Source: Practice Index 25th April 2024
In recent years, the explosive growth of artificial intelligence or AI-powered applications has brought about a transformative wave across various industries, and healthcare is no exception. Indeed, AI is often touted as the ‘silver bullet’ that could save healthcare, freeing up time across the board, thanks to faster diagnosis, risk prediction and analysis. It’s no surprise, therefore, that numerous AI technologies are already in use in the NHS, and clinicians are trialling (and being trained in the use of) a long list of them. But what about Practice Managers? How are they using this much talked about technology?
Talking to Practice Managers up and down the country, it’s the rapid development of highly advanced generative AI chatbots, including ChatGPT and Gemini (formerly known as Bard), powered by large language models (LLMs), that are of most interest. A recent survey by healthcare leaders found that over a third (36%) of respondents were using ChatGPT in their daily work – with many more planning to enter the world of LLMs shortly. But what exactly are using these tools being used for?
“For me, it’s all about speeding up writing, whether that’s responses to complaints, emails to staff or pulling together posts for social media,” one Practice Manager from a large practice in south-west London told us. “At the end of a long day, when I’m trying to find the right words for Facebook, ChatGPT can be really helpful.”
Complaint handling
Almost every Practice Manager we spoke to who used generative AI was using it to handle complaints. It’s clear that the chatbots’ responses need editing, but in terms of providing the structure and outline of a response, they can save a significant amount of time. One Practice Manager estimated they were saving a day a month, which can be a considerable relief in their busy schedules!
For common complaints, one Practice Manager explained, “We simply ask the AI chatbot to draft a response, which we then review and edit before sending it back. We need to provide more specific instructions for more complex queries, but the chatbot still provides a solid starting point that we can customise. This process saves us significant time compared to writing a response from scratch. Of course, we always remove any personal data before using the chatbot.”
Further uses of LLMs in practice management
Our conversations clearly show that Practice Managers are using LLMs to help with numerous tasks. Some use the tools to automate various administrative tasks such as form processing and data entry. Other uses include:
Training and onboarding: Chatbots can be used to train new staff members, providing them with instant access to information about practice procedures, protocols and frequently asked questions. This can streamline the onboarding process and ensure consistent knowledge across the team.
Feedback collection and analysis: Chatbots can assist in collecting patient feedback via surveys or feedback forms. They can also help to analyse this feedback to identify common themes and areas for improvement in the practice.
Documentation and reporting: The tools can help to generate reports and documentation, such as patient summaries, inventory reports or financial reports, by structuring data and drafting initial versions of these documents.
Staff scheduling and rota management: Chatbots can assist in creating and managing staff schedules, considering various constraints and preferences to ensure that the practice is adequately staffed at all times.
Mental health support for staff: On a supportive note, Chatbots can serve as a first step in providing mental health support to staff members by offering resources, self-care tips and guidance on when to seek professional help.
Project management: Outside of LLMs, Practice Managers are also adopting AI-powered apps. These include Todoist AI, which helps to manage projects, and the various transcription apps that are available.
The limitations of LLMs
While LLMs hold great promise and have the potential to revolutionise clinical practice, several barriers impede their success. One of the most notable limitations of LLMs surrounds privacy and data security.
Security concerns remain a significant obstacle to the immediate application of LLMs in clinical care. Currently, information inputted into publicly available LLMs lacks anonymity, leaving patient confidentiality and privacy unprotected. At the time of writing (the world of LLMs is moving at pace), there’s no option to redact any entries, so what’s entered stays entered.
There are also question marks over GDPR compliance. We received different answers to our questions about ChatGPT and GDPR compliance from legal firms while researching this article. Therefore, ensure you’re not sharing personal details.
Finally, LLMs such as ChatGPT are known to produce logical but factually incorrect outputs. Clinicians must be mindful that current non-domain-specific LLMs are not designed to provide medical advice, and any medical interpretations should be fact-checked against prevailing clinical evidence/guidelines.
“Great tools, but use wisely”
Overall, tools such as ChatGPT and AI-powered apps have the ability to save substantial amounts of time for practice teams, but only if used correctly. As one Practice Manager told us: “We’ll inevitably use AI chatbots more and more. It’s another tool that can save us time, so why wouldn’t we? However, we must actively embrace and educate ourselves about the capabilities and limitations of AI and use the tools accordingly. After all, a builder carries more than just a hammer!”
Click here for the source.
