0.001	(2.30-3.06)
Gastrectomy	2.72	<0.001	(2.29-3.23)
Pancreatectomy/Pancreaticoduodenectomy	4.05	<0.001	(3.49-4.69)

---

Back to 2014 Annual Meeting Abstracts