China's Insurance Regulator to Surgical Robot Makers: Stop Chasing Volume and Showmanship

Global surgical robot installations keep climbing. China's laparoscopic surgical robot installs are projected to reach 136 units in 2026, up from 97 in 2025 — an increase of roughly 40%.

The market has grown from 3.9 billion yuan to 4.4 billion yuan, while domestic systems surpassed 41% of new installations for the first time, priced 30–40% below imports. Every metric points upward.

Then the National Healthcare Security Administration convened seven manufacturers and five hospitals — and opened with an uncomfortable message: stop using surgical robots as window dressing, avoid performing extra robot-assisted surgeries just to generate fees, and stop chasing the world's highest annual per-unit surgery volume. It was the bluntest passage of the symposium, and no industry representative pushed back. There was nothing to push back against.

The core tension in surgical robotics is not whether fees are too high or whether insurance covers them. That was already resolved early this year: navigation capped at 3,600 yuan, assisted execution at 12,000, precision execution at 26,000, and remote surgery at 37,000, with Hunan and Guangdong leading implementation.

In early September, version 3.0 of diagnosis-related group payment created nine separate groups for robot-assisted surgery, opening the reimbursement channel to the medical insurance fund. Both administrative gates — pricing and payment — swung open within six months.

Yet once the gates opened, the flow was smaller than expected. The real bottleneck sits with surgeons. In July, The BMJ published a rapid response on robotic pancreaticoduodenectomy, noting that a very low 2.4% conversion rate in trials reflects the "volume and selectivity" of seven high-volume Chinese centers, whose surgeons operate "at the apex of their learning curves."

Outside those elite institutions, the learning curve becomes a genuine safety risk. The article raised a question the industry collectively avoids: hardware is spreading globally, but structured training pathways have not kept pace. In plain terms — the machines are here, but the people who can use them are not.

The autonomy grading of surgical robots makes this clear. As of 2025, of nearly 50 FDA-cleared systems, 86% remain at Level 1 — tremor filtering and master-slave mapping, where the robot is merely the surgeon's hand and makes no decisions. Only 6% reach Level 3 conditional autonomy; not a single Level 4 or 5 fully autonomous system has been cleared.

In other words, today's surgical robot is essentially an expensive advanced tool whose performance depends entirely on the surgeon. If the surgeon's hands are unsteady, the robot is just a pricier endoscope.

China's situation is subtler. The industry has long operated on a "brand-binding" model: manufacturers build their own training systems, and what a surgeon learns depends on which brand their hospital bought. Standards differ by brand, so a doctor switching hospitals may not be able to use a different machine.

In July, Beijing Friendship Hospital led a training course that, for the first time, placed five systems — from Johnson & Johnson, MicroPort, Weijing, Jingfeng and Suruix — in one venue for surgeons to rotate through. The industry called it breaking the "brand islands," which only highlights how long those islands have existed.

At the symposium, manufacturers and hospitals agreed that surgical robots remain at an early stage of innovative development. Against a backdrop of intensive pricing policies, that is a rare honesty. The NHSA has fixed the pricing channel, built the payment groups, and established remote-surgery pricing — then told the industry: the road is paved, but you must learn to drive first.

What is worth watching next is not how many provinces adopt pricing standards, nor whether domestic installation share reaches 50%. It is whether the cross-brand training course becomes a regular mechanism rather than an annual event, and whether any Chinese hospital actually delivers the "transparent reporting of learning curves" The BMJ called for.

A surgical robot sits in the operating room. Its most expensive part is not the robotic arm — it is the person standing beside it who has not yet practiced enough.