Reducing Human Error
It is no understatement to say that the pharmaceutical industry is currently undergoing a sea change in attitude and approach to reducing errors caused by personnel, particularly in the operational areas. This area is broadly known as human error reduction (HER). Human errors can be expensive, lead to accidents, and risk product quality. Historically, there have been too many of them in the industry and the emerging approach for responding to them is promising to be fundamental and far reaching.
When something goes wrong in our bioprocess supply chains that could have an impact on product quality, the business is obliged to record and investigate the incident. This is required by industry regulators and accepted by the industry. Current practice is for a manufacturing deviation report to be issued and effective corrective and preventative measures to be implemented. Failure to see this process through (within a reasonable timeframe) can result in regulatory concerns and, in the worst cases, regulatory observations and Warning Letters.
Analysis of the root causes of deviations is most commonly traced back to human error (more than 50% according to a recent survey of BioPhorum member companies). Where human error is the root cause, the most common corrective action set is to retrain the people involved. In most cases, training involves having staff read the relevant SOP and then having a supervisor or trainer validate that training by checking that the procedure is correctly understood by the staff. Unfortunately, evidence suggests that this approach does not eradicate the problem and the errors are repeated (47% of the time, in fact according to the same BioPhorum survey).
Given the pressure to close corrective and preventive actions within a reasonable timeframe and the consequences of not doing so, it is tempting to think that the primary motivation is to get people working in compliance within quality policies and SOPs rather than to develop effective learning processes. These two motivations are not necessarily mutually exclusive, but there is a risk that our GMP compliance culture is preventing us from getting to true root causes. In other words, the corrective part is easier to do rather than the preventive part.
When properly investigated, the incidence of human error is actually lower than normally cited. Is an ambiguous SOP that can easily be misinterpreted an operator error or a document error (written by another human)? One scenario might lead to the operator being retrained (expect a repeat) while another scenario might lead to the SOP being rewritten.
Although this is the status quo today, there are two drivers changing the landscape. First, there is pressure to improve operational performance, which means it is no longer acceptable to bear the cost of repeated errors that can sometimes run into millions of dollars. Second, regulators are realizing that there is a better way to respond to human error. One only has to look to the aerospace and nuclear industries to learn that errors can be driven down to levels that the biopharmaceutical industry can only dream of. For examples closer to home, one can turn to the medical device and chemical manufacturing sectors.
Dissatisfaction with current practice is clearly building. HER is relatively new to our industry despite the extent of manual intervention in the biomanufacturing process, but it is rapidly gaining boardroom attention. But if the current Retrain–Read SOP–Check cycle is not fit for purpose, what are the essential elements of this new landscape? And how should the transition be managed?
Below is a top-10 list of best practices gleaned from a recent BioPhorum collaboration event in which 14 of the top biopharmaceutical manufacturing companies came together to benchmark their approaches and experience for handling human error.
1. Change the perspective and the message. Advanced practitioners talk about increasing human performance and creating a high reliability workforce as opposed to reducing human error. But developing a positive, as opposed to removing a negative, is a subtle yet important difference. The development of these capabilities in a systematic and programmed way is underway in 75% of the top 20 biopharmaceutical companies.
2. Educate people in the science of HER. The nature and causal factors of human error are largely behavioral and psychological. Only when these factors are understood can effective tools and techniques be developed and adopted. There is some great reference material available and specialists are now working in the industry.
3. Encourage a climate of revealing errors as opposed to hiding them. This best practice means removing a blame culture if one exists in the company and replacing it with an open culture that focuses instead on transparency of exactly what happened. This means getting the operators, engineers, whoever was around to recall their observations as the error occurred.
4. Track human errors in a way that underlying trends can be identified and analyzed. Record near misses as an additional source of useful information and to help prevention.
5. Adopt effective investigation techniques, thereby establishing the true causalities. This practice may require the internal Quality organization to change the way it manages and measures some investigations. The 30-day rule-of-thumb target for closing investigations may not be sufficient if an issue is complex and difficult. However, an average closure time well below 30-days across all manufacturing deviations can be easily achieved.
6. Use internal specialists and experts in root-cause analysis. These experts need to be viewed as independent and should be trusted so there is no fear of a comeback. The approach needs to have a light touch, recognizing that middle managers may be defensive about discussing mistakes made in their plants.
7. Create effective lesson-learned processes and feed them across communities of practices. Companies are now leveraging their knowledge-management systems to capture single point lessons as well as broad lessons-learned case studies. In one case, "operations alerts" are formally written up and signed off by a vice-president. The implementation of the learning across the network of facilities is mandatory.
8. Transform the role of trainers from tactical focus to business focus. Trainers need to think more about performance improvement overall than pure skills training. The emphasis must be to developing capabilities that make a real difference.
9. Invert the pyramid, so that there is a "servant leadership" philosophy to support operators and the people at all levels. Senior managers need to ensure everything around operational staff is in place to enable them to get it right the first time. For example, procedures need to be fit for purpose, the work environment needs to be free from distractions, and the operating approach needs to be well thought through.
10. Establish a vision and goals that help move the organization and culture away from reactive HER to active human error prevention. Many of the companies involved in the collaboration event boiled down the above top-10 list to three supporting pillars. They are: better training, open reporting, and rigorous root-cause investigation.
The industry is at the beginning of its long journey to be able to claim it has human errors driven down to a level comparable with other safety critical industries. The early signs are encouraging. The BioPhorum members who gathered recently to share best practices are reporting greater than 50% reduction in human errors following implementation. Of course, statistics need to be used carefully. Tracking human errors through the lens of HER usually leads to more errors being counted because not all are picked up by the traditional means. Counter to this, many errors are recategorized when properly investigated because they are revealed to be system, equipment or organizational issues rather than human errors.
From the HER practices being developed in the leading companies, the industry should be expecting and seeing a fundamental change in over the next two years. New best practices will change the landscape so completely that lagging companies will not be able to rely on Retrain–Read SOP–Check any longer. Time to fetch out the sea-sickness tablets. ?
Simon Chalk is director of the BioPhorum Operations Group, firstname.lastname@example.org