You are leaving Medscape Education
Cancel Continue
Log in to save activities Your saved activities will show here so that you can easily access them whenever you're ready. Log in here CME & Education Log in to keep track of your credits.

Respiratory Infections: Ventilator-Associated Pneumonia and Healthcare-Associated Pneumonia


Respiratory infections represent a growing challenge for pulmonologists and intensivists. Along with changes in the epidemiology of respiratory infections, the microbiology of these infections continues to evolve. Classically, infections were segregated into those that were nosocomial in origin and those that arose outside the hospital. The theory was that this distinction captured the discordance in pathogens between syndromes, such as traditional community-acquired pneumonia (CAP) and hospital-acquired pneumonia (HAP). With the diffusion of healthcare technology and delivery beyond the hospital, along with increasing rates of antibiotic-resistant pathogens in the community, this approach has been modified. At present, the American Thoracic Society and the Infectious Diseases Society of America recognize 3 (not 2) pneumonia respiratory infection syndromes.[1] Traditional HAP remains unchanged, and within this, one finds ventilator-associated pneumonia (VAP). In order to address the impact of the diffusion of healthcare delivery and the increased frequency of resistant pathogens, such as methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa (PA) in infections that present to the hospital, a new concept was created: healthcare-associated pneumonia (HCAP).[1] The purpose of HCAP is to identify a cohort of patients presenting to the hospital who nonetheless face an increased risk for respiratory infection with organisms, such as MRSA and PA. This abstract session at CHEST 2008 dealt with new research into VAP and HCAP.

The initial presentations focused on issues in the diagnosis of VAP. The diagnosis of VAP has always been challenging given the nonspecific signs and symptoms associated with this condition.[1] In addition, chest x-rays can be notoriously difficult to interpret in ventilated patients, particularly those with underlying acute lung injury. Because of this, many efforts have been made to develop both sensitive and specific biomarkers for the diagnosis of VAP. In Europe both C-reactive protein and procalcitonin have been studied and are employed by some.[2] However, their true diagnostic sensitivity and specificity have not been adequately evaluated. Soluble triggering receptor on myeloid cells (s-TREM) represents a novel and promising biomarker for the diagnosis of VAP. s-TREM is upregulated in the presence of infection with both bacteria and fungi. In an early study of s-TREM, Gibot and colleagues[3] reported that measurement of s-TREM on bronchoalveolar lavage (BAL) fluid had excellent sensitivity and specificity compared with BAL cultures.

Baker and colleagues[4] attempted to extend work with s-TREM into trauma patients. Diagnosing VAP in trauma patients may be particularly difficult because trauma itself can create a proinflammatory state mimicking infection, whereas blunt chest trauma can make interpretation of the chest x-ray difficult. Hence, there is a pressing need for a tool to facilitate the rapid and accurate diagnosis of VAP in trauma patients. To explore the role for s-TREM, these investigators compared s-TREM levels in BAL fluid in trauma patients suspected of having VAP. Their cohort included 19 patients, and BAL cultures served as the gold standard. Slightly more than half of patients were diagnosed with VAP. Not surprisingly, neither demographic nor clinical features separated those with VAP from those without VAP. In addition, the clinical pulmonary infections score (CPIS), which some advocate as a diagnostic scoring tool for VAP, was similar between the VAP and non-VAP cohorts.[5] However, s-TREM levels were significantly higher in persons with VAP (25.3 vs 5.7, P < .05). There appeared to be some overlap, though, between persons with VAP and those without VAP. For example, 3 subjects without VAP had mild elevations in their BAL s-TREM levels. Of interest, these subjects were subsequently diagnosed with infections outside the lung. Although these findings are encouraging, the sample size of this study is exceedingly small. In addition, the study authors made no effort to create a receiver operating curve to assess the true screening characteristics of s-TREM. Similarly, given that delay in therapy is a major predictor of poor outcome in VAP, it is important that screening tests for this disease err on the side of being overly sensitive rather than specific. Otherwise, a clinician may inadvertently withhold antibiotics from a patient truly needing them.

A second study exploring s-TREM examined a different type of patient. Investigators from The State University of New York at Buffalo measured s-TREM in subjects with aspiration syndromes.[6] Aspiration pneumonia can represent either a true bacterial infection or, more commonly, a chemical pneumonitis. Aspiration pneumonia occurs when a bacterial process usually arises from dysphagia, whereas aspiration pneumonitis generally results because of depressions in levels of consciousness. Clinically, these syndromes are difficult to distinguish. In turn, physicians tend to use antibiotics liberally in scenarios when faced with aspiration, even though antibiotics have no role in treating chemical aspiration and may in fact promote harm through driving resistance rates. Researchers led by Ali El-Solh, a noted investigator in pneumonia, hypothesized that s-TREM would differentiate the 2 aspiration types.[6] They measured s-TREM in BAL fluid in 71 subjects with aspiration and in 15 controls. They excluded patients who had a nosocomial syndrome, had been on antibiotics previously, or who were immunosuppressed. Most of their patients had either recently had a stroke or had taken a drug overdose. As with the trauma study, approximately half of the subjects had a culture from a BAL consistent with bacterial infection. Clinical characteristics were similar between those with infection, those with no infection, and the control population. Additionally, C-reactive protein levels were also equal between the groups. s-TREM measures in BAL were significantly higher in patients with positive cultures. Levels in the noninfected aspiration group and the controls were similar, as one would expect. With a cut-point of 250 pg/mL, s-TREM had a 66% sensitivity and a 92% specificity. The area under the receiver operating curve was very high (0.87), indicating that s-TREM performed quite well as a screening test for bacterial infection. The study authors concluded that alveolar s-TREM was a potential biomarker for BAL+ infection.[6] Unlike the earlier study, strengths of this analysis included its sample size, methodologic rigor, and appropriate analytic approach. With similar confirmatory studies, s-TREM may emerge as a useful biomarker in the management of suspected bacterial pneumonia whether in aspiration or in ventilated patients.

As noted above, HCAP represents a relatively recent concept in the approach to pneumonia. It remains unclear how valuable HCAP will be at correctly identifying patients infected with highly resistant pathogens or even the outcomes of patients with HCAP. Researchers from Washington, DC, and St. Louis, Missouri, attempted to retrospectively validate the HACP concept.[7] In a group of over 600 persons presenting to the emergency department with pneumonia (culture-positive), they examined how the prevalence of pathogens differed between those with CAP and persons with HCAP. They defined HCAP on the basis of the following: admission from a nursing home, recent hospitalization, chronic dialysis, underlying immunosuppression. Subjects meeting any one of these criteria were categorized as HCAP. Subjects with none of these factors were defined as CAP. Nearly two thirds of patients met the criteria for HCAP, and of the entire cohort, one fourth were infected with MRSA and 20% were infected with PA. Demographic characteristics did not differ between patients with HCAP and CAP. The concept of HCAP was fairly sensitive as a screening tool for identifying persons infected with a resistant pathogen (87%), but the specificity was less than 50%. In a logistic regression model to predict the presence of a resistant pathogen, HCAP itself was not an independent predictor of this type of pathogen. Rather, the 4 independent variables associated with pneumonia due to organisms, such as MRSA and PA, included recent hospitalization, nursing home residence, chronic dialysis, or need for admission to the intensive care unit. In persons with more than 2 of these risk factors, the prevalence of resistant pathogens exceeded 60%. Although limited by its single-center and retrospective nature, the study authors suggested that HCAP -- of and within itself -- may not be a valuable concept, rather clinicians must look at their local data to understand the prevalence of resistance. More importantly, these findings underscore the prevalence of resistant bacteria in pneumonia that presents to the hospital. As such, emergency department-based protocols for antibiotics must be modified to reflect this change in pneumonia epidemiology.

Utilizing this same dataset, these researchers reviewed outcomes in persons with HCAP and CAP as a function of their initial antibiotic therapy.[8] Appropriate antibiotic therapy, defined as administration of a therapy to which the culprit pathogen is in vitro sensitive, is a key determinant of survival in VAP and other infectious syndromes. Pneumonia presenting to the emergency department, however, represents a heterogeneous syndrome, with some patients needing mechanical ventilation (MV) and others being less severely ill. Hence, understanding the interaction between treatment choices and severity of illness becomes important because the effect of poor antibiotic choices may not be the same regardless of severity of illness. In a stratified analysis of predictors of hospital mortality, the investigators reported that the prevalence of resistant pathogens did not differ on the basis of eventual need for MV. Persons with HCAP, as opposed to those with CAP, were less likely to receive appropriate therapy. In the less severely ill subjects not needing MV, suffering from HCAP was independently associated with death (adjusted odds ratio, 4.5; 95% confidence interval [CI], 1.4-14.9). Receipt of inappropriate therapy doubled the risk for death as well. In the group that was suffering from respiratory failure and needing MV, inappropriate therapy also independently increased the probability of eventual mortality. Of note, suffering from HCAP compared with CAP did not correlate with death. This observation suggests that HCAP affects mortality in pneumonia differentially. The major issues are severity of illness and antibiotic management. Hence, taken with the earlier presentation, it appears that the value of HCAP as a concept is limited. If it neither properly identifies persons with high-risk resistant pathogens nor correlates with outcome, what role should it play? At present it appears that HCAP may help to get physicians to think about MRSA and PA in pneumonia presenting to the hospital. However, it likely represents just the beginning of our efforts to grapple with this problem.

Moving from diagnostics and epidemiology, the final abstract reported results for a novel antibiotic in VAP. MRSA accounts for nearly 1 in 6 cases of VAP in the United States, and current treatment options include vancomycin (VAN) and linezolid.[1] Outcomes with VAN historically have been poor, and new guidelines suggest more aggressive VAN dosing, although little clinical evidence supports this recommendation.[1] Telavancin (TLV) is a novel lipoglycopeptide that is highly active in vitro against MRSA. TLV was compared with VAN in 2 RCTs of HAP that included ventilated patients (ie, VAP). In these studies, TLV was given as a 10-mg/kg once-daily dose, whereas VAN was dosed 1 g every 12 hours with dose adjustments per local protocol. In the 2 original trials that included over 1500 subjects, TLV was noninferior to VAN. This abstract presented results from a subgroup analysis of patients with S aureus (SA) VAP.[9] The subgroup included 241 patients (130 TLV, 111 VAN), and more than half were infected with MRSA. Many also had a coinfection with a Gram-negative pathogen. Overall cure rates were equivalent in each arm (51.5% vs 52.3%). In those with MRSA VAP alone (ie, no other coinfection), cure rates with TLV equaled 66.7% vs 42.9% with VAN (P = .04). Although the TLV and VAN cohorts seemed similar in terms of multiple characteristics, polymicrobial infection with a resistant Gram-negative pathogen (eg, PA) was more common in the TLV arm (38% vs 25%, P = .04). Mortality rates at 28 days post enrollment were similar between those randomized to TLV vs VAN among patients with MRSA VAP. Mortality also did not differ in those with VAP purely due to SA (either MRSA or methicillin-sensitive SA). Among the entire 241 subjects, including those with MRSA and potentially polymicrobial infection, 27% of TLV patients died compared with 20% of VAN-treated subjects (P = NS). In a logistic regression, the 2 independent predictors of mortality were age and severity of illness. The adjusted odds ratio for 28-day mortality with TLV measured 1.85 (95% CI, 0.97-3.55). This analysis, however, did not adjust for rates of inappropriate therapy for Gram-negative pathogens because this has not yet been appropriately adjudicated in the clinical trial database. There were numerically more renal-related adverse events in the TLV arm, but the overall frequency of renal adverse events was small, and a blinded adjudication panel concluded that there did not appear to be a relationship between drug exposure and this outcome.

In summary, this session at CHEST 2008 dealt with an area of active interest to practicing pulmonary physicians and intensivists. The epidemiology of respiratory infections is clearly changing, and newer diagnostic modalities are urgently needed. Furthermore, given the growing burden of MRSA, particularly in pneumonia, development of novel agents is imperative.

  • Print