Bots cannot be cured: Document warns against AI self-diagnosis | Lucknow News

AI News


Bots cannot be cured: Document warns against AI self-diagnosis

LUCKNOW: When a 42-year-old private company employee developed persistent fatigue and mild abdominal discomfort, instead of seeking medical attention, he turned to an AI chatbot for answers.Based on online suggestions, he began self-medicating. After a few weeks, his condition worsened and he was forced to seek emergency treatment. Doctors say Ramesh's case is no longer an anomaly but part of a growing trend in the city.Doctors and clinics across the city are seeing a steady increase in the number of patients whose health deteriorates because they rely on artificial intelligence (AI) chatbots and online tools for medical advice.Doctors say 40 to 50 of these patients now visit clinics each month, many with complications caused by self-diagnosis, incorrect drug doses, discontinuation of prescribed drugs, or starting new treatments without consulting a qualified physician.Dr Sunil Verma, an internist at a well-known city hospital, said more and more patients are coming to the hospital after searching for symptoms online and starting medication themselves. “I often tell patients: Trust real doctors, not Google or AI. Medicine is a science of probability, not certainty. Unlike engineering, diseases don't always follow a set pattern,” he says.Quoting renowned physician William Osler, Dr. Verma said that only 5 to 10 percent of patients present with classic, textbook symptoms, while nearly 90 percent have variable or atypical symptoms.“This is why doctors rely on clinical experience, detailed patient interaction, and careful assessment before deciding to investigate,” he said, adding that clinical evaluation helps reach a diagnosis in about 80% of cases, but the remaining cases require further testing and observation.Dr Verma cautioned against over-reliance on AI, saying such tools work entirely on pre-fed data. “Illnesses like malaria can manifest in many different ways, from low-grade fever to changes in consciousness. Blanket AI-based conclusions can be risky,” he said.Another physician, Dr. Ajoy Tewari, highlighted how online searches create unnecessary anxiety, with patients panicking over slight changes in blood reports or imaging findings and fearing serious illnesses such as cancer with no clinical correlation.Professor KK Saulani, from the Faculty of Medicine at King George Medical University (KGMU), said the number of patients relying on online medical information and self-interpretation of symptoms is steadily increasing. “There is a tremendous amount of information on the Internet, but not all of it is accurate or applicable to all patients,” he says.“Sometimes the information is correct, but more often it can be misleading or incomplete. “You can't physically examine a patient on a computer,” he said, explaining that online tools do not take into account factors that doctors routinely consider before prescribing a drug, such as age, pre-existing conditions, organ function, and the drug's safety in conditions such as kidney and liver disease.He noted that many patients panic after reading about serious illnesses online or stop taking their medications after seeing a long list of side effects. “Even rare side effects are being mentioned online. If patients focus only on side effects, they may refuse treatment altogether,” he said, adding that people respond differently to medicines and medicines need to be individualized.



Source link