Microbiological issues in pharmaceutical plants can result in plant shutdowns and products withdrawn from the market. Failures are common in environmental monitoring, media fill, sterilisation, aseptic practice and gowning. Introducing lean practices can determine optimal batch size and lead to a decreased number of test sessions for bioburden, endotoxins and sterility testing, while taking a risk-based approach to determining objectionable micro-organisms may ease the cost and burden of unnecessary testing but requires a better understanding of the processes and risks involved.
The annual Pharmig conference in Nottingham, UK, revealed how greater understanding of microbial risks has increased the role of microbial testing labs, but also highlighted why there is a need for practice improvements. Susan Birks reports.
Microbiology services are crucial for trouble-free operation in pharmaceutical plants as microbiological issues can be hugely significant, resulting in plant shutdowns and products withdrawn from the market. Recent recipients of the FDA’s 483 forms show that no company is too big to fail. With this in mind, the two-day Pharmig conference held in Nottingham in November offered several best practice presentations on microbiology. From a full and informative programme three presentations have been chosen to highlight some common failings and some of the new methods for achieving improvements for the sector.
Neil Raw, GMP Inspector at the MHRA, highlighted some of the microbiology related issues raised during EU GMP inspections. While microbiological issues do not make it into the top 10 most common failings found during inspections, he said, they do occasionally give rise to concern.
Raw’s talk covered various operations, such as environmental monitoring, media fill, sterilisation, aseptic practice and gowning, where microbiology failures were common.
For example, the main failings on environmental monitoring that he listed, included: sample points not being representative of worst case locations; an absence of documented rationale to support sampling positions; sampling positions not adequately documented; insufficient trending and trends not identified; in-house media overcooked; the incubation of plates in sealed bags (where poor gas exchange may affect growth); and anaerobes not looked for where nitrogen is used.
His overriding message to delegates on environmental sampling was that, if the microbiological department has a procedure, make sure it has a rationale behind it and that it has been documented.
He then looked at problems of contamination, validation and cleaning issues. For example, in terms of out of spec results, he said many microbiology labs are lacking in clear procedures of what to do. “Chemistry labs on the whole have good diagrams and systems for dealing with out of spec results. But in my experience microbiology labs often don’t,” he said. Other failings in this area included: exotic bacteria being identified and then accepted without further questions; the air sampler for aseptic areas being stored in the microbiology lab; or media and incubation conditions not being suitable for testing the growth of fungi.
Failings in the use of media fills and sterilisation practice included: non-production personnel not being included in routine media fill programmes; sterilised equipment not being stored in Grade A areas; sterile and non sterile items not being readily identified.
Inspections of gowning areas picked up the following omissions: no gowning assessments had been undertaken, step-over benches were not disinfected; and gowns were being worn outside of production areas.
The potential for cross-contamination seen during inspections resulted from errors such as: live cultures being stored next to media in the fridge; non-sterile hand sanitisers being used before entering Grade A and B areas; and staff not removing lab coats when moving between dirty and clean areas.
Problems associated with infrastructure and equipment can also give rise to issues. For example, drains and sinks were found in inappropriate areas, and not sanitised or monitored; and clean tanks and containers were left damp.
His list was long and varied, making it clear that it is easy to breach good practice and instances are not uncommon. To avoid this, specialised training is key and needs to be continuous, the systems in place need to be practical and well understood and breaches made more difficult and obvious.
Valerie Dunne of Pfizer, Grange Castle, Ireland, looked at how lean techniques and operational efficiencies can be applied to microbial labs. The Pfizer labs in question are involved in the testing of bioburden, endotoxins, microbial identification, and viability-, purity- and sterility testing.
Dunne said that as a result of introducing lean practices optimal batch size determination has led to a decreased number of test sessions for bioburden, endotoxins and sterility testing. Workflow mapping has also resulted in improved efficiencies with test sessions. Scheduling optimisation has reduced non-testing preparation activities and gowning by 35%. And ‘role cards’ have been implemented that schedule the activities that need to be performed, allowing for dedicated time for cross-training and project work.
Pfizer partnered with outside consultants to implement lean labs and after initial consultations, teams were formed of QC chemistry and microbiology staff who were to be leaders of the operation excellence programme. Training was conducted that included both team and area management.
One of the first steps in the Design and Preparation involved creating a ‘levelling strategy’. The incoming sampling variability was analysed and a strategy selected to meet the required level of demand quickly. This is achieved by developing repeating sequences of testing (Rhythm Wheels) that move the samples through all the required tests and reviews quickly. This reduces the 'throughput' time and incoming samples can then be held in a 'levelling queue' at the start of the process. Trains (defined sequence of tests without fixed repeat interval) are also developed. To do this tasks were separated out into routine and non-routine and non-added value techniques. Any constraints on testing were also identified and removed.
Equipped with this strategy, the teams were able to optimise the way the tests were carried out, collect the task times for this, confirm customer demand (such as when the tests were needed) and optimise the tasks for the roles.
The use of role cards – cards with schedules of activities that need to be performed – were introduced. The analysts were trained to carry out all the tasks on the role cards so that the cards could be rotated. Shift huddles – short routine meetings using visuals such as white boards – were introduced to manage and monitor labs and to ensure good communication between shifts.
The white boards could, for instance, list test methods in rows with name status and action in the columns. Dunne suggested there should be a ‘parking lot’ on the board to show when there are enough samples for ‘wheels’ and ‘trains’ to go. For roles that require more than one day, the white board can show the progress so the person who picks up the next day knows where to start.
Using simple observation of the movement involved around test stations during tests led to the removal of unnecessary equipment and all necessary equipment was then given a designated location and label.
Other simple practices such as defined location and labelling of samples in sample bins in refrigerators enabled samples that were out of date to be located easily and removed for disposal. In the microbial identification lab there was a lack of organisation by priority, so they created a visual management system and queue so that samples could be stored in designated areas and labelled based on priority and type of sample.
time and cost savings
According to Dunne, the process was a big success and saved 5,795 hours per annum. The company was also able to reduce headcount significantly. The challenge was, she said, how to sustain it, and this involves both organisational and mindset changes. ‘It required a shift in mindset from cycle time to schedule adherence,’ Dunne said. Staff needed training for multi-skilled roles and supervisors on coaching and improvement. Individual performance also has to be tied into team-based success: communicate achievements and goals, review metrics and celebrate when they are achieved.
She added: ‘Flexibility is key; do not adhere to the original lean system. If tests change then go back to the beginning.’
How to take a risk-based approach to determining objectionable micro-organisms was discussed by Dr Stephen Rawling, principal microbiologist, GSK. The risk-based approach to safety regulations is meant to ease the cost and burden of unnecessary testing on companies but this less prescriptive approach by the regulatory authorities does require a better understanding of the processes and risks involved. The FDA regulations that cover this concept are:
Rawling said to comply with these regulations and take a risk-based approach you need to know what micro-organisms you are looking for, you need a sampling plan (it is no good testing for the sake of testing and not understanding why you are doing it), and you need to consider who should be involved in determining whether an organism is objectionable.
The first step is to establish what might be considered to be objectionable organisms to your processes and products. ‘Generate your own lists of anything that can pose a risk,’ he advised. It is crucial that the right people are involved in the process and not just the microbiologist. For example, he suggested that it could involve a medical team if there is one, a legal team if there is a legal position, management needs to be involved to ensure they understand why the decision is important, then the regulatory authority and quality teams need to be involved.
The next step is to consider factors such as product use, product nature, the method of application, the user or intended recipient, and the presence of extenuating factors that can make microbes more of a risk, such as the presence of disease or wounds.
Rawling then outlined the four main aspects to the risk assessment: hazard identification, hazard characterisation, risk characterisation and an exposure assessment.
Hazard identification – aims to characterise the importance of the organism to the product and should give the name of the organism, number and prevalence, resistance to intrinsic factors, virulence and toxigenesis.
Hazard characterization – aims to evaluate the nature of the adverse effect associated with the microbiological hazard. For this, consider the consumer and distinctive sub-groups, the severity of hazard, dose response assessment and hazardous level.
Exposure assessment – evaluates the level of organism at the time of application or consumption. This encompasses looking at its occurrence in raw materials, the effect of processing, the occurrence of toxins, potential for recontamination, the effect of formulation, packaging, product storage, consumer use, and the effect of an open shelf life.
Finally, whereas the exposure assessment gives the likelihood of exposure to the organism on use of the product and hazard characterisation gives the likelihood of adverse effects on exposure, the risk characterisation is an estimation of the likelihood of adverse effects occurring and their severity based on hazard identification and exposure assessment.
With such complex concepts, Rawling reminded delegates that the important thing was to ensure that the results of the risk assessment accord with common sense and experience.
For details of future Pharmig meetings and training workshops visit www.pharmig.org.uk.