Published in the July 2006 issue of Today’s Hospitalist
When it comes to the medical literature, 2005 was an interesting year for hospitalists. From ominous data about the growing threat of C. difficile to new insights into the relationship between antiplatelet therapy and GI bleeding, the year in research offered some salient lessons for hospitalists.
At the American College of Physicians’ annual meeting, which was held in Philadelphia earlier this year, two hospitalists reviewed the most important journal articles for hospitalists published in 2005.
To choose the articles that made the final cut, the physicians looked for research that would impact the practice of hospitalists, both because of their clinical significance and their relationship to quality improvement and patient safety.
“We surveyed opinion leaders from across the country,” explained Preetha Basaviah, MD, clinical associate professor of medicine at Stanford University School of Medicine and one of the two presenters. “Final studies were selected based upon evi¬dence-based medicine and clinical significance.”
Here’s a review of six of the studies identified during the presentation by Dr. Basaviah and Shaun Frost, MD, a hospitalist and regional medical director for Cogent Healthcare Inc.
A more virulent “and more prevalent “form of C. difficile
New data indicate that an old problem “Clostridium difficile “may be poised to play a bigger role as the infection becomes more virulent, more prevalent and more resistant to therapy.
C. difficile colonizes the colon in only 3 percent of healthy individuals, Dr. Basaviah said, compared with 20 percent to 40 percent of hospitalized individuals. The rates and severity of the disease are on the rise both in the hospital and the community, she added, suggesting that physicians may be facing a more virulent and resistant strain of the infection.
An article in the Dec. 8, 2005, issue of the New England Journal of Medicine makes exactly that point. In that study, McDonald and colleagues identified 187 isolates from patients at eight facilities in the United States with C. difficile-associated diarrhea (CDAD) outbreaks between 2000 and 2003.
They found that a new strain of the infection, BINAP1, showed a 16 to 23 times increase in the production of toxins. The new strain resulted in higher rates of toxic megacolon, the need for colectomy, leukemoid reaction, shock and death.
The new strain of C. difficile produces a range of nasty effects, but researchers are also concerned about the role of certain antibiotics in inducing this resistance strain. While historic strains of C. difficile demonstrate little if any resistance to fluoroquinolones, the McDonald study found that the new strain was 100 percent resistant to gatifloxacin and moxifloxacin, and about 80 percent resistant to clindamycin.
Prevalence of the new strain is also a concern. While older strains of C. difficile accounted for only 17 percent of infections before 2001, researchers found that a new strain accounted for 50 percent of the infections since that time.
In another article in the Dec. 8, 2005, New England Journal of Medicine, researchers found that both the prevalence of “and the mortality from “C. difficile infections is a significant and growing threat.
In a prospective study of 12 Quebec hospitals in 2004, Loo and colleagues found that incidence and mortality rates of C. difficile were strikingly high. The overall incidence of CDAD was 22.5 per 1,000 admissions. That’s a four-fold increase compared with a 1997 study of 18 Canadian institutions.
In addition, the 30-day attributable mortality rate of these infections was 6.9 percent, higher than the 0.5-5.5 percent rate reported in the literature. For older patients, the incidence jumped as high as 74 per 1000, and the 30-day attributable mortality climbed as high as 14 percent..
“Clearly, a more virulent strain of C. difficile has emerged,” Dr. Basaviah said. “We need more surveillance and more studies on the prevalence, virulence and outcomes of C. difficile-associated diarrhea.”
She added that until those studies are done, clinicians in the hospital need to increase their vigilance for C. difficile. “Greater antibiotic stewardship and strict infection control measures are also needed,” Dr. Basaviah explained. “Health care workers should be aware that the alcohol-based waterless hand agents do not kill C. difficile spores.”
Diagnosing IVD-related bloodstream infections
Catheter-related bloodstream infections are associated with a recognized morbidity and mortality, but determining which catheters are the source of bacteremia without removing the devices is often challenging. A new meta-analysis, however, sheds some light on which diagnostic tests for catheter-related infection are sensitive, specific and practical.
Although more than 200,000 cases of catheter-related bloodstream infections are reported annually, 80 percent of catheters that are removed as a suspected source of infection are subsequently determined not to be the source. Physicians need a way to diagnose these infections that does not involve removing the catheter, Dr. Basaviah said, but clinical findings are all too often unreliable.
The most common findings “fever with and without chills “are nonspecific, while specific findings “inflammation or purulence around the IV site “are not sensitive. As a result, the CDC recommends against culturing and removing catheters unless there is both inflammation around the site and systemic signs of candidemia or bacteremia.
In a meta-analysis in the March 15, 2005, Annals of Internal Medicine, Safdar and colleagues sought to identify the best way to diagnose an IV-related bloodstream infection. Researchers compared eight commonly used diagnostic tests and calculated the sensitivity, specificity and likelihood ratios for each.
While paired quantitative peripheral and IV-drawn blood cultures were the most accurate overall, they are costly, complicated and subject to contamination. Most of the other strategies, such as qualitative culture of catheter segments and qualitative culture of catheter-drawn blood, had acceptable sensitivity and specificity (mid to high 80s to low 90s on average) as well as high negative predictive values (greater than the 99th percentile).
Further, researchers found that there may be a practical alternative to quantitative assays: the differential time to positivity analysis. Most labs monitor blood cultures continuously and record the positive tests every 10 to 20 minutes. If the blood cultures from the central line become positive two hours before the peripheral culture, then it is highly likely that the line is the source of the infection.
A study done in 2004 in patients who had short-term and long-term catheters supported this conclusion. “The differential time to positivity adds no additional laboratory cost, it has similar sensitivity, though slightly lower specificity, and it utilizes a blood culture system already in place,” explained Dr. Basaviah.
Antiplatelet drugs and GI bleeding
New data may help change how physicians think about strategies to avoid GI bleeds in high-risk patients who need antiplatelet therapy.
While aspirin is commonly used to prevent cardiovascular events, even low doses of aspirin (75 mg/d) can double the risk of upper GI bleeding. Adding a PPI may reduce that risk of aspi¬rin-induced ulcer bleeding, and the strategy has been advocated for high-risk patients.
Clopidogrel has been advocated for inpatients with acute coronary syndromes who cannot take aspirin due to GI intolerance. While a secondary analysis of the Caprie Trial demonstrated that the aspirin group bled 1.3 times more often then the clopidogrel group, the dose of aspirin used was high (325 mg). It is known that the risk of gastrointestinal bleeding with 325 mg aspirin a day is almost double than that with 81 mg.
To assess the risk of bleeding in patients receiving antiplatelet therapies, Chan and colleagues compared two therapies: clopidogrel vs. a combination of aspirin and a PPI (in this case, esomeprazole). Researchers then compared the two strategies’ effectiveness in preventing recurrent ulcer bleeds in patients at high risk for ulcer.
The study, which was published in the Jan. 20, 2005, New England Journal of Medicine, found an 8.6 percent bleeding rate in clopidogrel group. By comparison, there was a 0.7 percent bleeding rate in the aspirin-plus-PPI group, a difference that was statistically significant. Further, the anti-ischemic efficacy was equivalent in the two groups.
There was no difference in the lower GI bleed incidence, an observation that Dr. Basaviah noted is consistent with expectations about how PPIs affect the GI tract.
Dr. Basaviah said that the Chan study gives strong credence to the notion of using aspirin and a PPI in patients who need antiplatelet therapy but are at high risk of GI bleeding. “In patients with previous aspirin-induced ulcer bleeding,” she said, “the addition of PPI to aspirin was safer than changing to clopidogrel.”
Cardiac resynchronization in CHF
Current evidence on the effects of cardiac resynchronization in patients with moderate to severe CHF suggests that biventricular pacing improves ventricular function, exercise tolerance and quality of life while reducing the need for hospitalization. Until recently, however, there has been little good evidence about the effect of cardiac resynchronization on mortality.
In a study published in the April 14, 2005, New England Journal of Medicine, Cleland and colleagues examined more than 800 patients with class III and IV congestive heart failure despite medical therapy. The research, known as the Cardiac Resynchronization-Heart Failure (CARE-HF) Study, randomized patients to receive standard medical therapy or a combination of standard medical therapy and a bi-ventricular pacemaker.
Researchers found that overall, cardiac resynchronization therapy helped reduce death from any cause or unplanned hospitalizations for a major cardiovascular event by a relative risk reduction of 28 percent when compared to medical therapy alone. Dr. Frost said that the study also showed a 32 percent relative risk reduction in all-cause mortality with cardiac resynchronization.
The bottom line? As Dr. Frost explained, “Patients with New York Heart Association Class III or IV congestive heart failure exhibiting significant intraventricular conduction delay and ventricular derangement should be referred for cardiac resynchronization therapy.”
Warfarin vs. aspirin for intracranial arterial stenosis
Recent research offers new information for physicians trying to choose between warfarin and aspirin to protect patients who have recently had symptoms referable to atherosclerotic intracranial arterial stenosis.
To shed light on the issue, Chimowitz and colleagues compared warfarin to aspirin. The goal was to determine which agent is more effective in preventing adverse events among patients with symptomatic atherosclerosis of a major intracranial artery.
In a paper published in the March 31, 2005, New England Journal of Medicine, investigators gave patients warfarin or aspirin at a dose of 650 mg twice a day. The primary endpoint was the composite incidence of ischemic stroke, brain hemorrhage or death from vascular causes other than stroke. Because of safety concerns, the study was discontinued before randomization was completed.
Researchers found no difference in the incidence of ischemic stroke, brain hemorrhage or death from vascular causes other than stroke between the aspirin and the warfarin treated subjects. Data also showed no significant differences between the therapies for several secondary endpoints.
The study did find, however, that death was statistically more likely if patients were treated with warfarin, as demonstrated by a hazard ratio for mortality for aspirin compared to warfarin of 0.46. Major hemorrhage was also statistically more likely with warfarin, with a hazard ratio for bleeding for aspirin when compared to warfarin of 0.39.
“The study shows that warfarin and aspirin are equivalent for stroke prevention among patients with symptomatic atherosclerosis of a major intracranial artery, and that aspirin appears to be the safer of the two treatments,” said Dr. Frost. “Because of this, the study’s investigators concluded that aspirin is the preferred therapy.”
He cautioned, however, that an analysis of the effect of INR on outcomes suggested that if managed meticulously to maintain therapeutic levels of anticoagulation, warfarin may reduce stroke incidence compared to aspirin.
Perioperative beta-blocker therapy
While beta-blockers are widely used to reduce perioperative cardiovascular complications during noncardiac surgery, there have been few studies that showed the effect of perioperative beta-blockade on in-hospital mortality in routine clinical practice. A recent study, however, raises new questions about their effectiveness and safety.
To examine the perioperative use of beta-blockers, Lindenauer and colleagues designed a retrospective study of 663,000 major, noncardiac surgery patients. The research was published in the July 28, 2005, New England Journal of Medicine.
Researchers analyzed the association between beta-blocker use and in-hospital mortality using the revised cardiac risk index. The index is a classification system that uses a scale of 0 to 5 to predict perioperative cardiovascular risk. Higher scores represent a higher risk of adverse events.
Patients who had the lowest scores on the index for perioperative cardiovascular complications appeared to face a significantly increased risk of death when treated perioperatively with a beta-blocker. Patients who had a score of 2 derived little or no benefit from perioperative beta-blockade, but they also appeared to suffer no harm from the therapy.
When patients had a score of 3 or higher, however, beta-blockers were associated with a significant decrease in the risk of death.
When patients had a score of 3, the number needed to treat to prevent death was 62. When patients had a score of 4 or higher, the number needed to treat to prevent death fell to 33.
Dr. Frost said that while the trial showed that beta-blockers reduce mortality in high-risk patients, their effects on other patients should give hospitalists cause for concern. “Beta-blockers conferred no benefit for intermediate-risk patients,” he explained. “However, they appeared to be harmful in low risk patients.”
Michael Krivda is a freelance writer specializing in health care. He is based in Perkasie, Pa.
