INTERNAL TRAINING · CHINA HEALTH INSURANCE POLICY

China’s 36 Comprehensive Consultation Price Items: Clinical Practice, Payment-Method Behavior, and the 2026 Inspection Risk Map

Audience: hospital insurance offices, pricing & finance, medical-records coding, clinical department heads and head nurses, operations, discipline inspection · Compiled 2026-07-04 · Based on public policy documents and authoritative reporting; local implementation always governs. Primary sources are in Chinese.

Understanding what a price item says is only the first layer of compliance. Understanding how it reshapes behavioral incentives under each payment method is the second. Anticipating which data signatures regulators will use to test it is the third. This guide places China’s 36 comprehensive consultation price items inside a four-layer institutional frame — listing, pricing, payment, supervision — and works each high-risk item through real settlement logic and the 2026 unannounced-inspection regime. A trainee who finishes this material should be able to answer, unaided: may I bill this service, under which settlement logic will it be paid, and through which lens will a regulator look at it.

Policy status as of 4 July 2026: DRG/DIP 2.0 remains the current implementation baseline. Version 3.0 is still in grouping, data-validation, and publication preparation; NHSA previously indicated publication around July 2026 and planned implementation in January 2027. This article does not treat consultation-stage 3.0 content as an effective rule.

Chapter 0 · The Complete 36-Item Base Table

The English labels are explanatory, unofficial translations for professional readers. The Chinese names remain controlling. This table establishes the item boundary before the article moves into payment incentives, clinical behavior, and inspection analytics.

Chapter 1 · The Institutional Frame: Every Service Passes Four Gates

The most common training error is to equate “the national guideline listed this item” with “we may bill it” or “insurance will pay for it.” Four layers must be kept apart; none can be inferred from another:

The frame’s practical value is error localization. When a clinician protests “the state listed it, why can’t we bill it,” layer ① is being mistaken for ②③. When finance asks “pricing approved it, why did insurance deny it,” ② is being mistaken for ③. When management wonders “our line items are compliant, why were we inspected anyway,” they have missed that layer ④ owns an independent data perspective: supervision judges not only whether a single charge is correct, but whether the statistical signature of behavior looks wrong. Chapter 4 develops this in full.

Chapter 2 · Settlement Behavior: One Item, Three Different Things Under Three Payment Methods

Comprehensive consultation items — consultation fees, bed fees, integrated nursing bundles — share one property: they are the most time-accumulating component of an inpatient bill. Each additional day adds another day of bed, consultation, and nursing charges. That property is exactly what makes their nature invert across payment methods. This chapter is the core of the entire curriculum.

2.1 Fee-for-service: revenue units, and the risk is over-counting

Under fee-for-service, every billed line is direct revenue, so incentives point toward “one more day is one more day of income,” and the rulebook concentrates on counting and mutual exclusivity: bed fees follow “count admission day, not discharge day”; ER observation, intensive-care, laminar-flow and special-protection beds may not be billed alongside any other bed fee; the neonatal incubator fee may not be stacked on the neonatal bed fee; the hospice care fee may not coexist with inpatient consultation or graded-nursing fees. The training point: all of these exclusivities are structured rules, which means claims-audit systems can screen the full database at zero marginal cost — under fee-for-service, every violating line is self-incriminating evidence already sitting in the settlement data.

2.2 DRG/DIP bundling: cost elements, and the risk migrates to coding and splitting

Once cases are paid by diagnosis-related group (DRG) or big-data diagnosis-intervention packet (DIP), the case payment decouples from length of stay: bed fees turn into cost elements inside the bundle, and each extra day burns the case’s margin. The consultation items themselves rarely star in violations here, but settlement behavior tells us pressure does not vanish — it relocates:

• Upcoding — steering diagnosis and procedure codes toward higher-paying groups; the first target of big-data supervision models in this regime;
• Split admissions — one clinical episode broken into two settlements whose combined bundles exceed one; re-admission-interval distributions are the classic screen;
• Cost-shifting — pushing in-bundle costs to outpatient visits, external prescriptions, or patient self-pay; this maps directly onto the inspectors’ attention to abnormal self-pay ratios;
• Under-treatment and patient dumping — the policy answer is the special-case deliberation mechanism: for cases unsuited to standard bundled payment because of long stays, high cost, new drugs/devices/technologies, or complex critical illness and multidisciplinary care, hospitals may self-file for deliberation, capped in principle at 5% of DRG discharges or 0.5% (5‰) of DIP discharges; expert panels review monthly or quarterly and may pay the case by item or adjust its standard.

Two teaching points for clinical departments. First, special-case deliberation is a right, not a favor — national briefings note that many regions file far below the cap; proactively filing complex and critical cases protects both legitimate departmental income and the willingness to treat, and the insurance office should push a “fileable-case profile” down to every ward. Second, policy expressly prohibits hospitals from using the DRG/DIP group payment standard as a quota to evaluate clinicians or tying it directly to performance pay — operations teams must treat this line as itself inspectable management conduct.

2.3 Per-diem payment (hospice, psychiatry, etc.): fixed daily units, and the risk is admission, bed-parking, and dilution

Per-diem payment turns “a day” into a fixed revenue unit — the incentive inverts again and longer stays become profitable once more. Local practice answers with three gates, using hospice care as the example (from published local schemes):

• Admission gate: end-stage diagnosis, prognosis and functional assessment, and documented consent must follow the local pilot document; thresholds from one locality cannot be generalized nationally;
• Duration gate: local schemes commonly use a stay limit, interval reassessment, and an exit mechanism, but the number of days and extension conditions require locality-specific verification;
• Rate-and-conduct gate: daily rates and facility tiers are local payment rules and cannot be exported across regions; common risks are bed-parking, split admissions, artificial extension, shifting included services to self-pay, and admitting ineligible patients.

The behavioral inference: a per-diem ward’s compliance lifeline is its assessment documentation chain — admission assessment, interval reassessment, symptom-control records, and nursing/psychosocial care records must substantiate the reality and necessity of every billed day. Inspectors read the data first: the right tail of the length-of-stay distribution (case pile-up at the locally defined boundary), the plausibility of assessment scores — then verify on site with bed checks and record review. The distribution itself testifies: if the ward’s length-of-stay histogram spikes at the cap, no individual chart narrative can outweigh the statistical picture.

2.4 The migration matrix: three regimes on one page

Keep this table as the curriculum’s memory anchor. Its deeper law: payment reform does not eliminate violations; it changes their shape — supervision migrates with them, and hospital internal control must migrate in step. A hospital that simultaneously runs DRG/DIP for ordinary inpatients, per-diem for its hospice ward, and fee-for-service for residual items needs three parallel control logics, not one generic “billing check.”

2.5 The clinical–settlement–inspection linkage

Chapter 3 · High-Risk Items in Three Dimensions: Clinical × Settlement × Supervision

This chapter works the four most exposure-heavy item groups through one fixed frame: how the clinical scene arises → what the settlement rule says → what the regulator’s data view sees → what the internal-control action is.

3.1 Hospice care fee: the “exclusivity radius” of an integrated item is exam question one

Clinical scene: an end-stage patient is admitted to the hospice ward; daily work spans condition assessment, symptom control (pain, dyspnea, nausea), graded nursing, psychological and spiritual support, family conferencing and bereavement counseling. Before listing, much of this real labor — humanistic care, family meetings — had no billing outlet; after listing, it is packaged wholesale into the price composition of the hospice care fee.

Settlement rule: because physician review and graded nursing are inside the price composition, the fee may not be billed together with any inpatient consultation fee or graded-nursing fee. Where the region pays hospice by per-diem, even the hospice fee itself dissolves into the daily rate — any itemized add-on for in-scope services then amounts to monetizing bundle content twice.

Regulator’s view: exclusivity co-occurrence is a structured rule and is screened in full by smart audit; per-diem regions add penetration checks on admission-assessment authenticity (scores sitting “exactly” on the threshold), cap-hugging stay distributions, and occupancy verification.

Internal control: hard-code the exclusivities as real-time billing-engine blocks rather than retrospective audit; run an “assessment–service–record” ledger in the ward; have the insurance office self-profile hospice cases quarterly on stay and score distributions.

3.2 The bed-fee family: “count-in, not-out” hides hospital-scale error

Clinical scene: bed fees look trivial, but ward transfers, bed changes, same-day admission and discharge, and ICU round-trips constantly generate counting boundary events.

Settlement rule: admission day counts, discharge day does not; ER-observation, ICU, laminar-flow and special-protection beds are mutually exclusive with any other bed fee; a neonatal bed fee may coexist with the mother’s bed fee, but the incubator fee may not stack on the neonatal bed fee; single rooms are hospital-priced with filing and disclosure duties.

Regulator’s view: bed-fee days exceeding inpatient days, discharge-day charges, and double bed fees on ICU-transfer days are all findable hospital-wide with a single query on claims detail — small per line, large in aggregate, and indisputable in nature: the cheapest wins an inspection team can book under “duplicate/above-standard billing.” The single room connects to a second thread: its charges land in the patient’s self-pay ratio, and an abnormal self-pay ratio is an explicit trigger for 2026 pinpoint inspections — steering patients into single rooms surfaces in the data as the combined signature “abnormal self-pay ratio + abnormal single-room utilization.”

Internal control: generate bed-fee counts exclusively from the system, ban manual back-entry; trigger automatic fee-type switching on transfer events; require electronic informed consent before single-room charges, and put single-room utilization and self-pay ratios on the departmental monthly dashboard.

3.3 Home-visit fees and home hospital beds: fine-grained stacking rules, exclusive funding channels

Clinical scene: primary-care institutions serve disabled, very elderly, chronically ill and handicapped patients at home. The 2026 Home Hospital Bed Service Guideline (Trial) defines nine service classes and sets staff-qualification floors (e.g., physicians with three or more years of independent practice; nurses with five or more years and corresponding titles).

Settlement rule: the home-visit fee is billed per “visit·person” and may stack with the medical services delivered; the home-bed establishment fee is billed once per establishment and may not be double-billed with the home-visit fee; and — the clause primary care most often misses — services already funded through basic public health programs, family-doctor contracting, or long-term care insurance may not charge a home-visit fee again. This is an exclusivity clause between fiscal channels and the price channel.

Regulator’s view: cross-system matching is the key move — joining claims data with public-health funding rosters, family-doctor contract lists, and LTC-insurance benefit data directly surfaces “paid twice for one service.” The 2026 supervision plan explicitly explores LTC-insurance thematic inspections; the intensity on this seam will only rise.

Internal control: run a home-visit ledger with a pre-billing check of the patient’s contracting and LTC status; count “persons” by professionals actually present, evidenced from the same source as the service record.

3.4 Inactive items and transition windows: the most regrettable violations come from jumping the gun

The internet first-visit consultation fee is flagged inactive in the national annex, pending separate rules from health authorities; provinces map the national guideline onto local catalogs at different speeds, creating windows where the national text exists but the local price document has not switched. The iron rule of the window: the billing basis is always the locally effective price document. Building a charge item in the HIS from a screenshot of the national guideline is the most common — and most avoidable — violation in every pricing reform round. The pricing office must keep a three-level mapping ledger (national listing → provincial adoption → in-hospital activation); nothing unmapped goes live.

Chapter 4 · The 2026 Inspection Regime: The Regulator’s Three Pairs of Glasses

The publicly announced 2026 supervision posture can be summarized as “three inspection types plus a five-piece toolkit.” Understanding each selection logic lets a hospital predict through which door it would enter the regulator’s field of view.

4.1 Selection logic of the three inspection types

• Annual inspections: breadth coverage across all provincial-level regions, targeting pooling regions with high fund-risk indicators — abnormal hospitalization rates, payment ratios, self-pay ratios, fund deficits, concentrated tip-offs and data leads, weak past rectification — and institutions with large fund usage; the 2026 notice names orthopedics, oncology, diagnostics and testing, ophthalmology, dentistry, general surgery, and neurology as focus fields.
• Thematic inspections: problem-driven, aimed at issues of broad social concern, with LTC-insurance thematic inspections being explored.
• “Pinpoint” inspections: lead-driven and fast, triggered directly by big-data anomalies, complaints and tip-offs, and abnormally high self-pay ratios — with no preparation window for the institution.

Training point: the three types mean three exposure paths — annual inspections read your indicator profile (deviation of hospitalization, payment and self-pay ratios from peer institutions), thematic inspections ask whether you sit on a social hot spot, pinpoint inspections ask whether a concrete lead about you has been captured. The indicator profile is exposed all year round — which is why compliance cannot be a pre-inspection sprint and can only be daily data health.

4.2 The five-piece toolkit, mapped to hospital actions

4.3 One belief to correct: inspection is not “auditing the books” — it is data profiling plus on-site penetration

Legacy preparation assumes checks begin with the ledgers. In reality the team arrives with the profile already built: which departments, items, physicians and patient groups deviate from statistical normality — then pulls the matching charts for penetration testing. Two corollaries: first, tidy paperwork means nothing, because distributions do not lie — fabricated records leave sharper statistical anomalies (operation times clustering on the hour, assessment scores that repeat); second, the best preparation is data that is healthy every day, not ledgers made presentable before the visit.

Chapter 5 · A Typology in Eight Drills: Turning Rules into Reflexes

All scenarios are teaching fictions. Apply the four-step method — clinical facts → settlement rule → regulatory characterization → internal-control remediation — before opening each analysis.

Chapter 6 · Internal Control: Three Lines of Defense, Concretely

Line 1 · System rules (make the violation impossible)

• Build a billing exclusivity rule base: encode every exclusivity, counting rule (count-in/not-out) and inactive flag as hard HIS validation; version the rule base against provincial and municipal documents with a change log;
• Generate all time-accumulating counts (bed, consultation) exclusively from the system; manual back-entry requires dual authorization with full audit trail;
• Connect to the pre-event alert system and surface alerts at the physician workstation.

Line 2 · Data self-profiling (see the problem before the regulator does)

• Quarterly, replicate the inferable supervision screens hospital-wide: exclusivity co-occurrence, day-count checks, discharge-day charges, re-admission intervals, per-diem right-tail distributions, departmental self-pay and single-room rankings;
• Close the loop on self-inspection — screen, verify, characterize, refund, rectify, re-check — with documentation; proactive refunds and rectification records are material inputs to credit management and leniency;
• Track special-case deliberation filing rates in the insurance office’s monthly report: under-filing is a management failure on par with over-billing.

Line 3 · Personal accountability (everyone knows whose score it is)

• Payment-qualification scoring goes into all-staff and onboarding training, with in-hospital linkage rules to performance, titles and honors made explicit;
• Department heads and head nurses own their unit’s settlement-data health; operations provides department-level dashboards, not just hospital aggregates;
• Hold the policy red line: the DRG/DIP group payment standard may not be used as a quota to evaluate clinicians or tied directly to performance pay — route internal schemes through compliance review before launch.

Chapter 7 · Assessment (Scenario Judgments · Answer Keys)

1. Day 3 in the hospice ward, the family asks to add a psychological-counseling charge. Billable as a separate item? No — psychological and spiritual support is inside the hospice fee’s price composition.
2. A DRG case runs far above the payment standard and involved multidisciplinary care. The department’s correct move? File for special-case deliberation (capped at 5% of DRG discharges) — not cost-cutting refusal or admission-splitting.
3. Single-room fees are hospital-set. Filing and disclosure required? What does an abnormal self-pay ratio trigger? Yes to both duties; an abnormal self-pay ratio is a named pinpoint-inspection trigger.
4. What exactly does “count-in, not-out” mean? Admission day counts as one day; discharge day is not counted.
5. May a home wound-dressing visit be billed a home-visit fee for a patient in active LTC-insurance coverage? No — services funded through the LTC channel may not charge again; verify status before billing.
6. Which 2026 inspection type is triggered directly by leads such as abnormal self-pay ratios, with no preparation window? Pinpoint inspections.
7. May a hospital set the group payment standard as a departmental cost cap tied directly to physician pay? No — expressly prohibited.
8. The national guideline is out but the provincial price document has not switched. May the hospital create the new charge item? No — the billing basis is the locally effective price policy.
9. What is the single most important compliance artifact in a per-diem ward? The closed documentation chain: admission assessment — interval reassessment — symptom-control and nursing records.
10. Beyond legal duty, what direct benefit follows full traceability-code uploading? It is an explicit tilt factor for insurance fund prepayment — better cash flow.

FAQ

Related Reading

• 中文版：综合诊查类36项专业深度解读
• Medical Service Pricing · Insurance and Hospital Management