A reply

We thank Drs Hughes and Rai for their interest in our paper [1]. Hughes and Rai used the Ambu aScope in 20 patients and decided that it would be unethical to test this device formally on patients in a clinical trial. Yet they exposed 20 patients to what they subsequently believed to be a poorly performing, potentially dangerous device; we would suggest that all of the features they describe could have been tested in the laboratory on a manikin. Furthermore, obtaining fully informed consent from patients when a clinician wishes to use a newly introduced device for the first time is difficult, and consequently, the first-time evaluation of a newly introduced device in the clinical environment could itself be considered unethical. The primary outcome measure in our study was time to placing the fibrescope. Tip surface collision count (secondary outcome and not the ‘key issue’) has been used before as an outcome measure of fibrescope handling skills in patients [2] and manikins [3, 4]. The crossover study design used in our study allowed participants to apply their skills to each of the tested devices in order to determine which one of the two handled better. When each participant’s level of skill, as measured by the tip surface collision count, is applied to two devices, then the collision count associated with handling each device is a useful indicator of the comparative ease of handling. Therefore, we still consider it perfectly reasonable to have used tip surface collision count as an outcome measure in our study. We consider that manikin-based research is an absolutely necessary first-line research exercise to obtain initial information on a newly introduced device [5, 6], as this does not expose patients to unnecessary risk. Only when devices have been shown to have a similar performance to other established devices in manikins should they then be used in clinical trials on patients. Of course, researchers should apply the same rigorous research design criteria to manikin-based as they do to patient-based studies. We are aware of at least two examples where manikin-based studies exposed a badly designed and inefficient piece of equipment [7, 8]. The recently published 4th National Audit Project (NAP4) [9] reported significant patient harm from devices for which the potential for harm was established in previously published manikin studies of those devices [7, 10]. When evaluating a new device, manikin studies should come before human ones since they may prevent poorly performing and/or dangerous devices from being used on patients.