Daniel M. Keller, PhD

June 13, 2013

PHILADELPHIA, Pennsylvania — A large study spanning 2 decades and 3 countries shows that the location of an infection predicts mortality in patients with septic shock. The study has implications for the design of future clinical trials in this area.

"Indeed, there is a wide spread in hospital mortality, from about 21% to about 80%, with sources such as a urinary infection with urinary tract obstruction on the low end and various abdominal sources on the high end," Peter Dodek, MD, from the University of British Columbia in Vancouver, Canada, told Medscape Medical News.

Also at the high end were what the researchers called disseminated infections, such as bacteremia without a clear-cut source.

This retrospective cohort study, led by Aleksandra Leligdowicz, MD, from the Center for Health Evaluation and Outcome Sciences at the University of British Columbia, involved 7974 patients with 20 anatomic sites of infection hospitalized with septic shock in Canada, Saudi Arabia, and the United States.

The most prevalent predisposing comorbidities were hypertension, congestive heart failure, and coronary artery disease (32.9%), and diabetes (26.1%).

The highest crude mortality rates (76% to 85%) occurred with disseminated infections, ischemic bowel, and spontaneous bacterial peritonitis. The lowest crude mortality rates (21% to 34%) were associated with hydronephrosis, enterocolitis, diverticulitis, and pyelonephritis.

For the cohort, the crude hospital mortality rate was 52.4%. The most prevalent sites of infection were the lung (40.1%), abdomen (31.3%), and genitourinary system (10.6%).

On regression modeling, after adjustment for predisposing factors such as age, sex, and geographic source (from services within the hospital or elsewhere), there was still a wide spread in the odds ratios for death according to anatomic site of infection.

When investigators included downstream factors, such as severity of illness, Acute Physiology and Chronic Health Evaluation (APACHE) II score, and the use of adjuncts like steroids and activated protein C, "the downstream factors still did not completely adjust away the effects of the anatomic site," Dr. Dodek reported.

Each of these regression models normalized mortality risk to that for lung infection. To better consider the absolute mortality associated with each site of infection, the researchers did a predicted standardized mortality. "We still found a range from about 20% to 80%, even after adjustment for predisposing and then predisposing plus downstream factors," he explained.

Essentially, predisposing and downstream factors had little influence and were largely overshadowed by the site of infection as a hospital mortality risk.

When asked by Medscape Medical News whether there was any interaction or effect modification by the type of organism causing the infection at each site, Dr. Dodek said that the researchers did not do that analysis.

"That is an interesting question because there are some data about organisms tracking with certain sites, certain comorbidities, and even with race," said Greg Martin, MD, from Grady Memorial Hospital in Atlanta, Georgia, who was not involved in the study.

He said the findings from this study might help refine the study population for clinical trials looking at the treatment of septic shock. "If you're trying to study a population where the mortality is particularly high," he noted "then studying people who have [central nervous system] infection, bacteremia, disseminated infection, or ischemic bowel would make the most sense, because then you have a mortality far in excess of 50%."

Such a study might have more power to show an effect in that population. However, Dr. Martin cautioned that it could limit the generalizability of the treatment to broader populations with septic shock.

There was no commercial funding for this study. Dr. Dodek and Dr. Martin have disclosed no relevant financial relationships.

American Thoracic Society (ATS) 2013 International Conference: Poster J68. Presented May 20, 2013.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.