BOSTON – Initial antibiotic therapy failed in nearly 25% of 3,535 adults with acute bacterial skin and skin structure infections in U.S. hospitals between 2000 and 2009, according to a retrospective analysis of data from multiple hospitals.
Initial treatment failure was associated with significantly higher mortality, longer hospital stays, and higher costs of care, Xing-Yue Huang, Ph.D., reported at the annual meeting of the Infectious Diseases Society of America. The case-fatality rate was 1.3% in the 532 patients in the early treatment failure subgroup and 0.2% in the 3,003 successful treatment patients, a significant difference. The mean length of stay and total hospital charges were also significantly increased, at 8.1 days vs. 4.6 days and at $23,383 vs. $12,393, respectively.
The costs and benefits of initial treatment with broad-spectrum and narrower-spectrum regimens needs to be evaluated, Dr. Huang said. "Broad spectrum regimens are generally most effective but their benefits have to be measured against their higher cost. Narrower-spectrum regimens are less expensive but they may pose a higher risk of treatment failure."
Dr. Huang and colleagues used a database to identify all adult patients hospitalized for acute bacterial skin and skin structure infections (ABSSSI), such as abscess, cellulitis, and surgical site infection, between Jan. 1, 2000 and June 30, 2009. All patients in the analysis had at least one positive isolate from skin, wound, or blood cultures within 24 hours of initial clinical presentation. All received parenteral antibiotics for at least 48 hours beginning within 24 hours of hospital admission.
"More than half of the patients received initial regimens – usually multi-drug – that provided coverage for MRSA."
All antibiotics received within 24 hours of admission were considered initial therapy, said Dr. Huang of Forest Research Institute in Jersey City, N.J. "We limited our attention to patients who received any of the 40 regimens most commonly used in calendar year 2009," he explained, noting that, for all of the ABSSSI patients identified, 79% received 1 of the 40 regimens.
Patients were excluded from the study sample if they had a secondary diagnosis of infections of other body sites or organs, except septicemia/bacteremia and systemic inflammatory response syndrome; necrotizing fasciitis; gangrene; ecthyma gangrenosum; osteomyelitis; complications of pregnancy, childbirth, and the puerperium; or impetigo, or if they had plasmapheresis or hemoperfusion performed, he said.
Per study protocol, initial treatment was considered a failure if a patient received a new parenteral antibiotic more than 24 hours following hospital admission, excluding substitution of a similar or narrower spectrum regimen or if a patient underwent drainage, debridement, or amputation more than 72 hours after admission, said Dr. Huang.
The researchers used multivariate logistic regression analysis to predict failure of initial therapy and for significance testing for differences in mortality. They used covariance model analysis to compare hospital length of stay and total inpatient charges. All models were adjusted for covariates, including patient age, sex, comorbidities, clinical status at admission, and hospital characteristics, Dr. Huang stated. All analyses were repeated with initial therapy failure limited to the 72-hour post-admission window to characterize "early" treatment failure, because later failure could be the result of nosocomial infections in patients with long hospitalizations, which would skew cause and effect, he said.
Of the 3,535 patients (mean age 52 years) included in the study sample, 797 (22.5%) experienced initial treatment failure, 66.8% of which occurred within 72 hours of admission, Dr. Huang reported. The most commonly identified pathogens were Staphylococcus aureus (of which 68.4% were methicillin resistant), Pseudomonas aeruginosa, Streptococcus agalactiae, and Escherichia coli. The most frequently prescribed initial antibiotic regimens were vancomycin (22%), cefazolin (14%), and ampicillin/sulbactam (12%). "More than half of the patients [56%] received initial regimens – usually multi-drug – that provided coverage for MRSA [methicillin-resistant Staphylococcus aureus]," he said.
Significant predictors of initial treatment failure were older age, peripheral vascular disease, septicemia/bacteremia, leukopenia, uremia, and hospitalization within the previous 30 days, Dr. Huang reported. Regarding the clinical and economic consequences, initial antibiotic failure was associated with a sevenfold higher rate of death, 5.1 additional hospital days, and additional hospital charges of $13,731, he said.
Dr. Huang acknowledged some of the study’s limitations, including its retrospective design, the inclusion of only those patients receiving the 40 most frequently used regimens, the exclusion of patients with ABSSSI of unknown etiology, and the fact that initial treatment failure could reflect multiple factors, such as the presence of causative organisms that are resistant to the antibiotics that were chosen, the severity of the infection, compromised host immunity, or adverse events.