OBJECTIVES: Given the importance of emerging health technologies in patient care, the Lung Artificial Intelligence-enabled Digital Solution (Lung AID) Pilot Study was designed to evaluate the feasibility of conducting a future randomized study using a digital intervention in real-world clinical settings. As study enrollment was a key barrier during study implementation, we conducted a qualitative analysis to identify and address the barriers to enrollment.

METHODS: Lung AID is a prospective, observational study targeted to include 100 patients with lung cancer at 10 German cancer centers implementing the Kaiku® Health platform. This platform enables patients to report their health status in real-time to their healthcare providers who may react promptly to health status changes. Investigator meetings were held to identify recruitment barriers and proposed solutions. Further, site visits were performed for motivational reasons, and to identify site-level recruitment barriers. These data were systematically analyzed using the fishbone analysis for root cause identification.

RESULTS: Identified barriers were categorized as technological readiness by patients and clinics, study adoption at clinics, inconvenience of study for patients, and lack of eligible patients. Based on identified barriers, changes were made to the study design including expansion of eligibility criteria, addition of two sites with expanded pre-implementation support, and enhanced patient education materials. Enrollment improved after these changes were implemented; however, recruitment remains low.

CONCLUSIONS: While it is feasible to implement a digital health application into the patient treatment journey, site engagement and patient recruitment remain challenging. Regular assessment and communication with sites and research teams are required to address emerging challenges. Digital health research requires a flexible design and considerations of the real-world setting and barriers that may be encountered by site staff and patients. Addressing these implementation barriers early and on an ongoing basis is critical for future research.