The modern consultation sometimes begins with a new form of second opinion: “Doctor, I checked Google.”
More recently: “Doctor, I asked ChatGPT.”
Soon perhaps: “Doctor, my watch, my phone and my AI all disagree with you.”
Welcome to general practice in the digital age.
Doctors should not be offended by this; patients looking for information are not the problem.
The real questions are whether information has become understanding, and who carries the responsibility when the two are confused.
ALSO READ: AI can’t answer this crucial question for patients
The advantages of technology
Technology is already changing primary care.
Patients measure blood pressure at home.
Smartwatches can record heart rhythms and flag possible abnormalities.
Glucose sensors generate continuous streams of data.
Electronic records remember what human beings forget.
Teleconsultations save travel and waiting time.
Artificial intelligence (AI) can summarise notes, organise information, suggest possible diagnoses and support clinical decision-making.
Used properly, these tools can make primary care considerably better.
A patient with hypertension (high blood pressure) can show weeks of home readings instead of one anxious measurement in a clinic.
A person with diabetes can share trends rather than isolated numbers.
A son living hundreds of kilometres away can join his elderly mother’s consultation.
A working parent may receive appropriate follow-up without sacrificing half a day to traffic and waiting rooms.
Technology can bring the doctor closer to the patient – that is its promise.
But there is a danger: more data does not automatically create better judgment.
Medicine is more than data
A blood-pressure application knows the number.
It may not know that the patient stopped taking medicine because he lost his job and is trying to stretch the remaining tablets over two months.
An algorithm can generate 20 possible diagnoses.
It does not necessarily know which possibility is dangerous enough to require immediate physical assessment.
An electronic medical record may contain years of consultation notes.
It cannot guarantee that anyone truly listened to the patient.
The challenge is especially important in general practice because symptoms often begin vaguely.
Abdominal pain may be indigestion; it may also require urgent assessment.
Dizziness may be benign, but then again, it may not be.
Chest discomfort may come from reflux, muscle strain, anxiety or something considerably more serious.
A photograph may show a rash; it does not always show how sick the person carrying the rash has become.
Medicine is not simply pattern recognition.
It is pattern recognition combined with context, examination, probability, uncertainty and responsibility.
This is why one of the most important sentences a digitally-competent general practitioner (GP) must be prepared to say is: “I cannot safely decide this online.”
That is not failure of telemedicine; that is good telemedicine.
Convenience has genuine value, but it cannot become a substitute for safety.
Where responsibility lies
Potentially serious or rapidly-worsening symptoms require appropriate clinical assessment, and sometimes, urgent physical examination, not casual reassurance through a screen.
AI creates the same challenge on a larger scale.
AI can assist doctors.
It can prompt possibilities we might otherwise overlook.
It can organise complex information.
It may reduce administrative burden and potentially give doctors more time for human interaction.
We should use those advantages.
But AI must not become a convenient place to deposit responsibility.
“The computer suggested it” cannot become the final explanation for a poor medical decision.
For the doctor using AI to make a clinical decision, professional responsibility does not disappear simply because an algorithm was involved.
Malaysia is not approaching this issue from a blank page.
The Malaysian Medical Council (MMC) now has specific ethical guidance on AI in medical practice, which emphasises human oversight, patient autonomy, privacy, safety and the continuing responsibility of doctors for their clinical decisions.
This also changes what doctors themselves must contribute.
For generations, part of the doctor’s value came from access to knowledge that patients could not easily obtain.
That information advantage is rapidly disappearing.
A patient can now obtain pages of medical information in seconds.
Trying to compete with a computer on who can recall more facts is increasingly pointless.
The future value of the doctor lies elsewhere.
What doctors can do
Can we distinguish important information from irrelevant information?
Can we recognise when the obvious diagnosis is wrong?
Can we understand why a technically-correct treatment is unrealistic for this particular patient?
Can we communicate uncertainty without creating panic?
Can we recognise when not to investigate?
Can we decide when technology has reached its limit?
Most importantly: Will someone take responsibility for the decision?
That is why Malaysia’s move towards interoperable health records is potentially so important.
The Health White Paper envisages secure electronic medical and lifetime health records capable of interoperability between healthcare providers.
If implemented well, that could transform continuity of care.
A private GP could know what happened during a recent hospital admission.
A hospital doctor could see important previous treatment.
Duplication of medications could become easier to identify.
Repeated investigations might be reduced.
Follow-up could become more coherent.
But digitalisation must not merely convert fragmented paper trails into fragmented screens.
Interoperability matters.
Cybersecurity matters.
Privacy matters.
Consent matters.
Clinical governance matters.
And good design matters.
The computer should demand less of the doctor’s attention so that the doctor can give more of it to the patient.
That should be one measure of successful health technology.
Understanding and responsibility
The GP of tomorrow therefore needs to become a hybrid physician:
- Digitally capable, but clinically grounded.
- Accessible, but not reckless.
- Efficient, but not mechanical.
- Data-informed, but not data-blinded.
- Modern, but still human.
AI may outperform doctors in an increasing number of narrowly-defined tasks.
That possibility should not frighten us.
It should force us to clarify what the irreplaceable part of medicine actually is.
And that is when a worried human being finally asks: “Doctor, what should I do?”
Someone must understand the evidence.
Someone must understand the patient.
And someone must be prepared to take responsibility for the answer.
Technology can support that responsibility; it should never erase it.
That brings us to the final question in this series, to be addressed in our next article: What must never be for sale when medicine, technology, regulation, insurance, patient expectations and business pressures all enter the same consultation room.
Dr Carollyn Kek Chee Yen is president of the Private Medical Practitioners’ Association of Selangor and Kuala Lumpur (PMPASKL), Dr Chang Chee Seong is the honorary secretary, and Dr Eugene Chooi is the immediate past president. This is the fourth in a five-part weekly series on the future of general practice in Malaysia. For more information, email starhealth@thestar.com.my. The information provided is for educational and communication purposes only, and should not be considered as medical advice. The Star does not give any warranty on accuracy, completeness, functionality, usefulness or other assurances as to the content appearing in this article. The Star disclaims all responsibility for any losses, damage to property or personal injury suffered directly or indirectly from reliance on such information.
