ESBL Detection: Screening, the Combined-Disk Confirmatory Test, and the AmpC Trap

A woman is admitted with a febrile urinary tract infection. The lab reports her E. coli as resistant to ceftriaxone, and the team switches to a carbapenem. It works. Three beds down, an identical E. coli UTI is reported as susceptible to ceftriaxone, so it is prescribed. The patient does not improve. The difference was not the organism. It was whether the lab looked past the routine result and asked one more question: is this cephalosporin resistance being caused by an extended-spectrum beta-lactamase? That question, and the simple plate test that answers it, is what this article is about.

What an ESBL is

An extended-spectrum beta-lactamase (ESBL) is an enzyme, produced mainly by Gram-negative bacteria of the family Enterobacterales (especially Escherichia coli and Klebsiella pneumoniae), that hydrolyzes the oxyimino-cephalosporins (cefotaxime, ceftazidime, ceftriaxone) and the monobactam aztreonam, and is inhibited by clavulanic acid. Two features define it, and both matter for detection: it defeats third-generation cephalosporins, and clavulanic acid restores their activity.

An ESBL does not effectively hydrolyze cephamycins (such as cefoxitin) or carbapenems. Most ESBLs are Ambler Class A, Bush-Jacoby group 2be; a less common subset are OXA-type (Class D). The full framework is covered in the beta-lactamase classification article.

Why ESBLs matter clinically

ESBL genes are plasmid-mediated, which drives both their spread and their danger. The same plasmid that carries the ESBL gene frequently carries resistance genes for other drug classes, aminoglycosides and fluoroquinolones among them. So an ESBL-producing organism often defeats not just the first-choice cephalosporin but the expected fallback as well, which is why ESBL infections escalate from routine to carbapenem-requiring so readily.

ESBL producers cause UTIs, bloodstream infections, pneumonia, and healthcare-associated infections, and they spread by contact with infected fluids, contaminated hands, or contaminated equipment such as catheters. This is why the detection question is not academic: a missed ESBL means a patient may receive a cephalosporin that will fail.

The two-step logic: screen, then confirm

ESBL detection is a screen-then-confirm sequence, and understanding why it has two steps is half the learning. Screening is sensitive but not specific: it flags anything suspicious. Confirmation is specific: it proves the suspicious result is actually due to an ESBL and not something else. A screen-positive isolate is a question; a confirm-positive isolate is an answer.

Step 1: Screening. Using standard disk diffusion, reduced zone sizes around an indicator cephalosporin flag possible ESBL production. With cefotaxime, a zone of ≤27 mm is suspicious; with ceftazidime, ≤22 mm. Because no single cephalosporin catches every ESBL, testing more than one indicator drug increases sensitivity. A suspicious screen does not diagnose an ESBL; it triggers confirmation.

Step 2: Confirmation, the combined-disk test. This is the core method. The principle is elegant: if resistance is caused by an ESBL, then adding clavulanic acid (which inhibits the ESBL) should restore the cephalosporin's activity, and the inhibition zone should grow. The test compares each cephalosporin alone against the same cephalosporin plus clavulanate:

• Cefotaxime (30 µg) vs. cefotaxime-clavulanate (30/10 µg)
• Ceftazidime (30 µg) vs. ceftazidime-clavulanate (30/10 µg)

Both pairs are tested because some ESBLs hydrolyze one cephalosporin far more than the other. After 16 to 18 hours at 35 ± 2 °C in ambient air, measure all four zones.

The reading rule (the single most exam-relevant number): a ≥5 mm increase in zone diameter for either cephalosporin in the presence of clavulanate, compared with that cephalosporin alone, confirms ESBL production. Only one of the two drug pairs needs to show the ≥5 mm jump.

Controls (do not skip these)

Every ESBL confirmatory run needs both controls, or the result is uninterpretable:

• Positive control: Klebsiella pneumoniae ATCC 700603 (a known ESBL producer, should show the ≥5 mm increase).
• Negative control: Escherichia coli ATCC 25922 (no ESBL, should show no significant increase).

The AmpC trap (why a real ESBL can read negative)

Clavulanic acid inhibits the ESBL, but it does not inhibit an AmpC enzyme. If an isolate co-produces both an ESBL and an AmpC, the AmpC keeps destroying the cephalosporin even after clavulanate has neutralized the ESBL, so the zone never grows by 5 mm and the confirmatory test reads falsely negative. In this situation, the ESBL is present but hidden.

A useful clue appears on the routine susceptibility plate. If the isolate is resistant to cefoxitin (a cephamycin), it suggests the presence of AmpC because ESBLs alone do not cause cefoxitin resistance. When AmpC is suspected, the ESBL confirmatory test can be performed on media containing cloxacillin, which inhibits AmpC. Another option is to use cefepime, an AmpC-stable cephalosporin, with and without clavulanate to reveal the hidden ESBL. This is a good example of how one enzyme can mask another, making it difficult for a routine automated report to detect the ESBL correctly.

How this fits the workflow

An ESBL result changes reporting. Historically, a confirmed ESBL prompted the lab to report all penicillins, cephalosporins, and aztreonam as resistant regardless of the individual zone sizes. Current CLSI guidance leans instead on using the revised cephalosporin breakpoints directly, but ESBL detection remains essential for infection control, for local resistance surveillance, and for the antibiogram that guides the next patient's empirical therapy.

How to Remember

The whole method in one image: clavulanate is the ESBL's off-switch. Flip the switch (add clavulanate) and if the zone jumps open by 5 mm or more, an ESBL was doing the damage. If the zone does not budge, either there is no ESBL, or something clavulanate cannot switch off (an AmpC) is also in the room.

The ≥5 mm rule, made sticky: picture the two disks side by side. The plain cephalosporin has a small clear zone (the enzyme is winning). Add clavulanate and the zone breathes out by at least a nail's width, about 5 mm. That breath is the ESBL confessing.

The AmpC trap in one line: clavulanate silences the ESBL but not the AmpC, so a co-producer keeps the zone shut and the ESBL hides. If cefoxitin is also resistant, suspect the trap.

Key exam facts in one table

Where Students Get Confused

"An ESBL hydrolyzes everything, so nothing works." No. ESBLs spare carbapenems and cephamycins. That is precisely why carbapenems are the reliable treatment and why cefoxitin susceptibility is a diagnostic clue.

"A negative confirmatory test means no ESBL." Not always. A co-produced AmpC can mask a real ESBL and give a false-negative. The cefoxitin clue and cloxacillin-containing media exist to catch exactly this.

"Screening positive means the isolate is an ESBL producer." No. Screening is sensitive, not specific. Many screen-positive isolates are not confirmed. The combined-disk test is what makes the call.

"Any zone increase with clavulanate counts." No. The threshold is a defined ≥5 mm increase, and it must be measured against the same cephalosporin tested alone, with valid controls on the plate.

"ESBLs are Class A, full stop." Mostly, but not entirely. The majority are Class A (group 2be); OXA-type ESBLs are Class D. The exception is worth knowing.

References

1. Clinical and Laboratory Standards Institute (CLSI). Performance Standards for Antimicrobial Susceptibility Testing. M100, current edition. Wayne, PA: CLSI.
2. Bush K, Jacoby GA. Updated functional classification of beta-lactamases. Antimicrob Agents Chemother. 2010;54(3):969-976. doi:10.1128/AAC.01009-09
3. Rawat D, Nair D. Extended-spectrum beta-lactamases in Gram-negative bacteria. J Glob Infect Dis. 2010;2(3):263-274. doi:10.4103/0974-777X.68531
4. Tille PM. Bailey & Scott's Diagnostic Microbiology. 15th ed. St. Louis: Elsevier; 2022.