Large hospitals are losing revenue from common illnesses, but gaining more weight for complex and severe cases under the DRG 3.0 system, which marks the end of the era where they could "handle everything" but does not necessarily make their situation worse.
On the morning of September 2, the National Healthcare Security Administration held a press conference to officially release the 3.0 version of the Diagnosis-Related Group (DRG) payment and Disease Insurance Payment (DIP) grouping scheme, which was more than a month later than the originally planned release in July.
The DRG core grouping has increased from 409 in version 2.0 to 492, and the number of subgroups has risen from 634 to 825. Meanwhile, the number of DIP core diseases has been compressed from 9,520 to 5,125, a reduction of nearly half. The implementation timeline has also been postponed, with local areas required to complete preparations for the transition by December 31, 2026, and apply the changes by the end of March 2027.
The Notice on Issuing the 3.0 Version of the Payment Scheme by Disease Group and Disease Type and Doing a Good Job in the Implementation of Relevant Work, recently released by the National Healthcare Security Administration, includes for the first time a national list: 158 basic disease types at the grassroots level.
It includes 31 DRG basic disease groups, covering 27 internal medicine disease groups such as hypertension, diabetes, and respiratory system infections, as well as technically mature first- and second-level surgeries like appendectomies, inguinal hernia repairs, and orthopedic fixture removals.
DIP covers 127 basic disease categories, including 97 conservative treatment groups and 30 surgical operation disease categories, covering common diseases in the respiratory, digestive, circulatory, and urinary systems.
There is a four-character rule outside the list: same disease, same payment. For the same insurance region, these diseases, regardless of which hospital, whether primary, secondary, or tertiary, treats them, will be settled by insurance according to the same standard.
In the past, the same disease had different settlement standards between hospitals of different levels - local authorities would set different coefficients according to hospital levels, with the gap calculated by local authorities themselves.
In recent years, top-tier hospitals have been taking high-profile, difficult cases while also dominating a large number of common and frequent illnesses, "taking all" and squeezing grassroots medical resources and healthcare funds from both ends.
This is the first time at the national level to clearly define common illnesses at the grassroots level, and it also marks a shift in policy thinking: no longer relying solely on administrative requirements to "guide patients to seek treatment at lower-level hospitals", but rather directly changing the economic incentives for hospitals to treat different illnesses, making it increasingly unprofitable for large hospitals to treat minor illnesses.
The in-hospital drug market for 158 diseases, including high blood pressure and diabetes, will migrate to county and grassroots levels along with the diseases.
The document also draws another red line: the payment standards for diseases cannot be used as a quota to assess departments or medical personnel, nor can they be linked to performance distribution indicators.
However, the release of the list does not necessarily mean that common diseases will return to grassroots levels. Whether grassroots hospitals can handle them, whether large hospitals are willing to let go, and who will take charge of moderately difficult diseases remain uncertain.
The more detailed the grouping, the more it tests a hospital's coding, cost accounting, and disease management capabilities. Can version 3.0 resolve the inertia of the new round of disease elimination and internal friction?
No text to translate.
Heavyweights are concentrated in the top three, can grassroots make a comeback with a "list"?
The 3.0 version of the grouping logic took shape in two scenario introduction events held at the end of March and early April, with multiple clinical experts from top tertiary hospitals in fields such as respiratory, urology, oncology, general surgery, and cardiovascular surgery, respectively, justifying the specialized grouping.
After the policy was implemented, the direction was confirmed - new additions include combined tumor treatment, critical care, pain treatment, and robot-assisted surgery, while rehabilitation and high-difficulty combined surgery were further subdivided, and painless childbirth was set up as an independent payment category for the first time.
Bi Jianbin, Director of the Department of Urology at the First Affiliated Hospital of China Medical University, summarized at a previous introduction event that this hierarchical logic is shifting from "disease-oriented" to "surgical procedure-oriented".
He cited the example of "bladder tumor surgery" tagged procedures, where radical cystectomy and transurethral bladder tumor electroresection have completely different resource consumption and technical difficulties - clinical resource consumption often depends on the surgical method, and grouping is adjusted accordingly.
Tumors are the area that has been most thoroughly unbundled in this round.
DRG has made a systematic refinement to the tumor-related treatment groups: mainstream treatment paths such as radiotherapy, chemotherapy, targeted immunotherapy, and interventional ablation are each grouped separately, and a new combined treatment group has been added.
Hematological tumors have been separated from solid tumors for the first time - in the past, leukemia and lymphoma were labeled as "malignant tumors" along with solid tumors, but their treatment paths and resource consumption differ greatly.
On the DIP side, synchronization is becoming more refined: the treatment of malignant tumors is divided into categories such as solid tumors, hematological diseases, and lymphomas, and hematopoietic stem cell transplantation alone is divided into different initial subgroups based on whether it is syngeneic or allogeneic.
In the field of critical care, the DRG has added a new disease group for invasive ventilators combined with CRRT, while DIP has included ECMO, invasive ventilator support for more than 96 hours, and artificial liver in the preliminary grouping, with separate calculations - and hematological tumors and critical care are precisely the core businesses of the top three Grade A tertiary hospitals.
According to Zhou Le, director of the China Health Industry Development Research Center at Yi Shiye, the 3.0 version incorporates different treatment pathways into group consideration, which is a significant benefit for tumor hospitals and large comprehensive hospitals with tumor centers - as the differentiated payment for different treatment pathways will more accurately cover the real costs of radiotherapy and chemotherapy.
However, Wu Min (pseudonym), an administrator at a provincial cancer hospital in central China, is not willing to simply describe it as a "major benefit".
She believes that version 3.0 has significance for comprehensive tumor treatment, or more accurately, "grouping is more standardized." In the past, due to unclear grouping, some cases of chemotherapy combined with targeted and immunotherapy may have resulted in a situation where "hospitals that know how to group benefit, and those that don't are at a disadvantage." The new version further refines the granularity of comprehensive treatment, and the actual impact may be two-way, helping to correct deviations.
Meanwhile, Zhou Yan believes that DRG 3.0 will have a different impact on hospitals at different income levels:
Top-tier hospitals have a patient base and technical capabilities, allowing them to focus on high-value disease treatments; regional leading hospitals with revenues of 3 billion to 5 billion yuan are under pressure, squeezed by provincial hospitals from above and pursued by county-level hospitals from below, placing them in a middle layer of pressure; grassroots and weak tertiary hospitals that continue to rely on low-value disease treatments can only compete in terms of quantity in low-value inpatient services.
This prediction has been confirmed in the official version - robot-assisted surgery has been set up with related grouping; the special case negotiation mechanism has been refined, and the number of special case negotiations in the DRG region is generally not more than 5% of the total number of discharged cases, leaving a reasonable compensation channel for difficult and severe cases and new drug consumption and new technologies.
Beijing YuFang Hospital Management Center's chief consultant, Qin Yongfang, believes that the DRG 3.0 version is essentially rearranging the deck chairs on a limited plate, "after all, there's only so much money, if the weight of high-complexity diseases is increased, the space for other disease groups will be compressed."
The grassroots disease list is precisely the official response to this squeeze — using flattened payment standards to push basic disease categories currently held by tertiary hospitals back down to the primary level. The notice also calls for accelerating province-wide standardization of grassroots disease categories.
However, the premise of "returning" is that the grassroots level can handle it and large hospitals are willing to let go.
Wu Min reminded that the transfer of basic diseases to grassroots levels is not necessarily more cost-effective. Many regions have implemented the same price for the same disease, "for the same disease, grassroots hospitals have lower personnel, equipment, and research and education costs, and under the same price, grassroots hospitals seem to have an advantage, with higher profits. However, if the technical capabilities of grassroots hospitals are insufficient, and there are many complications and longer hospital stays, the costs may not be lower than those of large hospitals. Large hospitals have mature treatment pathways and skilled doctors, which can lead to faster discharge and potentially lower average hospitalization costs.
Liu Ming, a pseudonym for a healthcare insurance administrator in the southeastern coastal region, broke down the issue further: it is reasonable for complex and high-value cases to be concentrated in top-tier hospitals, as "top-tier hospitals should naturally push the boundaries of medical technology"; however, if these top-tier hospitals continue to monopolize common diseases and only pursue revenue growth, they will destroy the resource base of grassroots healthcare providers.
Regardless of where the disease ultimately goes, a relatively consistent judgment in the industry is that after the implementation of 3.0, a new batch of diseases will be eliminated.
According to Zhou Le, cases without clear indications for hospitalization, where the value of bed days cannot cover the cost of bed days, and primarily reflect nursing or caregiving needs, will find it increasingly difficult to remain in the inpatient system - cases of type 2 diabetes without severe complications, mild upper respiratory tract infections, and ordinary surgeries that can be handled on an outpatient basis should be transferred to outpatient, daytime, or home management.
Hospital restructuring has already begun. Wu Min predicts that after rounds of restructuring, only two types of departments will remain: those with the largest market share and strong bargaining power in their region, and those that can only survive by controlling costs. Departments without technological advantages, cost advantages, and that rely heavily on inpatient volume to maintain revenue may be proactively abandoned by hospitals.
Taking Beijing and Shanghai as examples, after DRG/DIP has been implemented for several years, some hospitals have gradually become specialized: each hospital focuses on the diseases it treats well, and other hospitals no longer compete directly. However, there is a very important premise - large hospitals cannot continue to expand. If they still follow the old expansion model, top hospitals will not naturally become specialized, and instead may absorb more common diseases.
Notably, while the DRG grouping is being expanded, the core disease categories of DIP are being consolidated, with the boundaries between the two systems becoming increasingly blurred - the notice explicitly proposes "supporting qualified regions in perfecting grouping rules and piloting the integrated development of DRG and DIP". Wang Pingyang, Vice Chairman of the Zhejiang Province Medical Security Research Association, predicts that DRG and DIP will continue to converge in versions 4.0 and 5.0, ultimately forming a more unified disease-based payment system, with no clear distinction between the two.
February
A finer ruler cannot measure more wealth.
Beyond the issue of grouping, the more fundamental problem is how the money is divided.
In early June this year, at a medical insurance management experience-sharing session attended by nearly 200 people, a clearing table displayed by a policy expert from a provincial medical insurance bureau prompted a sea of phones to be raised across the audience.
During the six years the province implemented DRG, many pooling areas saw payment rates exceed 100% in the early stages, leaving hospitals with surpluses. As coverage and the budget pool expanded, policy dividends were exhausted, and some areas swung into deficit, with the annual payment gap in a single pooling area reaching as high as 800 million yuan.
Medical management expert Chen Pulan once wrote in a public account article about the questions that hospital administrators repeatedly asked but often ignored: As the grouping rules become more and more precise, are the local budget mechanisms adjusted accordingly? If the budget mechanism remains unchanged, to what extent can payment reform actually change the behavior of hospitals?
A chief accountant at a top-tier Grade-A tertiary hospital once consulted her: "Our orthopedics department has introduced robotic surgery, and the CMI has gone up. So why hasn't the amount we get reimbursed increased noticeably?"
The answer lies in the rates. The total regional amount remains unchanged, but the more complex diseases treated by tertiary hospitals and the more work done by doctors, the more likely they are to exceed their budgets - the numerator (the total budget for the medical insurance fund in the region) remains the same, while the denominator (the total weight of treatments covered) increases, causing the rate to depreciate. "A disease with a weight of 10,000 yuan may only be settled for 9,500 yuan in reality," said a seasoned medical management expert, noting that the shrinkage of point values and depreciation of rates exist in many places.
The pressure of involution ultimately falls on doctors.
Five years ago, Dr. Fang Hua (pseudonym), an oncologist at a top-ranked county-level hospital in southeastern China, posted a handwritten table on his desk with the total cost limits for different lengths of hospital stays for cancer patients undergoing supportive treatment and maintenance treatment. He would glance at it every time he determined a treatment plan, reminding himself not to exceed the budget.
The numbers on the desk have been constantly adjusted downward: in 2021, the total cost cap for maintenance treatment for tumor patients hospitalized for more than 30 days in the area was 21,780 yuan; last October, this standard was lowered to 19,830 yuan.
To increase revenue, many hospitals have taken the risk of linking DRG surpluses to the performance-based salaries of doctors or departments. The hospital where Fang Hua works gives doctors a "commission rate" of 10% - for every 1,000 yuan in surplus, the doctor receives a reward of 100 yuan; he has learned that some hospitals once raised the proportion to 50%, but then adjusted it back to 10% because the costs decreased too quickly, ultimately affecting the surplus.
Under this mechanism, doctors are incentivized to cut back on examinations, consumables, and hospital stays—but once the regional average cost drops, the next round of insurance pricing may lower payment standards again. The industry jokes this is "whipping the fast ox."
Combined with fee rate depreciation and tightening constraints, according to Fang Hua's statistics, after the implementation of DRG, his and his colleagues' overall wages dropped by 30%.
Under the impact, doctors' behaviors are also changing.
Zhejiang Province Medical Security Research Association Vice Chairman Wang Pingyang pointed out that one manifestation is excessive medical treatment, such as low-indication hospitalizations, where patients with mild symptoms who do not need hospitalization are admitted; another manifestation is insufficient medical treatment, such as shortening hospital stays to reduce costs to meet the standard, or even transferring some high-risk patients. Both behaviors will affect patient interests.
To address this issue, the 3.0 version of the notice reiterates that a standardized mechanism for managing leftover funds from disease treatments should be established, and that these leftover funds can be used as business income for medical institutions, for purposes such as discipline development and personnel performance-related expenditures. Medical institutions are not allowed to use disease payment standards as a cap for evaluating departments or medical personnel, nor can they link these standards to performance distribution indicators.
Top-level design for budget mechanisms is also being adjusted.
A local medical insurance department head revealed that the draft for comments' Chapter 4, Section 2, explicitly allowed regional authorities to establish a "vertical total budget for medical institutions," which would reasonably determine the annual budget for each designated medical institution based on the regional total, taking into account the institution's historical medical service provision and medical insurance costs, as well as comprehensively considering changes in service capacity and other factors.
The final version of the notice, however, did not adopt that approach. Instead, it kept the funds within the regional pooling pool and adjusted the fee structure: the total amount can be split into existing and incremental portions, or diseases can be categorized into severe and mild cases, with fixed and floating rates set for each. The former governs hospitals, while the latter governs disease types and incremental growth.
But in Liu Ming's view, these are merely optimizations to the existing architecture, not substantive solutions.
When the growth rate of medical services exceeds that of the medical insurance fund, devaluation will occur regardless of how the services are grouped. More complex is the fact that the momentum of hospital expansion still exists. For a long time, hospital management has revolved around scale, revenue, beds, projects, and rankings. Reforms to the payment model have attempted to change this logic, but the incentives between health authorities, medical insurance, hospital performance, and local finance are not always aligned, resulting in the inability to resolve the negative effects of policies.
Another highly anticipated outlet is commercial health insurance. In 2024, China's commercial health insurance premium income reached 977.3 billion yuan, approaching the total funding level of resident medical insurance for that year; since the beginning of this year, many provincial-level top-grade tertiary hospitals have set up commercial insurance service centers.
Wu Min was informed by the person in charge of the commercial insurance company she is in contact with that the medical insurance fund in a certain province is approximately 150 billion yuan, while commercial insurance is around 50 billion yuan - "This scale is not small, and it also means that commercial insurance will have increasing say." However, in her view, commercial insurance will also restrain excessive medical treatment, require directories and pathways, and for hospitals, it merely provides an additional financing channel, rather than a fund pool that can be drawn from at will.
Version 3.0 offers a finer ruler—492 core groups and 825 subgroups that weigh each case's burden with greater precision. Mechanisms like same-disease-same-payment, special-case review, and flexible rates attempt to patch the gaps beyond the ruler's reach. But a ruler can measure weight; it cannot measure growth.
The way the cake is divided has changed, but the size of the cake remains the same. The elimination of inferior varieties and the internal friction of competing for resources will continue – the answers to these problems do not lie in the grouping scheme.