IDEMUDIA ELOGHOSAVBUMWEN ANGEL

Background: Artificial Intelligence (AI) is transforming global healthcare by enhancing diagnostics and clinical workflows. However, in resource-constrained settings like Nigeria, the integration of AI remains uneven, often hindered by infrastructure deficits and limited training. While awareness of AI is growing, there is a significant gap between knowledge and actual clinical implementation among healthcare professionals. Aim: The study aimed to assess the uptake and level of utilisation of artificial intelligence in clinical support among healthcare professionals at the University of Benin Teaching Hospital (UBTH). Specifically, it determined the level of knowledge, attitudes, and factors influencing AI adoption within the institution. Methods: An analytical cross-sectional study design was employed, involving 409 healthcare professionals including doctors, nurses, pharmacists, medical laboratory scientists, and physiotherapists. Participants were selected using a multistage sampling technique. Data were collected through structured, pre-tested, self-administered questionnaires comprising sections on socio-demographic characteristics, level of knowledge, attitude, uptake and level of utilisation,and factors influencing AI use. Knowledge, uptake and level of utilisation scores were categorized based on a 70% cut-off, while attitud was assessed using a 5-point likely scale which was grouped into appropriate and inappropriate responses and scored using a cut-off of 70%. Data were analyzed using IBM SPSS version 27.0. Descriptive and inferential statistics were used to identify predictors of AI uptake and utilisation. Statistical significance was set at p< 0.050, and 95% confidence interval. Results: Among the 409 healthcare professionals surveyed, the majority were aged 20–29 years(50.4%), female (63.3%), Christians (97.3%), and single (61.9%). Nurses constituted the largest professional group (47.4%), followed by doctors (39.9%), while most respondents were junior staff (55.7%) with less than 10 years of work experience (85.3%). All respondents (100%) were aware of AI, with 61.1% demonstrating good knowledge. While more than half (51.3%) had ever used an AI tool, predominantly ChatGPT, routine clinical utilisation remained low. Slightly more than half (51.1%) of the respondents expressed a negative attitude toward AI in clinical assessment. Positive attitudes (OR = 1.59; 95% CI: 1.034–2.447; p = 0.035) and higher educational qualifications (OR = 3.169; 95% CI: 1.040 9.651; p = 0.042) were significant predictors of AI uptake and utilisation. Major barriers identified included unclear ethical guidelines, patient’s attitude towards AI use, infrastructurallimitations (such as unstablepower and internet), and concerns regarding patient data privacy. However, patients’ attitude was the only significant predictor (p = 0.049) Conclusion: While healthcare professionals at UBTH have relatively high awareness and initial uptake of AI, sustained and routine utilisation remains constrained by negative attitudes and perceived patient’s attitude. These perceptions appear to shape hesitancy in fully integrating AI into clinical workflows. To address this, there is an urgent need for structured institutional training, clear ethical frameworks, and improved digital infrastructure to shift attitudes and support safe, routine and effective integration of AI into clinical practice. Keywords: Artificial Intelligence, Clinical Assessment, Healthcare Professionals, Uptake,Utilisation, UBTH, Nigeria