Hospital-wide nosocomial pneumonia has become more common than surgical wound infection, and is numerically second only to urinary tract infection. The majority of nosocomial pneumonias develop in ICU settings. The overall mortality approximates 35 per cent and may be as high as 50 per cent in tertiary care centers. The mortality of nosocomial pneumonia is related to the type of organism and the concomitant presence of bacteremia (the overall incidence is about 10 per cent, but mortality increases threefold in cases complicated by bacteremia). Mortality from pneumonia caused by Gram-negative organisms (most commonly Pseudomonas,
Enterobacter, or Klebsiella) may be at least twice as high as that from pneumonia caused by Gram-positive organisms (usually Staph. aureus). Mortality is particularly common in Pseudomonas pneumonia, with as many as 70 per cent of affected patients dying.
Critically ill patients are at particular risk of the development of nosocomial pneumonia ( Table...!). Both short-term endotracheal intubation and long-term mechanical ventilation increase the risk because protective reflexes are lost. Since the nasopharyngeal 'filter' is bypassed, tracheobronchial mucociliary clearance of particulates becomes essential, but it is suppressed by the presence of the tube. Other risk factors include increasing age and immunosuppression. Pre-existing chronic pulmonary disease may pose a risk independent of prolonged ventilator dependency. Although endogenous flora and lapses in disinfection protocols as caregivers move from patient to patient can explain the higher incidence of pneumonia due to enteric Gram-negative bacilli, other mechanisms may be responsible for the high incidence of pneumonia from Pseudomonas, which is not normally present in the human gastrointestinal tract.
K'SWitJl psttfl mjrttf ng fWfl^dmKfarol trtijltfi Loii Of (WWCL-v* Sftiiy tf-JWt COTTO
Table 1 Intensive care interventions associated with nosocomial pneumonia
The colonization of airway surfaces with Gram-negative bacilli is central to the pathogenesis of nosocomial pneumonia. Although nosocomial pneumonia may follow bacteremia, the majority of these pneumonias appear to be due to aspiration of pathogens that have colonized the upper airway. The oropharyngeal flora of hospital in-patients changes rapidly to reflect a predominance of Gram-negative organisms, even if the patient does not receive parenteral antibiotics or spend much time in the ICU.
Hospital admission alone leads to a 23 to 45 per cent incidence of bacterial colonization of the pharynx with aerobic Gram-negative bacilli (compared with an isolation rate of less than 2 per cent from healthy individuals), and 45 to 60 per cent of patients admitted to the ICU are colonized within 1 week of admission. Nosocomial pneumonia subsequently develops in about 25 per cent of colonized patients compared with only 3 per cent of patients whose oropharyngeal flora remains normal.
The diagnosis of nosocomial pneumonia poses both clinical and microbiological challenges, particularly in the intubated patient. The classical findings of fever, purulent sputum, leukocytosis, and a new pulmonary infiltrate actually represent infection in only about 45 per cent of bacterial pneumonias in ventilator-dependent patients. Clearly, better methods of detecting nosocomial pneumonia are needed. This issue is fundamental because it can be very difficult to distinguish colonization from infection, or bronchitis from pneumonia, in patients who are infected. The analysis of tracheal sputum aspirates collected via a conventional suction catheter is notoriously unreliable. Such specimens are often contaminated with upper airway flora and may not be representative of flora deep within the lung. Numerous invasive techniques have been described to circumvent these shortcomings. The advent of fiber-optic bronchoscopy has led to the development of additional techniques. The simple aspiration of sputum after introduction of the bronchoscope via the endotracheal tube or pharynx harbors the same potential for the confounding identification of colonizing organisms. The protected specimen brush circumvents many of these difficulties, with sensitivity rates for the diagnosis of pneumonia reported to be between 70 and 90 per cent in patient studies. Newer directional catheters may obviate the need for bronchoscopic introduction of the brush. Accuracy has been enhanced by using quantitative bacteriology to distinguish colonization from invasive infection. Finding at least 10 3 colony-forming units/ml by quantitative microbiology is a reasonable indicator of invasive infection by protected specimen brush. However, the very small sample size (estimated to be only 0.01 ml) is a potential shortcoming.
Bronchoalveolar lavage has increased the overall accuracy of bronchoscopic diagnostic techniques, particularly when combined with quantitative cultures. Instillation of 10 ml of saline into a pulmonary segment distal to a 'wedged' bronchoscope may sample as many as 106 alveoli. Such a comparatively large sample appears to be responsible for enhanced overall accuracy. The technique has been reported to be 88 per cent sensitive for the diagnosis of bacterial infection of the lower respiratory tract when bacterial counts exceeded 104 colony-forming units/ml, and no more than 1 per cent squamous epithelial cells were present in a differential cell count of the fluid.
Was this article helpful?