Published in the March 2009 issue of Today’s Hospitalist
How effectively is medical science containing pneumonia, a condition that’s the leading infectious cause of death in the U.S.? According to one of the nation’s top hospitalist experts on the disease, the short answer is “not so hot.”
The risk of dying in the hospital from pneumonia is still 50% greater than from any of the next 10 most common diagnoses, said Scott A. Flanders, MD, associate professor at the University of Michigan and director of its hospitalist program. That number is particularly sobering, he noted, because it has not budged since 1988.
During a presentation at the annual hospital medicine conference held last fall at the University of California, San Francisco, Dr. Flanders said that part of the problem rests with certain aspects of how hospitalists approach treating pneumonia. That’s particularly true, he explained, when it comes to antibiotic regimens much fanfare “don’t help much in preventing pneumonia.
In outlining what hospitalists should start, stop or consider when treating pneumonia, Dr. Flanders said that some recent advances may bring good news. There’s now an innovative system to clarify and predict which pneumonia patients need ICU level care, for example, as well as new insights into the causes of highly fatal cases of community-acquired pneumonia (CAP), including S. aureus, CAP and pseudomonas.
According to Dr. Flanders, identifying CAP patients who need ICU care or vasopressor support is easier thanks to a new scoring system that assigns points for the following factors: low systolic blood pressure, multilobar chest X-ray involvement, hypoalbuminia, respiratory rate, tachycardia, confusion, hypoxia and low pH. The system “which is known by its acronym, SMARTCOPS ” is “useful at its extremes,” Dr. Flanders said.
Patients with a SMART-COPS score of less than three, for example, have a very low risk of needing intensive respiratory or vasopressor support. Once patients score more than seven points on the scale, however, they are more likely to need ICU level care. In a study in the Aug. 1, 2008, Clinical Infectious Diseases, two-thirds of patients with more than seven points needed a trip to the ICU.
For individuals who fall somewhere in the middle and have a score of three or four points, the risk was more moderate, while a score of five or six points indicated high risk.
MRSA and other bugs
New information is also now available on CAP caused by MRSA and by pseudomonas.
The CDC published a surveillance study that examined several cases of what turned out to be MRSA-associated CAP. Dr. Flanders said that 85% of those cases had a PVL gene, “which is a cytoxic gene associated with very serious disease in young people.” These patients often have positive sputum and blood cultures and usually end up in the ICU.
“Very few people have MRSA risk factors, so it’s hard to predict who’s going to get this,” he explained. Risk factors include past skin infection (abscesses or boils, not cellulitis), IV drug use and influenza. Presentation typically includes severe necrotizing infection, hemoptysis and leukopenia, and high fever. Vancomycin or linezolid, not daptomycin, should be used for this condition.
Other bugs that hospitalists should worry about include gram negatives, including pseudomonas. Gram negative predictors are what Dr. Flanders referred to as “the four A’s”: aspiration, prior admission, prior antibiotics and advanced pulmonary comorbid disease.
“COPD was one of the biggest predictors of pseudomonal infections,” he said. “Most recent guidelines strongly suggest initial empiric pseudomonal coverage for patients with severe COPD and pneumonia, whether they’re in the ICU or out on the wards.”
Antibiotics: less is more
Evidence is growing, said Dr. Flanders, that atypical antibiotic coverage “adding macrolides and/or fluoroquinolones ” delivers no reduction in mortality or in clinical failure rates.
“Amazingly,” he pointed out, referring to a Cochrane Review meta-analysis first published in 2005 and updated in 2008, “when patients actually had atypical organisms isolated, atypical coverage did not add any statistically significant benefit.” (That was not the case with Legionella, Dr. Flanders noted.) However, guidelines continue to endorse adding atypical coverage, and it remains the standard of care.
Studies also increasingly show better results with short-course antibiotic therapy, Dr. Flanders said. That’s good news because less treatment translates into fewer side effects.
He typically treats patients with antibiotics from five to seven days, rather than 10 to 14. He stops antibiotics when patients have had no more than one pneumonia-associated instability in the previous 24 hours, such as altered mental status, hypoxia or fever, which typically occurs after five days of therapy.
The timing of antibiotic therapy continues to be controversial. The latest iteration of the core measure from the Centers for Medicare and Medicaid Services calls for administering initial antibiotic therapy for CAP within six hours of arrival at the hospital, up from four. But according to Dr. Flanders, it’s unclear what this six-hour end point is based on “or if having a specific time cut-off really works.
Giving antibiotics early is associated with antibiotic overuse, said Dr. Flanders, citing one study that found that “a lot of patients were getting antibiotics for CHF exacerbations in non-pneumonia-related pulmonary disease.”
While patients should receive antibiotics before they leave the ED once they have been diagnosed with pneumonia, it may take longer than six hours to pin down the diagnosis, he said. Dr. Flanders suggested that it would be more helpful to create “time bands of performance, rather than reporting absolute percentages, which can lead hospitals to try to game the system.”
Statins and more
Dr. Flanders flagged statins as one possible bright spot for treating pneumonia. Retrospective studies show that patients who are hospitalized with pneumonia and are taking statins may be less likely to stay in the ICU.
“I don’t start people on statins when they come in for pneumonia,” he said. “I think that’s a bit premature. But I darn well make sure that they continue to get their statin when they’re hospitalized. The data clearly support that.”
Prevention has received a boost from studies suggesting that some drugs may lead to higher pneumonia risk. Proton pump inhibitors (PPIs), which lower acidity, increase the risk for pneumonia by allowing for bacterial colonization of the stomach. A recently started regimen of PPIs is linked to the highest risk. Because of these data, Dr. Flanders recommends taking patients admitted for pneumonia off their PPIs if they do not have a strong indication for continued use.
The same goes for antipsychotic medications, which have been linked to an increased risk of developing pneumonia when the drugs are first started. Dr. Flanders suggested counseling patients who are on these drugs to be aware of the signs of pneumonia.
And the jury is still out on whether steroids lead to less shock, lower mortality and lower length of stay for pneumonia patients. One study that showed benefit had “issues,” said Dr. Flanders, and a new study is just underway.
Dr. Flanders also dispelled some common thinking on the following topics:
- Pneumococcal vaccine: Several randomized, controlled trials show that pneumococcal vaccines have no effect on bacteremia on patients over age 65 with comorbid illness. However, case control studies indicate that the vaccine may reduce invasive disease. One study found that patients who had the vaccine and were then hospitalized with pneumonia had reduced mortality or less need for ICU care.
However, to prevent one episode of bacteremia, 20,000 people would have to be vaccinated; to prevent one death, at least 50,000 people would need to receive the shot. While the vaccine may ultimately have a positive effect, “Don’t expect this to reduce pneumonia in your institution,” Dr. Flanders noted.
- Flu vaccine: Although some studies have strongly suggested that flu vaccine can prevent pneumonia, a more recent study showed that influenza vaccination did not prevent hospitalizations for CAP. (See “How effective are flu shots in your elderly patients?” in the January 2009 issue of Today’s Hospitalist.) The study looked at the impact of the flu vaccine on the risk for developing pneumonia when there was no flu around.
- Discharge: A new study found that the most common cause of readmission for patients with CAP was an exacerbation of comorbid illness, like COPD or CAD, rather than a pneumonia-related cause. During discharge, hospitalists should focus on comorbidities, not exclusively on pneumonia, for patients with those illnesses. Doctors also should devise intensive discharge interventions for those patients to prevent readmissions, Dr. Flanders advised.
Paula S. Katz is a freelance health care writer based in Vernon Hills, Ill.
NEW STUDIES ARE SHOWING that health care-associated pneumonia (HCAP) is looking more like hospital-acquired pneumonia (HAP) with multidrug resistant pathogens than like community- acquired pneumonia (CAP).
While HCAP develops outside the hospital, it is common in people who have had contact with the health care system, such as home therapy; hospital or dialysis clinic care in the past 30 days; hospitalization of two or more days in the past 90 days; or nursing home or long-term facility care. (HAP, on the other hand, develops in the hospital after 48 hours.)
HCAP patients tend to have higher rates of staph, MRSA and pseudomonas than CAP patients, and greater lengths of stay and less pneumococcus, according to Scott A. Flanders, MD, associate professor at the University of Michigan and director of its hospitalist program.
About 30% of HCAP patients are receiving too narrow a spectrum of therapy, which leads to higher mortality because they’re not being identified appropriately.
At the annual hospital medicine conference held last fall at the University of California, San Francisco, Dr. Flanders said that HCAP should be treated with a fairly broad regimen that consists of an anti-pseudomonal beta-lactam plus an aminoglycoside or linezolid.
A narrow treatment spectrum, on the other hand, may be indicated in some patients with nursing home-acquired pneumonia. One study found that patients with a mean age of 85 experienced lower admission rates to the hospital, fewer hospital days, lower mortality and lower costs with a care pathway that included oral levoxacin vs. the usual care.
Meanwhile, an older study on these patients is getting a new look. That trial makes the point that HCAP, especially in nursing homes, should be treated for multidrug resistant organisms only if two of the following three criteria are met: severe pneumonia, prior antibiotic use or poor functional status.
Dr. Flanders noted that HAP should be treated with multidrug resistant coverage if certain risk factors are present. Those include instances where a family member has a resistant organism or cases in which the patient is undergoing immunosuppressant therapy.
The problem is that there is no gold standard for diagnosing HAP. Studies on clinical diagnoses; CDC criteria, which could be overly inclusive; and sputum.
The result? “You end up giving a lot of people antibiotics,” Dr. Flanders said, “and perhaps overtreating some.” Solutions include identifying low-risk patients and reviewing your local flora to best treat your high-risk patients, then tailoring antibiotic therapy as culture results become available or the patient improves.