In this feature, a panel of IDSA members identifies and critiques important new infectious diseases studies in the current literature that have a significant impact on the practice of infectious diseases medicine.
No Association Between Antibiotic Prophylaxis for Dental Procedures and Prosthetic Knee or Hip Infection
Reviewed by Christopher J. Graber, MD, MPH
Dental procedures, and the lack of antibiotic prophylaxis prior to them, were not associated with increased risk of prosthetic joint (knee or hip) infections (PJI) in a case-control study published in the Jan. 1 issue of Clinical Infectious Diseases. The study enrolled 339 case patients hospitalized at the Mayo Clinic in Rochester, Minn., with PJI from 2001 to 2006 and 339 control patients with prosthetic knees or hips who were hospitalized on the orthopedic service for non-infectious issues. Researchers collected putative clinical risk factors for infection and dental records up to two years prior to admission, along with data that controlled for the propensity that each particular patient would visit a dentist.
Among dentate patients, 57 percent of PJI cases and 47 percent of controls underwent low-risk dental procedures; 48 percent and 34 percent, respectively, underwent high-risk dental procedures. Lack of antibiotic prophylaxis among procedures performed in the six months (OR 1.1, p=0.77 for low-risk; OR 0.8 p=0.60 for high-risk) or two years (OR 0.6, p=0.11 for low-risk; OR 0.8, p=0.56 for high risk) prior to admission was not associated with PJI. Only 35 patients (10.3 percent) had infection due to organisms of potential oral origin; no association was seen between low-risk or high-risk dental procedures or antibiotic prophylaxis in this subset. Notably, a trend for lower risk of PJI was seen for patients having at least one dental hygiene visit (OR 0.7, p=0.07).
This study does not support the expansive recent statement by the American Academy of Orthopedic Surgeons that “clinicians consider antibiotic prophylaxis for all total joint replacement patients prior to any invasive procedure that may cause bacteremia,” and instead suggests that maintenance of adequate dental hygiene may be of benefit in preventing PJI.
(Berbari, et al. Clin Infect Dis 2010;50:8-16.)
back to top
The Epidemiology of Bacterial Colonization and Infections in Long-Term Care Facilities and ICUs
Reviewed by Rachel Simmons, MD
Two recent point-prevalence studies increase our understanding of the epidemiology of bacterial colonization and infection in two important settings: long-term facilities and intensive care units (ICUs).
In the first study, published in the December 2009 issue of Infection Control and Hospital Epidemiology, researchers obtained bacterial cultures from residents, residents’ rooms, surfaces in common rooms, and health care workers at a Boston long-term care facility. Of 161 residents, 37 (22.8 percent) were colonized with gram-negative bacteria resistant to three or more antibiotics. Approximately 19 percent of the multidrug-resistant (MDR) gram-negative isolates were resistant to meropenem. Several identical MDR gram-negative strains were found in multiple residents on different wards. These bacteria were also isolated from three surfaces (1.7 percent)—two in a common room—and from the hands of one health care worker (7.7 percent). Colonization with MDR gram-negative bacteria was associated with resident stays longer than four years, fecal incontinence, and antibiotic exposure.
The high rate of carbapenem resistance among drug-resistant gram-negative isolates is concerning and underscores the importance of antibiotic stewardship. The presence of MDR gram-negative bacteria in shared spaces and on the hands of a worker, and the identification of the same strain on different wards highlight the importance of infection control measures at these facilities.
The second study, from the Dec. 2 issue of the Journal of the American Medical Association, examined the prevalence of infections in 1,265 ICUs in 75 countries on one day in 2007. Just over half of the 13,796 adult patients were considered by their treating physicians to be infected. Seventy percent of the infected patients had positive cultures. Sixty-two percent of the isolates were gram negative, 47 percent were gram positive, and 19 percent were fungi. Staphylococcus aureus was the most commonly isolated microorganism (20.5 percent of total).
The rate of infection, particularly with S. aureus, Acinetobacter, Pseudomonas, and Candida species, increased with length of stay in the ICU. Both ICU and hospital mortality rates were significantly higher in infected patients (25.3 percent and 33.1 percent respectively) compared to uninfected patients (10.7 percent and 14.8 percent respectively). Infection was independently associated with hospital mortality in a multivariable logistic regression. These results show that globally, infections are highly prevalent in the sickest patients and are associated with significant excess mortality.
(O'Fallon et al. Infect Control Hosp Epidemiol 2009; 30(12): 1172-1179 and Vincent et al. JAMA 2009; 302(21): 2323-2329.)
back to top
Adjunctive Therapy in Pediatric Meningitis: Back to the Drawing BoardReviewed by Christian B. Ramers, MD
Bacterial meningitis is a devastating disease that inflicts substantial morbidity and mortality, especially in children. Hearing impairment is an important sequelae of this condition with long-lasting developmental effects. A study in the January 2010 issue of Pediatrics examines the use of adjunctive therapy—in addition to appropriate antimicrobial therapy—to prevent this outcome.
The authors conducted a randomized, double-blind, placebo-controlled trial in Latin America including 383 children age 2 months to 16 years with bacterial meningitis. In addition to ceftriaxone, children were randomized to receive IV dexamethasone (n = 101), PO glycerol (n = 95), IV dexamethasone + PO glycerol (n = 92), or placebo (n = 95). Hearing impairment at different levels (40dB, 60dB, and 80dB) was assessed and validated by an external audiologist.
Most children had a bacteriologically confirmed diagnosis (72 percent), and of these, Haemophilus influenzae type B was most common (52 percent), followed by S. pneumoniae (25 percent) and N. meningitidis (19 percent). Regardless of the threshold of hearing impairment used, no adjunctive treatment improved hearing outcomes above placebo. The only factors predictive of poorer hearing outcome were young age and low Glasgow coma scale on presentation.
As one of the largest randomized controlled trials of adjunctive therapy for bacterial meningitis ever conducted in children, this study adds considerably to the debate in this area. Although trials in adults have demonstrated a mortality benefit and decreased hearing sequelae with the adjunctive use of dexamethasone, the benefit in children is less clear. A meta-analysis concluded that corticosteroids showed a protective effect on severe hearing loss in children in high-income countries, but no single trial has demonstrated this benefit. The results of the present study suggest that no currently available adjunctive therapy helps prevent hearing impairment from bacterial meningitis. Rather, the most predictive clinical variables are age and mental status on presentation to care.
(Peltola et al. Pediatrics 2010; 125(1):e1-e8.)
back to top
Virological Outcomes Improve As Incidence of HIV-1 Drug Resistance Decreases
Reviewed by Ed Dominguez, MD
The success of highly active antiretroviral therapy (HAART) in reducing morbidity and mortality from HIV-1 infection is well-documented. However, the effects of HAART on the incidence of antiviral drug resistance and on HIV viral load in a population are not well-established. The results of such a study, from researchers in British Columbia, appear in the Jan. 1 issue of Clinical Infectious Diseases.
In this province-wide study, plasma viral loads and genotypic resistance were determined in patients receiving treatment since 1992. Because resistance testing didn’t start routinely in the program until July 1996, annual incidence was reported only from 1996 through 2008. In all, 24,652 resistance tests were performed from 5,422 participants. The primary outcome was resistance to one or more of the three categories of antiretroviral drugs (protease inhibitors, nucleoside/nucleotide reverse-transcriptase inhibitors, and non-nucleoside reverse-transcriptase inhibitors) that developed during the surveillance period.
In 1996, resistance was identified in 571 cases during 39,188 cumulative person-months of exposure per year. By 2008, there were only 71 cases identified during 52,638 cumulative person-months of exposure per year. Expressed as incidence rate, the overall rate decreased about 12-fold, from 1.73 cases per 100 person-months to 0.13 cases per 100 person-months. This represented an exponential decline and was seen in all three drug classes when analyzed individually. Concomitantly, investigators saw an increase in the proportion of people with undetectable viral loads, from 64.7 percent in 2000 to 87.0 percent in 2007 (P<.001). By 2007, the median lowest plasma viral load recorded was below the limit of detection of the assay (i.e., < 50 copies/ml).
Although this observational study cannot confirm a causal relationship between decreased antiviral resistance and decreases in viral load, it does suggest that improved accessibility to HAART may reduce HIV-1 viral load in a population without incurring drug resistance.
(Gill et al. Clin Infect Dis. 2010; 50:98-105.)
back to top
Pre-Operative Skin Preparation With Chlorhexidine-Alcohol Prevents More Surgical Site Infections Than Povidone-Iodine
Reviewed by Sara Cosgrove, MD
Surgical skin preparation with chlorhexidine-alcohol reduces the risk of subsequent surgical site infection by 41 percent compared to povidone-iodine, according to a study published in the Jan. 7 issue of the New England Journal of Medicine. In the study, patients at six university-affiliated hospitals undergoing clean contaminated gastrointestinal, biliary, thoracic, gynecologic, and urologic procedures were randomized to receive chlorhexidine-alcohol or povidone-iodine skin prep.
The primary endpoint of the study was the occurrence of a surgical site infection in the 30 day post-operative period using definitions from the Centers for Disease Control and Prevention (CDC). Site investigators who were blinded to study arm determined when these infections occurred. Four hundred and nine patients who received chlorhexidine-alcohol and 440 who received povidone-iodine were included in the intent-to-treat analysis. Surgical site infections developed in 39 patients (9.5 percent) who received chlorhexidine-alcohol and 71 (16.1 percent) who received povidone-iodine. The difference was seen with superficial and deep incisional infections, but not with organ space infections.
This is the first randomized trial to demonstrate the greater efficacy of chlorhexidine-alcohol compared to the more commonly used povidone-iodine in reducing surgical site infections after clean contaminated surgery. A potential limitation of the study is the apparent lack of assessment of the techniques used to apply the agents to the skin at the different study sites. Historically, there has been a tendency to fail to allow povidone-iodine to dry or to wipe it off, rendering it less effective; it is possible that if it were more fastidiously applied that the difference in efficacy may be smaller. In addition, external validation of surgical site infection cases would have strengthened the conclusions. Finally, the efficacy of newer povidone-iodine plus alcohol skin prep agents relative to chlorhexidine-alcohol remains to be studied.
(Darouiche et al. NEJM 2010; 362:18-26.)
back to top
< Previous Article | Next Article >