Pretty Good Processes

December 1, 2004
Ronald S. Waife

Ronald S. Waife, MPH, is president of Waife & Associates Inc.

Applied Clinical Trials

Following the examples of these six companies might change your organization for good.

Following the examples of these six companies might change your organization for good.


In keeping with this month's resource issue, the editors asked me to reflect on 2004's best practices in clinical development process, as seen in real life situations. Being somewhat of a non-believer in best practices (who's to say any practice is "best"?, and why is someone else's "best" necessarily best for you?), I did conclude that I could cite a number of "pretty good" process examples worthy of emulation. All of the following are actual examples; the company identities have been omitted to protect their competitive advantage. If we were giving Best Process Practices Awards this year, these would be the winners.

Merging boldly
Mergers are rampant in our industry among companies large and small, and as far as most investors can tell, they run smoothly, even if they don't produce dramatic improvement in the flow of new drugs. But as anyone who has been inside a merger knows, the operational impact can be devastating, and at the very least, highly disruptive. Merger failures include poor understanding of where the strengths lie in the new organization, missing the opportunity to make real operational improvements, trying to keep people happy without regard to business effectiveness, keeping unproductive functions alive,and so on.

One company took a different approach. It decided to act swiftly to analyze strengths and weaknesses in the merged operations, and where necessary, create a new organization chart with re-articulated roles and responsibilities before any final management changes were announced. The analysis was rapid but intense, and informed with extra-company experience. Major re-assignments were made, even of line managers, and in some cases new units were formed for tasks the new company would be much in need of. Instead of the usual postmerger behavior, where a company just looks at the management players and leaves the rest up to them- often resulting in years of slow, distracting upheaval-this company sought to get as much change as possible accomplished quickly. And not change for change's sake, but change that was incisive and potentially transforming. An excellent approach to exploit the operational opportunity mergers present.

Investing in change governance
We have written before how critical it is to govern change effectively-that there must be empowered leadership, with money and executive backing, and an infrastructure to handle the myriad of tasks necessary to make process improvement successful. Too often in clinical development, we retreat to leaderless teams or matrixed relationships; major change is cast as an "initiative," which leaves individuals and managers free to choose whether they will commit. And without resources, both human and monetary, even strong visionaries cannot succeed.

A large pharma introducing a major change in clinical development made a significant investment in change governance this year. It wanted to move quickly but knew this change was mission-critical with high stakes. The company created a new department, which would operate globally to manage this multiyear process transition. It is a permanent unit, with permanent members formally transferred in. Former line operations staff are now fulltime change agents. Top executives are fully informed and committed to its success. The department is fully budgeted and is resourcing ahead of the curve of need, instead of behind-in itself a major accomplishment at most companies. The company has sought to staff all the roles necessary for change management, from study team mentors to investigative site trainers, and expects to achieve successful, compliant process change in much less time than most of its big pharma peers. The investment in governance is paying off.

Learning from mistakes, sincerely and concretely
Many clinical development groups routinely talk about "lessons learned." They properly seek to understand what went right or wrong with a project or particular trial, so they can avoid repeating past mistakes and remember to do again what they did right.

Unfortunately, such efforts are usually little more than one long meeting, a flipchart pad full of observations, and a bunch of action items that individuals are supposed to remember for next time. Too often, the "learning" dies on the spot. There is no formal means of documenting the discussion, no one keeping track of actions taken, and no protection from that learning just walking out the door in the head of a departing employee. Worse, the learning is often incomplete: messengers of bad news are afraid of being shot, or the attempt to seek information may not be cast widely enough. Equally common is that the lessons are articulated too vaguely, in words and opinions too familiar to be truly heard or acted upon.

One biopharma with many lessons from a just-concluded pivotal trial decided that learning as usual was not enough. Instead, it used a formal exercise designed to comprehensively analyze and document what it had learned about this trial's conduct. The examiners looked for root causes of problems they encountered, which enabled them to identify broad but meaningful themes whose improvement will impact future trial conduct. These root causes are actionable, not theoretical, and provide a foundation for a diverse set of clinical operations process improvement steps.

Measuring the vital few
We have written many times that measuring our performance can be so useful to companies understanding where and how to devote process change energies. As we have described, those companies who get "metrics religion" often go into overdrive and end up with a vast array of data, and pages of reports, which managers ignore and staff resent. Another flaw in most companies' metrics programs is a focus only on time intervals (last patient visit to database lock, for example), which are determined by too many contributing variables to derive correct conclusions.

An international biopharma has demonstrated that in one area of its operation, clinical data management, it can apply metrics for resource planning, performance monitoring, and technology assessment without falling into either of these traps. By focusing on the "vital few" metrics which are both reasonable to collect and have operational significance to know, it has been able to minimize the practical burden of performance data collection while increasing the usefulness of the results. As importantly, it is focusing on the units of work which best describe operations, rather than abstract time intervals. Using units of work translates much more accurately into management decision-making.

Knowing it's never too soon for compliance
Emerging biopharmas who are just beginning to see their first drug candidate approach Phase III usually continue the pattern of operations which got them there: use a lot of CROs, focus on the science, and ignore clinical development or postmarketing infrastructure. They follow the same philosophy they followed through discovery: I'll buy that next piece of equipment or hire that next person I need the day I need it and not a moment sooner. That can work in discovery, but having the organizational infrastructure for a successful submission and postapproval support is not something one can buy in a moment-it takes time, anticipation, skill, and practice.

Just such an emerging biopharma has demonstrated visionary practice in its small clinical group as the company gets close to submission. It has instituted a broad and deep review of its clinical SOPs, and fleshed them out in the many areas where they were thin. These SOPs also serve to ensure that the company's clinical program is robust right now, as it expands the scope of its candidate's indications, instead of continuing to rely on outsiders' standards. And it has also recognized the importance of creating a legitimate pharmacovigilance function, with the appropriate tools, so that it is prepared to handle internally the safety monitoring that an approved drug will require. In both cases-SOPs and pharmacovigilance-the company internalized a compliant infrastructure at modest cost for many long-term benefits.

Using the right reasons for vendor selection
My last example for the year's Pretty Good Practices is a company who has gone about technology vendor selection more efficiently than most companies do, because it focused on the right reasons for choice. Instead of spending months of intense effort to develop, define, and document the functional requirements for a clinical IT software application (i.e., how should the commands appear on tab 12), this company recognized that such actions would be reinventing the wheel. Clinical IT applications are so well defined in purpose and use that developing functional requirements for them de novo is akin to specifying how a word processor should work.

This company instead focused on business requirements: that is, what did the company need from the software and its vendor, rather than how the software should look or act. It focused on the imperatives it was facing, the skills of its staff, and the financial and time constraints of its particular development program at this point in time and the foreseeable future. These characteristics are not the same among all biopharmas; indeed they can vary greatly. And having identified what it felt was important, the company found specific, meaningful differences among the products and vendors that a functional review would never have revealed.

Four common qualities
There are common elements in these Pretty Good Practices: boldness, speed, simplicity, and honesty. These are qualities sorely lacking in most pharmaceutical development operations. When you see them, good things are likely to follow. Whether or not the specifics of these stories apply to your company at this moment in time, these qualities will always serve you well.