Nearly 60% of all protocols used in clinical trials for new drugs are amended during the study, but one-third of those changes could have been avoided, according to a new study.

Completed protocols across all clinical trials incur an average of 2.3 amendments, with each requiring an average of 6.9 changes to the protocol, and this leads to significant unplanned expense and delay, says the study, which was conducted by the Tufts Center for the Study of Drug Development (Tufts CSDD) in order to quantify the impact of protocol changes on clinical trials.

"Although amendments to protocols of clinical trials are sometimes necessary to optimise study results and ensure patient safety and ethical treatment, study sponsors can minimise the number of protocols through better initial study design and improved recruitment of study volunteers," commented Ken Getz, who carried out the study. Mr Getz is a senior research fellow at Tufts CSDD and assistant professor at Tufts University.

He forecasts that a high incidence of protocol amendments is likely to continue, given the analysis' finding that the mean number of amendments was positively and significantly correlated with the rising number of procedures per protocol, study length and number of investigative sites participating in each clinical trial.

The research is reported in the September/October issue of the Tufts CSDD Impact Report and notes that until now, the incidence, causes and impact of protocol amendments have never been systematically quantified across the pharmaceutical and biotechnology industry.

Among the study's other findings are that: 
- across all phases of clinical trials, 43% of protocol amendments occur before first patients are enrolled, with amendments more likely to occur in Phase I studies;
- more than half of all protocols require one or more amendments, with Phase III studies requiring the highest number of changes per amendment, at 8.5;
- one-third of all amendments relate to protocol description and patient eligibility criteria; and
- the median total cycle time to identify and resolve a protocol problem is 61 days.