Thursday, April 30, 2009
Wednesday, April 29, 2009
I’ve never understood Twitter, nor have I ever tweeted. But a colleague of mine, Phil Polgreen, recently told me about an interesting research project he is working on with Alessio Signorino, using real-time tweets to track public interest in swine flu. Some of the background and details of the project can be found here.Open up this webpage, and you’ll see a map that will begin to light up all over with tweets—scroll across the tweets to read them! Right now it only works with Firefox or Safari, not Internet Explorer....
Also, WHO has raised the pandemic alert level from 4 to 5.
What this means is that in addition to the 91 cases in 10 states that are now being reported, we can assume that S-OIV is now widespread. As more cases are reported with no links to Mexico or to other confirmed cases, the epidemiologic criteria for testing outlined in prior posts becomes increasingly obsolete. Testing will become more and more widespread, and case counts will skyrocket. The epidemic will appear to be expanding rapidly, but most of the changes will reflect increased testing and recognition of prior spread. We won’t know what is really happening now (in real time), for at least a couple weeks. I believe we’ll see more deaths and severe clinical presentations, but that the current epidemic will fizzle out over the next month—to return in the Fall as one of the circulating seasonal strains. Hopefully there will be an effective vaccine by then.
In the meantime, we need to prepare for more pessimistic scenarios than the one I believe to be most likely.
I realize there is now some controversy brewing about what is the most accurate and yet most politically/diplomatically correct way to refer to this virus. For the time being, I will continue to refer to it as swine flu or swine influenza (H1N1), or in shorthand as swH1N1. I predict the media and public health officials will eventually settle on “North American flu”.
Addendum: For today at least, the name appears to be "swine-origin influenza".
Tuesday, April 28, 2009
The opening paragraph of the CDC website for clinicians states,
“Clinicians should consider the possibility of swine influenza virus infections in patients presenting with febrile respiratory illness.”That sounds pretty general. If we tested everyone with that presentation, we’d quickly overwhelm our public health labs (since they are currently the only labs capable and recommended to handle these samples). Or, as our state public health department is recommending, should we only test those who meet the CDC criteria for a suspected case, as outlined here:
A suspected case of swine influenza A (H1N1) virus infection is defined as a person with acute febrile respiratory illness with onsetOnce a case is confirmed in your locale, of course this question becomes moot. But for those in states that haven’t yet seen confirmed cases, it looms fairly large.
- within 7 days of close contact with a person who is a confirmed case of swine influenza A (H1N1) virus infection, or
- within 7 days of travel to community either within the United States or internationally where there are one or more confirmed swine influenza A(H1N1) cases, or
- resides in a community where there are one or more confirmed swine influenza cases.
Monday, April 27, 2009
Sunday, April 26, 2009
I won’t be blogging all the developments as they occur, since you can find the latest updates at the CDC and WHO swine flu websites. I will only comment when I have time or to address specific controversial issues.
One interesting issue comes out of the CDC guidance for infection control in healthcare facilities. While this guidance calls for standard + droplet + contact precautions for suspected/confirmed cases, it also expresses a preference for N95 masks during direct patient care and negative pressure rooms, which sure sounds a lot like airborne isolation to me! I’m recommending the same type of isolation we’ve recommended for avian flu or SARS, which we call combination precautions (essentially a combination of standard, airborne, and contact precautions).
The CDC infection control guidance for swine flu is a bit maddening, since it uses different distance parameters than traditional droplet precautions (6 feet rather than 3), and in other respects doesn’t fall neatly within either airborne or droplet categories. The reasons for this are twofold: (1) we don’t yet know enough about the transmissibility of the new swine flu virus, and (2) we know that not all respiratory pathogens fall neatly within droplet or airborne transmission categories. It would nice if pathogens either traveled for long distances on air currents or fell to the earth in large droplets within 3 feet of expectoration. Unfortunately, as we learned with SARS, the situation is far more complicated.
The CDC site I pointed to yesterday is still the best source of up-to-date information, and contains links to other sites on infection control issues, antiviral use, WHO information on the situation in Mexico, etc. As of today we have two additional cases in Kansas (a husband and wife who recently traveled to Mexico), and as surveillance becomes widespread across the country I predict we’ll be adding other states to the list very soon.
For those of us on the front lines of hospital infection prevention, this is another chance to test bioemergency preparedness as we ensure that our hospital wards, clinics and labs can respond effectively to the arrival of this new influenza strain.
Friday, April 24, 2009
So now that it’s almost May, we are suddenly running headlong into the real story of this season, which is the emergence of a new swine influenza virus (SIV, of the H1N1 variety). This story is moving very quickly right now, so by the time I post it will likely be out of date. We’ve gone from 2 U.S. cases yesterday to a total of 7 today, in California and Texas. The details are here as they are updated.
Human SIV cases have become a bit more common, but of the 12 cases reported between December 2005 and January 2009, 11 were in persons with some pig contact (either direct contact, or being “near” pigs). So the report of all these cases in those with no pig contact is suggestive of human-to-human transmission.
We also now know that a much larger SIV outbreak, responsible for up to 60 deaths, is ongoing in Mexico, preliminary details here. I can’t find whether this virus has been confirmed to be the same as that found in TX and CA, but I’d wager that it is.
So the CDC has new interim guidance for SIV, here, and the WHO is convening an expert panel to determine whether to raise the global pandemic alert level (based upon sustained human-to-human transmission of a new virus). Fasten your seatbelts…..
Wednesday, April 22, 2009
Monday, April 20, 2009
I found this article from today’s NY Times to be very interesting, both as a critique of our ability to detect foodborne illness, and as a corollary to problems inherent in the detection and reporting of healthcare-associated infections (HAIs).
The gist of the piece is that when a state has an excellent foodborne illness surveillance and investigation system (e.g. Minnesota), that state will find a disproportionate number of foodborne illnesses and outbreaks (now you might be saying, “Duh!”). But if one looks at the foodborne disease data without any knowledge of disparities in quality of surveillance, detection and investigation, Minnesota looks like a cesspool of foodborne disease! In fact, Minnesota may be the safest place in the U.S. to eat, because they are more likely than any other state to quickly detect and investigate foodborne disease outbreaks.
Similarly, those hospitals that do the best surveillance are more likely to detect HAIs when they occur. As just one example, the best way to avoid detecting Legionella infections is to fail to test for Legionella among patients with healthcare-associated pneumonia. Because Legionella won’t grow on the routine culture media used for respiratory specimens, a concerted effort must be made to request specific testing (using enriched selective media or the urinary antigen test).
Other examples are more nuanced, but suffice to say there are also many ways to fail to detect ventilator associated pneumonias, catheter-associated bloodstream infections, and surgical site infections.
How to address this problem? One way is to perform external validations (or audits) of each hospital’s surveillance data. For example, the CDC is working with states to develop validation methods for data submitted to the National Healthcare Safety Network (NHSN), which is extremely important since states with mandatory reporting legislation usually utilize NHSN as the reporting mechanism. But such validation is complicated, time-consuming and expensive.
An alternative approach would be to directly monitor data from microbiology laboratories or pharmacies, to correlate with (and validate) reported infection rates—or even as a surrogate for those rates. For example, if a hospital reported zero nosocomial bloodstream infections but exceeded a certain rate of blood culture positivity (for cultures drawn > 48 hours after hospital admission), that might trigger further investigation…..similarly, parameters could be established to compare reported SSI rates with post-operative antimicrobial use data.There is no simple and inexpensive solution, but addressing this problem will become increasingly important as HAI rates become public across the country.
Sunday, April 19, 2009
Mike’s post on the “Getting to Zero” movement ended with a reference to the race between humankind and the formidable microbe. For a vivid example of the adaptability, and survivability, of microorganisms, read this new Science paper describing a microbial ecosystem that has survived for over a million years beneath the glacial ice of Antarctica. The organisms have adapted to this harsh environment by deriving energy from sulfur and iron.
None of the organisms detected in this study appear to be human pathogens, but if I were treating an infection acquired after submersion in a glacial pool beneath the Antarctic ice cap, I might include coverage for Desulfobulbus and Thiobacillus….
Although not related to infection control, but of potential interest to the readers of this blog, is Abraham Verghese’s new novel, Cutting for Stone. It’s an absolutely beautiful and riveting story that I can’t quit thinking about. Dr. Verghese, an infectious diseases doctor and writer, is currently Professor of Internal Medicine at Stanford.
Friday, April 17, 2009
Well, having practiced as an infectious diseases physician for nearly 20 years, it seems obvious to me that the elimination of healthcare associated infections is not in the realm of possibility. As I make rounds in an academic medical center, the patients I see are ever more critically ill. For many of the patients in the ICU, we have bypassed nearly every natural barrier to infection for prolonged periods of time. More of our therapies involve high levels of immunosuppression. Despite our best efforts at reducing environmental contamination, any space where humans are found will be inhabited with the organisms that live on and in us. Thus, my biggest problem with the GTZ concept is that I believe that it’s inherently dishonest, and the dishonesty is unfair to patients and their families. It implies that HAIs can be completely eliminated (doesn’t zero mean no infections?).
APIC is now developing a series of HAI elimination guides. This sets up its members in hospitals across the country for failure and places them under enormous pressure from administrators when zero remains elusive. The end result is that when an infection occurs, there is the implication that someone must be at fault. It creates a culture of blame. And it sets the stage for more legislation. I think APIC's perseveration on the zero concept demonstrates what I have suspected for quite some time based on my feedback from many APIC members—there’s a real disconnect between frontline ICPs and APIC leadership. Just this week I was invited to give my GTZ talk at another regional APIC chapter and we have received many positive comments about this blog from APIC members.
I am not arguing that we can’t improve our infection rates. Peter Pronovost’s group shattered the glass floor, showing us how low bloodstream infection rates could be driven when the process of central line insertion is highly standardized. I used to believe that only a minority of HAIs were preventable; now I believe that most can be prevented. But there’s a big difference between most and all. Importantly, Pronovost’s group did it the right way—they conducted good studies and published their results in the peer reviewed literature. They didn’t view their work as a product to market and sell.
The medical ICU in my hospital has not had a case of ventilator associated pneumonia for the past 15 months. I’m enormously proud of that unit and the wonderful people who work there, but we’re not having a zero celebration and believe me, we haven’t eliminated VAP. Zero is ephemeral because in the race between man and microbes, the bugs usually win. They were here long before us, and no doubt will be here after we're long gone.
Thursday, April 16, 2009
Following on Mike’s post about measles in Maryland, we now also have reports of measles from Pennsylvania and from my state of Iowa. Ugh. If you are interested, you can read the details of the last big measles scare in my state (focusing on the overall costs of the public health response to a single case). My fondest memory from that 2004 episode was spending a long weekend afternoon in our emergency department while a child with a compatible clinical illness was being evaluated--compiling exposure lists, reassuring healthcare workers, and communicating with our on-call state epidemiologist about setting up vaccine clinics.
Details are still emerging, but the Pennsylvania cases appear to be associated with unvaccinated children, the common theme in these situations (lack of vaccination, that is).
Wednesday, April 15, 2009
Now that I have your attention, let’s talk about the greatest emerging antibiotic resistance challenge in healthcare settings: multiply drug resistant gram negative rods (MDR-GNRs). Reports keep coming in about the alarming rates of MDR-GNR carriage and infection in long term care facilities (LTCFs). I just ran across one in a journal I don’t often read, the Journal of Gerontology. This study, by Erika D’Agata’s group in Boston, found that MDR-GNRs (defined in this study as resistant to 3 or more antibiotic classes) were more prevalent in clinical cultures than were MRSA or VRE at one large LTCF in Boston. Another recent report from Anthony Harris’ group in Maryland found that carriage of Acinetobacter baumannii was more common than MRSA among long-term acute-care residents.
Compared with the MDR-GNR universe, MRSA is very simple: one gene (mecA), one altered target, one resistance phenotype. For MDR-GNR? Not so much. Detection and prevention approaches remain a tremendous challenge, which is why Acinetobacter outbreaks are more likely to require closure of units to new admissions than are outbreaks due to any other organism.
In this context, I think what we really need are more laws that force hospitals to focus their antibiotic-resistance prevention resources on MRSA. Don’t you agree?
4 persons have developed measles in Rockville, MD, a child in Idaho has died of pertussis, a Minnesota man has died of polio (vaccine strain), and a Chinese woman is suspected of dying of SARS on a train in Russia.
Tuesday, April 14, 2009
The Joint Commission has released a 200-page monograph on measuring hand hygiene adherence. I was given the opportunity to review and comment on this document prior to publication, and I highly recommend it as a comprehensive resource—not just about adherence measurement, but about hand hygiene campaigns in general. The authors address but generally sidestep one question we’ve already discussed on this blog: what should be the “goal” of our hand hygiene campaigns? What is achievable? The closest they come to answering that question comes on page 118: “the ultimate goal should be to demonstrate sustained improvement over time”.
Given the lack of reproducibility and comparability of adherence measurements at different healthcare facilities, this seems like an eminently reasonable statement. Let’s hope all JCAHO surveyors find the time to read this monograph and arrive at our hospitals armed with a genuine understanding of the complexity of this issue.
The crux of the paper is that individuals have various levels of predisposition to hand hygiene, which is influenced by numerous factors, such as repetition during training, role modeling and institutional culture. However, the most important factor influencing the predisposition to hand hygiene was direct vivid experience, such as personal involvement in an outbreak. Of course, translation of intention to actual behavior is influenced by situational factors, such as convenient placement of alcohol products, visible soiling of the hands, smells, signs and posters. The authors conclude that vivid vicarious experiences might improve hand hygiene, but somehow I don't think that showing surgeons scary pictures will do the trick. Figuring out how to get doctors to wash their hands remains the greatest enigma in healthcare epidemiology. It drove Semmelweis crazy. And 150 years later, many of us are following in his footsteps.
Monday, April 13, 2009
Given the unproven benefit of MRSA screening, such legislative mandates are unfortunate. But this is the end result of the hype surrounding MRSA that has been perpetuated by a relatively small number of hospital epidemiologists, as well as the Association of Professionals in Infection Control and Epidemiology (APIC), which fronted a series of seminars across the country sponsored by a company that makes MRSA testing kits. As I have noted in a previous posting, the role of industry in the MRSA active surveillance debate has received little scrutiny. I also think it's a failure of the hospital epidemiology community for collectively idly standing by while all of this occurred.
Sunday, April 12, 2009
Mike’s post points to the cost savings that may be achieved through reduced blood culture contamination rates, a strong argument for phlebotomy teams. The article Mike cited focused on emergency departments, but phlebotomists offer even more benefits in the hospital, including improved bloodstream infection surveillance. Since coagulase-negative staphylococci (CoNS) are both the most common cause of central line associated bloodstream infection (CLABSI) and the most common blood culture contaminant, high contamination rates lead to a lot of misclassification.
The problem of contaminants being classified as CLABSIs has been reduced by the change in NHSN surveillance definition, which now requires 2 or more cultures positive for common skin contaminants (like CoNS) to define a CLABSI. The most likely form of misclassification now is probably failure to identify true CLABSIs due to skin contaminants—either because 1 positive out of 2 cultures still has a predictive value of 20% for CLABSI, or because only one culture was obtained (often through a central line). Incorrect blood culture practices like this are much more likely when trained phlebotomists are unavailable, leaving blood cultures to be obtained by busy nurses, residents or medical students.
So let’s hope that tighter hospital budgets don’t result in short-sighted decisions to cut back on phlebotomy services—the costs of increased blood culture contamination rates and substandard blood culture collection practices are likely to far outweigh any savings……
So let’s hope that tighter hospital budgets don’t result in short-sighted decisions to cut back on phlebotomy services—the costs of increased blood culture contamination rates and substandard blood culture collection practices are likely to far outweigh any savings……
Saturday, April 11, 2009
An interesting study from Parkland Hospital has been published in the April edition of the Journal of Clinical Microbiology which shows that blood cultures drawn by phlebotomists in the ED were significantly less likely to be contaminated than those drawn by non-phlebotomists (roughly 3% vs. 6-7%). The authors then compared median hospital charges between the patients with contaminated blood cultures and those with negative blood cultures and ascribed the $8,720 difference in charges to the contaminated blood culture. This methodology has a great deal of potential for bias since the patients with contaminated cultures were not matched in any way to those with negative blood cultures. The authors contend that avoiding just 5 contaminated cultures would pay for one full time phlebotomist. While I do worry about this conclusion, even if the authors erred by a factor of 10 in their estimates of attributable charges, reducing the rate of contaminated blood cultures from 6% to 3% in an ED that draws 13,800 cultures per year (as does Parkland), would still pay for 9 phlebotomists at $40,000 annually for each. The findings were strong enough for Parkland to hire round-the-clock phlebotomists in their ED, and because of this paper I hope to convince my hospital to do the same.
Friday, April 10, 2009
I’ll try to focus my response on the main themes of their comment:
The claim that I demand randomized controlled trials (RCTs) is a straw man. I called for “well-designed, controlled trials.” Farr and Jarvis are correct that RCTs are expensive and can be difficult to adapt to some infection control questions. But there are many study designs other than RCTs that qualify as well-designed controlled trials! The study by Harbarth et al is perhaps the best example of a well-designed controlled trial of active surveillance and contact precautions. It is in fact the largest such study ever performed. Check it out. To demonstrate my objectivity, I am also linking to the study by Huang et al, which I believe to be the strongest published evidence supporting active surveillance for MRSA.
I won’t dignify the criticisms of the NIH-funded RCT of active surveillance with a detailed response. That study isn’t published yet! When it is, we’ll have plenty of time to debate the methods and the findings. The fact that some people began criticizing this study pre-emptively says more about their data-proof fervor for active surveillance than it does about the study.
The entire last half of their comment proves my point about poorly conceived, poorly designed studies of active surveillance. The last paragraph of my post was (I thought!) an obviously tongue-in-cheek parody of the kind of study Farr or Jarvis might cite to support active surveillance, if the numbers were reversed. Of course it is wrong to claim that the difference in publicly reported BSI rates at U.Va. and MCV is related to their respective approaches to MRSA control! And Farr and Jarvis needn’t convince me that U. Va. is a wonderful hospital—I'm proud to say that I did my residency training there, I'd be happy to receive care there (provided I'm shown to be MRSA-free), or to work there (OK, maybe that last part is out of the question now...).
It’s quite simple: (1) the data to support use of active surveillance/contact precautions to reduce morbidity and mortality from MDROs, in the absence of an outbreak, remain inconclusive, and (2) the burden of proof is on those who wish to force hospitals to adopt this approach to the control of MRSA.
Our professional societies (SHEA and APIC) have already weighed in: we shouldn’t be mandating active surveillance.
Partners HealthCare, which includes Massachusetts General Hospital and Brigham and Women’s Hospital, is implementing a new policy to better regulate the ties that their physicians have with industry. The new rules forbid free lunches but allow for companies to fund educational events with oversight. Participation of physicians in speakers bureaus will also be restricted. The Globe notes that the new policy is not as restrictive as the policy at Stanford, and some critics cited in the article argue that the new policy still allows too much industry influence.
The other article focuses on an outbreak of MRSA infections that involves nearly 40 newborns and mothers over the past several months at Beth Israel Deaconess Hospital. The infections were primarily skin and breast infections, and the article implies but doesn’t directly state that the infecting strains were community acquired strains (CA-MRSA) that may have been transmitted in the hospital. The article does point out, however, that in some cases transmission of the organism may have occurred after hospital discharge since the mean incubation period was 12 days post-discharge. Given the increasing prevalence of MRSA in the community, introduction of the organism into hospitals from healthcare workers, visitors, and patients colonized or infected prior to admission poses a very difficult problem for hospitals. The greatest hope we have for controlling this is scrupulous attention to hand hygiene and environmental decontamination.
Of note, Paul Levy, the CEO of Beth Israel Deaconess, wrote about the problem in his blog yesterday. He’s the gold standard when it comes to transparency.
Thursday, April 9, 2009
I’ve been thinking a lot about whether sustained excellence in hand hygiene performance is achievable, and I wonder if, over time, we’ll see views on this change just as they did about the “preventable fraction” of healthcare-associated infections. I’ve never been a big fan of “get to zero” language (for several reasons--the pathophysiology of some infections is not amenable to available preventive approaches, and our diagnostic tests and definitions are not perfect). But I’ve never understood how someone could push a “zero” message but remain nihilistic about hand hygiene adherence. If 60-70% really is the best we can do in hand hygiene adherence, we’ll never eliminate healthcare associated infections! So I think elimination paradigms must assume that sustained excellence in hand hygiene can be achieved.
The other thing I’ve noted is that our hand hygiene improvement coincides with improvement in several other areas—for example, over 85% of our staff accepted flu vaccination this season, we’ve gone months without a VAP or CLABSI, etc. If I thought I knew what led to this “tipping point”, I’d tell you. The major facilitator, if I had to guess, has been increasing administrative engagement and resource allocation (examples include the provision of a robust reward system for units achieving high flu vaccination rates, and hiring of a dedicated hand hygiene observer).
Wednesday, April 8, 2009
Tuesday, April 7, 2009
An obvious example? Legislative mandates for active MRSA screening, and the Veterans Affairs system-wide directive to screen all admissions for MRSA. Given the lack of scientific consensus on the effectiveness of active MRSA screening approaches, these mandates amount to a huge natural experiment, with every patient admitted to the involved hospitals becoming a study subject. Worse yet, the conclusion is foregone—as invasive MRSA infection rates fall for other reasons, those states or healthcare systems that enforced screening mandates will claim that this intervention was key to their success.
This would all be fine if active screening for MDROs such as MRSA, combined with increased use of isolation precautions for asymptomatic carriers, had no associated risks. Alas, this is certainly not the case. It’s not just bad public policy, it’s harmful and wrong.
Monday, April 6, 2009
While active surveillance proponents could argue that some of these healthcare workers may have avoided colonization had all colonized patients been identified and isolated, the avoidable proportion is unknown. However, even in hospitals with aggressive active surveillance programs, unless patients are isolated until proven to not be colonized, the possibility of transmission could still occur prior to the institution of contact precautions. And just because a patient is in contact precautions does not mean that the potential for transmission to the HCW is eliminated. Thus, on many levels the active surveillance-contact precautions approach falls short.
As the AIDS epidemic unfolded in the mid-1980s, CDC wisely never recommended that all patients be tested for HIV and precautions applied to only those patients who were infected. Rather, assuming that all patients could be infected was the foundation for universal precautions, an enormously successful public health intervention. We should use similar thinking in our approach to multidrug resistant organisms. Rather than culturing patients for every conceivable MDRO, focusing on excellent infection control practices in the care of all patients (the so-called “horizontal” approach to infection control) reduces the potential for transmission of all organisms that are transmitted via direct or indirect contact.
The increasing proportion of MRSA colonized HCWs should serve as a reminder that colonized patients are not the only source of MDROs in hospitals. Given the rising prevalence of MRSA in the community, transmission of MRSA in the hospital will increasingly occur from colonized HCWs and visitors. This is just another reason that we will never be able to culture our way out of this problem.
Saturday, April 4, 2009
Thursday, April 2, 2009
In the Virginia report each hospital’s individual ICU CLABSI rates for July through December 2008 are rolled up into a composite rate. This was done because some hospitals have a single combined medical-surgical ICU, while others have multiple specialized units, such as burn and neurosurgery. While it gives a better overall picture of the situation at each hospital, the downside is that the rates cannot be compared to CDC benchmarks.
There are many unanswered questions regarding publicly reported HAI data in addition to validity. How can the data be risk adjusted so that hospitals that care for the sickest patients are not penalized? Will consumers use the data to choose their venue of care? And even if they want to change hospitals based on the data provided, are they able to given the control exerted by third party payers? Will hospitals with high rates attempt to game the system to make their rates appear lower? Will public reporting increase inappropriate antibiotic use?
Despite the difficulties, the principles that underlie public reporting, transparency and accountability, are so vital to maintaining trust with our patients that the onus is on us in healthcare epidemiology to continue to improve our methods so that validity can be maximized.
While most of us are concerned first with adequate staffing of our infection control programs, hiring freezes and personnel shortages in other areas will probably have greater consequences for infection prevention. Three obvious areas are in nursing, housekeeping, and clinical laboratory support. The literature linking nurse-patient ratios to infection risk is convincing, data keep accumulating about the critical role of environmental disinfection in infection prevention, and our labs serve as the most important source of surveillance data to guide prevention efforts.
So we are at risk for a double (triple? quadruple?) whammy—loss of capacity to monitor outcomes and processes of care, and the loss of front line personnel who are so important to infection prevention. We need to be at the table during these discussions, to help hospital administrators make difficult budgetary decisions without compromising patient safety.