Matthew Schreiber

Bacteremia in Staphylococcus aureus pneumonia: Outcomes and epidemiology ☆ ☆☆

PURPOSE Staphylococcus aureus represents a major cause of pneumonia in critically ill patients. Although bacteremia may complicate S aureus pneumonia, the epidemiology of and outcomes associated with bacteremia in this syndrome are poorly described. MATERIALS AND METHODS We retrospectively identified (January 2005-December 2007) all patients admitted to the hospital with S aureus pneumonia necessitating mechanical ventilation. All subjects underwent lower airway and concurrent blood cultures. The prevalence of bacteremia served as a primary end point. We assessed the impact of bacteremia on mortality and length of stay via either logistic regression or a Cox proportional hazard model, respectively. In both models, we controlled for multiple covariates (eg, demographics, severity of illness, comorbidities, and appropriateness of initial antibiotics). We subsequently developed a prediction rule to identify subjects likely to have concurrent bacteremia based on variables assessed at the time of presentation. RESULTS The cohort included 59 patients (mean ± SD age, 58.0 ± 17.4 years; 55.9% male, 59.3% methicillin resistant, 39.0% crude mortality). Bacteremia complicated nearly 20% of cases. The mortality rate in those with bacteremia was 39.1% vs 8.3% in persons without bacteremia (P = .007). Three variables were independently associated with mortality in S aureus pneumonia: age, need for vasopressors, and concurrent bacteremia. Bacteremia independently conferred a 6-fold increase in the risk for death (adjusted odds ratio, 5.96; 95% confidence interval [CI], 1.08-33.10). Bacteremia also correlated with a longer length of stay. The adjusted hazard ratio for remaining hospitalized if bacteremic was 2.65 (95% CI, 1.14-6.18). For the clinical prediction rule for concurrent bacteremia, we assigned points as follows: 2 points if the patient had received prior antibiotic therapy and 1 point each for acute lung injury and for the need for vasopressors. As the total score increased, the prevalence of bacteremia increased (P < .001). As a screening test for bacteremia in S aureus pneumonia, the scoring system had good predictive value. The area under the receiver operating curve measured 0.83 (95% CI, 0.72-0.94). CONCLUSIONS Bacteremia often arises in S aureus pneumonia and is associated with both increased morbidity and mortality. Several simple clinical factors to determine clinical features identify patients with S aureus pneumonia likely to have simultaneous bacteremia.

Challenges and opportunities in the treatment of ventilator-associated pneumonia

ABSTRACT Introduction: Ventilator-associated pneumonia (VAP) is a distinct clinical entity characterized by an onset after 48 hours of the application of mechanical ventilation (MV). Protocols exist to aid in the prevention of VAP, but this infection carries a devastating impact on patient morbidity and potentially mortality. Areas covered: In this review we present key concepts from existing guidelines to aid clinicians. Challenges remain in defining this disease and, most importantly appropriate empiric antimicrobial treatment is the main determinant of outcome. We highlight that the selection of initial antibiotics is critical, as VAP can by caused by a broad array of drug resistant organisms (DROs), the appropriate duration of treatment for VAP is an evolving concept, but may, in part, be guided by biomarkers, and provide focus on diagnostic challenges, initial therapies and treatment strategies for VAP. Both traditional and novel antimicrobials are presented, including developments in the modes of delivery. Expert commentary: The clinical approach to VAP continues to evolve. Recent evidence regarding the changes in microbiology, diagnostics approaches, and treatment strategies for VAP are important for clinicians to remain informed of to provide optimal patient care.

Response to: ‘letter to the editor: “Challenges and opportunities in the treatment of ventilator-associated pneumonia”’

We appreciate the comments by Drs Metersky and Kalil regarding our observations concerning the risk for adverse outcomes with short-course therapy for non-glucose fermenting gramnegative pathogens [1]. We further appreciate the opportunity to respond to their observations so as to expand the argument on this important topic given that Pseudomonasspp. are implicated as the causative pathogen in nearly 20% of cases of ventilator-associated pneumonia (VAP) [2]. Drs Metersky and Kalil suggest that two points support their claim that short-course therapy for these troublesome pathogens is appropriate. First, they argue that time bias confounds the conclusions of clinical trials showing higher rates of VAP recurrence with short-course therapy when the pathogen in a non-glucose fermenter [2]. Second, they claim that the recurrences that are noted in these reports, and aggregated in metaanalysis, are of uncertain clinical significance [3,4]. We respectfully disagree. With respect to time bias, the letter writers misunderstand the concept of disease density. Of course there is more opportunity for recurrence to be observed with a longer period of observation. Nonetheless, these recurrences are discrete events and these recurrences tend to appear late after the discontinuation of antibiotics in the short-course therapy arm. If recurrences occurred soon after antibiotic cessation in the shortcourse population, then we would concur with their point. Moreover, the clinical import of these recurrences may, indeed, be significant. In fact, the abstract of the meta-analysis they cite does indeed question the meaning of the noted recurrences. The authors of this analysis, however, expressly note in their manuscript that ‘persistence may be clinically significant (as) bacterial load in mechanically ventilated patients colonized with P. aeruginosa, but without meeting the clinical criteria for VAP, may predict outcome...with a significantly greater 28-daymortality.’ [3] Furthermore, in the trial examining doripenem vs imipenem, demonstrates that the clinical pulmonary infection scores, which integrate multiple clinical factors (as opposed to microbiologic ones) for the two groups separated after completion of therapy in the 7-day treatment arm suggesting ‘duration and not antibiotics themselves contributed to the difference in outcomes.’ [5] Finally, the failure of either individual studies or a metaanalysis to note a difference in mortality is not reassuring [3,4]. The sample sizes are small and even the largest meta-analysis is clearly underpowered to truly provide confidence about the safety of short-course therapy when P. aeruginosa is the causative organism. Given that the risk for the patient is potentially severe, the burden of proof for short-course therapy in this scenario clearly lies with those advocating for this approach. In other words, ‘the absence of proof is not proof of an absence.’ Until larger, adequately powered studies are conducted, we suggest that short-course therapy is not appropriate in all cases of VAP.

A 36-year-old Haitian man with coma, acute kidney injury, lactic acidosis, and respiratory failure.

A 36-year-old man was found diaphoretic and unresponsive. Two days of headaches and back pain preceded his presentation. The patient had lived in Haiti all his life, working as an agricultural officer, and arrived in the United States 2 weeks earlier. Physical Examination The patient was afebrile but tachycardic (133/min) and tachypneic (40/min). His BP was 119/63 mm Hg, and his arterial oxygen saturation was 93% on 2 L/min of oxygen via nasal cannula. He had no spontaneous eye opening, verbal response, or motor movements, but withdrew from noxious stimuli. His pupils were 3 mm, symmetric, and reactive to light. The results of the remainder of his physical examination were normal. Laboratory Findings Serum chemistry profile revealed the following values: sodium 124 mmol/L, bicarbonate < 5 mmol/L, blood urea nitrogen 89 mg/dL, creatinine 6 mg/dL, glucose 93 mg/dL, and lactic acid 20 mmol/L. His measured serum osmolality was 315 mOsm/kg with an estimated osmolar gap of 29 mOsm/kg. His arterial blood gas study revealed a pH of 6.8, Pco2 of 12, and Po2 of 185 on 2 L/min of oxygen. His aspartate aminotransferase level was 106 U/L, alanine aminotransferase level was 99 U/L, and bilirubin level was 5 mg/dL. His total WBC count was 24,300/μL with 58% bands and 24% segmented neutrophils. The hematocrit was 32.6%, and the platelet count was 22,000/μL. International normalized ratio was 1.6. Partial thromboplastin time was 93 s. Calcium oxalate crystals were present in the urine. HIV serology was negative. He was intubated and started on broad-spectrum antimicrobial therapy. Chest radiograph revealed prominent interstitial markings (Fig 1). A CT scan of the head without contrast was normal. A lumbar puncture was not done, given coagulopathy and thrombocytopenia. Evaluation for ethanol, salicylates, and acetaminophen returned negative. His blood smear is shown in Figure 2. Figure 1. Admission chest radiograph with prominent interstitial markings. Figure 2. Thin blood film showing numerous burr cells, several schistocytes, and intraerythrocytic parasites (arrow). What is the diagnosis? Diagnosis: Severe Plasmodium falciparum malaria