Clinical Trials

Clinical Trials: A Practical Guide ■❚❙❘General considerationsSource of patientsOne of the first considerations when designing a clinical trial is to establish wherethe patients will be recruited from. The sample size and the clinical outcomes tobe measured will need to be balanced against practical considerations. For studiesthat involve a large number of very detailed observations, it may only be practicalto enroll a small number of patients.Phase I and II pharmaceutical studies usually involve small numbers of patientsand are therefore frequently performed in specialist centers (eg, asthma laboratoriesin hospitals with dedicated research units). Phase III studies, however, typicallyrequire larger numbers of patients. Enrolment for these studies might be conducted,for example, through a primary care physician’s surgery, although the extent ofpatient evaluation might be restricted as a consequence.When researchers select centers to be used to recruit patients, there is oftena tendency to concentrate on ‘centers of excellence’. This is because theinvestigators at these centers are often experienced in clinical trials or experts inthe disease area. It is important to bear in mind, however, that patients at ‘centersof excellence’ can be highly selected, with different patterns of referral from moregeneral centers. It is also possible that more difficult or complex cases are handledmore skillfully, and therefore efficacy and safety might be overstated; this is oftenseen in surgical studies, where there is usually a learning curve associated withnew surgical techniques.Investigators and centersIn order for recruitment to be successful, it is important that patient selection isperformed at centers with enthusiastic investigators who are convinced of thescientific importance of the trial. Investigators must have an understanding ofwhat is required for the trial so that they can clearly communicate this informationto potential participants.Investigators also need to have access to the facilities and equipment that arerequired to perform the trial-related tests and measurements (eg, echocardiographymeasurements in a study of heart valves). Patient selection can be compromisedif investigators are performing competing studies; it will certainly be compromisedif the investigators do not have enough dedicated research staff to screen patientsand perform the trial-related procedures.49