Prevention of Nosocomial Infection in the Intensive Care Unit

The most common nosocomial infections in medical ICUs are urinary tract infections (31%), followed by pneumonia (27%) and primary bloodstream infection (19%).^[81] VAP is defined as pneumonia developing in mechanically ventilated patients after more than 48 hours of intubation, with no clinical evidence suggesting the presence or likely development of pneumonia at the time of initial intubation.^[82] VAP is a leading cause of morbidity and mortality in mechanically ventilated critically ill patients. In addition to these morbidities, VAP increases hospital length of stay and costs. ^{[83] [84]} Additional hospitalization costs due to an episode of VAP have been demonstrated to be as high as $40,000.^{[85] [86]} The incidence of VAP is estimated at 10% to 25%,^[87] and the estimated mortality rate from VAP is 5% to 27%.^{[84] [88]} The pathogenesis of VAP probably involves microaspiration of oropharyngeal or gastric secretions contaminated with bacteria. Risk factors include exposure to prior antibiotic therapy and the presence of invasive devices. ^[89]

The diagnosis of VAP is challenging. Often, a presumptive diagnosis of pneumonia is made when the patient develops a fever, leukocytosis, purulent secretions, a new infiltrate on chest radiography, and when bacteria are isolated by nonquantitative analysis of endotracheal aspirates.^[90] These nonspecific diagnostic criteria may lead to unnecessary antibiotic use, increased hospital costs, emergence of resistant microorganisms, and a potential delay of diagnosis of the true cause of fever. These problems led several investigators to propose an invasive management strategy for the diagnosis of VAP.

Fagon and colleagues^[91] used fiberoptic bronchoscopy to obtain protected specimen brush samples or bronchoalveolar lavage samples for quantitative cultures. Patients were considered to have VAP if at least 10³ colony-forming units (CFU)/mL of bacteria grew from the protected specimen brush sample or at least 10⁴ CFU/mL of bacteria grew from the bronchoalveolar lavage fluid. The patients in the invasive management group had reduced mortality at day 14 (16.2% versus 25.8%, P = .022) and increased number of antibiotic-free days (5.0 ± 5.1 versus 2.2 ± 3.5 days, P < .001). This study makes a compelling argument for a more definitive diagnosis of VAP before initiating antibiotic therapy. Some studies^{[92] [93]} have shown that blind placement of protected brush catheters or protected bronchoalveolar lavage is just as sensitive as directed bronchoscopy in detecting bacteria in the lungs.

Prevention of Ventilator-Associated Pneumonia

As in all nosocomial infections, prevention is the most efficacious approach. Excessive use of gastric pH-altering medications for stress ulcer prophylaxis increase gastric pH and increase the risk of VAP.^[94] The use of sucralfate, an agent that does not increase gastric pH, may be preferable to H₂ -receptor antagonists or proton pump inhibitors.

2797

Pooling of secretions above the endotracheal tube cuff may increase the volume of bacteria that enter the airways. Removal of these secretions by continuous aspiration in the subglottic region requires the use of a specialized endotracheal tube with a second lumen that permits a suction catheter to exit proximal to the endotracheal tube cuff.

Three randomized, controlled trials have examined the efficacy of aspiration of subglottic secretions. Mahul and associates^[98] studied 145 patients randomly assigned to prevention of aspiration by hourly subglottic secretion drainage or prevention of gastric colonization using sucralfate or antacids. Subglottic secretion drainage was associated with a lower incidence of VAP (13% versus 29.1%), a prolonged time of onset of VAP (16.2 days versus 8.3 days), and a decrease in the colonization rate of subglottic secretions (2.1% versus 33.4%). Valles and colleagues^[99] randomized 153 patients to subglottic aspiration or usual care. The intervention group had a reduced incidence of VAP (19.9 episodes/1000 ventilator days versus 39.6 episodes/1000 ventilator days). They also found that the episodes of VAP occurred later (12 ± 7 versus 5.9 ± 2.1 days, P = .003) in the continuous aspiration group.^[99] The third trial by Kollef and coworkers^[100] randomized 340 patients undergoing cardiac surgery to continuous aspiration of subglottic secretions. VAP was seen in 5% of the interventional group versus 8.2% of the control group (P = .238), but episodes of VAP occurred statistically later among the patients receiving subglottic aspiration (5.6 ± 2.3 days) compared with those who did not (2.9 ± 1.2 days) (P = .006).^[100] Unfortunately, none of the studies was able to show a mortality benefit from continuous aspiration of subglottic secretions. At this time, aspiration of subglottic secretions looks to be promising for the prevention of VAP but cannot be recommended because of the mixed results from randomized studies.^[101] Table 74-4 lists strategies for the prevention of VAP.

Prevention of Central Venous Catheter Infection

It is estimated that more than 5 million central venous catheters are inserted every year in the United States.^[102] Unfortunately, more than 15% of patients have complications from these lines.^[103] Infectious complications are reported to occur in 5% to 26% of patients with central venous catheters.^{[104] [105]} The attributable mortality rate for catheter-related bloodstream infections ranges from 12% to 25%,^{[106] [107]} and the attributable cost is estimated to be between $3700 and $29,000^{[107] [108]} per episode.

Antibiotic-Coated Central Venous Catheters

Several studies have been conducted to examine the efficacy of antimicrobial-impregnated catheters in an effort to decrease the incidence of catheter-related bloodstream infections. Central venous catheters coated with the

TABLE 74-4 -- Evidence-based prevention of ventilator-associated pneumonia

Intervention	References
Semirecumbent positioning	[97] [205] [206]
Sucralfate	[94] [95] [207] [208] [209] [210]
Aspiration of subglottic secretions	[98] [99] [100]

antimicrobials chlorhexidine and silver sulfadiazine or impregnated with the antibiotics minocycline and rifampin are commercially available. Chlorhexidine/sulfadiazine-coated catheters lowered the rate of bloodstream infections from 7.6 to 1.6 infections per 1000 catheter days (P = .03).^[109] A cost-effectiveness analysis conducted by Veenstra and associates^[110] concluded that these catheters would decrease costs by $196 per catheter inserted if the institutional catheter-related bloodstream infection rate was higher than 2%.

Later evidence suggests that minocycline/rifampin-impregnated catheters are even more efficacious. In a randomized trial performed in 12 university hospitals comparing minocycline/rifampin catheters with chlorhexidine/silver sulfadiazine catheters, the minocycline/rifampin catheters were one third as likely to be colonized as catheters impregnated with chlorhexidine or silver sulfadiazine (7.9% versus 22.8%, P < .001) and one half as likely to lead to a catheter-related bloodstream infection (0.3% versus 3.4%, P < .002).^[111]

The use of antimicrobial-impregnated catheters should be considered in all critically ill patients who require long-term (at least 3 days) indwelling central venous access. Immunocompetent patients who will be catheterized for less than 3 or 4 days may not need these catheters, because infection is rare during this period. Additional studies are required to further elucidate the benefit of minocycline/rifampin-over chlorhexidine/silver sulfadiazine-impregnated catheters.^[112] Although theoretically possible, no studies have shown the development of significant antibiotic resistance with these catheters. However, continued surveillance for resistance is recommended as part of the increased use of these devices.^{[105] [111]}

Central Venous Catheter Insertion Techniques

Optimal sterile technique is a cost-effective way to reduce the incidence of central line colonization and catheter-related bloodstream infections. Use of maximal sterile-barrier precautions such as mask, cap, sterile gown, sterile gloves, and a large sterile drape as opposed to small towels has been shown to reduce the rate of catheter-related bloodstream infections and to save $167 per catheter inserted.^{[113] [114]} The use of chlorhexidine solution for skin decontamination reduces the risk of catheter colonization and therefore is preferred over the use of povidoneiodine solution. ^[115]

With the development of portable, inexpensive, ultrasound machines, use of real-time ultrasound guidance has been promoted as a technique for reducing the complications with central venous cannulation. Ultrasound guidance has been shown to reduce the number of

2798

The risk of infection with catheterization stays relatively low until about the fifth to seventh days and then increases exponentially.^{[119] [120]} Several trials have studied scheduled replacement of catheters over a guidewire or in a new site, but none of these strategies was able to show an improved outcome.^{[121] [122]} Optimal catheter insertion technique and maintenance can minimize infectious risks, but the most efficacious maneuver is to remove central venous catheters as soon as they are no longer required. The need for central venous catheter placement should be reassessed daily, and unnecessary catheters should be removed because the probability of catheter-related infections increases with time. The need for central venous access should be reassessed again before discharging the patient from the ICU, because this is often an opportune time to remove any indwelling lines, including bladder catheters. The evidence-based strategies for decreasing central line infections are listed in Table 74-5 .