Microbial air samplers: selection, efficiency and qualification

A microbial air sampler must provide a sample suited to the programme’s question. A flow calibration certificate addresses only part of the problem: it does not by itself demonstrate particle collection, preservation of the biological response and suitability in actual use.

Start with the result you need. This guide concerns instrument selection and control; the overall environmental monitoring strategy remains a separate activity.

1. Define intended use and what the result means

Specify environment, location, operational state, period to be covered and the decision supported by the data. Active sampling draws a measured volume; settle plates collect what deposits during exposure. Their results do not become equivalent simply by changing the unit.

Total particle counting, often called “non-viable”, does not determine particles’ biological viability. A culture method detects units capable of forming colonies under the conditions used: CFU is neither an exact cell count nor a measure of every viable microorganism. State what the result can and cannot represent.

2. Compare collection principles

Agar impaction collects directly onto culture medium; filtration retains material on a substrate analysed according to the method; liquid collection adds further sample-handling steps. Each option changes the path from sampled air to the final signal.

Compare compatibility with the intended analysis, losses along the path, duration effects and additional operations. No principle is superior for every purpose. Do not automatically transfer performance demonstrated with one substrate or configuration to every accessory marketed under the same name.

3. Separate physical efficiency from biological recovery

Physical efficiency describes particle collection under the test conditions, including its dependence on particle size. Biological recovery concerns what remains detectable after sampling and analysis. Read how it is defined: against which reference, with which material, collection substrate and final measurement.

A 2025 primary microsphere study concerns physical capture of surrogates; it does not demonstrate recovery of viable organisms. A 2004 comparative study also shows that the medium can affect comparisons between samplers. Neither supports a universal ranking.

Look for configuration, conditions, reference, replicates and limitations in the test report. A percentage without a denominator or context is insufficient. A collection curve or cut-off diameter does not replace relevant biological evidence; satisfactory flow rate replaces neither.

4. Translate use into verifiable requirements

Assess flow and volume over the range used, materials, compatibility with cleaning, disinfection or biodecontamination, battery endurance, noise and data handling. Consider footprint, head orientation, air exhaust, access and necessary interventions. Placement must be practical without compromising the process.

EU GMP Annex 1, 2022, §§9.22 and 9.29 connects sampling methods, airflow interference and recovery-supporting data. Translate these aspects into checks for the intended use.

5. Distinguish qualification from calibration

Manufacturer testing characterises the model under stated conditions. Local qualification verifies configuration, installation and performance relevant to use. Flow calibration relates indication to a reference with its associated uncertainty. Link each document to unit identity, accessories, range and status; avoid assigning a correct certificate to the wrong configuration.

The official BS EN 17141:2020 record also describes sampler verification. The ISO 14698-1:2003 record identifies that standard as withdrawn. Check edition, adoption and applicability; public records do not replace the details of authorised texts. WHO good practices, 2011 provide the general framework for controlling microbiological equipment.

6. Connect assembly, volume and count

Before use, check identity, status, assembly, specified head and substrate, available controls and authorised settings. Record location, times, actual volume, events and the link to analysis. A programmed volume must not conceal an interruption.

Simulated arithmetic example, not a prescription. Assuming a constant actual flow of 80 L/min for 5 min, the volume is 400 L, or 0.400 m³. A count of 6 CFU, already processed according to the applicable method, corresponds to 6 ÷ 0.400 = 15 CFU/m³. This is neither an acceptance limit nor a recommended volume. If flow varies, multiplying the setpoint by time may not represent actual volume.

Apply statistical positive-hole corrections only where specified for that technology and geometry, retaining the original count, rule and result. Do not use another head’s correction table. Absence of colonies does not demonstrate absolute absence of contamination. For subsequent stages, see incubation and sample history and colony counter verification.

7. Manage replacements and comparisons over time

Simulated case. A laboratory replaces an impactor with a filtration system while retaining the same flow rate. The substrate, subsequent processing and potential losses change: the flow certificate does not make the results interchangeable.

Before the transition, microbiology and QA define which decisions depend on the historical series, which evidence transfers and which local comparisons are needed. The change remains identifiable in the data, and differences are assessed without arbitrarily resetting programme limits. Bridging tests must represent actual use and have justified criteria defined before evaluation.

Manage cleaning, maintenance, checks, relocations and head replacements through unit-linked records. After a significant change, decide which verifications to repeat and whether earlier results need review. Programme continuity depends on documented history as well as a serial number.

Sources and access limitations

• EU GMP Annex 1, 2022, and WHO TRS 961 Annex 2, 2011: public texts linked above.
• BSI and ISO: official records consulted on 30 September 2026; BS EN 17141:2020 listed as current, under review; ISO 14698-1:2003 withdrawn. No threshold is attributed to an unread clause.
• Zhu and colleagues, 2025: public physical collection study using surrogates. Shintani and colleagues, 2004, PMID 14706270: public abstract, healthcare comparison. Scope-specific evidence, not qualification of the local sampler.