Previous Abstract | Next Abstract
Printable Version
A4147
October 24, 2017
43032.635417 - 43032.697917
Room 154
Fluid Administration and Hospital Practice in Major Abdominal Procedures Results from Examination of a Large Administrative Database
Thomas J. Hopkins, M.D., Timothy Miller, M.D., Jennifer Sahatjian, Ph.D., Julie Thacker, M.D., Walter Linde-Zwirble, B.S., Douglas M. Hansell, M.D.
Duke University Medical Center, Durham, North Carolina, United States
Disclosures: T.J. Hopkins: Consulting Fees; Self; Cheetah Medical. T. Miller: Consulting Fees; Self; Cheetah Medical. J. Sahatjian: Salary; Self; Cheetah Medical. Salary; Self Relationship; Employee. J. Thacker: None. W. Linde-Zwirble: None. D.M. Hansell: Salary; Self; Cheetah Medical. Salary; Self Relationship; Employee.
Introduction: Intraoperative fluid therapy may impact perioperative outcome. While some institutions have adopted a goal-directed approach based on dynamic physiologic parameters, unguided intraoperative fluid administration remains the standard of care at many hospitals. The aim of this study was to better understand the impact of unguided Day of Surgery Fluid (DSF) use and associated patient outcomes following abdominal surgery, using a large administrative database.

Method: We examined DSF within the 2013 Premier database. The study population consisted of patients ≥ 18 years who underwent a major abdominal procedure (MAP) during the index hospital admission. Hospitals that reported a <1L median DSF and individual cases with a reported DSF of <1L were excluded. To adjust for potential confounds, a DSF propensity model was constructed including factors that could predispose an individual to receive fluid during surgery; including patient characteristics (age, comorbidity, acute organ dysfunction and diagnosis) and disease management factors (day of surgery, procedure, location, surgical approach and duration). Hospital fluid use groups (FUG) were characterized as Low, Medium and High (mean DSF <3 L, 3.00-4.99 L and 5+ L, respectively). Additionally, risk models were constructed for: hospital mortality, post-op pulmonary complications (PPC), post-op cardiac complications (PCC), any post-op complications (POC) and hospital length of stay (LOS).

Results: The analysis set consisted of 36,252 discharges from 393 hospitals. The mean patient age was 62.9 and 45.5% were male, with no substantial differences in patient characteristics or outcomes between included and excluded hospitals. The mean DOS fluid use was 4.2 L, with 21% receiving 6L or more. While patient case mix and surgical characteristics did not vary greatly between hospitals, there was significant variation in DSF administered between hospitals. The proportion of patients receiving 6+L DSF varied greatly by FUG, where only 5.8% of cases at Low hospitals (22% of hospitals) received 6+L DSF, High hospitals (26% of hospitals) exposed 40.5% of patients to 6+L DSF. The propensity model accounted for slightly more than a quarter of the variation in DSF (R2=27.2%). However, the exclusion of FUG resulted in R2 dropping to 6.1%, indicating that most of the variation was came not from patient or procedural details, but simply from hospital practice. Notably, above DSF of 5 L the actual mortality was significantly greater than expected (p< 0.05 at 6L DSF). Those having 6+ L DSF had an observed hospital mortality rate 48% higher than expected (5.6% vs. 3.6%, p<0.0001, Figure 1). The same pattern was found across all patient, disease and surgery related subgroups; and was found in other outcomes as well including PPC (p<0.0001), and PCC (p<0.001), POC (p<0.001) and LOS (with an addition of 0.65 days per patient, p<0.001).

Conclusions: Study results suggest that DSF of 6+L is associated with harm to patients across a number of outcomes and that this level of DSF is primarily driven by hospital practice and not patient or disease factors. This relationship was observed in the raw outcomes and maintained when patient and surgical details were included in the risk adjustment. Having more hospitals adopt a goal oriented approach to DSF management could have a substantial impact on patient outcomes and perioperative inpatient resource utilization.
Figure 1

Copyright © 2017 American Society of Anesthesiologists