Doctor Apologizes After AI Makes Error About Patient's Medical History

A mistake in a patient's medical record said she took psychedelic mushrooms.

When Rebecca Green went to see a urologist, she was asked if she agreed to have the appointment recorded by AI.

She agreed without hesitation.

Rebecca did not want to be a difficult patient and take up more of the doctor's time.

But this decision caused her a lot of stress because the AI made a false claim that she was taking illegal drugs.

AI tools are becoming more common in healthcare to help busy doctors.

However, there are concerns about AI being used in medical settings because it can make mistakes.

Rebecca found the mistake after her kidney stone surgery when she read a letter from her specialist.

The letter said she took mushrooms, which could have caused bleeding around her kidneys.

Rebecca was scared because the letter said something that was not true and could have serious consequences.

She felt like she was not getting the care and attention she needed from her doctor.

Rebecca discovered the error in her medical record after reading a post-operative letter.

She was stressed because she was receiving workers' compensation and any mention of illegal drugs could affect her case.

The regulator of medical practitioners said doctors must check AI output for accuracy.

Rebecca's doctor did not check the letter, which is why the mistake was not caught.

If the doctor had checked the letter, she could have corrected the mistake before it was sent.

Some patients are saying no to AI being used in their healthcare.

After Rebecca complained, the urologist sent a letter apologizing for the mistake.

The doctor said she took the accuracy of her documentation very seriously and was sorry for the error.

The doctor did not know how the claim about mushrooms got into the record, but it seemed to be a mistake during transcription.

The urologist corrected the mistake and said she would review how AI was used in her practice.

The doctor did not respond to questions about what changes she made after the incident.

AI can confuse which breast has cancer.

Rebecca feels lucky she found the mistake, but she is worried others might not know about errors in their records.

This could have serious consequences for their health, career, or insurance.

Private health insurers are asking for patients' full medical history, which is causing concern.

AI scribes are common in healthcare, with 40% of GPs using them regularly.

The peak body for GPs estimated that 40% of GPs use AI scribes at work.

Research shows that AI can reduce the administrative burden for GPs and allow them to focus on patients.

But there are risks to using AI in healthcare.

A report found that Australians are being exposed to unregulated AI medical scribes, which can lead to errors.

Examples of mistakes include a scribe recording the wrong breast in a cancer diagnosis.

The CEO of the Consumers Health Forum said she often hears complaints about AI mixing up words.

Patients find it frustrating to get their records corrected when errors are made.

Elizabeth Deveny says it is frustrating for patients to fix errors in their records.

Dr. Deveny is involved in a research project on AI medical scribes and says doctors may be too trusting of AI.

There is evidence that people check AI output carefully at first, but then stop checking.

A GP who has been using AI scribes for three years said it is the doctor's responsibility to check every transcript.

He believes the technology is becoming more accurate, with fewer corrections needed.

Sean Stevens says AI transcriptions are getting more accurate, but doctors still need to check them.

Dr. Stevens is a big advocate of technology, but he admits AI has weaknesses and can produce incorrect results.

He said using AI allows him to run on time more often and produce more thorough referrals.

Patients need to consent to AI being used in their healthcare.

There are huge discrepancies in how consent is obtained, with some doctors only putting up a sign in the waiting room.

Other times, consent must be given when the appointment is booked online, and some patients are refused appointments if they do not consent.

Dr. Deveny said consent must be enthusiastic and fully informed, and patients should feel comfortable asking questions.

The Consumer Health Forum recommends asking questions at your next medical appointment.

The DRW report found that people might not be forthcoming about sensitive issues when they know AI is listening.

Dr. Stevens said it is essential for patients to disclose all relevant information and know they can say no to AI.

The TGA has acknowledged that some AI scribes may be in breach of the law.

The TGA is reviewing AI scribes and will release its findings soon, and any breaches will be acted upon.

The CEO of Patients Australia believes all AI scribes should be regulated.

Dr. Stevens has concerns that extending regulation could stifle innovation.

The DRW is calling for the AI Safety Institute to test AI scribes and publish a list of approved tools.

This would be similar to New Zealand's approach, where two AI tools were approved for use by clinicians.

Ms. Deveny said ensuring doctors use AI safely could also be enforced through compulsory checks and balances.

The error on her medical record has made Rebecca reconsider giving consent to AI transcription again.

Rebecca said her trust in AI has been damaged by what happened, and she would only give consent if she had a established relationship with the doctor.

See more ABC coverage in your search results.

Promotion.

Top Stories.

BREAKING.

LIVE.

Topic: Foreign Affairs.

Topic: State and Territory Government.

Topic: Air Transport Industry.

Top Stories.

BREAKING.

LIVE.

Topic: Foreign Affairs.

Topic: State and Territory Government.

Topic: Air Transport Industry.

Just In.

BREAKING.

Topic: Animals.

Topic: Australian Rules Football.