Risk estimation matrix
REM is a simple risk assessment tool that assumes failure of each critical parameter and uses the likelihood and severity
of that failure to determine the overall risk. REM is based on a 3 x 3 matrix, similar to a heat map. Like CM, REM is limited
in that it is not a formal risk assessment tool; hence, it does not have the level of detail and rigor that more complex systems
and processes may require.
The process for REM is as follows:
Table II: Risk action level.
|
1. Determine qualitative scales for likelihood and severity rankings. Develop an action level table (see Table II).
2. Identify critical parameters for the system under review.
3. Brainstorm potential failures for each critical parameter.
4. Rank each potential failure for likelihood and severity using the criteria established in Step 1.
Table III: Risk matrix.
|
5. Determine overall risk using the risk matrix (see Table III). Propose mitigation for unacceptable risks.
In order to preserve objectivity and ensure consistency of the risk assessment to follow, the first step in the REM methodology
involves the establishment of risk ranking scales. Two qualitative scales will be developed, each containing three potential
scores. The likelihood scale addresses how likely is it that the failure will occur, given the current controls in place.
This scale includes options ranging from remote (unlikely) through average (likely) to certain (very likely or unknown). The
severity scale addresses the question: If that failure did occur, how severe would the consequences be? The severity scale
ranges from minor (insignificant impact) through moderate (moderate impact) to critical (significant impact).
The final scale that must be established is an action level table that dictates the acceptability for overall risk, including
whether mitigation measures are required. Low-risk items may not require any mitigation activities or resource expenditure,
whereas high-risk items will require additional risk control measures to reduce risk to an acceptable level.
Table IV: Risk estimation matrix: hypothetical scale-up of saline solution.
|
Returning to the hypothetical saline solution scale-up, the risk team would first brainstorm potential failures associated
with each critical parameter for the saline solution process. For example, the batch could fail the bioburden specification,
the closed aseptic system could be breached, the new material may not be biocompatible, or the vessel capacity may be insufficient
for production needs (see Table IV). Each of these potential failures is then ranked for likelihood and severity and the overall risk identified using the risk
matrix in Table III.
Focusing on the new product-contact material, it may be difficult to assign a likelihood score if there is no available data
on the biocompatibility or extractable/leachable profile of this material. In such cases, it is best to take a conservative
approach and assign a likelihood score of "certain" to the lack of biocompatibility. Based on the potential patient impact
of this failure, the severity would be given a score of "critical." The intersection of "certain" and "critical" in the risk
matrix shows this risk to be high. Thus, the risk of changing the vessel type to a new material is not acceptable, and additional
risk control measures must be taken. Because in this example the overall risk is driven primarily by a lack of data, mitigation
efforts would focus on biocompatibility testing to better understand the implications of the new material on the product.
Once this action is taken, it is expected that overall risk would then be reduced to an acceptable level.
Conclusion
To ensure that the quality system and associated processes remain in control over time, every company must understand how
their risk exposure is affected as validated systems evolve. The application of quality risk management principles and tools
facilitate this understanding, allowing for more comprehensive strategy development and informed decision-making. It is not
always, however, necessary to perform lengthy, formal risk assessments to reach these goals. For simple systems and processes
as well as for changes that are well understood, less-formal tools such as the comparison matrix and risk estimation matrix
provide a comprehensive picture of the associated risk in an easily applied format. The consistent use of these tools can
enable the pharmaceutical industry to prioritize resource expenditure and provide only the highest quality products to patients.
Kelly Waldron is principal continuous process improvement analyst, Global Quality Risk Management, at Genzyme, a Sanofi company, kelly.waldron@genzyme.com
. Marissa Gray is field marketing manager, Sterile Filtration at EMD Millipore, Marissa.Gray@Merckgroup.com
.
Submitted: July 6, 2012. Accepted: July 18, 2012.
Reference
1. ICH, Q9 Quality Risk Management (Nov. 2005).
|
