Go backSearch Session number: 1132

Abstract title: Hospital Care for Low-Risk Patients with Acute Upper GI Bleed: A Comparison of Practice Settings

Author(s):
IM Gralnek - VA Greater Los Angeles Healthcare System
GS Dulai - VA Greater Los Angeles Healthcare System
TT Oei - University of Southern California
AM Kilbourne - VA Pittsburgh Healthcare System
D Chang - UCLA-Harbor Medical Center

Objectives: The proportion of patients admitted to the hospital with acute upper gastrointestinal hemorrhage (UGIH) at "low-risk" for adverse outcomes appears to be substantial. The process of care for this low-risk population likely varies across practice settings, but has not been extensively studied. The aims of this study were to evaluate and compare: (1) incidence of Rockall "low-risk" admissions to a community hospital (Santa Monica Hospital, SMH) and neighboring tertiary care university hospital (UCLA Center for the Health Sciences, CHS) during calendar years 1997 and 1998, (2) incidence of adverse outcomes in these low-risk cohorts, and (3) level of healthcare resource utilization as a measure of the process of care.

Methods: This is a historical cohort study using existing medical record data. The UCLA Offices for the Protection of Research Subjects approved this study. Potential cases were identified via electronic search of an administrative database containing data on all consecutive adult patients (>18 years of age) admitted to SMH and CHS during calendar years 1997 and 1998. These years were chosen as they followed the publication and dissemination of several papers demonstrating the safety of early discharge for low-risk UGIH patients. ICD-9-CM codes for primary discharge diagnosis were used to identify potential cases. Medical record data for all potential cases were abstracted by two investigators (TO and DC) using a standardized data collection form. Data abstractors were blinded with respect to the study's purpose and Rockall Risk score calculation. Data were collected from three distinct time periods associated with the bleeding episode: the peri-admission period (within 24 hours of the documented time of hospital admission), the hospital course, and the 30-day period immediately following hospital discharge. The Rockall risk score is a simple, validated predictive index that may serve as a useful clinical decision tool for assessing the risk of subsequent adverse outcomes in patients with acute UGIH. "Low-risk" cases were defined by Rockall Risk scores of <2.

Results: The low-risk cohorts consisted of 49/187 (26%) SMH cases and 53/175 (30%) CHS cases (p=0.40). Rebleeding was uncommon 3/49 (6%) at SMH and 2/53 (4%) at CHS (p=0.64). No deaths occurred. 35/49 (71%) at SMH vs. 26/53 (49%) at CHS were admitted to an ICU/monitored bed (p=0.04). 45/49 (92%) at SMH vs. 30/53 (57%) at CHS were prescribed IV H2 blockers for the acute bleeding event (p<0.001). Low-risk patients had a mean hospital length of stay of 3.3 + 2.4 days at SMH vs. 2.6 + 2.1 days at CHS (p=0.15).

Conclusions: The proportion of acute, low-risk, non-variceal, upper GI hemorrhage admissions to neighboring community and tertiary care medical centers was high (26% and 30% respectively), yet adverse outcomes, rebleeding and death, were minimal or none. Utilization of healthcare resources appears greater in the community hospital. This observed variation in the process of care for populations with similar disease severity and outcomes suggests an opportunity for evidence-based interventions aimed at improving efficiency of care.

Impact statement: VA has placed as top priorities the quality of health care delivered to its veterans and patients' satisfaction with the care they receive. VA has also emphasized delivering patient-centered quality of care as part of its total quality improvement focus. However, there are no specific or standardized VA policies, clinical guidelines or "best practices" existing for the management of VA patients with acute UGIH. Studies within VA are urgently needed to further assess this patient population within VA and to establish an evidence-based "best practices" VA model for the management of "low-risk" patients with acute UGIH.