Written and maintained by CASRAI Editorial Board
Last updated
Sterility Assurance Level (SAL) is the probability that a single viable microorganism survives on a unit after it has gone through a sterilization process. The number attached to almost every terminally sterilized medical device is a SAL of 10-6 — a one-in-one-million probability of a surviving organism per sterilized unit. That number gets cited constantly in device labeling, sterilization validation reports, and audit findings, and just as often gets misread as a guarantee of sterility. It is not. SAL is a statistical target built into how a process is designed and validated, not a measured property of any individual item coming out of the chamber.
What SAL 10-6 Actually Means
SAL is expressed as a probability, always as a negative power of ten, and it describes the theoretical population of survivors across a batch, not a claim about any one unit. A SAL of 10-6 means that if you sterilized one million identical units under the validated process, the expected number of units with a surviving viable microorganism is one — or, framed the other way, the probability that any single unit is non-sterile is one in a million. It is a modeled endpoint of an exponential microbial inactivation curve, extrapolated well past the point where any surviving organism could actually be detected by a practical test.
That extrapolation matters. No sterility test can confirm 10-6 directly — testing a statistically meaningful sample at that resolution would require destroying essentially the entire batch. Sterility assurance is instead inferred from process data: the demonstrated microbial lethality of the cycle (time, temperature, gas concentration, or radiation dose, depending on method) combined with a known or worst-case starting bioburden. This is precisely why FDA’s Quality System Regulation treats sterilization as a process that “cannot be fully verified by subsequent inspection and test” and therefore must be validated (21 CFR 820.75) rather than confirmed retrospectively on every load. For the surrounding validation framework this sits inside — IQ/OQ/PQ, the method-specific ISO standards, and revalidation triggers — see Sterilization Validation: IQ/OQ/PQ Across Steam, EtO, and Radiation Methods.
How a Cycle Is Validated to Reach a Target SAL
Every sterilization method — steam (moist heat), ethylene oxide (EtO), gamma or e-beam irradiation — kills microorganisms following roughly the same pattern: an exponential decline in the surviving population as exposure (time, dose, or concentration) increases. Validating a cycle to a target SAL means generating enough microbiological and physical data to model that decline curve and show it crosses the target probability with an adequate margin.
The D-Value and Log-Reduction Math
The core unit of that math is the D-value: the exposure (time at a given temperature for steam and EtO, or absorbed dose for radiation) required to reduce a microbial population by one log — 90% — under defined conditions. A D-value is specific to an organism, a method, and a set of process conditions; it is generated experimentally, not assumed. If a biological indicator’s population and D-value are known, the number of logs of reduction a given exposure delivers can be calculated directly, and that log-reduction figure is what gets compared against the target SAL. Reaching SAL 10-6 from a starting population of, say, 106 organisms requires roughly 12 logs of reduction under the standard convention used across the method-specific ISO standards (ISO 17665 for steam, ISO 11135 for EtO, ISO 11137 for radiation) — six logs to reach zero detectable survivors, plus another six-log margin to reach the 10-6 probability floor.
Biological indicators (BIs) — standardized preparations of a known, unusually resistant spore-forming organism, with a certified population and D-value — are the direct microbiological evidence used to generate this data. Geobacillus stearothermophilus is the standard steam and VHP challenge organism; Bacillus atrophaeus is standard for EtO and dry heat. ISO 11138 specifies BI requirements (population, resistance, species) per method. For the practical, day-to-day mechanics of running and interpreting a BI test on an already-validated autoclave — incubation, process challenge devices, what a positive result requires — see Autoclave Spore Testing: Biological Indicators Guide.
Overkill vs. Bioburden-Based Validation
There are two recognized ways to build the required margin into a cycle, and the choice affects both how much bioburden data a validation needs and how aggressive the resulting cycle is:
- Overkill approach. The process is designed to deliver at least a 12-log reduction of a highly resistant BI regardless of the product’s actual bioburden, which for any realistic clinical or lab bioburden (rarely above a few hundred to a few thousand CFU) delivers a comfortable safety margin past SAL 10-6. This is the dominant approach for steam sterilization and is common for EtO, because it avoids the ongoing cost of bioburden testing every product lot and is more forgiving of bioburden variability. Its tradeoff is a longer, more thermally or chemically aggressive cycle than a product’s actual bioburden would strictly require — workable for heat- and chemical-tolerant items, unworkable for delicate or heat-sensitive products.
- Bioburden-based approach. The product’s actual pre-sterilization bioburden is measured directly (population and resistance characterized), and the cycle is set to deliver only the reduction that specific bioburden requires to reach the target SAL, plus a defined safety margin — rather than always assuming a worst-case resistant challenge. This is the standard approach for radiation sterilization under ISO 11137, most commonly via the VDmax dose-setting method or AAMI TIR-based dose-substantiation methods, and it is also used for EtO where a shorter, less aggressive cycle is needed. It requires routine bioburden monitoring as an ongoing input to the validated state, and a bioburden excursion is a direct trigger for revalidation.
Neither approach is inherently more rigorous than the other — both are recognized paths to the same target SAL under the relevant ISO standard, and the choice is driven by the product’s material tolerance and the practicality of routine bioburden testing, not by a difference in the sterility claim itself. For how EtO’s aeration and residual-gas requirements factor into the overall validated cycle, and where low-temperature methods sit relative to steam, see Low-Temperature Sterilization Method Selection: EtO vs. VHP vs. Ozone.
Why SAL 10-6 Specifically for Sterile Body Cavities
SAL 10-6 is not an arbitrary universal constant — it is the target tied specifically to devices that contact normally sterile tissue, the vascular system, or breach the skin or mucous membranes. This traces directly to the Spaulding classification, the framework that sorts a device into critical, semicritical, or noncritical based on how it contacts the patient (see High-Level Disinfection and the Spaulding Classification for the full decision logic). A device classified as critical — one entering sterile tissue, a sterile body cavity, or the vascular system, such as a surgical instrument or an implant — carries the highest consequence for a failure: a surviving organism introduced past the body’s normal barrier defenses has a direct route to cause infection, with no immune surveillance layer (intact skin, mucus, normal flora competition) standing between it and a sterile compartment. SAL 10-6 is the accepted margin against that consequence.
By contrast, a less stringent SAL — historically 10-3, a one-in-a-thousand probability — has been accepted for devices whose only contact is with intact skin or non-sterile mucosal surfaces, where the body’s own barrier and immune defenses provide additional protection against the residual risk a less exhaustive process leaves. FDA’s own sterility guidance for 510(k) submissions reflects this same distinction directly at the labeling-claim level: a device labeled sterile generally needs a stated SAL of 10-6, with an explicit carve-out for devices that contact only intact skin. The practical takeaway is that SAL is a risk-proportionate target, not a single fixed bar — the sterilization cycle, and the validation burden behind it, should be matched to what the device actually touches, which is exactly the judgment the Spaulding classification exists to make explicit before a sterilization method is even selected.
SAL Is a Statistical Target, Not a Guarantee
It is worth stating plainly, because the framing gets flattened constantly in casual use: SAL 10-6 does not mean a device is proven sterile, and it does not mean zero risk. It means the process that produced the device has been validated, with documented physical and microbiological evidence, to reduce the probability of a surviving organism to that modeled level under defined, controlled conditions. A process that was correctly validated but is then run outside its qualified parameters, or applied to a load with bioburden well outside what the validation assumed, no longer carries that assurance regardless of what the original validation report says — which is exactly why routine biological and physical monitoring, and change-triggered revalidation, are treated as inseparable from the initial validation rather than optional follow-up. A package that has since lost integrity carries no sterility assurance at all, independent of the SAL the original cycle achieved; see Event-Related Sterility and Sterile Storage for how that boundary is actually managed after the cycle ends.
Frequently Asked Questions
What does a sterility assurance level of 10-6 mean?
It means the validated sterilization process is modeled to leave, at most, a one-in-one-million probability of a viable microorganism surviving on any given sterilized unit. It is a statistical property of the process, established through microbiological and physical validation data, not a directly measured or guaranteed property of any single item.
Why is SAL 10-6 used instead of a lower or higher number?
10-6 is the consensus target for devices contacting normally sterile body sites, vasculature, or breached tissue, where a surviving organism has a direct route past the body’s normal barrier defenses. It represents roughly a 12-log reduction from a realistic starting bioburden under the standard convention used across ISO 17665, ISO 11135, and ISO 11137 — six logs to eliminate a nominal bioburden population plus a further six-log safety margin.
What is the difference between the overkill and bioburden-based validation approaches?
The overkill approach sets the cycle to deliver a fixed, high log-reduction of a highly resistant biological indicator regardless of the product’s actual bioburden, which is simpler to validate but produces a more aggressive cycle. The bioburden-based approach measures the product’s real bioburden and sets the minimum cycle that specific population requires to reach the target SAL, which allows a gentler cycle for heat- or dose-sensitive products but requires ongoing bioburden monitoring as part of the validated state.
Is SAL 10-3 ever an acceptable standard?
Yes, historically for devices whose only patient contact is intact skin, where the body’s own barrier defenses provide additional protection and FDA’s own sterility labeling guidance carves out a less stringent expectation than the 10-6 standard applied to devices contacting sterile tissue or the vascular system.
Does achieving SAL 10-6 mean a device is definitely sterile?
No. SAL is a modeled probability derived from validated process data, not a per-unit sterility test result. A correctly validated process run outside its qualified parameters, applied to an unexpectedly high bioburden load, or followed by a compromised sterile package no longer carries the assurance the original validation established, regardless of the SAL figure on file.
Which ISO standard governs SAL for my sterilization method?
ISO 17665 covers moist-heat (steam) sterilization, ISO 11135 covers ethylene oxide, and ISO 11137 (in three parts) covers radiation sterilization, including gamma, e-beam, and X-ray. ISO 14937 sets the general cross-method framework all three implement, and ISO 11138 specifies the biological indicators used to generate the microbiological validation evidence for each method.
