Randomised Clinical Trials. David Machin
Чтение книги онлайн.
Читать онлайн книгу Randomised Clinical Trials - David Machin страница 36
![Randomised Clinical Trials - David Machin Randomised Clinical Trials - David Machin](/cover_pre950308.jpg)
3.12 Organisational structure
The contents of this section will depend to a large extent on the size and complexity of the design of the trial. For example, protocol SQCP01 (2000) involves only two clinical centres but different countries with very different first languages, and which are geographically quite distant. Because this trial is organisationally complex, and involves many clinical disciplines, including specialists in surgery, orthodontics, and speech therapy each with roles to enact over a very long period of 17 or more years, keeping track of the individuals concerned is a major challenge. However, each centre has long experience of dealing with such complex issues so the protocol runs with four named craniofacial/plastic surgeons, four orthodontic coordinators, three speech therapy coordinators and two research coordinators. At the statistical centre a medical statistician and clinical project coordinator are designated to the trial. The protocol contains full names, addresses, telephone, fax and email addresses of these individuals. As may imagined, the actual individuals concerned will no doubt change as the trial progresses forward in time.
In contrast, the SQNP01 (1997) was conducted within a single centre and three clinical coordinators were identified, one representing radiation oncology, the other two medical oncology. The conduct of this trial reflected the day‐to‐day management practices of the centre concerned. At the statistical centre, a medical statistician was designated and a nurse coordinator had shared responsibilities for the trial within both the statistical centre and the clinic from where the patients were recruited. The protocol contains full names, addresses, telephone, fax and email addresses of these individuals.
In situations where the demands of a trial are somewhere between these two extremes, the protocol development team must ensure that the necessary organisational structure is in place and each component thereof knows of their individual responsibilities. One important role of the trial office is to help maintain this functionality throughout the life of the trial.
If at all possible, the protocol should fit as closely as possible within the confines of current practice in the centres concerned, with the proviso that the aims of the trial are not compromised by so doing. This facilitates acceptance of what is new in the protocol from the local team and thereby should help with the smooth running of the trial and hopefully maximise recruitment rates.
3.13 Publication policy
In most clinical trial groups, there will inevitably be several members of the team and should the trial be multicentre then the full team of collaborators may be very numerous. It is useful to have stipulated a clear policy as to who the authors of the final publication will be and in what order they appear on the title page. Provided this is clear, and agreed by all concerned (including latecomers to the trial once it is ongoing) this need not be in included in the protocol itself. If a large number of collaborators are involved then it may be more sensible to publish under a group name with a full list of the investigators included as an appendix to the report but again, this should have been discussed and agreed from the outset.
Example 3.27 Protocol SQOLP01: Comparison of steroid with cyclosporine for the topical treatment of oral lichen planus
PUBLICATION POLICY
The results from the different centres will be analysed together and published as soon as possible. Individual clinicians must not publish data concerning their patients which are directly relevant to the questions posed by the trial until the main report is published. This report will be published under the name of the Asian Lichen Planus Collaborative Study Group listing all members of the group and any others contributing investigators.
In the event, the editor of the journal publishing the eventual report of Poon, Goh, Kim, et al. (2006) refused publication under a group name although a full list of contributors was permitted. Thus, collaborative groups may need to verify the policy of the target journals before stipulating a formal policy in this respect.
Example 3.27 also underlines an important requirement that individual groups should ‘not publish data concerning their patients which are directly relevant to the questions posed by the trial until the main report is published’. An important reason for this is that such publication can only (at best) refer to a subgroup of the total number of patients recruited to the trial. This number is consequently less than that stipulated by the design and so any analysis will be underpowered for the hypotheses under test. So, for example, such an analysis may report ‘no statistical difference’ in situations where the whole trial data may conclude the opposite. Premature publication by an individual group may also jeopardise the acceptance for publication of the report of the final trial results.
3.14 Trial forms
As we recommended earlier, since recording the patient data is integral to successful trial conduct, inclusion of the trial forms into the protocol itself is often desirable, even when they are quite simple in structure. However, if these are web rather than paper‐based how these are to be presented to any protocol review committee may depend on local circumstances.
The forms should be developed in parallel with the protocol, and may (depending on local regulations) have to be submitted for approval with the trial protocol itself in any event. The number, structure and complexity of the forms required for a trial will be very trial‐specific but, as a minimum, there will be forms containing subject‐specific information relevant to the registration and randomisation process including the intervention assigned, those encapsulating eligibility and other baseline characteristics of those recruited, and a form for the endpoint assessment. In almost all circumstances there will be many more than this and most trials will include special forms for recording details of, for example, any surgical procedures undertaken or unexpected adverse events should they arise.
In general terms, the forms for the clinical trial should focus on essential detail that is necessary to answer the question(s) posed by the design and should not be cluttered with irrelevant items. This focus keeps the clinical teams aware of the key issues and takes them less time than having to record inessential details. Consequently, there are likely to be fewer errors. The completed form also becomes easier to check if there are fewer items and thereby reducing the data management processes and speeding up checking. In this way, any problems remaining can be fed back to the clinical teams more rapidly – which again reduces the workload at the clinical recruiting centre. The briefer the forms, and indeed the simpler the trial procedures, the easier it becomes for collaborators and the more rapidly they are likely to recruit the patients required. But this must be balanced against the need to ensure that the forms do contain all the necessary information that will be required for the analysis. However, the experience of many groups indicates that most trials collect far too much information that is then never analysed nor reported on.
Forms may need to include patient management details, such as the date of the next follow‐up visit, or to confirm if an action has been taken such as the despatch of a laboratory specimen or the completion by the patient of a quality‐of‐life questionnaire. It is best if these are kept to a minimum, and located in a distinct part of the form (perhaps the last items) so that when the forms are received for processing at the trials office these items can easily be distinguished from variables that must be included in the trial database.
3.15