NHS staff recognise the potential of AI, but need assurance to confidently implement AI-enabled services

This summer we carried out a survey of healthcare decision makers’ perceptions on the adoption of AI-enabled services. 

The respondents were predominantly NHS staff from community, primary and secondary care with around half of these being clinicians (graph1 ).  

Graph 1. Respondents segmented by healthcare sector. 

Which tasks could AI help with most?   

Confidence levels were highest for administrative services such as appointment booking and management, patient notes and letters, and staff rota administration. 

Within clinical contexts, respondents felt positively about image analysis and first line triage to support clinical decision making. 

In which care settings could AI-enabled services have the most impact? 

Respondents believe the greatest impact of AI is likely to be in primary and secondary care, versus only 50% who saw the opportunity in community care. 

What are the concerns around AI-enabled services? 

The perceived obstacles to widespread AI adoption include bias and explainability,  data protection/regulation, quality and safety, clinical errors and ethical use and transparency. These concerns were echoed by low confidence levels in how to use AI-enabled services safely.

 Graph 3. Barriers to AI adoption weighted by prevalence in respondent feedback 

 Our conclusion 

Healthcare decision makers seem to share a guarded optimism for the potential of AI. However, several key areas of concern must be addressed to build confidence. 

Effective governance, responsible evidence generation, strategic implementation and ongoing clinician education can all play a part in building confidence in AI-enabled solutions. 

At Visiba we believe another crucial factor is working closely with clinicians to co-create AI-enabled solutions. We are a clinically-led organisation with clinicians embedded in all stages of our product lifecycle. We also operate a unique feedback loop which harnesses insights from the clinicians using our product to continually improve its accuracy.

We are currently working with a number of vanguard sites to test, refine and grow the confidence needed for the safe use of our AI-enabled services. 

If you have any additional thoughts or queries you’d like to raise with us, then do get in touch.