Specific high-level opportunities for Lean implementation include (1) reducing lead time by dedicating smaller teams to one
project instead of having larger teams working on multiple projects; (2) reducing information transfer by maintaining project
team memberships for successive project stages; (3) capturing lessons learned and organizational knowledge to avoid rework;
(4) using screening evaluations (e.g., a standardized set of initial experiments) to quickly identify process problems; (5)
miniaturization of process development experiments in scale-down models; and (6) investing in improvements that have a high
probability of success. Detailed examination of staffing variations for similar projects can identify improvement targets.
Specific practical Lean approaches are evident in the industry. In one example, an accelerated deviation management process
was implemented, reducing lot disposition time from an average of 85 to <50 days with low variability.8 In another case, lead time for atypical resolution was reduced by eliminating procedural redundancies and by co-locating
quality personnel at the processing site for in-process and prompt postexecution batch record reviews.9
Just in Time
The "just-in-time" concept of Lean aims to supply what is needed, when it is needed, in the right amount; in other words,
supplying the right "part" at the right time in the right amount. The "part" can be a process stream, development data, or
material for clinical trials. The effort uses the minimum materials, equipment, labor, and space required to meet the required
quality attributes.
The 3P Framework: Designing for Manufacturability
The production, preparation, and process (3P) framework develops an appropriate system in the shortest time to satisfy design
quality, production, and cost requirements.10 This concept is also known as designing for manufacturability
8
or producibility.11 Once such processes are commercialized or even initially scaled up, they perform reliably with minimal deviations.8 This framework emphasizes the line-of-sight to the manufacturing facility as the eventual customer of the process development
group. Critical to its success is the simple documentation of key development knowledge (e.g., one-page technical summaries,
trade-off curves showing safe operating regions).12
Six Sigma: Reduce Variation
Variation is a key concept of Six Sigma, which uses the "DMAIC" approach to define, measure, analyze, implement, and ultimately
control variation. To identify improvement targets, typically variability is quantified using at least 30 data points. For
biopharmaceutical processing, in which the number of events may be low (e.g., campaigns per year, batches per suite), campaign
subtasks (e.g., raw material release, batch record review) generate sufficient data in a faster timeframe. Quantification
of variation highlights areas for additional control. In one process, for example, yield was improved through tighter control
of two process variables.6 Control charts and other statistical tools have been used to compare performance level and variability between successive
product campaigns.13
A few other examples are documented. In one case, after calculating resources used to address unplanned events, a comprehensive
root-cause analysis was developed. Most of the wasted time was waiting for non-core operations such as equipment availability,
batch record signatures, or computer data entry.6 Others replaced retroactive incident investigations with proactive investigations for potential process problems.14 Another company implemented the operational excellence concept of "Quick Response Manufacturing" by scheduling to avoid
weekend holds (i.e., running cultivations over the weekend) and reducing purification step variability, and dropped lead times
for protein production from 11 days to <5 days and increased throughput by 60%.15 In another example, root-cause investigation was used to identify improvements that made it possible to reduce contamination
rates several-fold.16
|
