A single-center, retrospective cohort analysis of 639 patients done by Micek et al yielded similar culture differences between CAP and HCAP patients. In this study, criteria for HCAP were defined as hospitalization in the past year, immunosuppression, nursing-home resident, or hemodialysis. The study authors found that a significantly higher percentage of HCAP patients were infected with MRSA (30% vs. 12%), Pseudomonas aeurginosa (25% vs. 4%), and other non-fermenting gram-negative rods (GNR) (10% vs. 2%). HCAP patients again were noted as having significantly fewer infections with S. pneumoniae (10% vs. 40%) and Haemophilus influenza (4% vs. 17%).
In addition to showing a difference in the bacteriology of CAP and HCAP, the Kollef study also evaluated mortality rates, length of stay, and hospital charges. Mortality rates for HCAP (19.8%) were similar to those of hospital-acquired pneumonia (HAP) (18.8%), and both of these were significantly higher than CAP (10%). Length of stay and hospital cost increased across the spectrum, from CAP to HCAP to HAP, with significant differences between each.3
ATS/IDSA Guidelines
In 2005, a joint committee of the ATS and ISDA updated its initial 1996 nosocomial pneumonia guidelines. The guideline update included the new HCAP category.2 The No. 1 goal of these guidelines was to emphasize early and appropriate antibiotics, followed by tailoring of the treatment regimen based upon culture and clinical data. To this end, HCAP risk factors were developed via extrapolation from observational data generated from HAP and VAP patients.5,6,7
The risk factors are summarized in Table 1 (see p. 19).2 Guidelines dictated that the identification of any of these risk factors in pneumonia patients at the time of admission indicates increased risk for infection with an MDR organism. These high-risk patients require placement into the diagnostic category of HCAP.
click for large version
click for large version
Once a patient has been diagnosed with HCAP, the guidelines recommended obtaining lower-respiratory-tract cultures and initiating broad-spectrum antibiotic therapy. Appropriate empiric antibiotic therapy was suggested to be the same as for HAP. This regimen requires coverage with two anti-pseudomonal agents, as well as an agent with activity against MRSA.
The rationale behind initial coverage with two anti-pseudomonal agents stems from the finding that pseudomonas has a high rate of resistance to many antibiotics, and that if two agents are empirically started, chances of appropriate coverage increase from the outset. This is important, as timely administration of appropriate antibiotics has been shown to decrease mortality in infections.8
Additional considerations for empiric antibiotic treatment include sensitivities of local microbiologic data, as well as any recent antibiotic regimens given to the patient. Following this broad primary antibiotic coverage, de-escalation was recommended based on results of lower respiratory cultures and clinical improvement.2
Evolution of Diagnostic Criteria and Empiric Antibiotic Coverage
Since the publication of the 2005 ATS/IDSA guidelines, the aforementioned risk factors for HCAP have been brought into question, as they have yet to be validated by prospective trials. There is a growing concern that these criteria may not be adequately specific and, therefore, might call for too many patients to be treated with a broader spectrum of antibiotic coverage, thereby increasing the likelihood of developing MDR bacteria.
In order to further analyze HCAP criteria, Poch and Ost wrote a review earlier this year examining the data behind each of the risk factors cited in the ATS/IDSA guidelines; they found considerable heterogeneity in magnitude of MDR infection risk for these criteria.9 The authors also reviewed studies looking at other risk factors for MDR infections in patients living in nursing homes or afflicted with CAP. They proposed that such additional factors as patient specific risks (including functional status and previous antibiotic exposure) and contextual risks (including nurse-to-patient ratio) be evaluated and possibly incorporated into criteria.