“The end of Clinical Breakpoints as we know them?…””


During the ASM conference in Boston, I attended a couple of talks on the process of anti-microbial susceptibility testing (AST). There are really only two main players (a duopoly of sorts) left standing with regards to setting standards for AST. These are the Clinical & Laboratory Standards Institute (CLSI) and the EUropean Committee on Anti-microbial Susceptibility Testing (EUCAST).

Representatives from each organisation spoke during the conference sessions. Reading between the lines,  I did not get the impression that a unification of the two organisations was imminent, mainly because I did not detect any warmth between the respective speakers…

There was considerable discussion on the setting of Clinical Breakpoints. Clinical Breakpoints are necessarily a “messy” compromise between Epidemiological Cut-off values (ECOFFs), pharmaco-kinetic & pharmaco-dynamic (PK-PD) data, and most importantly, clinical response. As long as this model continues, EUCAST and CLSI committees will continue to tinker with and argue about the breakpoints, but in the painful knowledge that a single value can never be agreed upon as being a definitive answer. There are just too many variables.

The problem is therefore ripe for an app-based solution…

I foresee a future in anti-microbial susceptibility testing where the main variables are taken much more into account than they currently are.

I see the time when all the microbiology laboratory will do will be to provide an organism identification and  MIC values for a restricted number of key antibiotics.

The clinician will then take his smartphone app, enter this lab information, but also add in key patient variables such as site of infection, renal function & body mass index, and degree of immunocompromise.

The app will then process all these variables together and give for each antibiotic the likelihood of clinical response of a certain antibiotic, not only for this particular organism, but for this particular patient.

And thus there will no longer be a need for concrete/single value clinical breakpoints, but rather the focus will be on personalising the lab information to each particular patient.

I can see such apps for MIC interpretation being developed in the next few years. 

And if I were a lead member of NCCLS or EUCAST I would be recruiting the services of a good IT applications specialist on to the steering committee, and sooner rather than later…



12 thoughts on ““The end of Clinical Breakpoints as we know them?…””

  1. thoughtful post again Michael, welcome back. Maybe this approach could be combined with advanced micro fluidics and next gen sequencing we might improve our prediction of response. But even still in think that agreement through commitee is going to be needed. Like resistance genotype a for HIV

  2. This is very interesting for me to consider, not solely as a GP but as one who in another role teaches a 100 level Microbiology course ( a pre-requisite for a Registered Nursing programme.)

    In the US it is more and more the Nurse Practitioner or Physician Assistant interpreting such test results and using algorithms to determine treatment (albeit algorithms provided by doctors)

    But, a smartphone app like you predict, Michael, would involve algorithms too. What is it worth a clinician knowing in terms of microbiology theory behind this kind of algorithmic “thinking”? Theory for responsible antibiotic stewardship surely. What is “good clinical judgement”? Could an algorithm be better for responsible prescribing (or non prescribing)? When should a clinician decide differently than an algorithm / app might suggest?

    How differently could teachers like me be teaching our clinicians of the future about antibiotic sensitivity testing and antibiotic prescribing?

    Looking forward to the future!

    1. Hi Geraldine. I think there will always be room for individual choice, and cliniciain input into these decisions. Very hard to protocolise, standardise and objectivise everything, and nor should it be…

  3. CLSI is bought and paid for by industry and academia, neither of which knows how to treat patients or consider the consequences of their inactions. At least EUCAST is transparent and willing to consider multiple points of view.

    1. Thanks for your comment David. I think things are going to change quite rapidly in this area. in the next few years, and I suspect the one who adapts the most quickly will survive.

  4. Posted on Behalf of Dr. Jean Patel, Chairholder and Dr. Mel Weinstein, Vice Chairholder of the CLSI Subcommittee on Antimicrobial Susceptibility Testing:

    Contrary to the comment that CLSI is “bought and paid for by industry and academia”, we can attest to the fact that CLSI has open meetings and strong COI policies. If Mr. Smith or Michael would ever attend a meeting of the CLSI Subcommittee on Antimicrobial Susceptibility Testing, they would find that multiple points of view are expressed and heard, committee votes are open and transparent to all, and voting members on the losing side are invited to indicate the reasons and rationale for their votes. Participants from industry or individuals who derive a major portion of their income from industry sources are not permitted to serve as voting members of the Subcommittee. Compared to other breakpoint setting organizations, CLSI’s commitment to transparency is unparalleled. Microbiologists, physicians, and other scientists who work on setting breakpoints, no matter what organization, are motivated to improve patient care and the treatment of infectious diseases. We continue to hope that there will be more collaboration between organizations in the future.

    1. Many thanks for your comments Tracy, much appreciated. I am glad you have posted this because there has been a couple of comments on this particular post that have been a bit offensive. That is not the point of this website. Dogma and current thinking can be challenged, but in a respectful way. I will block anybody who continues to cause offence.

  5. Dr dharmadra Sharma and I, during our last months in ras al khaimah developed an axel based program and fed in the max, t 1/2 as well as auc of different antibiotics as per label approved by fda. The user needs to select the antibiotic, the dose and frequency of administration and the software would calculate the three Pk PD parameters associated with antibiotic action and successful therapy. We offered the excel based software free for use by interested scientists with no strings attached.

    1. I would be very interested in looking at this Ashok. I will send you an email in the next couple of days with regards to this. Many thanks, much appreciated.

      1. i can be reached at drashokrattan@gmail.com Both Dr Dharmedra Sharma and I had decided that we will share the excel program with interested scientists without any strings attached. You are free to use it any way you want and add or delete as you deem fit.

Leave a Reply

Your email address will not be published. Required fields are marked *