The project developed a rapid, PCR‑based workflow for detecting antibiotic‑resistant Acinetobacter baumannii directly from patient swabs. The first step was to optimise DNA release from swab samples. Various resuspension buffers were tested, each containing 500 µL of either pure water, water with 1 % SDS, water with 1 % Tween 20, water with 10 µL Proteinase K, or combinations of Tween 20 and Proteinase K at different concentrations. The best performance was achieved with a simple solution of de‑ionised water, 1 % Tween 20 and 0.004 µg/µL Proteinase K, incubated for 10 minutes at 56 °C in a water bath. Swabs were first dried for five minutes at room temperature, then incubated in this buffer, and the resulting lysate was used directly in PCR without dilution. Fluorescence‑based detection on a real‑time instrument confirmed that this protocol yielded robust DNA signals. The commonly used Amies transport medium was also found to release DNA effectively, indicating the method’s compatibility with routine clinical sampling.

For the detection assay, a TaqMan real‑time PCR system was established using standardised reagents: a commercial master mix containing DNA polymerase, dNTPs and buffer components, white 96‑well polypropylene plates, and a real‑time PCR instrument. Primers and probes were designed from reference sequence databases to target a broad panel of resistance genes, including blaOXA‑51‑like, blaOXA‑23‑like, blaOXA‑24/40‑like, blaOXA‑58‑like, blaOXA‑143‑like, mcr variants (mcr‑1, ‑2, ‑3, ‑4, ‑5, ‑6, ‑7, ‑8, ‑9, ‑10), blaADC, blaNDM, blaGIM, blaVIM, blaIMP, blaGES, blaPER, blaTEM, blaSHV, blaVEB, blaCTX‑M groups, blaKPC, and a housekeeping rpoB gene for species confirmation. Initial single‑plex reactions were performed, followed by multiplex assemblies. Sensitivity was evaluated using a ten‑fold dilution series of positive controls ranging from 10 to 10 000 copies, while specificity was tested with negative samples containing 10 000 copies of unrelated DNA. The assay reliably detected all target genes in the dilution series, and no cross‑reactivity was observed in the negative controls. Because the microfluidic chip used in later work packages could accommodate only 12 reaction chambers, multiplex design criteria were applied to avoid primer dimers and cross‑reactivity, allowing simultaneous detection of two resistance gene groups with distinct fluorophores.

The microfluidic implementation was carried out at the Microfluidic Systems and Control Laboratory (MFCS). A chip design featuring 12 chambers was discussed in teleconferences and produced by MFCS. The PCR reactions were transferred from the 96‑well format to the chip, maintaining the same master mix and primer/probe concentrations. The chip successfully amplified the target genes, demonstrating that the established reactions could be integrated without further optimisation. Reagents were lyophilised at 37 °C and stored at –20 °C; stability studies over three years confirmed that primers, probes and polymerase remained functional, enabling long‑term stockpiling.

Software development at inno‑train produced a device‑embedded routine that automatically calculates cycle threshold (Ct) values and interprets them to indicate the presence of A. baumannii and the specific resistance genes present. Initial software runs revealed minor calculation errors, which were corrected by the MFCS team. The final software version can therefore report both species identification and resistance profile directly from the raw fluorescence data.

System integration and validation were coordinated with the Institute for Hygiene (IFH). The primer/probe mixes were transferred to the chip, and prototype chips were produced and tested at IFH. Validation at IFH confirmed that the integrated system met performance criteria for sensitivity, specificity and turnaround time. Throughout the project, inno‑train maintained regular communication with regulatory authorities to align the development with certification and market requirements.

The collaboration involved inno‑train as the project lead, MFCS providing microfluidic design and fabrication, and IFH conducting validation and integration. The work spanned multiple work packages (AP3 to AP10), covering assay design, reagent optimisation, microfluidic implementation, reagent storage, software development, system integration, and final validation. The project successfully produced a fully integrated, rapid, and reliable assay for detecting outbreak clusters of antibiotic‑resistant A. baumannii directly from patient swabs.