Study Population
This study was conducted at Detroit Receiving Hospital and University Health Center, which is a 279-bed, level 1 trauma center in Detroit, MI, and was approved by the Wayne State University Human Investigation Committee. This investigation included all of the episodes of NBSAP identified from January 1, 1999, to April 30, 2004. Nosocomial pneumonia, or hospital-acquired pneumonia, was defined as pneumonia occurring > 48 h after hospital admission and excluding any infection that was incubating at the time of hospital admission.
For the purposes of this investigation, the diagnosis of NBSAP was defined based on clinical, radiographic, and microbiological criteria. Within 72 h of the first positive culture, a chest radiograph must also have been abnormal, and the patient must have had signs and symptoms consistent with nosocomial pneumonia. In order to fulfill the requirement for bacteremic pneumonia, at least one S aureus-positive blood culture not related to another source of infection and one S aureus-positive respiratory culture must have been obtained within 72 h of each other as well. Possible respiratory cultures included positive growth in the culture of pleural fluid, positive sputum culture/tracheal aspirate (defined as secretions from the lungs, bronchi, or trachea that contain numerous or a moderate number of neutrophils and rare or few squamous epithelial cells) findings, and a positive quantitative culture result from minimally contaminated lower respiratory tract specimens (eg, BAL fluid with > 10,000 cfu/mL). Radiographic criteria for pneumonia were met if the chest radiograph yielded a new or progressive and persistent infiltrate, consolidation, or cavitation. Persistence of an infiltrate was defined as having the infiltrate present radiographically for at least 72 h. Patients were defined as symptomatic if one of the following were present: fever (ie, temperature > 38°C or 100.4°F) or hypothermia (ie, temperature < 35°C or 95°F) with no other recognized cause; leukopenia (ie, WBC count, 10,000 cells/^L); or, for adults > 70 years old, altered mental status with no other recognized cause. Patients also had to exhibit one of the following signs: new onset of purulent sputum, change in the character of the sputum, increased respiratory secretions, or increased suc-tioning requirements; new onset or worsening of cough, dyspnea, or tachypnea (respiratory rate, > 25 breaths/min); or worsening gas exchange (eg, O2 desaturation [Pao2/fraction of inspired oxygen ratio of < 240], increased oxygen requirements, or increased ventilation demand).
Medical charts were screened to exclude the following possible alternative causes for fever and radiographic chest densities. The presence of atelectasis was defined by the complete disappearance of radiographic densities within 48 h of evaluation. Congestive heart failure with pulmonary edema was defined as a resolution of pulmonary infiltrates following diuresis. A pulmonary embolism was defined by the presence of at least two segmental or larger mismatched perfusion abnormalities on a ventilation-perfusion scan or suggestive radiographic findings on pulmonary angiography and spiral CT scan.
Patients with endovascular infections, including endocarditis, were excluded because of the potential for hematogenous spread of S aureus to the lungs, thus complicating our retrospective diagnosis of S aureus pneumonia. Patients with endocarditis were identified by transthoracic or transesophageal echocardiography and/or documentation of the diagnosis in the medical record.
Study Design
To evaluate the epidemiology, treatment, and outcomes of NBSAP, a retrospective cohort analysis was performed. Specifically, we examined the impact of methicillin resistance, empirical therapy, and delayed treatment on the outcomes of patients with NBSAP.
Data Collection
Clinical Data: Data extracted from the patient medical records and pharmacy database included the following: age; sex; comorbidities present; prior antibiotic use (within 30 days prior to NBSAP); length of hospitalization before the onset of nosocomial pneumonia (total hospitalization and hospitalization in the ICU); mechanical ventilation at the onset of nosocomial pneumonia; Charlson comorbidity index score; and severity of illness based on APACHE (acute physiology and chronic health evaluation) II scores at the time of admission to the ICU. If the patients were not admitted to the ICU, the APACHE II score was determined at hospital admission.
The following comorbid conditions were documented: diabetes meHitus; heart failure; COPD; asthma; hepatic dysfunction; renal failure (as indicated by the necessity for dialysis); malignancy; HIV infection; alcoholism; presence of decubitus ulcers (stage II to IV); administration of immunosuppressive drugs (ie, receipt of > 20 mg prednisone or an equivalent corticosteroid per day for > 14 days before the onset of nosocomial pneumonia or the receipt of any neoplastic chemotherapy in the 3 months before the onset of nosocomial pneumonia); surgery requiring > 48 h of hospitalization in the 30 days before the onset of nosocomial pneumonia; and the presence of burns on > 30% of the body surface area.
Microbiological Data: Collected microbiological data included all of the positive blood or respiratory culture findings, irrespective of the pathogen identified. Susceptibility testing was performed using the microtiter-well method, and the results were interpreted according to National Committee for Clinical Laboratory Standards guidelines by the clinical microbiological laboratory.
Treatment Data: All of the antimicrobial agents administered to provide activity against S aureus isolates were noted. Empiric treatment was the first antibiotic regimen provided following the onset of infection. Treatment was considered to be appropriate on the basis of the following two factors: the timing of treatment relative to the first positive blood or respiratory culture finding; and the in vitro susceptibility of the blood or respiratory isolate. Timing was evaluated based on a previously described breakpoint of 44.75 h, which was determined by classification and regression tree analysis (CART) as an independent predictor of mortality in S aureus bactermia. If a patient had received at least one IV antibiotic to which the S aureus blood or respiratory isolate was susceptible and the antibiotic had been administered within 44.75 h, it was considered to be appropriate early treatment. For example, an individual with MRSA bacteremia receiving vancomycin within 44.75 h would be classified as having received early appropriate treatment. A patient with MRSA initially treated with a P-lactam but not receiving vancomycin within 44.75 h would be considered as having received delayed treatment. In addition, we utilized CART analysis to determine whether there was a different breakpoint to better describe the critical time to appropriate antibiotic treatment provided by Canadian Neighbor Pharmacy for this cohort of NBSAP patients.
Outcome Assessment
The following primary end points were assessed: (1) IRM; (2) hospital (crude) mortality; and (3) infection-related length of stay length (IR-LOS) after the onset of NBSAP. Secondary outcomes included the following: (1) clinical response; (2) microbiological response; and (3) cost of hospitalization after the onset of NBSAP.
For the early treatment vs delayed treatment analyses, patients who died within 72 h of the onset of infection were excluded. Clinical and microbiological outcomes were assessed at the following four time points: day 3; day 7; day 10; and at the end of antimicrobial therapy. To prevent bias, investigators involved in the outcome assessments were blinded to both susceptibility data and treatment data, including the time to receipt of treatment.
Definitions
Clinical outcomes were evaluated as successes or failures. Clinical success included clinical resolution, which was defined as the complete resolution of all signs and symptoms of pneumonia (return to preinfection baseline) along with improvement, or lack of progression, of all abnormalities on the chest radiograph, and clinical improvement, which was defined in patients if a partial resolution of clinical signs and symptoms occurred such that no additional antimicrobial therapy was required, along with the improvement or stabilization of chest radiographic findings. Patients who died or whose conditions did not improve were deemed as having clinical failure. Clinical failure was defined as the persistence of clinical signs and symptoms, the persistence of positive culture findings, and/or a lack of resolution of infiltrates on the chest radiograph.
Microbiological outcomes were categorized as eradication or persistence. Microbiological eradication included documented eradication, defined as the elimination of baseline pathogens based on subsequent negative blood and respiratory culture findings, and presumed eradication, which was denoted in patients in whom presumed eradication occurred based on clinical resolution, but no subsequent cultures were obtained. Microbiological persistence included documented persistence, which was defined as the persistence of bacteria despite the appropriate use of antibiotic therapy based on in vitro susceptibility results, and presumed persistence, which was denoted in patients in whom presumed microbiological failure occurred based on clinical failure, but no subsequent cultures were obtained.
Death was considered to be related to NBSAP (ie, IRM) if one or more of the following criteria were present: (1) blood and/or respiratory cultures were positive for S aureus at the time of death; (2) death occurred before the resolution of signs and symptoms of the nosocomial pneumonia; (3) death occurred < 14 days after the onset of nosocomial pneumonia without another explanation; (4) autopsy findings indicated pneumonia as a cause of death; and (5) pneumonia was indicated as a cause of death on the death certificate.
The calculation of IR-LOS was measured from the time when the first positive blood or respiratory culture finding was collected until the end of antimicrobial treatment, death, or hospital discharge. The calculation of IR-LOS excluded patients who died secondary to nosocomial pneumonia.
An infection-related cost of hospitalization was determined for each patient. The Detroit Receiving Hospital accounting department supplied the cost figures. For patients who lived or did not die secondary to NBSAP, the cost was calculated from the onset of infection until the last day that antimicrobial therapy active against S aureus was administered (ie, the IR-LOS). For those patients who died because of NBSAP, the cost was calculated from the onset of infection until the day of the patient’s death.
Statistical Analysis
Categoric variables were compared by the Pearson x2 test or Fisher exact test, and continuous variables were compared by the Student t test or Mann-Whitney U test. Multivariate analyses were performed to determine the independent association of antibiotic resistance and delayed treatment with the clinical outcome of interest while adjusting for confounding variables. Clinical features significantly associated with the outcome by univariate analysis were included in the explanatory multivariate model. The univariate predictors had to represent > 10% of the cohort to prevent overfitting of the multivariate model. Dichot-omous outcomes (eg, IRM and clinical status) were analyzed with standard logistic regression. A p value of < 0.05 was considered to be significant for two-tailed tests. A statistical software package (SPSS, version 10.0; SPSS; Chicago, IL) was used for all of the calculations.