Q. Can you offer any best practices for change management?
A. I have dealt with change management for many years at various pharmaceutical companies and I have seldom found a program
that is fully compliant and that works effectively and efficiently. Based on my experience and input from other PAREXEL consultants,
among the most common problems are impact assessments and documentation of the change, as described below.
Impact assessment: In many cases, proposed changes are not properly and thoroughly assessed for risk. Appropriate departments should always
evaluate changes to determine their potential impact on a validated process, critical equipment, and product quality, as well
as the potential regulatory impact.
Documentation of the change: When a change is being considered, the topic is often widely and frequently discussed, and is clearly understood among the
site staff. When it comes to the documentation, the scope of the change, rationale, and justification for the change, as well
as other details, little is written since "everyone knows what is going on." Then, when a third party (read: FDA Investigator)
reviews the documentation package, it lacks details, which gives the impression that the evaluation or risk assessment is
inadequate or incomplete. The documentation package should tell the complete story such that the third party can reach the
same conclusion without provoking many questions.
Below are a few additional best practices for running an effective and efficient change-control management program.
Zone in on: Regulation
Do not use "planned deviations."
Many companies use the erroneous concept of "planned deviations," which are not deviations at all, but rather, changes. These
changes are often temporary and handled within the deviation management system or similar program. Deviations are unexpected
events wherein some control is lost and the product quality impact of the event must be evaluated retrospectively. No one
wants to "plan to deviate" and deviations raise a red flag to regulators.
Changes on the other hand, are planned events where the potential impact is evaluated prospectively. This holds true even
for a temporary change where the change is for one batch–the potential impact of the change on that one batch must be evaluated
prior to the change being implemented, even if immediately before the batch is to be manufactured. Therefore, changes that
are referred to as "planned deviations" should be managed under a change-control program, which should provide for emergency
and temporary changes.
Avoid unauthorized changes.
Individuals sometimes implement a change outside of the change control mechanism. Such changes may be well-intentioned "good
ideas," but nonetheless are unauthorized and undocumented. A common example of an unauthorized, "good-idea" change may occur
with respect to modification when "like-for-like" parts are not available and a repair is made with a nonequivalent part.
In other cases, some change-control processes may be so cumbersome that changes cannot be approved on short notice. Implementing
a change prior to a full evaluation and approval, however, constitutes an unauthorized change.
Limit the number of change management SOPs.
Many companies have an excessive number of SOPs that relate to changes. Sometimes, each department has its own change control
procedure or each topic (e.g., processes, analytical methods, facilities, equipment, materials, documents, computer systems)
is managed under a separate SOP. These situations cause confusion, inconsistencies, overlap, and gaps. One company had 27
SOPs that addressed changes in one way or another, such that it was unclear where to even begin the change process.
Watch for cumulative changes.
Companies must consider that multiple individual changes categorized as "minor" on their own merit may have a cumulative effect
on product quality and the impact of multiple changes should be evaluated.
Follow-up and closure.
Companies often fail to follow-up after a change to verify that it was implemented as approved, provided the intended outcome,
is effective, and has not caused other "unintended" changes. The follow-up should provide for closure in a timely manner.
The appropriate timeframe for follow-up and closure can vary dependent of the type of change, but should be verified before
product affected by the change is released. Changes should not be left "open" for an excessive length of time.