Understanding the CLSI M100 Document
The CLSI M100 Document is the reference standard for antimicrobial susceptibility testing in clinical microbiology labs across the US and much of the world. It contains the interpreted category tables, quality control ranges, and testing methodology that labs depend on every single day. The document gets updated annually, and the changes matter. If your lab is relying on an outdated edition, you are probably interpreting results incorrectly without realizing it. I have seen this happen repeatedly with cefepime and Enterobacterales. The 2024 edition changed the interpretive criteria for several drug-organism combinations, and a handful of labs I consulted with were still using the 2023 print copy well into the year. They were misclassifying resistant isolates as susceptible on a few key agents until someone actually did a side-by-side comparison. You can download a digital subscription from the CLSI website. The print version is also available but most people end up paying for the subscription because it includes the mobile-friendly tables and automatic annual updates. A subscription runs somewhere around $700 to $900 depending on whether you get the base package or the bundled add-ons like M02 and M07. Some hospitals negotiate institutional pricing that brings it down significantly. If you are running a small reference lab on a tight budget, check whether your state or regional public health laboratory offers shared access. I worked at a lab where we split a single subscription with two nearby hospitals and it worked fine since nobody needed simultaneous access to the tables at the same time. Here is what the document actually contains. The main tables section is what most people look at first. It has interpretive criteria for every combination of organism and antimicrobial agent tested by disk diffusion, broth microdilution, and gradient diffusion methods. There are also quality control range tables for the standard reference strains. The methodology section covers how to actually perform the tests, though many labs treat M02 and M07 as the primary references for that. The SDD approach section has been expanded in recent editions with dose-dependent organism-drug pairings, which is something most people still find confusing when they first encounter it.
I ran into a real problem last year with Pseudomonas aeruginosa and ceftazidime. The isolate showed intermediate zone diameters right on the border, and the old table I had memorized did not match the current edition. I spent about twenty minutes digging through the SDD dosing recommendations and the breakpoint explanations before I realized the isolate actually met the criteria for dose-dependent susceptibility with high-dose therapy. The lab was about to report it as resistant and the clinicians would have lost an option we knew was still viable. The fix was straightforward once I confirmed the correct edition and checked the dosing tables, but it was a close call. It took maybe five minutes of actual lookup time, but the consequences of getting it wrong would have been worse. One thing beginners consistently miss is that the QC ranges in M100 are edition-specific. You cannot use 2023 quality control ranges with 2024 interpretive criteria and expect accurate results. The ranges shift slightly each year, usually because organism-drug pairings get reclassified or new data becomes available. Your lab needs to update both the interpretive tables and the QC tables together at the beginning of each calendar year. I have seen two different labs make this exact mistake in the same month, each blaming a reagent lot change when the real issue was a stale edition number. Another nuance that people overlook involves extended-spectrum beta-lactamase organisms and certain cephalosporin breakpoints. The M100 tables sometimes show different interpretive criteria depending on whether the organism is an Enterobacterales species or Pseudomonas. A result that reads susceptible in one genus might read resistant in the other, even when the zone diameter is identical. This is not an error. It is deliberate, based on pharmacokinetic and clinical outcome data, but it trips up anyone who assumes the tables work uniformly across all organisms.
The document does have limitations. The breakpoint tables reflect clinical and pharmacological data as of the publication date, which means they are always a few years behind newly approved agents. If a drug was approved in late 2025, you will not see susceptibility criteria for it until the 2027 edition comes out, assuming CLSI decides to include it. EUCAST sometimes publishes breakpoints for newer agents faster than CLSI does, and that creates a gap that reference labs notice immediately. There is also the question of whether CLSI breakpoints align with FDA-approved labeling. They do not always match, and when they diverge, hospitals have to decide which standard they follow for reporting. Most clinical labs follow CLSI, but ID physicians sometimes push back when the two disagree. If you need to access the tables quickly in a busy lab setting, the CLSI mobile app or the online portal is significantly faster than flipping through a printed book. I keep the mobile version bookmarked on the lab computer, and it loads the interpretive tables in about three seconds. Printing tables from the PDF works too but you will waste a lot of paper and the search function is useless compared to the web version. The subscription also lets you export tables to Excel if you need to build custom reports, which saves a lot of manual copying when you are preparing accreditation materials. I would recommend keeping a current edition bookmarked and visible for anyone in your lab who touches susceptibility results. Print out the QC tables and post them at the bench. Make sure the edition year is clearly marked on whatever copy you use. It sounds trivial but it is one of those things that quietly causes problems until something goes wrong and then nobody can figure out which edition was actually being used when the error happened.
Get the Full Details
