OR WAIT 15 SECS
Volume 35, Issue 10
This risk-management case study focuses on assessing empty capsules.
The following case study on an empty-capsule product deviation is the fourth of eight in a series put together by the Product Quality Research Institute Manufacturing Technical Committee (PQRI–MTC) risk-management working group. The series is meant to advance the understanding and application of the International Conference on Harmonization (ICH) Q9 Quality Risk Management guideline by providing actual examples of risk-management assessments used by the bio/pharmaceutical industry. The introductory article explaining the history and structure of the series, as well as the first case study on defining design space, appeared in the July 2011 issue of Pharmaceutical Technology (1). The second study addressed functional equivalence for equipment replacement, and the third addressed facility biocontainment and inactivation; they appeared in the August and September 2011 issues, respectively (2, 3).
A drug product site produces pain-free capsules, which are indicated as an anti-epileptic and for treatment of neuropathic pain. Multiple customer complaints of empty capsules were received. Lot ABC was fully distributed in the United States and no product remains within company control. There is no evidence of tampering.
Batch record review indicated that during processing, a loose dosator was replaced on the encapsulation equipment. Following replacement—and prior to resuming encapsulation, acceptance testing of capsules produced required by standard operating procedure was performed and product met requirements.
Further investigation revealed that the loose dosator caused empty capsules to be produced. The encapsulation system used a vacuum system to remove empty capsules. This empty capsule removal system includes a reservoir for holding empty capsules rejected during the manufacturing process. As a result of the loose dosator, an atypically high number of reject empty/low fill capsules were produced during the encapsulation operation, causing the reservoir to be filled and eventually overflow. The reservoir was physically located over the acceptable capsule flow. Therefore, it was determined during the investigation that, if the empty capsule chamber overflowed, there was potential for rejected capsules to fall back into the acceptable capsule exit chute and to be reintroduced to the lot.
Sealed bottles were obtained from remaining inventory of the lot. Of 30,661 capsules examined, 151 empty and 5 low-fill weight capsules were found. During the inventory evaluation, one to five empty capsules were found in 46% of the bottles evaluated.
In the this case study, the manufacturer decides to use quality risk management to evaluate the potential impact of the deviation.
Risk question and risk-assessment method
The risk question developed for the subject case study is: Do a small number of potential low fill or empty capsules in a single batch of pain-free capsules pose an unacceptable risk to patients, and secondarily, to the company?
In this case study, the risk factors are more qualitative than quantitative. Failure-mode effects analysis (FMEA) is specifically designed to systematically study processes for possible failure modes and then to develop actions to mitigate these failure modes. An element that requires consideration in this case study is detectability of the defect. Is it possible for the pharmacist or the patient to readily detect empty capsules? The FMEA technique is an optimal tool for this application as the standard methodology includes all three risk components (i.e., probability, severity, and detectability). Therefore, the risk methodology selected for the subject case study is: FMEA.
Risk identification, analysis, and evaluation
The FMEA process used intends to identify potential risks that could result from empty capsules in the market and the possible consequences of each risk. It is understood that sufficient data may not always be available to reach conclusions, and identification of risks should be based on best available data, scientific knowledge, and historical experience. The rationale used to identify risks should be documented.
Examples of potential risks include the following: patients may receive empty capsules, patients may not have availability of medically necessary product, the company could receive an audit observation from an internal auditor or external regulatory agency, and so forth.Numerical ratings for the FEMA analysis are based on the following criteria:
A numerical ranking of 1–3 is applied to each evaluated hazard, as demonstrated in the example FMEA risk-score ranking table using a three-point ranking scale (see Table I).
Table I: Failure-mode effects analysis risk-score ranking.
In the FMEA analysis executed for this case study, the firm determined that, if there was potential for greater than a moderate risk-evaluation score, appropriate risk mitigating actions would be required to lower the risk to an acceptable level. Therefore, If the score for an evaluated hazard exceeds 9, corrective measures to the reduce risk of failure will be taken. If after attempting risk mitigation, the score could not be lowered below 9, the resulting risk would not be accepted. For those items with a score below the defined threshold, risks will be accepted. Conclusions are documented.
An example worksheet for calculating the Risk Evaluation Score for this analysis is presented in Table II.
Table II: Sample worksheet for calculating a risk-evaluation score.
In the current case study, risk-reduction and acceptance decisions centered on patient safety. Despite the apparently low number of empty/low fill capsules that may be present, there was no assurance that the numbers of empty capsules in the market was low. There was little ability for a pharmacist or patient to detect an empty or partially filled capsule. Most importantly, the potential medical consequence for some patients due to receiving and administering an empty or partially filled capsule was severe. As a result of the potentially high risk associated with some patients taking empty/low-fill capsules, and the inability of the firm to implement appropriate risk mitigating actions to lower this risk (i.e., severity and ability to detect) to an acceptable level, the decision was taken to initiate a product recall.
Risk documentation and communication
Site procedures require preparation and Quality division endorsement of a deviation report. The risk assessment output was incorporated into a number of existing work processes and their associated documentation. To prevent reintroduction of rejected product to the product, an automated sensor was incorporated into the empty-capsule reject reservoir. When activated, this sensor automatically shuts down the encapsulation machine should rejected capsules in the reservoir reach the sensor level, thereby preventing overflow. The overflow sensor with automated equipment shutdown was applied to all encapsulation machines at the firm. Additionally, the results of the FEMA analysis and recommendation for product recall were presented at an internal-management notification process meeting and minutes of risk analysis and associated conclusions were documented. Affected regulatory agencies were formally notified of the recall decision.
Risk review and conclusion
Evaluation of product complaints, adverse-event reports, and further examination of product received from the product recall may potentially be used to gain additional understanding of the extent of the defect that reached the market.
Ted Frank is with Merck & Co; Stephen Brooks, Kristin Murray,* and Steve Reich are with Pfizer; Ed Sanchez is with Johnson & Johnson; Brian Hasselbalch is with the FDA Center for Drug Evaluation and Research; Kwame Obeng is with Bristol Myers Squibb; and Richard Creekmore is with AstraZeneca.
*To whom all correspondence should be addressed, at firstname.lastname@example.org
1 . T. Frank et al., Pharm. Technol. 35 (7), pp. 72–79.
2. T. Frank et al., Pharm. Technol. 35 (8), pp. 72–75.
3 . T. Frank et al., Pharm. Technol. 35(9), pp. 82–84.