Amending a protocol
A change to a running study reaches the protocol record at a level that matches its urgency, from an urgent safety measure to a note to file. Protocol amendments are common, have recurring causes and delay a study's implementation.
Urgent safety measures, substantial changes and minimal changes
Where a disrupted trial continues, its protocol amendments are documented according to whether they are urgent safety measures, substantial or non-substantial.
An urgent safety measure is needed when a change has to be immediate, and it does not need prior approval. For example, life-saving medicines can be moved from administration at the clinic to administration at home without waiting for approval. The competent authorities and the ethics committee still have to be informed immediately, and each country may have its own way of doing this. The general principle is to make the immediate change, notify, and then document the change in a substantial amendment.
Significant changes are also substantial changes, but they may be able to wait. For example, participants who still have tablets for a couple of days can wait until the approval is implemented, so the change is not urgent. Such a change still has to be captured as deviations. Guidance encourages aggregating these deviations and their reporting into one larger amendment where that is feasible.
Minimal changes, such as a pause in recruitment, also need to be documented, but without over-reporting them, and a note to file is used where one can be.
Protocol waivers are not allowed in the EU, and they were not allowed during the COVID-19 pandemic either. Participants' rights have to be protected at all times, and the documentation has to be kept properly.
How often protocols are amended
Amendments to clinical trial protocols are widespread, and they can increase costs and delay a study's implementation. The Tufts Center for the Study of Drug Development and 17 midsize to large pharmaceutical and biotechnology companies measured the incidence, causes and repercussions of amendments in protocols approved between January 2006 and December 2008, across a range of therapeutic areas and phases. The 3,410 protocols provided data on 3,596 amendments, and 54% of the protocols were phase I studies, 18% phase II, 13% phase III and 15% phase IIIb/IV.
Across all phases, 58.8% of completed protocols had at least one amendment, and 43% were amended before the first subject's first visit. Each amended protocol had an average of 2.3 amendments and required an average of 6.9 changes, and later-stage phase II and III protocols had slightly more amendments on average, at 2.7 and 3.5. Cardiovascular and gastrointestinal protocols had the most amendments and changes. Larger studies and longer treatment durations were significantly positively correlated with more amendments.
In an analysis of just under 3,500 protocols, phase 1 protocols needed just under two amendments each and phase 3 protocols just over three, almost four. Almost 40% of all the amendments in that analysis occurred before the first study volunteer received a first dose, so the protocol had to be amended before the first patient's first visit, and in phase 1 studies the share was over 50%. Some of those amendments may have been caused by ethical or regulatory evaluation.
What amendments change and why
In the 2006–2008 study, the commonest adjustments were to the population description, including the inclusion and exclusion criteria (16%), and to the safety assessments (12%). The commonest causes were new safety information about the drug (19.5%), requests from regulatory agencies to amend the study (18.6%) and changes in study strategy (18.4%), and design flaws and difficulty recruiting were also commonly cited.
The time an amendment adds
Each amended protocol in the 2006–2008 study took an average of four additional months to put into action, and about half of that time was spent determining what changes needed to be made. Each protocol amendment has been calculated to add 61 days to the conduct of a clinical trial.
Avoidable amendments
Amendments are often necessary, particularly when they affect patient safety, but the results of the 2006–2008 study suggest ways in which some can be avoided, with possible savings in time and cost. The two analyses differ on how many could have been avoided: the study of 3,410 protocols considered one third of amendments somewhat or completely avoidable, and the analysis of just under 3,500 protocols considered almost one in four partially or completely avoidable.
The number and cost of amendments raise the question whether the planning phase of the protocol could have been performed better to avoid them. An avoidable amendment means that the protocol as initially approved was never practical, and that is where protocol writers, and everyone who takes part in planning a protocol, need to contribute to better quality in clinical research.