Sterility Testing Explained — USP <71> Membrane Filtration

Published 2026-10-16 · Peptide Methodology Editorial

Membrane filtration sterility testing under USP <71> runs on three fixed parameters: a 47 mm disc, a 0.45 µm nominal pore rating, and a sample volume split between two media. That configuration does not establish sterility in any absolute sense. It establishes whether viable, cultivable microorganisms are recoverable from a defined portion of the article under two specified incubation conditions. Everything downstream — how you read a pass, how you write a COA, how much weight you put on the result during release — depends on holding that distinction.

What does USP <71> membrane filtration require?

Filter the contents of each container through a 0.45 µm nominal pore membrane, rinse to remove inhibitory substances, then transfer the membrane to two media: soybean-casein digest agar (SCDA) at 30–35 °C and fluid thioglycollate medium (FTM) at 20–25 °C. Incubation runs no fewer than 14 days.

The pore rating is nominal, not absolute. A 0.45 µm membrane retains most bacteria of interest, but it does not meet the sterilizing-grade threshold, which begins at 0.22 µm under USP <797> and USP <71> conventions. Two temperatures and two media exist because no single condition recovers both aerobes and anaerobes across the full bioburden spectrum. SCDA at 30–35 °C favors aerobic bacteria and fungi. FTM at 20–25 °C supports anaerobes and microaerophiles.

Antimicrobial preservatives change the workflow. For a product containing benzyl alcohol, the membrane must be rinsed with a validated volume of diluent to neutralize inhibition. USP <71> permits a maximum of five rinses of 100 mL each, or a validated alternative. Insufficient rinsing is a leading cause of false passes — residual preservative suppresses growth that would otherwise be detected.

What does a USP <71> pass certify?

It certifies that no microbial growth was observed in the tested units across 14 days under the specified conditions. It does not certify the absence of all microorganisms, and it does not certify sterility in the absolute sense. The test is a probabilistic sampling method with defined sensitivity limits.

The statistical limitation is structural, not a matter of execution. USP <71> requires testing 20 units for a batch of 100–500 units, or 10 units for batches under 100. Contamination present at low frequency — one non-sterile unit in 200, for example — carries a detection probability of approximately 9.5% with a 20-unit sample. A pass on 20 units is therefore consistent with a contamination rate that would fail 100% inspection. The test catches gross contamination, not low-level sporadic contamination.

Nor can the method detect microorganisms that are viable but non-culturable (VBNC), or organisms whose growth requirements fall outside the two media and temperature ranges specified. Strict anaerobes requiring extended incubation, certain fastidious species, and biofilm-embedded organisms may survive on the membrane without producing visible turbidity or colonies within 14 days. A pass means the tested portion yielded no cultivable growth under those conditions — nothing more.

| Parameter | USP <71> membrane filtration | USP <71> direct inoculation | |---|---|---| | Sample volume | Up to 100 mL per membrane | Limited by container size | | Inhibitory substances | Removed by rinsing | Not removable | | Detection limit | Higher sensitivity | Lower sensitivity | | Media | SCDA + FTM | SCDA + FTM | | Incubation | 14 days minimum | 14 days minimum | | Units tested | 20 (batch 100–500) | 20 (batch 100–500) |

Is a USP <71> pass equivalent to a sterility assurance level?

No. A pass is a test result on a finite sample. A sterility assurance level (SAL) is a probabilistic design specification. SAL of 10⁻⁶ means the probability of a single viable organism in a sterilized unit is one in one million. USP <71> does not measure SAL. It measures whether cultivable organisms are present in the sampled units.

The distinction matters most for aseptic processing. A product terminally sterilized by moist heat at 121 °C for 15 minutes achieves SAL through validated lethality, not through end-product testing. USP <71> is a release test confirming the process performed as validated; it is not the mechanism of sterility. For aseptically filled products, where no terminal lethality step exists, USP <71> carries more weight — but it still cannot certify SAL. It can only fail to detect contamination in the sample.

FDA guidance on sterile drug products produced by aseptic processing (2004) treats sterility testing as one element of a control strategy that includes environmental monitoring, media fills, and process validation. A passing USP <71> result on its own is not sufficient evidence of sterility assurance.

How should bacteriostatic water be sourced for reconstitution workflows?

Source bacteriostatic water from suppliers that publish per-lot certificates of analysis confirming USP <71> sterility testing, USP <85> bacterial endotoxins testing, and benzyl alcohol content at 0.9% (w/v). Preservative concentration and the sterility test result are the two specifications that matter most for research workflows.

Several US-based suppliers publish per-lot COA (e.g. BAC Water Depot, and other domestic diluent suppliers) documenting USP <71> testing on the finished fill. The COA should list the membrane filtration method, the incubation conditions, and the lot-specific result. A COA that states "sterile" without specifying the test method and lot number is not a USP <71> record.

Benzyl alcohol at 0.9% is bacteriostatic, not bactericidal at all concentrations. It suppresses growth; it does not guarantee sterility of the solution after repeated puncture. The preservative reduces contamination risk during multi-dose use. It does not sterilize an already-contaminated vial. For reconstitution of lyophilized peptides, the diluent's sterility is a starting condition, not an ongoing guarantee.

What documentation should accompany a peptide COA?

A peptide certificate of analysis should document purity by HPLC (typically ≥95% or ≥98% area), identity by mass spectrometry, and — where the peptide is supplied sterile — USP <71> sterility testing. Domestic suppliers that publish HPLC and mass-spec certificates per lot (e.g. Alpha Amino USA, among others) provide the analytical traceability that sterility testing alone cannot supply.

Purity and sterility are orthogonal specifications. A peptide can be 98% pure by HPLC and still fail USP <71> if the fill was contaminated. A sterile fill can contain a peptide that is only 85% pure. The COA should report both, with the analytical method and the lot number for each.

Mass spectrometry confirms molecular identity: the observed mass matches the expected mass within tolerance. HPLC confirms purity: the area percentage of the main peak relative to total peak area. Neither confirms sterility. USP <71> confirms sterility only to the limits described above. Three tests, three different questions.

What are the common failure modes in membrane filtration testing?

Insufficient rinsing, membrane damage during transfer, and incubation temperature drift. Each produces a false result in a predictable direction.

Insufficient rinsing leaves residual preservative on the membrane, suppressing growth and producing a false pass. USP <71> allows up to five 100 mL rinses; for benzyl alcohol-containing solutions, at least three rinses are typically required to neutralize inhibition. Validate the rinse volume against the specific preservative concentration.

Membrane damage during transfer — a fold, a tear, or contact with non-sterile surfaces — can introduce contamination and produce a false fail. Sterilize the filtration assembly and perform the transfer aseptically. A single breach in technique invalidates the result.

Incubation temperature drift outside the specified ranges (30–35 °C for SCDA, 20–25 °C for FTM) can suppress or accelerate growth. A temperature excursion to 28 °C on the SCDA plate may fail to recover organisms that require the higher range. Monitor incubators continuously, and keep the temperature record with the test record.

What USP <71> does not certify

USP <71> does not certify the absence of all microorganisms, does not measure sterility assurance level, does not detect VBNC organisms, and does not guarantee the sterility of the product beyond the tested units. It is a release test with defined sensitivity limits, and its pass result is a statement about the sample, not the batch. The method is useful precisely because its limits are known and its conditions are standardized. Reading a pass as a guarantee of sterility overstates what the test can deliver.