What does the term "Improbable" really mean in the context of risk acceptability?

When discussing probabilities in qualitative terms, there’s a common thread across all industries—regardless of the risk profile or quantity of units used per year . This commonality lies in the concept of the "IMPROBABLE," which refers to a single failure, complaint, or nonconformity. The "IMPROBABLE" acts as the common foundation upon which the QQ Chart is built.

Let’s explore how this foundational principle applies across various product types, from high-volume, low-risk items to low-volume, high-risk devices.

Low-Risk, High-Volume Products: The “IMPROBABLE” Criteria

For products with low complexity and a low-risk profile—like those produced in the millions each year —probability of failure plays a pivotal role as the criteria for risk acceptability. The "IMPROBABLE" risk level in this case is set at 1 x 10^-6, representing one failure, complaint, or nonconformity out of a million units sold. The product may be simple in design, but the sheer volume makes managing even the smallest risk critically important.

Moderate-Risk, Moderate-Volume Products

In the case of products with moderate complexity and risk, the "IMPROBABLE" risk acceptability criteria change accordingly. For items manufactured in hundreds of thousands of units per year, the "IMPROBABLE" threshold is set at 1 x 10^-5. This accounts for a slightly higher chance of failure due to the increased complexity and volume of the product, but it still maintains a high standard for quality and reliability.

Medium-Risk, Lower-Volume Products

As we move toward products with medium complexity and risk profile—typically manufactured in tens of thousands of units annually —the "IMPROBABLE" threshold increases to 1 x 10^-4. Here, the balance of risk and volume comes into play. While the number of units is smaller, the complexity of the device raises the probability of nonconformities, so the threshold adjusts accordingly.

High-Risk, Low-Volume Products

Lastly, for high-risk products or those with complex designs, typically sold in smaller quantities (thousands of units per year), the "IMPROBABLE" criteria is 1 x 10^-3. These devices have a higher likelihood of failure due to their complexity and the inherent risks involved in their use. This is reflected in the increased threshold, emphasizing the need for even more stringent quality controls despite the lower production volume.

The Role of "IMPROBABLE" in Probability

Once the base "IMPROBABLE" threshold is established, other probability ranges follow suit, growing by one order of magnitude. However, it's crucial to remember that relying solely on percentage values to make risk-based decisions can be misleading. For instance, consider two different scenarios:

• One nonconformity in 1,000,000 units translates to a probability of 0.0001%.

• One nonconformity in 10,000 units translates to a probability of 0.01%.

Though both are categorized as "IMPROBABLE," the percentage difference might make it seem like the risks are far apart. But when dealing with vastly different products—such as a single-use tongue depressor and an infusion pump—the raw number of nonconformities is the true metric to focus on, not just the percentage.

Why Numbers Matter More Than Percentages

Focusing on the number of failures, complaints, or nonconformities gives a clearer picture of the risk acceptability associated with a product. Percentages alone can be misleading and might cause confusion when comparing products from different industries or with varying levels of complexity. So, in risk-based decision-making, always consider the actual number of failures, complaints, or nonconformities instead of relying on percentage values alone.

In conclusion, understanding and applying the "IMPROBABLE" concept to probability criteria can lead to better decision-making, clearer communication of risks, and more effective quality control across various industries. By focusing on the actual numbers of nonconformities rather than percentages, businesses can ensure they’re managing risk accurately and effectively.