Title: Quality Risk Management (QRM) Application to Establishment of Weighing Device Performance Testing Intervals | |||||
| Guidance Number: 89 | |||||
| Prepared by: | Date: | Supersedes: | |||
| Checked by: | Date: | Date Issued: | |||
| Approved by: | Date: | Review Date: | |||
Quality Risk Management (QRM) Application to Establishment of
Weighing Device Performance Testing Intervals
Introduction
Application of Quality Risk Management to performance checks for weighing devices such as balances and scales is intended to provide a tool for determining the acceptability of decreasing the frequency of verifying the performance of a weighing device from the current frequency (e.g. daily) to an alternative schedule.
This guidance document is applicable to weighing devices used in all aspects of operations. The phrase “performance verification” is representative of testing conducted using certified, traceable weights representing the upper and lower ends of the established working ranges for a specific weighing device. Performance verification typically includes determination of accuracy (a weighing device’s ability to accurately determine the mass of a weight having a certified mass) and linearity (accuracy across the established working range of the device). The term “weighing device” refers to any instrument whose purpose is to render a mass determination of material placed on its weighing surface or pan. The term “weighing device” will be used universally throughout this document to refer to weighing instruments regardless of the mechanism or technology employed, i.e. analytical balances, top loading balances, comparators, floor scales, etc.
Practice Summary
This document provides guidance in assessing the risks associated with increasing the interval between executions of performance verification testing on a weighing device. Consideration will be given to the following factors:
- Regulatory requirements and cGMP’s
- Application of the weighing device (how used)
- Performance Verification History (accuracy and linearity of the specific weighing device)
- Maintenance/Calibration History of the specific weighing device
- Environmental Conditions to which the weighing device is exposed, and
- Business Considerations
The quality risk management approach as applied to the evaluation of reduction of performance verification frequency illustrated in this guidance not only identifies the different risk factors to consider when performing the evaluation but also demonstrates a simple tool (depicted in tabular format) for determining how to group potential risks into low, medium, and high categories. For the purpose of this evaluation, two risk factors, probability and severity, will be examined for each perceived risk associated with the defined risk scenario. From this evaluation of individual perceived risks the cumulative risk profile associated with a potential change in frequency will be devised. Through application of a simple tool coupled with requisite background knowledge it is expected that this assessment will serve as a model to a GMP site to standardize the evaluation of changes to the frequency of weighing device performance verification testing.
Recommendations & Rationale for Recommendations
Risk Question
In this case the proposed change drives the creation of the risk question. Our risk question becomes, “what are the potential risks associated with changing the frequency of weighing device performance verification testing from the current schedule (e.g. daily) to an alternate, longer period”.
Risk Assessment Tool
Given the nature of the data to be used for the assessment the Risk Ranking and Filtering (RRF) method has been selected to aid in the assessment of risks associated with a change in testing frequency. RRF focuses on two separate risk factors, probability and severity, associated with each potential risk relevant to an issue.
Risk Assessment -Identification, analyses, and evaluation of potential risks
The potential risks associated with decreasing the frequency of weighing device performance verification testing were derived through completion of a brainstorming exercise. The risks identified should be relevant to most weighing devices and operations. The following areas were identified as having risks related to a change in verification testing frequency:
Regulatory Expectations – the formalized requirements pertaining to control of balances should be reviewed and understood to determine the potential risk of non-compliance. Risks may vary from one market to another, it is suggested that the expectations for the most stringent market served be used for the assessment when multiple markets are involved.
cGMP Expectations – the unwritten expectations that are generally accepted as “standard practice” should be considered. Many times these expectations are verbally expressed by regulatory inspectors during facilities inspections. As with Regulatory expectations the assessment should be based on the most restrictive expectations.
Environment – the potential for change in environmental factors that could affect the performance of the weighing device and potentially go unnoticed during a more prolonged test interval should be evaluated. These factors generally include static, vibration, temperature, humidity and the potential for physical disturbance. Controls such as temperature and humidity monitoring systems in place within the work space should be considered to determine if the potential risk is mitigated or eliminated as a concern. Typically a thorough IQ should address the risk of exposure to unacceptable levels of vibration or static; however, these variables should not be overlooked. There should be procedural controls in place to evaluate the need for additional performance verification when the environmental conditions or the physical location are changed such that it is inconsistent with the IQ.
Measurement – the impact of loss of accuracy due to device drift should be assessed. The criticality of the weighing operations performed, i.e. standards, samples, actives, excipients, media, etc. should be considered in the assessment. As should the type of operation supported, i.e. intermediates, API or finished product. The impact of weighing inaccuracy may be lesser for an intermediate since it is tested at the API stage and again as part of the finished product. The probability of the weighing device encountering drift during the proposed period should be based on the operating history of the device. This review should include the frequency of unscheduled maintenance over the previous two year period and the device’s performance history from, minimally, the last five calibrations. If the idealized running history, respective of maintenance and calibration, is not available for a specific weighing device, the history of a like model instrument deployed under similar conditions may be used for reference.
The most critical measure is of the performance testing history. This assessment should be based upon a statistically significant number of consecutive measures; a minimum of 30 consecutive measurements (representing the current verification testing frequency, e.g. daily) is recommended to minimize the influence of statistical error. All data points for each weight employed during the performance verification test should fall within ±3 σ of the overall mean for each weight. Any points falling outside the ±3 σ range should be investigated.
Business – although decreasing the test frequency may save time and money in normal practice, it must be recognized that the inverse would be true when a failure is observed during less frequent (e.g. weekly) testing. A greater number of weighing operations may be termed “suspect” and subsequently require repeat. This would represent greater impact to cycle times, costs of goods, and man hours required for investigations and repeat operations. Therefore as the frequency of use for a weighing device increases, so does the potential impact of a performance failure.
For each of the above stated risks related to a change in frequency of performance verification testing the individual risk components must be assessed. As identified previously each potential risk has an associated probability and a severity. Table 1 below represents the suggested scale for use in the assessment; it reflects the differentiations in harm resulting in each area of risk relative to the scenario. During the assessment the risks will be scored based on the worst applicable harm associated with a given risk.