Purpose: Multiple studies have reported an unchanged mortality rate from upper gastrointestinal hemorrhage (UGIH) over the past five decades despite new endoscopic and medical therapies. Our aim was to determine the incidence, mortality and resource utilization for UGIH over the past two decades in a nationally representative database.
Methods: We used the Nationwide Inpatient Sample (NIS) to calculate outcomes every 5 years from 1989 to 2009. The NIS is the largest nationally representative inpatient database in the United States. Patients were identified using ICD-9 CM code indicating a primary diagnosis of UGIH. Annual incidence was calculated using the July 1 population census estimate for each year. For the annual case fatality rate, subgroup analysis included variceal and non-variceal UGIH. Procedures were identified using the corresponding ICD9-CM codes. All charges were adjusted for inflation using the consumer price index. Comorbidities were measured using the Charlson index.
Results: The annual incidence of UGIH did not show a clear trend and overall appears stable (see Table). The case fatality rate for total, variceal and non-variceal UGIH steadily decreased from 1989 to 2009. The length of stay also progressively decreased. In contrast, the proportion of inpatient diagnostic and therapeutic endoscopies, the median hospitalization charge and the overall economic burden of UGIH increased over time to 7,636 million U.S. dollars in 2009. Although the median age was almost constant until 2004, it decreased in 2009. The median Charlson comorbidity index was essentially stable over time.
Conclusion: The inpatient mortality rate from upper gastrointestinal hemorrhage (total, variceal and non-variceal) is decreasing over the past two decades, despite an essentially unchanged patient age and comorbidity. There has been an increase in the diagnostic and therapeutic endoscopy rate. The overall direct economic burden of UGIH in 2009 has risen to over 7.6 billion dollars.

Upper gastrointestinal hemorrhage outcome trends between 1989 and 2009

	1989	1994	1999	2004	2009
Incidence (per 100,000 persons)	83	110	90	95	79
Inpatient mortality
Total upper gastrointestinal hemorrhage	4.69%	3.43%	2.93%	2.43%	2.13%
Variceal upper gastrointestinal hemorrhage	10.70%	7.68%	5.31%	5.26%	5.61%
Non-variceal upper gastrointestinal hemorrhage	4.48%	3.36%	2.89%	2.39%	2.07%
Total inpatient endoscopy rate	69%	80%	85%	86%	85%
Percent therapeutic endoscopies among total inpatient endoscopies	2%	18%	22%	23%	27%
Median length of stay (95% C.I.)	4.52(4.43;5.56)	3.64 (3.58;4.69)	2.90 (2.86;2.95)	2.89 (2.85;2.93)	2.85 (2.81;2.90)
Consumer price index adjusted median total charge (95% C.I.)	9,249 (8,987;9,459)	10,641 (10,349;10,940)	11,050 (10,730;11,351)	16,326 (15,799;16,884)	20,370 (19,551;21,221)
Consumer price index adjusted economic burden in million U.S. dollars	3,279	4,988	4,357	7,130	7,636
Median age in years (95% C.I.)	68 (67-69)	69 (68-70)	70 (69-71)	70 ( 69-71)	67 (66, 68)
Charlson comorbidity index	0.68 (0.67-0.70)	0.70 (0.69-0.72)	0.75 (0.73-0.76)	0.83 (0.81-0.85)	0.93 ( 0.91-0.95)

95% C.I. = 95% confidence interval

Citation: . THE INPATIENT MORTALITY RATE FOR UPPER GASTROINTESTINAL HEMORRHAGE IS DECREASING IN THE UNITED STATES: A NATIONWIDE ANALYSIS OVER TWO DECADES. Program No. 2. ACG 2012 Annual Scientific Meeting Abstracts. Las Vegas, NV: American College of Gastroenterology.