From Dr. Google to Dr. ChatGPT: Digital Companionship During Cancer and Cyberchondria
In this article
Artificial intelligence and the internet are revolutionizing oncology practice: Are they a "superpower" for cancer patients or a dangerous "information dump"? The most comprehensive guide to safely integrating the digital world into your treatment, asking the right questions, and protecting yourself from risks.
Information Is the Most Important Medicine (When Used Correctly)
The moment we hear the word "cancer," our lives split into two: Before and After. In this new reality, another need arises that is just as vital as chemotherapy or radiotherapy: Information. In the past, our patients depended solely on a few words from their doctor or thick encyclopedias in libraries. Today, the phone in your pocket opens a door to all the world's medical libraries, the latest clinical trials, and the experiences of millions of patients.
Yet "accurate information" is not the only thing entering through this door. Fear, manipulation, commercial traps, and confusion also enter our lives unchecked. This report analyzes in exhaustive detail the transition from the era of "Dr. Google" to "Dr. ChatGPT" (Generative Artificial Intelligence), the extraordinary opportunities these technologies offer (empathy, rapid summarization, trial matching), and their potentially life-threatening risks (hallucination, data privacy).
1. The Evolution of Patients' Information-Seeking Behavior
The way we access information has changed dramatically over the past 30 years. We have evolved from passive recipients to active researchers and now to "interactive questioners." Here is a brief history of our digital health journey:
Information Source: Doctor Only
Trust: Very High
Access: Very Limited
Patient's Role: Passive Listener
Source: Search Engines/Forums
Problem: Information Pollution
Outcome: Cyberchondria (Digital Anxiety)
Source: LLM (ChatGPT, Gemini)
Opportunity: Personalization & Synthesis
Risk: Hallucination & Privacy
2. "Dr. Google" and Digital Anxiety: Cyberchondria
Research shows that 90% of cancer patients actively use the internet. However, search engines prioritize "what is popular," not "what is true." Seeing "advanced brain tumor" when searching for a simple headache can turn an innocent search for information into pathological anxiety. In the literature, we call this Cyberchondria .
Searching for Symptoms -> Encountering a Worst-Case Scenario (Catastrophizing) -> Increased Anxiety -> Searching Again (Seeking Confirmation) -> Conflicting Information -> More Anxiety and Distrust of the Doctor.
3. Dr. ChatGPT: Advantages and Opportunities
Artificial intelligence (Generative AI) has the potential to organize and personalize the information chaos created by search engines. Here are its concrete, demonstrated benefits in oncology:
A. Translating Medical Jargon into "Turkish"
Pathology and radiology reports can seem like riddles to patients. Terms such as "lymphovascular invasion" and "carcinoma in situ" are frightening. Artificial intelligence can simplify this technical language to a primary-school level within seconds. A study by the Technical University of Munich (TUM) demonstrated that patients understood AI-simplified reports much faster and more accurately.
"On chest CT examination, the cardiomediastinal silhouette is in the midline. A small pericardial effusion is noted. No lymphovascular invasion is seen. No pleural effusion is detected."
Heart: Your heart is in its normal position. A very small amount of fluid has accumulated around the lining of your heart (this is usually treatable, and your doctor will monitor it).
Spread: Great news! Cancer cells have not spread to the lymph nodes or blood vessels. There is no fluid around the lungs either.
B. A Surprising Fact: Digital Empathy
We tend to believe technology is "cold." Yet JAMA Internal Medicine published a striking study showing that, in responses to patients' questions, artificial intelligence (ChatGPT) was considered more "empathetic" than doctors. While doctors give "brief and clear" answers because of time constraints, AI always begins with "I understand your concern; this can be a difficult process..." and listens without judgment.
C. Access to Clinical Trials (Matching)
When standard treatment options have been exhausted, clinical trials (phase studies) may be lifesaving. However, finding an appropriate trial for a patient requires screening thousands of criteria. Artificial intelligence (such as Tempus AI) can review a patient's genetic profile and reports to find a suitable trial anywhere in the world within seconds.
4. The Dark Side: Risks and Dangers
Blind trust in technology can lead to life-threatening mistakes in oncology. The greatest mistake is believing these tools are "flawless."
A. "Hallucination" (Fabricated Information)
Large Language Models (LLMs) are not "search engines"; they are "word prediction machines." They can sometimes produce medical information that sounds very plausible and is beautifully written but is completely fabricated . They may even cite nonexistent articles and nonexistent doctors as references.
B. Privacy and Genetic Data (Critical Warning)
Many patients copy and paste their reports into systems such as ChatGPT to have them interpreted. Remember:
- Your genetic data (your DNA mutations) cannot be anonymized. Those data are your biological fingerprint.
- The data you upload to these systems are processed on company servers and may be used to train the model.
- In a future data breach, this information could be used against you (as a high-risk group) by insurance companies or employers.
5. Asking AI the Right Question (Prompt Engineering)
To get the right answer from artificial intelligence, you need to ask the right question. Here are safe and effective question templates for cancer patients:
EXAMPLES OF SAFE PROMPTS
Goal: Simplifying a Report
Goal: Preparing for a Doctor's Appointment
Goal: Nutritional Support
6. Become an Information Detective: The CRAAP Test
To assess the reliability of information you read online or through AI, apply the CRAAP Test, which is used at universities:
-
CCurrency: When was the information written? In oncology, information more than 5 years old may be "history." Do not trust undated articles.
-
RRelevance: Does this information apply to your specific cancer type, stage, and genetic subtype? General information may mislead you.
-
AAuthority: Who is the source? A university hospital or a cancer society? Or a website selling products or an "influencer"?
-
AAccuracy: Is the information supported by other reliable sources (for example, the Turkish Society of Medical Oncology)? Are references provided?
-
PPurpose: Is the purpose to inform you, or to frighten you into buying an herbal product or a "miracle cure"?
DRÖZDOĞAN Academy Commentary
Artificial intelligence and the internet are not the "forbidden fruit" of modern medicine. On the contrary, when used correctly, they are the most important "copilot" empowering the patient. I always support my patients in doing research, asking questions, and taking ownership of their journey. An informed patient is an empowered patient.
However, the golden rule you must remember is this: Artificial intelligence is an assistant, not a doctor. Do not hesitate to share information you find online with me. Asking "I read this online; what do you think?" does not undermine the trust between us; instead, it allows us to discuss whether that information applies to your situation. Use technology, but do not hand it the steering wheel. You and your doctor are always the captains of your treatment.
Scientific References
- Ayers JW, et al. Comparing Physician and AI Chatbot Responses to Patient Questions. JAMA Intern Med. 2023.
- Chen S, et al. Use of Artificial Intelligence Chatbots for Cancer Treatment Information. JAMA Oncol. 2023.
- TUM School of Medicine. AI helps cancer patients better understand CT reports. 2024.
- Uses and limitations of artificial intelligence for oncology. PMC.
- DISCERN: Instrument for judging the quality of written consumer health information.
- Tempus AI: Improving patient matching efficiency.
Was this article helpful?
Your feedback helps us improve our content.
Reader comments
Share your comment
Your name and comment are published only after administrator approval. Your email address is never shown to visitors.
Be the first to comment.