Jim Kling

October 22, 2012

LAS VEGAS, Nevada — The past 2 decades have seen a decrease in the inpatient mortality rate from upper gastrointestinal hemorrhage (UGIH), according to a study presented here at the American College of Gastroenterology (ACG) 2012 Annual Scientific Meeting and Postgraduate Course. The trend was seen despite the absence of changes in patient age and comorbidity. However, the overall economic impact of UGIH increased from $3.279 billion in 1989 to $7.636 billion in 2009.

Despite new endoscopic and medical therapies, previous studies have shown no change in mortality rate from UGIH over the past 5 decades. The researchers used the Nationwide Inpatient Sample (NIS), the largest nationally representative inpatient database in the United States, to determine mortality rates over the past 2 decades, along with associated costs.

"Despite a lot of improvements in all the different techniques available to us to treat upper GI hemorrhaging, no one has so far documented any decrease in mortality," Marwan Abougergi, MD, a gastroenterology fellow at Brigham and Women's Hospital and Harvard Medical School, in Boston, Massachusetts, who presented the research, told Medscape Medical News.

The researchers used NIS to calculate outcomes every 5 years from 1989 to 2009. They identified patients with a primary diagnosis of UGIH using ICD-9-CM codes. The July 1 population census estimate was used to determine annual incidence. The annual fatality rate was broken up into subgroup analyses for variceal and nonvariceal UGIH. Healthcare costs were adjusted for inflation using the consumer price index. They used the Charlson index to measure comorbidities.

The researchers found no change in the incidence of UGIH between 1989 and 2009, but there was a steady decrease in its overall mortality rate (4.69% in 1989 to 2.13% in 2009) for both variceal (10.70% to 5.61%) and nonvariceal (4.48% to 2.07%) UGIH. Hospital stay length also decreased (1989: 4.52 days, 95% confidence interval [CI], 4.43 - 5.56; 2009: 2.85 days, 95% CI, 2.81 - 2.90).

However, there was an increase in the proportion of inpatient diagnostic and therapeutic endoscopies (total inpatient endoscopy rate increased from 69% to 85%, and among total inpatient endoscopies, therapeutic endoscopies increased from 2% to 27%).

The overall economic burden of UGIH increased from $3.279 billion in 1989 to $7.636 billion in 2009. There was little change in the median Charlson comorbidity index.

Database Improves Study Quality

The difference between the current study and previous studies is attributable to the NIS database, according to Dr. Abougergi. "Most of the studies were not nationally representative. Usually they were looking at subgroups of people and [1 group] looked at 1 year, someone else looked at another year, and then they would compare studies that weren't really comparable. The good thing about this database is that it is published every year, and it includes the same [population] each year, so that each year can be compared to the others to get accurate comparisons," he added.

The lowered mortality rate is likely due to improvements in treatments for upper GI bleeding rather than general improvements in hospital infrastructure and patient care, because the mortality rate for upper GI bleeding is declining more quickly than for other conditions.

"We are doing something specific to GI bleeding. We believe this is [attributable to] advances in endoscopies and medical therapies like [the introduction] of protein pump inhibitors," said Dr. Abougergi.

Not everyone was convinced that the reported death rate was accurate. "The mortality is decreasing, the trend is correct," Grigorios Leontiadis, MD, PhD, assistant professor of gastroenterology at McMaster University, in Hamilton, Ontario, Canada, who attended the presentation, told Medscape Medical News.

But previous, smaller studies have shown higher death rates, and many of them were conducted at smaller, more specialized institutions where one would expect death rates to be lower. The current study includes many large hospitals that would likely have less optimal care, according to Dr. Leontiadis.

It is also possible that the study design can influence the observed mortality rate. Studies that utilize ICD-9 codes, as this one does, miss severely ill patients who were originally admitted for some other condition, such as cardiac surgery.

"These patients contribute more to the [overall] mortality," said Dr. Leontiadis. "All studies show that the mortality is decreasing, [but] I don't think we should be complacent, saying that the mortality rates are low in the US. I don't think this is the actual number."

Dr. Abougergi has disclosed no relevant financial relationships. Dr. Leontiadis has received research grants and honoraria from companies that produce proton pump inhibitors.

American College of Gastroenterology (ACG) 2012 Annual Scientific Meeting and Postgraduate Course. Abstract 2. Presented October 22, 2012.