BETHESDA, MD. — Evidence suggests that multidrug-resistant gram-negative organisms contribute to worse patient outcomes if patients are not severely ill, Arjun Srinivasan, M.D., said at an annual conference on antimicrobial resistance sponsored by the National Foundation for Infectious Diseases.
“When you factor severity of illness, extremely severe illness overrides the impact of multidrug resistance on mortality,” Dr. Srinivasan said in an interview.
In some studies of patients with gram-negative infections, multidrug resistance was not associated with significantly higher all-cause mortality. But many outcome studies in this research area have not differentiated between colonized vs. infected patients, or controlled for severity of illness, noted Dr. Srinivasan, a medical epidemiologist in the division of health care quality promotion at the Centers for Disease Control and Prevention in Atlanta.
Dr. Srinivasan and his colleagues conducted an outcome study of 96 patients with multidrug-resistant Acinetobacter in 2004. The researchers used validated definitions to separate colonized and infected patients and used Acute Physiology and Chronic Health Evaluation (APACHE) III scores and the Charleston comorbidity index to assess severity of illness. They compared the infected patients with two different control groups, one group of 90 patients with susceptible Acinetobacter infections and one group of 89 patients without Acinetobacter infections.
The resulting patient characteristics were “exactly what we had expected from the beginning,” Dr. Srinivasan said.
Patients who were infected with multidrug-resistant Acinetobacter were significantly sicker than were patients in the two control groups. They also had higher APACHE III scores, which indicated that they were more severely ill, and more comorbidities compared with both control groups.
In a univariate analysis, the patients with multidrug-resistant Acinetobacter had a significantly higher rate of mortality (27.1%) compared with both the susceptible Acinetobacter group (16.7%) and the no Acinetobacter group (12.4%). In addition, the mean length of hospital stay was significantly higher in the multidrug resistant group (27 days) compared with the other two groups (20 days and 19 days, respectively).
To assess the independent impact of resistance, the researchers used a multivariate analysis to control for the effect of severity of illness. Patients with multidrug-resistant Acinetobacter were twice as likely as were the controls to have a longer-than-average hospital stay and were 2–3 times more likely to have a longer-than-average ICU stay.
But the multivariate model showed no significant difference in mortality among the groups. The findings emphasize that drug resistance cannot overcome the impact of severity on mortality, although resistance was associated with a longer hospital stay.
Assessment of the role of illness severity continues to challenge researchers in outcome studies of drug resistance, Dr. Srinivasan said.
“The nature of the epidemiology of resistant gram-negative pathogens is a dynamic one. It's not steady, and it is going to continue to evolve,” Dr. Srinivasan commented. “The resistant pathogens are becoming a bigger and bigger problem, and our therapeutic options simply are not keeping pace.”
Five risk factors for infection or colonization with gram-negative organisms have surfaced frequently in studies: severity of illness, prolonged mechanical ventilation, prior antimicrobial exposures, prolonged hospital or ICU stay, and exposure to invasive medical devices.
Klebsiella pneumoniae carbapenemases, or KPCs, are perhaps the next big thing in gram-negative resistance. They cleave carbapenems, effectively conferring moderate to high levels of drug resistance to all drugs in the carbapenem class.