The healthcare industry is no stranger to Artificial Intelligence's (AI) transformative potential. From improving treatment plans to diagnosing diseases with remarkable precision, AI has already proven its value in clinical settings. Research has shown AI imaging technology has reduced the workload of clinicians by 30-45%.
Yet, despite these breakthroughs, the integration of AI in patient-facing experiences has been uneven and sluggish. What is holding the sector back from fully leveraging AI to improve patient care?
The chatbot cure? Why 90% of healthcare AI is stuck in the waiting room
For many years, the health service has struggled with long wait times, rising numbers of patients waiting for treatment, operational inefficiencies, and staff shortages, affecting the consistency and continuity of care patients receive.
To address these issues, AI has been deployed to automate basic tasks and alleviate some of the burden on healthcare workers. Generative AI tools, such as chatbots and virtual assistants, have been successful in automating appointment scheduling, prescription requests, and test result management. There are pockets of innovation in the NHS where chatbots are having a positive impact.
Yet, despite the evident value that AI can provide, adoption has been far from seamless. For instance, only 10% of patient interactions with conversational AI or chatbots fully resolve their queries without requiring escalation to a live agent. This is because many of the AI-enabled chatbots in use today are too simple and cannot make the step into more advanced, secure automation, ultimately providing sub-par patient interactions.
Smart tech, slow progress - why AI can't fix the NHS's 7.5 million patient backlog yet
The stakes and hurdles are much higher in a complex industry such as healthcare, hence the slow integration. Healthcare demands rigorous evidence and safeguards when introducing new tools in order to protect patient safety and ensure compliance with privacy policies such as GDPR and DSPT. However, AI systems learn by processing large volumes of data, which the NHS collects and stores, but the barriers to accessing it are understandably high, given its sensitive nature.
Training AI tools with large data sets lays the foundation for this technology. It leads to improvements in the accuracy of responses, reduces bias, and provides patients with more personalised interactions. If continuous, it will create a system that can handle complex, multi-step patient queries, such as referrals to specialists and appointment scheduling.
Currently, the NHS is facing one of its biggest challenges with a waiting list of 7.5 million, but properly trained and implemented chatbots can play a role in alleviating this.
Nevertheless, implementation is only part of the battle. Organisations must continually invest in their AI infrastructure to progress with advancing technology and proactively address emerging regulatory challenges. The issue is the pace at which the technology develops; governance simply cannot keep up. The resources and expertise required to monitor algorithms and fill regulatory gaps are sparse, and without sustained investment in both technology and talent, the ability to ensure AI remains safe, accurate and effective can be compromised.
Progress stalled, the investments AI needs to transform care
Despite these challenges, there are secure, compliant AI tools that can easily conform to industry regulations. It is possible to balance innovation with safety and privacy in the right way, offering healthcare organisations a way to integrate AI responsibly. Clait, Israel's largest and leading healthcare organisation, shifted its communication securely from phone to messaging with a unique AI-driven solution, automating 47% of patient requests and boosting patient satisfaction by 28%, showing improved productivity and patient experience is possible.
However, for the industry to evolve with AI as a whole, bigger commitments need to be made. Last year, the UK government announced that the NHS was one of the beneficiaries to receive a share of the £32 million in AI funding. Despite this being a step in the right direction, the lack of strategy and standardisation hinders the possibility of AI being appropriately scaled and adopted.
The solution requires developing a nationwide AI standard for the industry, including inputs from policymakers, healthcare professionals, AI developers, and patient advocacy groups. Ultimately, standardised guidelines will enable AI to be implemented more effectively, enhancing patient-facing services through improved efficiency, accuracy, and personalised care.
Secondly, the NHS has made significant strides toward more sophisticated and safe data sharing through developing Secure Data Environments (SDEs). Supporting initiatives like this will enable AI development and deployment to progress within the industry. For instance, these data storages uphold the highest standards of privacy and security, which are then used for research and analysis without the raw data ever leaving the SDEs. This approach not only safeguards sensitive information but also empowers the NHS to unlock the full potential of AI through correct algorithm training.
The future of care is connected
AI has shown great promise in healthcare, offering vast potential in many areas. However, its integration into patient-facing services is complex. Practical steps such as continuing investment in SDEs, supporting data sharing between NHS trusts and integrated care systems, and nationwide standards for AI adoption are all processes that need to be supported further. Conversational AI chatbots, in particular, have the potential to transform patient services by enabling seamless, real-time interactions, enhancing accessibility, and driving greater efficiency in healthcare delivery across the system.