Responses of local research ethics committees to a study with approval from a multicentre research ethics committee

Participants, methods, and results

We made 113 submissions on behalf of local investigators to 99 local research ethics committees between September 1998 and September 1999. We analysed the committees' responses to the first submission. A committee was classified as “fast track” if the administrator stated that the submission would be reviewed by an executive subcommittee, as recommended by the guidelines. Our main outcome measure was response time, defined as the number of days between arrival of the submission and the date on which written confirmation of the committee's decision was typed. We considered a response time of 21 days or less to be satisfactory since this was the upper limit suggested by the guidelines.1 In a survey of 26 committees, submissions arrived a median of three days (range 1-7) after they were sent. For the other committees, we took the date of receipt to be seven days after the documents were sent. We defined earlier submissions as those received before April 1999.

Submissions were classified as approved if complete or conditional approval was granted, even if requests for clarification were made to the multicentre research ethics committee, the trial steering committee, or the local investigator. Requests for opinions from third parties, failure to grant at least conditional approval, and requests for amendments to study documents were classified as non-approval. Requests for minor amendments to study documents (such as changes to letter headings) were classified as approval, except in two cases when the local committee asked to review such changes before granting full approval.

Fewer than half of the committees used a fast track system (table), with 21 (44%) of earlier submissions and 23 (45%) of later submissions being reviewed by an executive subcommittee. A third of the committees reached a decision within 21 days. There were no significant differences in median response times between standard and fast track committees, or between earlier and later submissions. Seventeen committees did not approve the study after the first review. One committee had not had a quorum for over six months and, when it did meet, requested an opinion from a third party. Another committee recommended several amendments which the multicentre research ethics committee did not consider important enough to merit global amendments to the study protocol. The resulting impasse was unresolved six months later. The required number of complete copies of protocols and documents from the multicentre research ethics committee was significantly lower for local committees that used a fast track system (Mann-Whitney rank-sum test: z=3.11, P<0.002). However, four fast track committees requested 12 or more copies.

Comment

The two tier system of ethical approval of multicentre research was intended to combine rigorous local review with expedient timing.2 We found that only a third of committees responded to submission of a study that had been approved by a multicentre research ethics committee within the recommended period of 21 days. A sixth of committees did not approve the study after the first review. There was no evidence of more efficient review or wider adoption of the NHS Executive's guidelines six months after these guidelines had been issued. Fewer copies of documents were required by committees using an executive subcommittee for fast track decisions, but these committees did not make faster decisions. Our findings echo the comments of other researchers, that the two tier system of ethical review retains the inefficiencies of the former system.3