Search Annals: Advanced search

Podcast Transcript - January 15, 2008

[ Return to Podcast Home Page ]

Hello, and welcome to this week’s Annals of Internal Medicine audio summary for our January 15, 2008 issue. I’m Michael Berkwits, Deputy Editor at Annals.

We have another scintillating issue for you this week, with articles on serum markers of mortality in patients with peripheral arterial disease; a new risk score for predicting the onset of hypertension; and the comparative effectiveness of medications for treating rheumatoid arthritis. Plus, I’ll summarize all the other articles in this week’s issue.

But our lead feature article this week is one about community-acquired multidrug resistant MRSA infection in communities of men who have sex with men. We’re releasing the article on-line and discussing in this summary, ahead of its appearance in the print journal, to notify the clinical and public health communities quickly about the potential spread of multidrug resistant MRSA strains and of possible risk factors for that spread. Here’s an in-depth summary.

Community-Associated Multidrug Resistant MRSA in Men Who Have Sex With Men (Time: 00:54)

MRSA first appeared in community settings in the year 2000, and since then has spread to become an important etiology of skin, soft tissue, and systemic infections. Most cases of community-acquired MRSA are resistant to beta-lactams and 1 or 2 other drug classes, but they’ve generally remained susceptible to drugs such as clindamycin, tetracycline, or trimethoprim-sulfamethoxazole. In 2006, the authors of this week’s report identified a form of MRSA resistant to those drugs too, that had been isolated from the wrist abscess of a 36 yo HIV positive man being cared for at San Francisco General Hospital. Since that report, studies have identified clusters of similarly multidrug resistant MRSA infection, also called MDR USA300 infection, among gay men and people with HIV in San Francisco and Boston.

In this week’s report, lead authors Binh Diep and Henry Chambers of the University of California in San Francisco extend those observations by quantifying the incidence MDR-USA300 infections in San Francisco, and by using clinic data to characterize possible risk factors for the infection.

Their report actually summarizes 4 smaller studies, which is a little unusual for an Annals article.

In the first of the four studies, the authors used a random sample of MRSA isolates from 9 of the 10 medical centers serving the city of San Francisco to calculate the incidence of MDR-USA300 infection in the city. They estimated an annual incidence of infection of 275 cases per 100,000 persons throughout the city, and found that incidences were higher in zip codes where census data indicated a higher proportion of male same sex couples lived. They found a peak incidence of infection in the Castro District, which also had the proportion of male same sex couples that was highest among all San Francisco zip codes, about 26%.

In the middle two studies, the authors retrospectively reviewed the charts of patients in San Francisco General Hospital’s HIV clinic and Boston’s Fenway Community Clinic to identify risk factors for the MDR-USA300 infection. They found that men were more likely to have the infection; and that being a man and having sex with men was a risk factor, as was HIV infection; having had MRSA infection in the past 12 months; and having been treated with clindamycin. The association between infection and male-male sex was reinforced by the observation that infections of the buttocks, genitals, and perineum were much more common with MDR than with non-MDR USA300 infection. Of interest, one Fenway clinic patient who reported frequent travel to the Castro District has an MDR-USA300 isolate that had a resistance-inducing plasmid identical to those identified in San Francisco, suggesting that frequent travel between the coasts, or between centers for same-sex sexual activity, may facilitate the spread of the clone.

Finally, the authors reviewed about 200 USA300 isolates collected as part of a previously reported survey of emergency departments, and tested 8 which showed clindamycin resistance. Only two carried the plasmid which conveys multidrug resistance, suggesting that multidrug resistant USA300 is currently uncommon in the general community.

The authors conclude that infection with MDR-USA300 MRSA is transmitted sexually in populations of men who have sex with men, and that it has spread rapidly in this population. They acknowledge that their estimates may not be quite accurate, because their survey study relied on a passive surveillance system which would likely lead to underestimation of infection incidence, and because the 2 clinic studies relied on retrospective chart reviews, which introduce a host of potential selection, referral, documentation, and other biases. Because the study relied on medical chart review, the authors could not determine the specific sexual behaviors facilitating spread of infection, which would be important to know for targeting prevention efforts. and the numbers of infections were few, so

Still, the sum of the evidence from the article’s substudies seem to converge on the real phenomenon of MDR-MRSA spread in , one which warrants the attention of clinicians, infectious disease specialists, the public health community, and the community itself.

Interview With Co-Authors Binh Diep, Phd, and Henry Chambers, MD (Time 05:06)

To get a better handle on some of the clinical issues that come out of this report, I called the study’s lead authors. Dr. Binh Diep is a postdoctoral fellow in the Division of Infectious Disease at the University of California in San Francisco, and Dr. Harry Chambers is a Professor of Medicine at UCSF and the Chief of Infectious Diseases at San Francisco General Hospital. They were good enough to take the time to explain to me some of the more practical implications of their findings.

Q: Dr. Diep and Dr. Chambers, thank you for talking to me.

HC: You’re quite welcome.

DP: Great to speak with you.

Q: This week’s article distinguishes 3 clones of community-acquired MRSA: multidrug resistant USA300; non-multidrug resistant USA300; and non-USA300 MRSA. That can get a little confusing for the non-specialist, so I’m wondering if you can tell us where those terms come from, and if you could go back to the beginning, say when MRSA was strictly a nosocomial infection, and walk us through the evolution of these strains and how that evolution has been manifest clinically.

BD: I think the MRSA nomenclature is best understood if we really start from the very beginning. The first MRSA was described in 1961; that’s only one year after the introduction of methicillin into clinical use. Then over the next four decades, the MRSA clone spread in hospitals worldwide and developed resistance to multiple antibiotics. These clones are called hospital-associated MRSA. And you read of MRSA emerging in 2001 in communities across the United States. These community-associated MRSA clones are commonly called USA300. Now, unlike the traditional hospital-associated MRSA, the USA300 clone is susceptible to one or two antimicrobial classes in addition to the beta-lactams. We refer to these prototype strains as non-multidrug resistant USA300. We had expected that the multidrug resistant form of USA300 would evolve deep in the hospital, where high rate of antibiotic consumption would select for the resistant clone. But beginning in early 2004, we noticed an increasing prevalence of the multidrug resistant USA300, not in the hospital, but actually in an outpatient clinic specializing in HIV/AIDS care. The HIV positive patients in the clinic frequently had recurrent infections. We used a very brute force approach of whole genome sequencing, and we discovered that the multidrug resistance was encoded by a conjugative plasmid called pUSA03. We define the multidrug resistant USA300 clone on the basis of its carriage of this pUSA03 plasmid.

Q: Table 1 of the paper suggests that the main distinction between multidrug and non-multidrug-resistant MRSA is resistance to clindamycin, mupirocin, and tetracycline. So those 3 drugs, with the possible exception of clindamycin, don’t seem like they’re that commonly used. So why is that a big deal - does resistance to those 3 additional drugs imply resistance to other drugs in other classes?

HC: The importance is the potential impact on therapy. For example, the two bodies that have made recommendations on use of antibiotics in treatment of community MRSA infections, IDSA and CDC, both recommend in addition to considering the use of trimethoprim-sulfa, clindamycin and tetracycline. So in this particular lineage, two of the three drugs that have been recommended for use in therapy, the strains are resistant to. So that is of therapeutic importance when selecting empirical therapy, and when a patient shows up, the therapy generally is going to be empirical until you get the susceptibility data back. With respect to mupirocin, that is an important drug because that’s been used to eradicate colonization, in the hopes of aborting future infections by ridding the patient of a colonizing strain. That remains to be seen that you can actually affect that outcome with mupirocin, but it’s a common practice, and here again the more resistant strain has eliminated that as a therapeutic option, in the hopes of preventing future infection by eradicating colonization..

Q: You write that most patients in the San Francisco HIV clinic population who had MRSA infections presented with skin and soft tissue infections, and those were most commonly abscesses, cellulitis, or folliculitis. Did these infections look different in any way when they were caused by MDR-MRSA than when they were caused by more susceptible strains?

HC: Overall, there was no difference in the distribution among the various types of infections – abscess, cellulitis, or folliculitis -- although the more drug resistant strains tended to involve the buttocks, genitals, and perineum. There are also a couple of studies that have compared community MRSA infections with skin and soft tissue infections caused by other clones. And they’re not distinguishable clinically, so I very much doubt that you would be able to, in this subclone of a clone, tell much of a difference clinically on initial presentation. Again, to the extent that antibiotic therapy makes a difference, the lack of response may be more commonly encountered in the drug resistatnt strain.

Q: In the two clinic populations you studied there were virtually no instances of MDR USA300 infection in men who did not have sex with men. So could it be said that this multidrug resistant clone of CA-MRSA is in some respects a newly recognized sexually transmitted infection in this population?

BD: Yes. I think there are two lines of evidence supporting a role of sexual transmission of MDR-USA300. First, the multidrug resistant infections affect men who have sex with men at body sites in which there is skin-to-skin contact occurs very frequently during sexual activites. Second, independent data from the Fenway clinic shows that 2% of men who have sex with men carry USA300 asymptomatically in the perianal area, and thus because it is carried perianally it can be spread during sexual contact. The risk factors for colonization include use of the internet for sex, having skin-abrading sex, using crystal meth, as well as a variety of other sexual risk factors. So taken together, I think it can be said that multidrug resistant USA300, as well as non-multidrug resistant USA300, can be sexually transmitted

Q: The paper seems very careful not to use the words gay or even homosexual as descriptors for male patients who had sex with other men and who you found had a higher risk for MDR-MRSA transmission. So I wonder if that was deliberate? Do you know how people in the clinic populations self-identified in terms of their sexual orientation, and do you think it will be important to MRSA epidemiology, as it has been to HIV infection, that men who self-identify as straight may be having sex with other men and possibly increasing their risk for infection, and transmission of this clone more generally?

HC: I don’t think we can answer that from the data in this study, simply because we had to rely on the way the history was elicited in the charts, so I couldn’t speak to a self-identified heterosexual population that might have risky sexual behavior. However, risky sexual behavior is correlated with other risky behavior, and to the extent that this behavior may reflect other risky behavior, and I’ll just give you an example, crystal meth use, or injection drug use, they could be factors in populations that don’t identify themselves as men who have sex with men.

Q: You couldn’t do an analysis that looked at HIV infection and male-male sex at the same time, but what do you think about the relationship between those two: is there any suggestion that MDR-MRSA might be some kind of new opportunistic infection in HIV positive patients?

BD: Well in general, it has been well demonstrated that HIV positive patients have increased risk for community associated MRSA infections. But unfortunately we could not assess the independent effects of HIV infection and male-male sex, because we could not identify any case of multidrug resistant USA300 in men who are straight. I think we can reasonably hypothesize that HIV positive men who have sex with men may have high rates of antibiotic consumption, that then favors the spread of the multidrug resistant form of USA300. So this may be a situation where the multidrug resistant form of USA300 evolves among HIV positive men who have sex with men and then spreads into the population of HIV negative men who have sex with men. In Boston, about 50% of the patients who have these multidrug resistant USA300 infections are HIV negative men who have sex with men.

Q: MRSA infection in the past 12 months and clindamycin use in the past 12 months were also risk factors for MDR-MRSA. What can you tell us about that?

BD: I think there is actually a plausible reason why these two are independent risk factors. Clindamycin is used as empiric therapy for treating suspected staph infection in our outpatient clinic. But clindamycin is ineffective for treating this multidrug resistant form of USA300. So that may lead to the recurrence of infection, and hence the prior MRSA infection that’s recorded in the medical record review. What’s even more, is that clindmaycin provides the antibiotic selection pressure that then favors the spread of the multidrug resistant clone. But I think it’s also important to note that antibiotic selection pressure is not at all essential to sustain the spread of multidrug resistant USA300. For example, less than 1% of the patients treated at Fenway in Boston had prior treatment with clindamycin. The finding has a very important implication and that is multidrug resistant USA300 can spread into the general population in which antibiotic usage is less frequent.

HC: And If I could throw in one more factor, the plasmid we’re talking about, this pUSA03, encodes determinants for each of the drugs that we have assigned the multidrug resistant phenotype to. So you don’t have to just use clindamycin; one could select with either tetracycline or mupirocin and still wind up with this clone because they are all linked. So a general increase of use of antibiotics can favor the selection of this clone in this population, you don’t have to necessarily have a specific drug that might be used in a patient’s individual therapy.

Q: So how do or should these findings change clinicians’ practices? When a skin or soft tissue infection doesn’t lend itself to culture, should practitioners assume multidrug resistance and be thinking of drugs like Bactrim and others in the bottom half of the paper’s Table 1 for the empiric treatment of infections in gay men and others who have male-male sex? Or is this still a phenomenon limited to tight clusters of men in these regional communities?

HC: It’s probably that we’re looking at a cluster of a particular clone in a fairly well-defined population. But this is a moving target, and one change that community MRSA has wrought is the need to culture skin and soft tissue infections. There are a variety of recommendations that could be made. You could culture everyone, you just culture those who fail, or, what I think is important is to periodically assess what the culture data are in your population because it allows you to detect the presence of a strain such as this by a shift in phenotype, then that would have an impact on your practice. So for example, in our clinic, we know that this clone is quite prevalent, on the order of 20 or 30%. So that is prevalent enough where you should culture, but it’s not so prevalent that I think you would avoid the other antibiotics. But it does affect your practice in that you have to realize that a substantial minority may be resistant to a drug that you would like to use empirically.

BD: I think the bottom line message is that community associated MRSA is really a preventable disease. Good scrubbing with soap and water may really be the most effective way to prevent this skin to skin contact transmission, especially after sexual activity.

HC: I would agree, Acquisition and transmission of Staph aureus, I call it a contact sport. And to the extent that hygiene is important after any contact, and this is all speculation because no one really knows, but in this population as well, I think that is something that could be easily accomplished, and might in fact be of benefit.

Dr. Diep and Dr. Chambers, thanks so much for talking to me.

HC: It’s been a pleasure

BD: Thank you.

That was Drs. Binh Diep and Henry Chambers discussing this week’s article on community acquired multidrug resistant MRSA in communities of men who have sex with men. The article, entitled “Emergence of Multidrug-Resistant Community Associated, Methicillin Resistant Staphylococcus aureus clone USA300 in Men Who Have Sex With Men”, will appear in the February 19 edition of the print journal.

Improving the Care of Patients At The End of Their Lives (Time 18:22)

Our other feature article this week is a systemic review of evidence about improving the care of patients at the end of their lives. Lead authors Karl Lorenz of the Veterans Affairs Healthcare System in Los Angeles, and Joanne Lynn of the George Washington University in Washington D.C., organized the review around 4 key questions.

First, they asked what people mean when they talk about the end of a patient’s life, and what those meanings suggest about how to identify patients who need or would most benefit from palliative care, which is care aimed at comfort rather than cure. They found many different definitions of end-of-life in the palliative care literature. Some studies used specific clinical characteristics or survival prediction rules, but many others relied on provider judgment, and what those providers judged was varied and not well-defined. For example, some studies relied on clinician assessment of patients’ having a condition that could cause their death with the very next complication, or within 6 months, while others relied on clinicians’ assessment of patients’ readiness to die. The authors also make the important point that the end of a patient’s life may be easier to recognize in some diseases than in others. For example, diseases like cancer are characterized by a short period of rapid functional decline before death, so that a patient’s impending death can usually be fairly easily anticipated. But other diseases, like heart and lung failure, are characterized by a long period of gradual functional decline, punctuated by frequent exacerbations that may look like the patient could die but they do not. In these instances, it’s very difficult to tell if a patient is close to dying or not. So in the end of this part of the review, the authors conclude that waiting to implement palliative approaches to care until clinicians are nearly certain a patient will die will lead to many missed opportunities to provide comfort care; and that offering palliative approaches to all people living with serious illness as an option will better identify those who would desire or benefit from that approach to care, even if the timing of their death is uncertain.

The review’s authors then asked what the literature suggests are the critical issues clinicians should address when caring for people who are nearing the end of the lives. They found qualitative research and expert opinion that identified pain prevention and symptomatic relief as important, along with support of families and caregivers; ensuring continuity of care; and providing information and options so patients can make informed decisions about their treatment options. They also found literature to suggest that attending to patients’ emotional well-being, including their spiritual concerns, and helping them sustain their function is also important.

Then the authors asked what treatment strategies work well for pain, dyspnea, and depression, 3 of the most prominent symptoms that affect patients as they approach the end of their lives.

They found strong evidence that NSAIDs, opioids, radionuclides, and radiotherapy are effective for treating cancer pain, and inconsistent evidence that bisphosphonates are effective for metastatic bone pain, but no evidence about treatments for pain in other disease, such as heart failure and dementia.

They found strong evidence from small and short-term trials that beta-agonists and opioids are effective for treating dyspnea caused by end-stage COPD, and strong evidence that pulmonary rehabilitation and oxygen relieve symptoms of dyspnea during activity in patients with COPD. For dyspnea caused by cancer, however, they found only weak evidence that opioids relieve symptoms, and weak evidence that oxygen may not be useful. They found no evidence about treatments for dyspnea in heart failure.

And in their review of treatments for depression, they found that the literature was limited to studies of treatment of depression in patients with cancer; but that that evidence strongly shows that psychotherapy, tricyclic antidepressants, and SSRIs are effective.

Finally, the authors asked what collaborative strategies might improve end-of-life care and relieve caregiver burden. They found strong evidence that the use of health care teams that manage the care of heart failure patients after they have been discharged from the hospital can reduce rehospitalization, but there was no similar evidence for similar effects in other conditions. And they found weak to moderate evidence that attention to caregivers can relieve some of the burden of caring for a sick relative or partner, although those effects were generally small.

Based on this review, the authors conclude that strong evidence supports interventions to improve end-of-life care, but that research is desperately needed to improve many aspects of dying. And an accompanying clinical practice guideline for palliative care issued by the American College of Physicians makes the additional recommendation that providers should help patients develop advance care plans. Those plans should address wishes for resuscitation and emergency treatment; wishes for long-term treatment, such as the use of tube feedings for patients with dementia and chemotherapy for cancer patients; and identification of surrogate decision makers to uphold their wishes and to make decisions if a patient is incapacitated.

So in summary, care that aims at comfort should be offered to all patients with serious illness; that care should involve accepted and effective treatments for pain, dyspnea, and depression, and should address other symptoms and attend to their emotional and spiritual needs; it should ensure continuity across sites of care, perhaps with the use of health care teams that include but are not limited to physicians; and it should incorporate patients’ wishes if known, and address the burden of caregivers responsible for the patient, especially in disease like dementia, where functional decline and the process of dying may occur gradually over years.

So that’s some of the science underlying the end-of-life, but we’ll close this segment with two alternative views. In our September 4 audio summary about the play Wit we talked about John Donne, who was preoccupied with the puzzle of Christian salvation and the promise of life after death; in this most famous poem, death itself dies, vanquished by the promise of eternal life. Here’s Death Be Not Proud, read by Kristin Hughes (poem available at http://www.bartleby.com/105/72.html). . And here’s another take on the topic, in which poet e.e. cummings distinguishes between “perfectly natural” dying, and “evil, legal” death. Here’s “dying is fine, but death, ” read by Bob Dylan.

(Poem available at www.internal.org/view_poem.phtml?poemID=378)

I’ll post web addresses for both of those poems in the transcript to this podcast at annals.org/podcast.

Other Articles In This Week’s Issue (Time 25:49)

Other articles in this week’s issue include a cohort study of patients with peripheral arterial disease, reporting that elevated levels and increases in the levels of D-dimer, C-reactive protein, and serum amyloid A levels -- the latter an inflammatory marker like C-reactive protein -- were associated with an increased risk for death within 1 to 2 years.

We have a report on the epidemiology of pain in adults with sickle cell disease, suggesting that patients with the disease have pain that is far more severe and frequent than most physicians suspect, and that because most patients obtain relief outside of the health care system, their symptoms are widely under-recognized and undertreated.

We’re publishing an easy-to-use risk score that combines information on age, sex, body mass index, smoking, blood pressure, and family history of hypertension to predict the short- and long-term probabilities of developing hypertension in people with normal blood pressures.

And we also have two other systematic reviews.

The first looks at the evidence for the comparative effectiveness and harms of drugs for rheumatoid arthritis. The available evidence, which is very limited, suggests that synthetic disease-modifying agents – specifically methotrexate, leflunomide, and sulfasalazine -- do not differ in their efficacy. Biological agents – specifically adalimumab, etanercept, and infliximab -- also don’t differ in efficacy. Biological agents improve radiographic but not clinical outcomes more than methotrexate; and combinations of drugs improve outcomes in patients who don’t adequately respond to treatment with a single drug. The review also found that there was insufficient evidence to draw conclusions about differences in rare but serious adverse events for the biological agents.

And the other review in this weeks issue looks at the evidence that releasing provider and hospital performance data improves the quality of health care. It found some evidence that publicly releasing performance data stimulates quality improvement efforts at the hospital level, but inconsistent evidence about the effects of releasing those data about providers, and they found no evidence that releasing data led to improvements in clinical outcomes or in patient safety.

Well, that’s it for today.

Keep telling your colleagues about Annals and about these summaries, and send feedback, suggestions, and comments about these summaries and the journal, by writing us at podcast{at}annals.org.

Continued thanks to Kevin Stahl and all our friends at WHYY, public radio and television of Philadelphia, who helped us develop these podcasts.

Technical support for this summary was provided by Andrew Langman, Neil Kohl, and Beth Jenkinson.

We’ll shuffle off our audio coil this week with a song in honor of our review and practice guidelines for end-of-life care. From Alan Lomax’s 1959 field Georgia Sea Island field recordings, reissued in 1998 by Rounder Records as volumes 12 and 13 of the Southern Journey series, here’s the incomparable Bessie Jones, belting out some advance care planning, with “O Death. ”

[Music]

Bessie Jones died in 1984 at the age of 82.

Check back in two weeks for our regularly scheduled February 5 issue; I’m Michael Berkwits, and thanks for listening.

[ Return to Podcast Home Page ]