A practical guide for clinical microbiologists and researchers evaluating antibiotic combinations with gradient diffusion strips.
This guide is for educational and scientific purposes only. It does not replace validated clinical laboratory protocols and should not inform therapeutic decisions directly.
Etest gradient strips (bioMérieux) are widely used to estimate Minimum Inhibitory Concentrations (MIC), and by combining two strips on the same agar plate, laboratories can also approximate the Fractional Inhibitory Concentration (FIC) index of an antibiotic pair without running a full broth checkerboard. Once both combined MICs are read, the result can be interpreted with the same FIC calculator used for checkerboard data. This guide compares the two main Etest synergy techniques step by step, reviews how well Etest results track the liquid reference method, and flags the antibiotic classes where visual reading is most error-prone.
This is the original configuration described by White et al. (1996). Each antibiotic's MIC is determined alone first. Both strips are then placed together on a freshly inoculated plate, perpendicular to each other, positioned so that their individual MIC graduations intersect at a right angle. Only one plate and one incubation cycle are needed, which makes the cross method the more economical option in terms of both agar and hands-on time. Its main drawback is at the reading stage: the intersection point is a small, often irregularly shaped area where growth is inhibited from two directions at once, which demands a steady hand and consistent lighting to read accurately.
Here the two strips are not applied simultaneously. The first strip is placed and left on the agar for a defined intermediate incubation period (commonly around 60 minutes) to allow partial diffusion into the agar, then it is removed. The second strip is placed directly along the groove the first one left behind, so its gradient diffuses through agar already exposed to the first drug (Doern, 2014). This sequential exposure can better approximate a real pharmacological scenario where one drug already occupies the tissue when the second is introduced, but it is more labor-intensive — it requires a timed removal step and careful strip repositioning — and consumes the same amount of plate real estate for a longer total protocol time.
A third variant, the MIC:MIC ratio (parallel placement) method, avoids intersection reading altogether by placing the two strips side by side at a distance proportional to the ratio of their individual MICs. In a comparison of all three gradient techniques against the checkerboard on 70 multidrug-resistant Pseudomonas aeruginosa isolates, the MIC:MIC ratio method produced the highest correlation with the reference method among the three (Okoliegbe et al., 2021).
The broth microdilution checkerboard remains the reference standard for FIC determination, but it is slow and labor-intensive, which is exactly why gradient-strip alternatives were developed. The evidence on how well they agree with the checkerboard is mixed and organism/drug-dependent:
Etest synergy testing should be treated as a practical screening tool, not a drop-in replacement for the checkerboard — particularly for high-stakes combination therapy decisions in multidrug-resistant infections, where a discordant or borderline FICI is worth confirming with a reference broth method.
Colistin is the antibiotic class where gradient diffusion methods, including Etest, are least reliable. In a comparative evaluation of six colistin susceptibility testing methods against carbapenem-nonsusceptible Klebsiella pneumoniae and Acinetobacter baumannii, Etest produced an unacceptably high rate of very major errors relative to reference broth microdilution (Dafopoulou et al., 2015). The likely explanation is colistin's poor diffusion through agar combined with the frequent presence of resistant subpopulations (heteroresistance) that a single gradient reading cannot detect. For colistin-containing combinations, an Etest-derived FIC index should be interpreted with particular caution and, where feasible, confirmed with broth microdilution.
Beta-lactam strips are prone to a distinct set of reading artifacts, most notably the "phantom" or rounded inhibition zone: a faint, ill-defined ellipse — sometimes with visible growth (haze or trailing colonies) within it — that can make the true endpoint difficult to pinpoint. This phenomenon was well documented for the Etest ESBL screen, where a rounded zone below the clavulanate-containing gradient, with or without a clear ellipse around the unprotected drug, complicates interpretation and was shown to depend partly on reader training and experience (Cormican et al., 1996). The same class of artifact can appear in beta-lactam synergy testing at the strip intersection, where two independent trailing-growth patterns can overlap and obscure the true combined MIC. Reading at a consistent angle against a dark background, and repeating any plate where the intersection zone is not a clean, well-defined ellipse, reduces this source of error.