The emperor of all maladies has stalked my family this year. From my father (multiple myeloma) to my cousins (multiple myeloma and throat cancer, respectively), and myself (melanoma), it’s been the year of cancer, and we are only halfway through it.
We’ve been bombarded with medical information in meetings with oncologists and doctors, some of it quite complex and scary. After emerging bewildered from one medical appointment, unsure exactly of what was agreed, I decided to record the following ones.
I feed the recordings into Otter.ai, a software application I use on my phone to automatically transcribe the interviews I do each week. Automatic transcription has transformed journalism, cutting hours out of a process that used to involve listening back to interviews, then laboriously typing out the interviewee’s quotes.
But in the past 18 months or so, Otter has also harnessed the latest developments in artificial intelligence to present conversation summaries alongside the transcriptions. It will tell you the key points discussed and highlight any decisions or action points that came out of the meeting. The AI-generated notes provide a reliable reference for discussing our medical issues.
General practitioners, overwhelmed with patients and admin, are doing the same thing. In February, I met South Auckland GP Karl Cole, who is an enthusiastic adopter of AI for recording and summarising his meetings with patients. “I don’t use the word lightly, but this has been a game changer, and probably stopped me from burning out,” he says.
Cole uses the Nabla Copilot app, and estimates it saves him about an hour a day of writing up clinical notes. It allows him to focus more on direct patient care and complex decision-making. The summary notes also empower patients to take a more active role in their healthcare.
My own GP doesn’t use AI yet, and Manage my Health, the online portal the practice uses, gives a fragmented picture of my health. There are test results that are hard to decipher, and no record of what was discussed during appointments. Hospital-related documents don’t necessarily go into Manage My Health.
Primary care providers such as GPs are relatively free to introduce AI tools, but a lot more regulation and red tape needs to be navigated to allow AI to be more widely used in the wider health system.
Cole sees much greater value being realised when AI-powered clinical notes are connected to your electronic health record so current and past medical conditions, treatments and drug regimes are included. The sooner we sort this out the better.
“We now have nearly 900,000 people over 65,” he says. “Some of them have three or four health problems. Some of them are slipping through the cracks.”
Judith Collins, the Minister for Digitising Government, sees healthcare as a key area where inroads can be made with AI. If the productivity improvements Cole gets from using AI can be applied across primary care, that’s hundreds of thousands of hours freed up each year to focus on patients. It’s a no brainer to use it.
That’s not to say we shouldn’t proceed with caution. Clear policies regarding the storage, use, and sharing of recordings must be established to protect patient privacy.
AI has tremendous potential to provide insights humans might not think of. An AI trained in diabetes, drawing on anonymised patient data, might be able to read meeting transcripts and suggest treatments for the doctor to consider.
Given the pressures on public health services, removing the admin burden on doctors will pay dividends for the nation’s health.
