Methods
All gynaecology RCTs with a date range from 2010-2011 were exported from the CGF specialised register (a bibliographic management database using a ProCite® platform). This database contains benign gynaecology and fertility trials. The gynaecology trials are coded in the database with all the varying conditions and treatments around gynaecology, and the fertility trials are coded simply with the term ‘subfertility’ (along with other conditions and interventions) so the search used a strategy of “does not contain” the keyword “subfertility”, and this provided the gynaecology cohort of trials.
A two-year time-period was a pragmatic decision and was chosen to give an indication of the scope of the problem. We considered that the time lag, from 2011 to present, should have allowed enough time for the trials to be incorporated into the appropriate SRs.
The list of selected RCTs was exported from a ProCite® database into EndNote® (reference management systems), then the text file was imported into an Excel® spreadsheet – the ‘master sheet’. Trials were then excluded if they were either not in the scope of this project or were an inappropriate publication type, these included letters, authors’ replies, and errata. The scope of this project includes RCTs that look at interventions for benign gynaecological health issues. The conditions of benign gynaecology included menopause, gynaecological surgery, polycystic ovary syndrome (PCOS), painful menstruation, endometriosis, adenomyosis, dysfunctional uterine bleeding, fibroids, premenstrual syndrome, disorders of menstrual cycle, chronic pelvic pain and hyperandrogenism. The interventions for these conditions included medicines, alternative therapies, lifestyle interventions, psychological and physical therapies and surgery.
The CENTRAL Register of Studies (CRS Web), a web-based repository of Cochrane trials, which records links between trials in the repository and Cochrane reviews, was searched by trial’s title and/or author’s name to discover whether it had been used in a Cochrane review.
Trials were coded as ‘used’ or ‘unused’, if the trial had been linked to a Cochrane review (used), we then noted if the trial was in the included, excluded, awaiting assessment, or the ongoing trial sections of the Cochrane SR. The ‘used’ or ‘unused’ decisions were double-checked by manually searching the reference sections of appropriate systematic reviews in the Cochrane library.
The unused trials were then categorised first by population (health condition), and then by specific interventions. Following the coding of trials based on population and intervention, trials were checked against existing Cochrane review titles to determine if they could be included in an updated version of the review, and if so, they were coded as ‘existing reviews”. The unused trials that were categorised as ‘out of scope’ (either not an RCT or not in the gynaecology scope) or ‘duplicate’ (either the same publication appearing in the database twice or a separate publication of the same trial, and in this case, we only used the primary publication) were excluded.
Unused trials that could be considered for new review titles, categorised into their health condition and specific intervention, were then used to formulate a list of potential new titles for CGF reviews. From this list, new topics were prioritised if they had three or more associated unused trials. The rationale for this for priority setting decision was that a SR would require analysis of least two RCTs allowing for one to be potentially excluded from the SR.
Results
Between 2010 and 2011, 740 published trials of interventions for benign gynaecology conditions were found by searching the CGFG specialised register. Of these 740 trials, we excluded 15, nine of these were secondary publications and six were fertility trials and therefore out of scope.
Of the remaining 725, CRS Web classified 159 trials as used, and 566 trials as unused (78%) in CGF SRs. Of these unused trials, a further 95 publications were excluded: 73 of these were subsequent publications of the same trials and 22 were found to be out of scope or protocols.
The 471 unused trials were then categorised into 11 gynaecological health conditions: menopause, gynaecological surgery, PCOS, painful menstruation, endometriosis, heavy menstrual bleeding, fibroids, premenstrual syndrome, cyclic disorders (i.e., amenorrhoea), chronic pelvic pain and hyperandrogenism and these conditions were then linked to six intervention groups: medical, alternative therapy, lifestyle, psychological, surgery and physical therapies (table 1). During this process a further 23 trials were excluded due to being out of scope or found to be used in a CGF SR. We also hand-searched the reference sections of Cochrane SRs with similar intervention and populations to double check these used/unused decisions, and we found that 26 of these trials were used by other Cochrane groups, most commonly in Cochrane Anaesthesia reviews, so these were then moved to the ‘used’ trials cohort. Therefore 422 trials of the total 725, were classified as unused (32%) by any Cochrane systematic reviews. 192 of these could be linked to an existing Cochrane SR and could be used if or when the review was updated (figure 1).
Grouping by population showed that the vast majority of unused trials were in the ‘menopause’ category followed by surgery for benign gynaecological conditions (table 2). Medical interventions (generally one medical intervention compared to another) and alternative therapy were the most common interventions in the menopause group, and medical therapy in the gynaecological surgery group (generally the use of different analgesics and anaesthetics) (figure 2).
There are 104 potentially new topics for Cochrane systematic reviews from the 230 unused trials to become SR titles (figure 3), however only 22 of these topics captured three or more trials, the number considered to be the minimum required to produce a useful SR. Menopause had seven new topics, with 3 or more associated trials, the largest being ‘Plant and herbs for menopausal symptoms. Gynaecological surgery had six new topics, the largest was ‘Pregabalin (pre-surgery) for analgesia post abdominal hysterectomy (table 3).