Monday, May 2, 2016

The importance of considering time when evaluating risks of base jumping (and maybe even antibiotics)

This is a guest post by L. Silvia Munoz-Price, MD, PhD. Associate Professor of Medicine at the Medical College of Wisconsin. Enterprise Epidemiologist at Froedtert Health. Milwaukee.

Over a month ago, Eli asked me to write this piece to discuss my recent CID paper on handling time dependent variables. I knew this had to be done with an analogy but after several weeks of mulling over this, I was still uncertain on how to colloquially explain this concept to you. So, as I was almost ready to forget about this post while on a plane to Miami, I had sudden inspiration as I was about to nap! I really hope this helps everyone understand this statistical concept. If not, then I'm not sure reading the CID paper will help you much either…stick to 2x2 tables (sorry!!).

Setting: Let’s imagine Eli and his wife invited my hubby and me to go base jumping in New Zealand for a week (See figure…Eli take note!!). So, now let’s observe our jumping habits: of course, I would jump once and be done with it for lifetime. My hubby would probably not jump at all and just enjoy watching crazy people jump. Let’s say Eli decides to jump every day (two days he jumps twice!) and his wife jumps three consecutive days.

Study design: Ok. Not to be morbid, but the easiest outcome to evaluate is mortality (binary variable; 1: dead or 0:alive) by the end of the vacation. The exposure variable of interest is base jumping. 

Option 1: The easiest way to look at this association is to construct a 2x2 table: Did you jump? (yes/no) Did you die? (yes/no). See, the problem with this analysis is that it ignores the intensity of the exposure as Eli, his wife, and I would be considered as a “yes” and only my husband would be a “no”. But, is it reasonable to analyze the exposures for Eli, his wife and I the same way? Intuitively, we probably could say no.

Option 2: A tad more elaborate way to look at this would be to count the number of jumps per person and enter these numbers in the analysis. So, Eli would have 9, his wife would have 3, I would have 1, and my husband would have 0. What is the problem with this approach? Well, it completely disregards time of exposures, correct? It is like having all those jumps in only one day. We need to ask: when was “that” day that all those exposures got summed? Was it at the beginning of the week or towards the end? Did the outcome happen at the beginning of the week, in the middle or at the end? Is it reasonable to analyze all those jumps clustered in time within a single day? Intuitively, I would say no. A similar problem happens with number of days that jumps occurred, especially for me. When did my one jump happen (at the beginning of the trip or towards the end?).

Option 3: A more elaborate way to determine the association between jumping and mortality is to account for the richness of the exposures. Not just taking into account the specific days the jumps occurred, but how many jumps occurred each day and from which different altitudes these jumps took place. Then we can calculate the hazard of dying on a daily basis based on the previous 24 hours of jumps. Let’s go over this a bit further. The hazard on day 1 would be calculated using 3 people. Assuming we all survived, on day 2 the hazard would be calculated only among the people that jumped (2). On day 3, assuming we all survived, the hazard would be calculated again only among the people who jumped that day (2 jumps). On day 4, assuming we all survived the hazard would be calculated only among the people who jumped (1). If any of us were to arrive to the outcome during the observation, then that person would be removed from the analysis. This is the concept of time dependent exposures. You measure the outcome as the exposure occurs over time. This is in contrast to what we usually do in our hospital epi studies: exposure treated as binary variable (yes/no) or exposure treated as number of days exposed (9 or 3 or 1) or even as number of jumps performed. More concerning, the outcome on the latter examples is fixed towards the end of the observation rather than measured as time progresses.

Bringing it home: ANTIBIOTIC EXPOSURES. Antibiotics are such rich exposures. Think about it. They can be given during many different days throughout the hospital stay and there are many types of antibiotics, with various doses and routes. Outcome variables, such as acquiring a multidrug resistant organism or even developing an infection by this organism also vary in time during hospitalization. Is it reasonable to analyze all those antibiotic exposures clustered in time within a single day or even worse as binary variables? Is it optimal to fix the outcome variable as happening at the end of hospitalization? Intuitively, I would say no to both. There are a couple of examples in the ID literature that compare these analyses. One of them by my co-author Marc Bonten. However, specifically for antibiotics it is not fully clear to me if the associations found would justify the additional cost and time of obtaining all this rich information about exposures and outcomes (note: think about relooking at your cohort datasets using this method).

Let’s end this post here [so that I can take a quick nap before landing] and see the feedback I get with this example. If the feedback is good then I will explain the biases that can occur by not accounting for time in your analyses, and maybe go over delayed effect of antibiotics. In the meantime, I will be sipping a mojito with my hubby while enjoying Miami. Salud!

Sunday, May 1, 2016

The delusion continues (part 3)

In response to my last post a reader emailed me the following: Very easy to sit at a keyboard and throw blog bombs...  I would be thrilled to hear your constructive suggestions for a solution(s).

Fair enough. I'll address that. But it's important, I think, to first say a few words about this blog, which is now in its 8th year. From the beginning, we wanted to make controversial issues a focal point, and the issue of the ID workforce (or lack thereof) is controversial and a topic of great interest to readers. In addition, we welcome comments and guest blog posts to offer alternative viewpoints. Eli, Dan, and I don't always agree with each other (as is evident in our posts). The only comments that are censored are those advertising black market erectile dysfunction drugs and other products. And all requests for guest posts have been honored unless the author has conflicts of interest with industry. So readers, please feel free to respond to our posts.

My comments on the workforce/compensation issue and IDSA's response are made in the context of my experience with these issues. In my former job as an infectious diseases division chief in an academic medical center, I had firsthand experience with the difficulties of recruiting fellows and faculty, the inequities that resulted from a purely RVU-based compensation plan, and the toll this took on teaching and morale. At the same time, I was observing a private health system across town crank through a multitude of infectious diseases doctors, each of whom left practice once their guaranteed salary expired and they one by one came to the realization that they couldn't generate enough RVUs to maintain their salaries. Several of these physicians became hospitalists. In my current position, I see my division chief struggling with trying to balance his budget, offer salaries that can compete with other hospitals and medical schools, deal with ever increasing consultation volumes and expectations for rapid responses to consult requests, while trying to minimize the stress all of this has on his fellows and faculty members. We now have starting salaries for brand new nurse practitioners that are within a few thousand dollars of junior ID faculty salaries. I'll be the first to admit that my experience may not be the same as others. In the IDSA compensation survey, one respondent reported a salary of $1.45 million, so obviously his situation is quite different than mine and his views on these issues probably are as well.

I did a little more research on salaries by looking at the AAMC data. The median salary for an infectious disease assistant professor is $152,000, while the median for a hospitalist assistant professor is $207,000. For a third year internal medicine resident, that's a huge difference. At the associate and full professor levels, hospitalists still earn more money than infectious diseases specialists. Moreover, hospitalists salaries are rising yearly at a higher percentage than ID's, so the difference continues to expand.

Another interesting finding is that of salaries for chairs of Departments of Internal Medicine. Unfortunately, if you're an infectious diseases doctor you'll earn significantly less than your chair peer who's an invasive cardiologist, a difference of about $350,000. And what do cardiologists learn in their fellowship about being a department chair that would explain that difference? I hate to sound like Donald Trump, but it's a rigged system. And it follows you throughout your career.

As for constructive suggestions for solutions, I've written about this in older posts, but here are a few:
  • Focus on the parity with hospitalists, since that's our biggest threat with regards to recruitment of residents into infectious diseases. Until ID salaries are at least as good as hospitalists', there's little reason to think that we will turn this around. 
  • Consider shortening the ID fellowship to positively affect the cost-benefit calculus of additional training. Do trainees who plan to enter private practice really need hands-on training in research or scholarly activities? 
  • Develop hybrid models of training to lessen the economic impact on trainees (for example, integrate ID training with hospitalist practice). Various models could be envisioned—such as one month hospitalist attending, alternating with one month ID fellowship. This would increase the fellow’s salary, and even if the total duration of training were extended, may entice more residents to consider ID training. Some would probably continue this model beyond training into employment. 
  • If IDSA is working hard to address these issues, it's not apparent from their website or communications with its members. Most importantly, in my view, IDSA needs to own the workforce issue and honestly deal with it. And that begins by calling it what it is--a crisis. A crisis, magnified by the many problems that are in the news every day, like Zika virus and antimicrobial resistance. I'm not a communications specialist, but it seems to me that these issues could be highlighted to help our cause. 
Unfortunately, the two articles and editorial published this week in IDSA's journals spin an unrealistic view of the problem. I doubt that the your-salary's-not-as-bad-as-you-think-it-is campaign will have much impact. Time will tell. In December, we'll see the results of the next Match. 

Saturday, April 30, 2016

The delusion continues (part 2)

A few hours after I posted yesterday on the survey of Internal Medicine residents on why they won't pursue training in Infectious Diseases, a reader alerted me to another new paper on ID physician salaries in Open Forum Infectious Diseases (free full text here with additional data available on the IDSA website requiring a membership password). This study was also sponsored by IDSA. The title of the paper is ID Physician Compensation--An Improved Perspective.

From this survey of nearly 1,900 ID physicians, we learn that the median salary of a full-time physician that is focused on patient care is $210,000. The data are sliced and diced in many ways, and are interesting to review.

The spin in the discussion (alluded to in the title of the paper) is that things are really not as bad as we thought. The authors point out that the often quoted Medscape survey has a small sample size and underestimates the true salary of ID docs. And while that is true, we still need to maintain perspective.

We know that our biggest competitor is hospital medicine. So let's take a look at hospitalist salaries. In 2013 (2 years before the IDSA survey), the overall median compensation for hospitalists was $254,000. So for an additional 2 years of training, the Infectious Diseases doctor will earn $44,000 less than his/her hospitalist colleague. And don't forget, full time for a hospitalist is, on average, 40 hours per week. How many ID doctors work 40 hours weekly?

Unfortunately, for most internal medicine residents, this decision is a no-brainer and I doubt this new survey will have much impact. Spin on, IDSA, spin on.

Friday, April 29, 2016

The delusion continues

There's a new paper in Clinical Infectious Diseases on the highly anticipated survey of Internal Medicine residents to solve the enigma of why no one is going into infectious diseases. It never seemed to be a mystery to me, but having data is always helpful, providing that you interpret it correctly. And here that caveat concerns me a great deal.

The survey results were divided into three groups: those applying or intending to apply to ID fellowship, those interested in ID but deciding not to apply, and those with no interest in ID. Of those who had an interest but didn't apply (the group that we should have some hopes of capturing), the number one reason for not going into ID was SALARY. Moreover, when asked what is the most important factor to increase interest in ID, all three groups said SALARY. For those going into ID and those who considered it, the percent responding salary was two-fold higher than the next most commonly cited factor (early exposure to the field of ID).

So no surprises here. But what was both surprising and alarming to me was that the discussion in the paper and the accompanying editorial both seemed to downplay salary as a factor in the current dearth of applicants to ID. The authors wax eloquently on career choice models, pedagogical techniques, the importance of mentors, etc. Salary is buried in half a paragraph of the eight paragraph discussion. The authors of the editorial even seem somewhat astonished that the top career choice of those who considered ID but didn't apply was general internal medicine, a specialty that they note "is not typically considered a high remuneration specialty." I think there's no surprise here either for two reasons: (1) hospitalists make significantly more money on a per hour basis than most infectious diseases doctors, particularly in the academic setting, and (2) whatever that difference in salaries is, it's magnified by the fact that two more years of training (at least) results in a lower salary. That is, you are punished economically for additional training, which many folks find too unpalatable to move beyond.

My bet is that the paper and editorial are nicely in line with IDSA's thinking since IDSA sponsored the study. And I suspect that IDSA will continue to pretend that all is well while the dumpster fire burns away. Once we get the microbiology courses in medical schools to stop making the students memorize so much, the students will come racing to ID!

Carry on then.

Thursday, April 28, 2016

More data support a common source for the M. chimaera outbreak

As we’ve suggested here and here, the information to this point strongly suggests that the M. chimaera outbreak linked to heater-cooler units (HCUs) is a “common source” outbreak, which has major implications for outbreak response. 

Another piece of the puzzle was published today by Haller and colleagues in Eurosurveillance. Read the whole thing for details of the German outbreak investigation, but the key additional findings are in the table above—brand new HCUs, and the water source at the manufacturing facility, grew M. chimaera. The genome-sequencing results are not included in this report, but read this key paragraph from the discussion below:
"Preliminary typing results indicate that the M. chimaera isolates detected by the authorities and the isolates from the manufacturer appear to be almost identical (unpublished data). The M. chimaera-positive environmental samples at the manufacturing site prompted the manufacturer to modify the manufacturing process, which now includes ethanol disinfection and an active drying of the HCU water circuit before shipment. ……According to the information provided by the manufacturer, HCUs manufactured before mid-August 2014 may have had environmental mycobacteria presence in the unit at the time of delivery [emphasis mine]. Our investigations could not elucidate if and until when contaminated HCUs may have been delivered to customers from this manufacturer." 
It is, of course, impossible to know for how long units were shipped “pre-contaminated” from this manufacturing site to users, but this now-published information only increases the rationale for removal of these HCUs from the operating room. Both Dutch and German authorities took this step, as the authors note. The US should as well—it shouldn’t take long to determine the impact on HCU function of extending the tubing sufficiently to allow this.

Wednesday, April 27, 2016

Here we go again: Active detection and isolation, C. difficile edition

An interesting study in JAMA Internal Medicine, likely to generate a lot of discussion, addresses the use of “active detection and isolation” (ADI) for control of C. difficile disease. This quasi-experimental, single-center study employed PCR screening (tcdB detection) of all patients admitted through the emergency department (patients admitted from other locations were excluded, as were “short stay” patients), and those that were found to carry toxigenic C. difficile were admitted into a kind of “quasi-isolation”—gloves were used, but not gowns or private rooms. So all-in-all, a very pragmatic (and somewhat idiosyncratic) intervention. Healthcare-associated C. difficile disease rates declined after the intervention, which students of prior quasi-experimental studies of ADI for MRSA and VRE will find unsurprising. 

At this point, I will outsource my blog post to Jon Otter and Martin Kiernan at the Reflections IPC blog. Go on, head over there for an excellent pro-con post about this study, and vote on the question posed at the end of the post. Then come back here to read my only additional observation…..I can wait (spoiler alert: I agreed with Jon).

OK, you’re back: the only thing I have to add to Jon and Martin’s excellent post is this: we’ve been here before. Recall the persuasive quasi-experimental studies (many single-center, some multicenter) of MRSA and/or VRE ADI published over the course of a couple decades. When better designed studies were eventually performed and published (e.g. STAR*ICU, REDUCE-MRSA, MOSAR, this one by Harbarth and colleagues that doesn't have a catchy acronym)—you know, studies that included concurrent control groups (control groups are for losers!), it became evident that ADI wasn’t the key to MRSA or VRE control. I think we’re headed down that road again, this time with C. difficile. Who’s going to step up and organize the multicenter, cluster-randomized trial we need to do now? Or perhaps better to ask: who is going to pay for it?

Tuesday, April 26, 2016

Peace, Love and Hand Hygiene

Sanjay Saint, Professor of Medicine at the University of Michigan and Chief of Medicine at the Ann Arbor VA Medical Center, discusses changing healthcare's culture through conformity, social learning (monkey see, monkey do) and mindfulness in a new TEDx talk. Worth a listen.