Tapping into AI

 
A human hand rests next to a robot hand on a computer keyboard.

Tapping into AI

 
A human hand rests next to a robot hand on a computer keyboard.

As artificial intelligence breaks new barriers, InMotion chats to a private practitioner who is embracing the new technology and applying it in the clinical setting.

In the ever-changing landscape of healthcare, technological advancements continue to reshape the delivery of patient care.

Among these innovations, artificial intelligence (AI) has emerged as a transformative force, revolutionising various aspects of clinical practice.

One area where the use of AI is forging a new path is clinical note-taking during patient appointments, traditionally a time-
consuming and labour-intensive task for practitioners. 

AI promises to streamline this process, offering unparalleled efficiency, accuracy and insights into patient care.

The use of AI in clinical note-taking marks a significant departure from conventional documentation practices, which often rely
on manual input and transcription. 

AI has algorithms capable of processing vast amounts of patient data in real time, giving practitioners access to sophisticated tools that facilitate comprehensive and accurate documentation while minimising administrative burdens. 

This shift not only optimises workflow efficiency but also allows clinicians to devote more time and attention to direct patient care, ultimately enhancing the quality and effectiveness of consultations.

However, as AI integration becomes increasingly prevalent in healthcare settings, it brings with it a myriad of considerations and implications. 

Privacy concerns regarding patient data security, the need for robust training and education to ensure proficient use of AI technologies and the potential ethical implications of algorithmic decision-making are just a few of the complex issues that accompany this technological paradigm shift.

While AI holds great promise in augmenting clinical decision-making and improving patient outcomes, it is essential to acknowledge its limitations and challenges, particularly in areas such as data accuracy, bias mitigation and interoperability with existing healthcare systems.

Despite these challenges, healthcare’s adoption of AI continues to have strong momentum, driven by the promise of increased efficiency, improved patient care and greater insight into disease management and prevention. 

As we delve deeper into the implications of AI-powered clinical note-taking and its broader impact on healthcare delivery, it becomes increasingly clear that this is a paradigm shift with the potential to reshape the future of healthcare in profound ways.

‘Everyone loathes paperwork. Traditional paperwork takes you away from building rapport with your client,’ says private
practitioner Darren Ross MACP. 

‘There was always a backlog of admin burden with clinical note-taking. 

'It would be lacking some detail; you had to complete it during lunch or in the evening. 

'It also takes attention away from the patients, from having that organic conversation flow.’

Darren is one of the directors at private practice Physica in Melbourne and a musculoskeletal physiotherapist in clinical practice for more than 25 years. 

He says that in his business development role at Physica, he regularly uses AI such as ChatGPT and Canva in the clinic. 

But for the past six months Darren has also been using AI for clinical note-taking, which he says is a significant advance on the SOAP [subjective, objective, assessment and plan] notes that many physiotherapists were taught to use at university.

‘We can do so much better than SOAP notes. I think that with AI, something amazing is on the horizon in terms of the way we’re going to track and record our clinical information,’ Darren says. 

‘We all know that we need to do our notes. 

'We have to create action plans, patient summaries and reports of what’s gone on in a consultation and we have to send that information either to referrers or to our medical peers as part of a multidisciplinary approach.

‘But we’re all time-poor and, as business owners, we’re always harping at our team and our team’s always telling us, “Yes, I know. I don’t have enough time. I need extra allocation or I’m staying back. It’s biting into my work–life balance.”’

Six months ago, Darren teamed up with a physiotherapist colleague and her husband, Sian and Jason Smale, to trial at Physica an AI product that Jason had created for Sian. 

Impressed with the offering, Darren jumped on board and now works to bridge the gap between the product and clinicians across a broad spectrum of the healthcare landscape. 

‘I’m helping clinicians get the power out of AI that the tech guys can deliver,’ Darren says.

Darren says Physica has now used the AI audio-to-text technology for note-taking in more than 3000 consultations with patients, all with prior two-party consent. 

As practitioners go through a consultation, the information is processed and uploaded to the secure management system. 

The program can be customised in terms of its output and has the capacity to distinguish between clinically relevant information and non-clinically relevant information, eliminating the recording of the chitchat between clinician and patient to focus instead on what is deemed to be of clinical significance. 

The captured data is later mined and used for referrals, medical writing, action plans, reports and professional compliance.

‘The program captures the audio of the consultation, transcribes it and then builds it into the clinical notes in the exact style and format that you want,’ Darren says. 

‘You might want your notes in a basic format or you might want a super detailed, drilled-in assessment. 

'For example, I might be having a chat with a patient and as they’re leaving the room, they say something pertinent to the initial part of the consultation. 

'The program builds it in and knows exactly where that information needs to be. 

'It’s like having a virtual assistant who knows just how you want your notes to be done. AI has the ability to reduce the administrative burden for practitioners as well as appeal to the younger, tech-savvy generations who have grown up adopting new technological developments.’

Advantages

Potential advantages of using AI in clinical note-taking include the following.

Efficiency—AI-enabled note-taking can significantly reduce the time spent on documentation, allowing physiotherapists to focus more on patient care during appointments.

Accuracy—AI systems can help ensure that clinical notes are comprehensive and accurate by analysing data in real time and flagging potential errors or inconsistencies.

Standardisation—AI-powered note-taking makes it easier to maintain standardised documentation practices across different physiotherapy clinics, ensuring consistency and quality of care.

Data analysis—AI algorithms can analyse large volumes of patient data to identify patterns and correlations, which can aid in treatment planning, outcome prediction and research.

Accessibility—AI note-taking tools can provide accessibility features such as voice recognition and natural language processing, enabling physiotherapists to create notes more efficiently, even in challenging environments.

Disadvantages

There are also potential disadvantages, including the following.

Privacy concerns—AI systems may store sensitive patient information, raising concerns about data privacy, security breaches and compliance with healthcare regulators such as Ahpra.

Dependence on technology—physiotherapists may become overly reliant on AI for note-taking, potentially leading to decreased critical thinking skills and clinical judgement.

Lack of personalisation—AI-generated notes may lack the individual touch and nuanced observations that physiotherapists can provide, potentially affecting the quality of patient care and rapport.

Initial investment and training—implementing AI systems for clinical note-taking requires a significant up-front investment in technology and staff training, which may be prohibitive for smaller practices or clinics.

Technical limitations—AI algorithms may encounter challenges in accurately transcribing complex medical terminology,
accents or dialects, leading to errors or misinterpretations in clinical documentation.

 

© Copyright 2024 by Australian Physiotherapy Association. All rights reserved.