Scope and magnitude

The threat that antibiotic resistance poses to human health seems even more dire when considered in the context of nosocomial infections. Ever since the 1840s, when Ignaz Semmelweis first demonstrated that hand hygiene by health-care workers could reduce rates of puerperal fever, there has been an awareness of both the importance of infection control and the impact of lapses in its application. Five percent of hospital patients, and 5-35 percent of patients in intensive care units (ICUs), acquire an infection during their stay (CDC, 1992; Roberts et al, 2003).

The CDC estimate that greater than 2 million people contract nosocomial infections annually in the US (CDC, 1992). Similarly striking statistics have been cited in Europe. As part of the European Prevalence of Infection in Intensive Care (EPIC) study, researchers conducted a one-day point-prevalence study of ICU-acquired infections, drawing case-reports from over 1400 ICUs from 17 countries. They discovered that 21 percent of the ICU patients had acquired infections during the course of their ICU stay; pneumonia, urinary tract infection (UTI), and bloodstream infection (BSI) were most commonly reported (Vincent et al., 1995). Although the crude rates of nosocomial infection have remained stable at 5-6 infections per 100 admissions between 1975 and 1995, progressively shorter average hospital stays over that time translate into higher incidence-densities of health-care associated infections (HAIs), increasing from 7 to 10 HAIs per 1000 patient-days in 1975 versus1995 (Weinstein, 1998).

In 1970, concern regarding HAIs prompted the CDC to establish the National Nosocomial Infection Surveillance (NNIS) System (now a part of the National Healthcare Safety Network - NHSN), a collection of nearly 300 hospitals that provided monthly data related to HAI rates. Most recently, the NNIS System reported average device-day incidence-densities for HAIs in US ICUs between January 2002 and June 2004. They documented 4.9 UTIs per 1000 urinary catheter days, 4.9 BSIs per 1000 central-line days, and 7.5 cases of ventilator-associated pneumonia (VAP) per 1000 ventilator days (NNIS System, 2004).

Many groups have demonstrated the clinical and economic impact of nosocomial infections. In the EPIC study, researchers showed that hospital-acquired BSIs, pneumonias, and clinical sepsis each acted independently to increase mortality (Vincent et al, 1995). Others have estimated that HAIs confer rates of attributable mortality ranging from 4 percent for UTIs with secondary bacter-emia to 47 percent for primary BSIs (Eggimann and Pittet, 2001). In addition to increasing mortality, HAIs increase cost. In a retrospective cohort study involving 139 non-infected patients and 25 patients with HAIs, researchers from Chicago's Cook County Hospital found that, after controlling for the confounding factors of severity of illness and ICU stay, an HAI diagnosis added over $15,000 in excess costs per associated hospital stay (Roberts et al, 2003). Using this cost estimate and the annual approximated incidence of 2 million cases, HAIs may cost $30 billion in the US each year.

Due to the selection pressure created by frequent antibiotic use in hospital environments, antibiotic-resistant organisms frequently cause nosocomial infections. From 1988 to 1994, the percent of hospitalized patients who received antibiotics increased from 32 percent to 53 percent (Pestotnik et al., 1996). The most recent NNIS System update reflects the consequences of increased use. NNIS reported antibiotic resistance rates for HAIs in ICU patients for 2003 and compared these with resistance rates for the previous five years, showing, for example, increasing methicillin resistance in staphylococcus and increasing resistance to third-generation cephalosporins in Klebsiella (NNIS System, 2004; Figure 9.12). The Surveillance and Control of Pathogens of Epidemiological Importance (SCOPE) study reported antibacterial resistance rates from over 24,000 nosocomial BSIs from 49 US hospitals from 1995 through 2002. They documented an increase in rates of MRSA, from 22 percent in 1995 to 51 percent in 2002. Ceftazidime resistance for Pseudomonas increased from 12 percent in 1995 to 29 percent in 2003 (Wisplinghoff et al, 2004).

Increases in fluoroquinolone use may be particularly problematic. Looking at over 35,000 Gram-negative ICU isolates collected between 1994 and 2000, Neuhauser and colleagues correlated the 10 percent decline in Pseudomonas susceptibility to ciprofloxacin (86 percent in 1994 to 76 percent in 2000) with increased national fluoroquinolone use (Neuhauser et al., 2003). More recent data have shown that rates of Clostridium difficile-associated diarrhea (CDAD) have increased 26 percent in the US between 2000 and 2001. Characterization of the latest outbreak strains of C. difficile revealed a predominant clone which produces higher quantities of cytotoxins A and B, secretes an additional "binary" toxin, may cause more severe colitis, and demonstrates markedly higher rates of

Vancomycin/enterococci

Methicillin/S.aureus

Methicillin/CNS

2003 increase in No. of resistance

3rd Ceph/K.pneumoniae**

Imipenem/P aeruginosa Quinolone/P aeruginosa 3rd Ceph/P aeruginosa

2048 12%

4100 3336 1355 1068

IZZl*31.9%

3rd Ceph/Enterobacter spp.

% Resistance * January through December 2003 □ 1998 through 2002 (+/- standard deviation)'

1825 2119 1411

Figure 9.12 Selected antimicrobial-resistant pathogens associated with nosocomial infections in ICU patients, comparison of resistance rates from January through December 2003 with 1998 through 2002. Percent increase in resistance rate of current year (2003) compared with mean rate of resistance over previous five years (1998-2002). From NNIS System (2004).

quinolone resistance than do historic C. difficile strains (McDonald et al, 2005). Multivariable regression analysis from a recent case-control study of CDAD from Maryland's Veteran's Affairs medical system showed that previous fluoroquinolone use conferred the greatest risk of developing CDAD compared with use of other antibiotics (odds ratio 12.7 for quinolones, 2.2 for clindamycin, 0.4 for cephalosporins; McClusker et al., 2003).

0 0

Post a comment