Background UCB Pharma have designed and developed an innovative electromechanical, universal injection platform for subcutaneous drug delivery (eDevice), providing full electronic management of the injection process via an interactive graphical user interface (GUI). Here, the eDevice was tested for use with certolizumab pegol (CZP).
Objectives To validate that patients (pts) can self-inject safely and effectively using the eDevice, and that healthcare professionals (HCPs) are able to configure the eDevice and demonstrate its operation by performing a simulated injection, in the expected scenarios and environments, using the instructions for use (IFU), eDevice GUI and training cassette.
Methods The study took place in Stockport and London, UK, with 15 pts with moderate to severe rheumatoid arthritis (RA), and 15 HCPs. An injection test scenario was performed in a room simulating a home (pts) or office (HCPs). Pts/HCPs were supplied with the eDevice, IFU, training cassette and a CZP cassette containing sorbitol, mimicking CZP viscosity. Using an artificial skin pad attached to the intended injection site (abdomen/thigh), pts were asked to perform 2 simulated self-injections of CZP loading dose (LD; 2 x 200mg) and determine the next scheduled injection date. HCPs were asked to configure CZP LD (2 x 200mg Q2W at Wks 0, 2, 4) and 2 maintenance regimens (200mg Q2W, 2 x 200mg Q4W), and perform 2 simulated pt injections. No training was provided; pts/HCPs could use only the IFU, GUI and training cassette for guidance.
An independent human factors researcher scored pts and HCPs against 18 and 22 predefined critical tasks, respectively. Pts/HCPs had 3 attempts for each task. Critical tasks were those required to safely and effectively use the eDevice, scored on the final attempt as pass, pass with difficulty, or fail. Pts/HCPs were then interviewed for 120 mins to identify root causes (RC) of difficulties/failures. RC analyses were reported as perception failure (misreading GUI/IFU), cognition failure (misinterpreting GUI/IFU), action failure (participant understood task but performed incorrectly), prior experience (assumed eDevice was same as other injection devices), test artefact (if pt said “I did not do X as it is not a real situation”) or medical condition (symptoms caused use error).
Results All 15 pts and 15 HCPs were able to inject the full simulated CZP LD. 330 critical tasks were assessed for HCPs, and 270 for pts. 4 task failures were recorded for HCPs and 8 for pts (Table). The 1 HCP failure to correctly configure the eDevice was due to poor English understanding, not device design (post-study interview RC analysis). 14 pt and 26 HCP critical tasks were passed with difficulty (Table); the most common reason being an inability to find information within IFU (RC analysis). There were no needle stick injuries or unexpected user errors.
Conclusions All study participants could perform a simulated injection and administer a full CZP LD safely and effectively. There were no incidences of potential over or under dosing, either by self-injection or incorrectly reading the next treatment date on the GUI. All HCPs able to read the instructions could configure the eDevice. Most difficulties encountered by pts/HCPs could be rectified by improving the IFU design.
Acknowledgement The authors acknowledge Costello Medical Consulting, funded by UCB Pharma, for writing and editorial assistance.
Disclosure of Interest B. Domanska Employee of: UCB Pharma, S. Poon Employee of: UCB Pharma, C. Pichon Employee of: UCB Pharma, O. Stumpp Employee of: UCB Pharma