This article is for surgeons and specialists who run a practice or lead a department, and for the marketing people who support them. It answers one question: if you spend any professional time on X, what should you actually do there, and what should you leave out?
The short version is that X is a poor place to broadcast to patients and a reasonable place to be known by peers. Referrals, conference invitations, journal club discussions and invitations to speak tend to come from other clinicians who have seen how you reason in public. That is a slower and smaller return than a Reel, and it is worth stating plainly before anyone spends time on it.
This is an opinion piece with a method attached. It does not claim any outcome, and the clinical content of anything you post must be reviewed by you or the treating team, not by a marketer.
Be honest about what X is for in India
X is a small platform in India compared with the ones patients use. DataReportal's Digital 2026 India report puts X's advertising reach at 22.2 million people in late 2025, about 1.5% of the population, against 481 million for Instagram and 500 million for YouTube. The same page warns that X reach figures can swing sharply and may include non-human accounts, so treat the number as an order of magnitude, not a precise audience.
We do not have a verified figure for how many Indian clinicians use X, and we are not going to invent one. What you can say from experience is general: the doctors who do use it tend to be academics, conference-goers, trainees and people who read journals closely. That is exactly why it works as a peer channel.
The practical consequence is that you should not measure X by patient enquiries. Measure it by the quality of the conversations, the people who start following your thinking, and the professional invitations that trace back to it. If your goal this quarter is bookings, put the time into Instagram or search instead.
The three conversations worth joining
Most useful specialist activity on X falls into three formats. None of them is a broadcast.
1. Conference threads
A conference produces hundreds of posts in a few days, most of them photographs of slides. The valuable contribution is rarely another photograph. It is a short, accurate thread that does one of three things: summarises a session for colleagues who could not attend, adds context to a result that the speaker did not have time to explain, or asks a good question in reply to someone else's summary.
Check the conference's own policy first. Some meetings ask delegates not to photograph slides or to wait for embargoed results. If you are covering the event for a hospital, our conference social coverage service page describes how a team can plan this properly. For a surgeon posting personally, the rule is simpler: attribute the speaker, quote the finding as the speaker stated it, and do not editorialise beyond what you can defend.
2. Journal discussions
When a paper is published, the useful reply is not "great read". It is a specific observation: the sample was drawn from tertiary centres, the follow-up period is short for this procedure, the comparator is not what we use in India. Link the paper from its publisher's page, name the limitation, and say what you would want to see next.
This is where peer authority is built. Colleagues learn how you read evidence. It also overlaps with what search engines look for when they judge expertise, which our E-E-A-T framework for healthcare websites covers on the website side. That article is about how Google reads authorship on your own pages. This one is about where your reasoning becomes visible to the people who will cite or refer you.
3. Live case-learning, without a case
Case discussion is the format surgeons value most and the one that carries the most risk. The way to do it safely is to discuss the decision, not the patient.
Instead of "a 54-year-old woman from Jaipur came to us with...", write about the decision pattern: "When imaging shows X and the patient also has Y, which factors would change your choice of approach?" Keep it a question you would ask at a departmental meeting, and keep it generic enough that no real patient could recognise themselves.
The patient-data line is stricter than it looks
On 6 October 2026 the NMC's Ethics and Medical Registration Board issued its Guidelines on Ethical Advertising and Public Communication, reported by Medical Dialogues as taking effect immediately. As reported, they say patient information, images, records or testimonials may be used only where the law permits and with consent and safeguards, that identifiable details must be cropped, blurred or removed, and that "before and after" material and publicising case results are not allowed except for scientific or educational purposes with anonymised consent.
Read the full notice with your compliance lead, because the details matter and this summary is not legal advice. Three practical points follow for X.
- De-identification is more than removing a name. A rare condition, an unusual date, a small town or a visible hospital wristband can identify someone. If a colleague at your hospital could work out who it is, it is not anonymous.
- Scientific discussion and promotion are different. A peer-to-peer comment on a paper is scientific discussion. A post that names a procedure, your hospital and an appointment link starts to look like promotion. Keep the two apart, and do not put a booking link in your academic threads.
- You cannot outsource the risk. The guidelines, as reported, say that content published through agencies, influencers or other third parties does not absolve a practitioner or institution that authorised or knowingly permitted it. If a social team posts for you, you remain responsible for what appears under your name.
Two further points from the same reporting are relevant. Promotional AI-generated content is not permitted, and permitted AI-generated material must be marked as such. So do not use a tool to generate "clinical insights" and post them as your own thinking. And buying followers, likes or views is prohibited, which rules out the shortcut some agencies still offer for "building authority" quickly.




