How to protect effective antibiotics: a conversation with Doc Ricky


I believe that the rapid spread of antibiotic resistance should be recognized as an urgent public health priority – possibly the #1 priority. E.g., CRE [1].

After reading Betsy McCaughey [2] “U.S. Lacks Will to Fight Superbugs” [3] I tweeted the citation for her op-ed

CRE the “nightmare bacteria”: U.S. Lacks Will to Fight Superbugs

Shortly Doc Ricky replied:

@stevedarden I’m kind of perplexed by the media repetition of “fighting superbugs” – what do ppl expect anyway? Some secret weapon?

What an excellent framing question! I replied with some suggestions:

  1. Transparency of hospital performance on sanitation standards.
  2. No excuses policy on resistance cases e.g. Israel
  3. Strict limitation of agriculture use to disease, no routine NTA dosing
  4. Transparency on physician prescribing by doctor

Shortly Doc Ricky replied:

“@drricky: @stevedarden but problem is most of these are preventative, what is expected when MDR {Multi Drug Resistance} is detected?”

I replied over several tweets: By #2. what I meant is that the CDC publishes “best practice” on procedures to execute upon every identified case of MDR[4]– beginning with effective quarantine and decontamination. The “best practice” level of response is mandated to be the minimum response. It should be the top priority of the hospital to eliminate the detected microbe from the institution. I appreciate that is a statistical goal, as we have no way to validate that “we killed it”.

For examples of such best practices consider the 2011 NIH Clinical Center response to a CRE outbreak [5], and of Israel 2006 (from Betsy McCaughey):

When CRE invaded Israel’s hospitals in 2006, public health authorities launched a military-style campaign requiring reports from all hospitals, which were ordered to test patients and undergo rigorous cleaning efforts. This reduced CRE by 70 percent in one year. Israeli researchers just announced a drug that may protect patients exposed to CRE from becoming infected.

My personal bias is that regulation is a blunt and ineffective tool in complex, fast-changing domains like this one. My question: How to incentivize hospitals to succeed?

Suggestion: first try transparency. E.g., if the Johns Hopkins data, such as MDR cases, hand hygiene and infection-control scores are published on the web every month – that is a powerful incentive to improve – to be ranked among the very best institutions globally.

Meanwhile Doc Ricky tweeted a critique of my first try on agriculture:

@stevedarden The agriculture issue is more nuanced than that, after all, how does one exactly limit the use?


@drricky Legislating detailed Rx rules not practical. How about transparency of farm usage per animal-KG? Is public shame effective?

I am thinking of the Denmark experience beginning 1999 where they succeeded to eliminate NTA use in agriculture. See my 2010 Denmark: results of stopping NTA (non-therapeutic antimicrobials)

Doc Ricky moves the discussion to the next level, biology:

@stevedarden only skirts around the real problem, which is biology. We culture animals with similar physiology to ours

The microbes shared by humans and pigs, chickens, beef are why we are so concerned about agricultural applications of antibiotics. Agriculture uses roughly 80% of the antibiotics effective in the human population – but in vastly larger quantities. If we were all vegetarians that would eliminate the whole worry about agriculture.

Doc Ricky is truly the expert in this topic – I’m looking forward to learning from him. We agreed to shift the conversation from Twitter to a long-form-friendly fora.

@stevedarden clearly a complicated topic, and hope you’ll continue to discuss.


  1. The CDC on CRE Carbapenem-Resistant Enterobacteriaceae.  ↩

  2. Betsy McCaughey founder of Committee to Reduce Infection Deaths  ↩

  3.  (…snip…) CRE was first uncovered in North Carolina in 1999. By 2008, it had spread to 24 states and was “routinely” seen in certain New York and New Jersey hospitals. But hospitals kept quiet. Now it’s in at least 43 states.
    (…snip…) Two months ago at a press conference, CDC Director Thomas Frieden dubbed CRE the “nightmare bacteria,” warning that “without urgent action now,” superbugs like CRE will prevent patients from getting joint replacements, cancer therapy and other treatments. The risk of incurable infections will make these treatments too dangerous. Yet, where’s the urgent action?
    The CDC doesn’t even have accurate data on how many CRE infections are occurring and where, because according to the director of the CDC’s Office of Antimicrobial Resistance, Steven Solomon, the government agency has never reached out to state officials to make CRE a reportable disease. Only 12 states require hospitals to report cases. Astoundingly, New York State did not require reports until July 2013, despite CRE menacing some of its hospitals for a decade.  ↩

  4. I am using the shorthand MDR to represent all the emerging multi-drug resistant microbes.  ↩

  5. This is what happened at the National Institutes of Health Clinical Center in Maryland in 2011. A 43-year-old woman known to have CRE was admitted from a New York City hospital. The NIH treated her, using CDC infection-control precautions, but three weeks later, a male cancer patient who had had no contact with her came down with CRE. Week after week, more and more patients contracted the infection introduced by the New York woman. Six of those patients ultimately died, one of whom was a 16-year-old boy. To stop the outbreak, NIH investigators double-cleaned rooms with bleach and misted hydrogen peroxide in measures far beyond what the CDC recommends.  ↩

7 thoughts on “How to protect effective antibiotics: a conversation with Doc Ricky

  1. Thank you for summarizing the interaction thus far. Discussing it in blog form is probably more effective than twitter, and we can follow up the history here. We have many points to discuss, but I would like to clarify one point first, which is what elicited my question to begin with: not to conflate the preventative (“protecting effective antibiotics”) from the reactive (“fighting superbugs”). I don’t disagree that preventative measures are sorely lacking – although perhaps calling them preventative is promising the wrong thing, since I don’t think we have evidence that they prevent antibiotic resistance more than just delay it. Many of the practices lumped into this issue stem from improvements of general infection control practices – that is, if we can improve total microbial control, we can minimize infection by antibiotic resistant bacteria.

    That said, I’d like to start by expanding my original question: while the media is aware of, and drums the mantra of preventative measures, worrying about that AFTER the spread has begun is hardly called “fighting”. Mind you, even the very diagnosis of MDR is no small feat – it is usually an operational definition. Some say at least two different classes of antibiotics, some say three – and the cutoffs for what is resistant and susceptible follows standards that are revised periodically, so what is considered resistant today may be susceptible by the standards in 5 years time.

  2. Thanks. I agree on all your points.

    calling them preventative is promising the wrong thing, since I don’t think we have evidence that they prevent antibiotic resistance more than just delay it.

    I think all the infectious disease people would underscore your point. Even at t-zero, when a new drug is released, there are likely already some microbes with some defenses for some of the pathways. Effective drugs promote selection of defensive traits. So “preventative” means “achieve a slower rate of loss of effectiveness.”

    Is it useful to think about this challenge broadly on two levels?

    1. What are the “switches” that we most urgently need to switch from FAIL (current setting) to SUCCESS POSSIBLE?
    2. For those, what are the process and behavior changes that are most likely to move the “switch”?

    For discussion I’ll nominate a step change in the rate of clinical release of new antimicrobials. My rationale is simply that unless we initiate this change pronto, then we have accepted defeat. And even if we achieve big increases in discovery the availability to clinicians is two decades out:

    1. Resistance is spreading faster than new drugs are discovered.
    2. Current time to release of a new molecule is 10 to 20 years.
    3. Current economic incentives for new drug discovery are mostly set to FAIL, because
    • An effective new drug will be put in the case labeled “break glass only in case of Presidents or Princes”, so no profits there.
    • Regulatory approval switches are mostly set to CAUTION, or “no thalidomide on my watch”.
    • The difficulty of discovery is much harder than we thought – consider the depressing experience of GlaxoSmithKline.

    I am not asking how to make this happen, rather is this one of the three-most-urgent switches to throw?

  3. We all need to reinspect the premise of the question of switches – it presupposes that there is just one break in the system, the limited nodes of high degree that can take care of the bulk of the problem. But it may not be the case – as we discuss, I hope to lay out the interconnected web of systems that antibiotics lie in, and the mismatch to what we expect of that web. So, let’s first address the issue of increasing the rate of clinical antimicrobials. First – clinical drugs of any sort take a long lead time to develop to market because of the gauntlet of tests required, from human safety to predictable efficacy. What we come to expect of any new antibiotic class is a very limited set of human side effects, broad spectrum effectiveness against microbes, and low cost. Even assuming there is a infinite set of possible compounds to meet these criteria (and we have no way of proving that they indeed exist), incentivizing the efforts to do so is a challenge to the private sector. Specially superficial changes in regulation involving approval is only effective if there is already a cadre of maturing drugs in the pipeline. And when it comes to antibiotics, that may not be the case.

    The common interpretation of antibiotics, and guidelines for its use, are hampered by an overgeneralized understanding microbiology. Even as we now are uncovering the vast array of possible commensal microbes – well, maybe commensal is too narrow a term, as the very ecology of microscopic organisms may be part and parcel of the human species – then, too, we must grapple with the blurry border between microbial friend or foe. Nonetheless, our antibiotic arsenal, borrowed and stolen usually from different microorganisms, are chosen and applied in the paranoid idea that all microbes are pathogenic. In short, our present crop and use of antibiotics are very blunt instruments, and that is the first problem. And we’ve used them as such – if we can make them more refined in use (and that doesn’t necessarily mean discovering new compounds, it may involve reformulating the protocols we use the current ones), we could save lives. But it would mean a dramatic increase in diagnostic tools, and reeducation of current clinical practices. Given how entrenched the systems are at the moment, no proposed change will come without a significant cost. And almost no one is willing to pay it.

    • I apologize for missing your comment (we lost our 3G internet for 1.5 days so it isn’t intentional rudeness). I need to think about your points before replying.

      Question: Does the GlaxoSmithKline report illuminate your points about refining the way we design and apply antibiotics?

    • (…snip…) if we can make them more refined in use (and that doesn’t necessarily mean discovering new compounds, it may involve reformulating the protocols we use the current ones)

      If you could cite an example or two of reformulating protocol – it would clarify for me the scope of what you have in mind.

    • We all need to reinspect the premise of the question of switches

      Possibly a bad choice of metaphor. It is also symptomatic of the analytical habits of an engineer-manager type. Which is first to attempt to prioritize where to expend energy. So instead let's go with your insights on where to focus.

      Given how entrenched the systems are at the moment, no proposed change will come without a significant cost. And almost no one is willing to pay it.

      Humans seem to have a powerful status-quo bias, and bureaucratic institutions even more so. Like antibiotics select for resistance traits, regulations select for rent-seekers. But there are techniques that can create new incentives that just bypass the vested interests. Paul Romer's Charter Cities strategy is a good example. Science prizes may be an example of bypassing the R&D bureaucracy.

      I don't have anything to add to your points. Question: is NCATS (National Center for Advancing Translational Sciences) relevant to the needed process changes? I've heard Francis Collins talk about it but don't know what contributions are emerging from this new NIH center.

  4. From a physician’s comments via email:

    (…snip…) I have two suggestions regarding outpatient over-prescription if anyone, say Kaiser, were to get serious about this problem. Review practitioners’ charts and when clear patterns of over-use are noted educate and/ or fine the practitioner until the behavior changes. This is particularly true of PAs and NPs.  Secondly, let it be known from the top levels of management down to the practitioners that if a patient complains because he did not receive an unecessary antibiotic that the management will support the practitioner and not the patient. That support would include educating the patient about inappropriate antibiotic usage and telling the patient that they were fortunate to have a physician who doesn’t over-treat them.As it is now, practitioners are sanctioned in some way for patient complaints, no matter how unjustified, but are not disciplined for inappropriate antibiotic use. Basic behavioral theory: If you want more of something, reward it.

    Your outpatient reforms would surely improve the appropriate dispensing. Can you imagine a path that would strongly motivate institutions to focus on such reforms?

    I keep orbiting back to transparency as incentive – because the management pyramid doesn’t like negative media coverage. And hospital managements evidently think they are competing (even if not).

    What about Dr. Private? Does he care if the NIH or FDA website shows that he is in the top quartile of “antibiotics per patient” or some more useful measure? How could we present that kind of data to patients to educate them that those Rx they are demanding have a heavy cost?

    As to patient education, that sure looks like classic “tragedy of the commons” logic. Immediate gratification (“it might help!”) vs. vague future cost – largely social, no obvious check-comes-due for the individual patient.

    (…snip…) about half of antibiotics prescribed in hospitals are inappropriate in that they are either prescribed when no antibiotic was necessary or that the wrong antibiotic was chosen.

    You’ve explained the perverse incentives for front-line physicians, PAs, etc. But those incentives don’t apply in the hospital setting do they?

Comments are closed.