Assessing AI Feasibility for Pain Evaluation in Hospitalized Cognitively Impaired Older Adults

Purpose The objectives for this study were to examine the acceptability and feasibility of using the artificial intelligence-driven application PainChek Ⓡ to identify pain within older adults with cognitive impairment in the inpatient setting within Ontario. Design A prospective cohort design was used. Methods Six nurses were recruited and completed standardized training. Following the recruitment and onboarding of nurse participants, 25 patient participants were enrolled and assessed using the application. Once patient data collection was complete, qualitative interviews with nursing participants were conducted. Descriptive statistics were used for patient and nurse participant data, and descriptive content analysis captured nurses' experiences using PainChek Ⓡ during the study. Results Participant demographics included 16% with delirium, 20% with cognitive impairment, and 64% with dementia. Average length of stay in the study was 7.8 days, with an average pain score while being assessed as mild pain. Thematic coding was conducted using the nursing metaparadigm as the guiding conceptual framework. Results were validated using member checking. Conclusions Our study demonstrates that it is acceptable and feasible to use PainChek Ⓡ within the older adult hospital setting; however, additional research is warranted to evaluate its feasibility and scalability in real-world environments.