Implementing patient-based costing to derive robust DRG weights for national reimbursement and hospital pathway management
Introduction
Diagnosis-Related Group (DRG) payment systems depend on relative weights that approximate the average resource intensity of clinically coherent inpatient groups. When weights are weak - because costing is inconsistent, feeds are incomplete, or coding is unreliable - national tariffs can misprice care, destabilize provider incentives, and reduce confidence in performance analytics. In this essay I try to synthesize established approaches to patient-level costing (PLICS/ABC/TDABC) and DRG tariff setting, drawing on documented national programs (e.g., England’s National Cost Collection, Australia’s AHPCS/NEP process, and Germany’s InEK cost accounting scheme), and propose an implementable framework to (1) produce reliable national reimbursement weights, (2) embed auditability via a practical controls checklist (General Ledger (GL) reconciliation, feed completeness, allocation governance, coding QA/QC), and (3) operationalize patient-level costs for internal hospital case-mix and pathway management.
1. Introduction
DRGs were originally developed as a case-mix classification to define hospital output and enable comparative measurement of utilization and performance (e.g., length of stay), with the foundational formulation described by Fetter and colleagues. In contemporary prospective payment systems, DRG weights are intended to represent relative resource use, so that a base rate multiplied by the DRG weight yields an expected payment.
Two realities motivate patient-based costing for weight-setting:
- A tariff is only as credible as its cost evidence. National weight schedules require consistent rules for which costs are in scope, how overhead is allocated, how activity is measured, and how outliers are treated. The German G-DRG system, for example, relies on a standardized patient-level cost accounting scheme and national plausibility checks; DRG cost weights are recalculated annually based on participating hospitals’ case-cost data.
- Hospitals also need patient-level costs for management. Patient-level costing is designed to connect resource consumption to clinical pathways (admission-to-discharge and beyond), enabling benchmarking and clinical engagement with cost drivers.
In this essay I want to deal with “patient-based costing” as a bottom-up or hybrid approach that traces resources to patient events wherever possible (feeder systems), and uses principled allocation methods where direct tracing is infeasible - consistent with national guidance in England and Australia.
2. Conceptual foundation: from ledger to patient episode to DRG weight
2.1 Costing approaches used in practice
Patient-level costing programs typically operationalize one of three families:
- PLICS (patient-level information and costing systems): patient-specific costing derived from tracing resources and allocating remaining costs using activity-based principles, explicitly positioned as an improvement over coarse top-down averages.
- Activity-Based Costing (ABC) and Time-Driven ABC (TDABC): TDABC estimates capacity cost rates and multiplies them by observed process times; it was proposed to address distortions in traditional ABC and has been widely reviewed in health care applications.
- Standardized national hospital costing standards that define minimum data, allocation logic, and assurance - e.g., Australia’s Australian Hospital Patient Costing Standards (AHPCS) and guidance, and England’s Approved Costing Guidance.
2.2 DRG grouping and coding dependency
Because DRGs are assigned from coded diagnoses/procedures plus patient factors and discharge status, weight-setting is inseparable from coding quality and consistent grouper logic. CMS (Centers for Medicare & Medicaid Services), for example, publishes MS-DRG grouper software versions and definitions manuals to standardize grouping logic. Coding rules are formalized in official ICD-10-CM and ICD-10-PCS guidelines.
Note on the U.S. CMS approach: CMS describes the data inputs it uses for MS-DRG relative weight methodology, including cost report data to produce cost-to-charge ratios (CCRs) and standardizing files, and publishes supporting datasets (e.g., After Outliers Removed (AOR)/Before Outliers Removed (BOR) variables used in relative weight calculations). This illustrates the broader principle: the weight model is a transparent pipeline from standardized inputs to published weights. In the MS-DRG system, CC (Complication or Comorbidity) and MCC (Major Complication or Comorbidity) are secondary diagnosis classifications used to measure patient severity and resource consumption. They significantly impact hospital reimbursement by allowing higher payment for cases with greater complexity compared to the base diagnosis.
3. Implementing patient-based costing for reliable national reimbursement weights
National weights require more than good hospital cost accounting; they require standardization + collection + validation + statistical construction of weights + publishable governance. England’s National Cost Collection explicitly combines aggregated and patient-level costs and uses submissions to inform national prices. Australia’s IHACPA publishes national pricing determinations with price weights and adjustments (e.g., by Australian Refined Diagnosis Related Groups (AR-DRG) version), supported by national costing standards. Germany’s InEK handbook defines a detailed standardized scheme for case-cost calculation and includes checklists and reconciliation expectations.
3.1 Step 1 - Establish a national costing standard and “in-scope” cost universe
A national authority (payer/regulator/commission) must define:
- Cost scope: clinical operating costs included in DRG weights vs excluded costs (e.g., capital, research/teaching) in line with the national policy model. (Germany’s G-DRG scheme explicitly treats certain cost categories as non-DRG relevant; the handbook defines the costing period and reconciliation frame.)
- Unit of analysis: admission/spell/episode definition aligned to the DRG grouper inputs and national discharge datasets.
- Minimum feeder expectations: what “patient-level traceability” means for key domains (pharmacy, implants, OR time, imaging, labs), and what must be allocated instead. Australia’s AHPCS guidance explicitly describes identifying feeder data and performing QA checks before costing.
Deliverable: a published “Approved Costing Guidance / Costing Standards” document with compulsory and optional requirements, versioned annually (as done in England).
3.2 Step 2 - Build the national data model and submission specification
At minimum, national weight-setting needs:
- Case-mix file: patient demographics, admission/discharge dates, discharge status, principal + secondary diagnoses, procedures, and any policy variables; coding per official ICD guidance (e.g. WHO ICD-11, WHO ICHI, USA ICD-10-CM/PCS).
- Case-cost file: total cost per case and (preferably) cost by standardized cost modules/cost center groups. Germany’s InEK approach uses a uniform cost-matrix structure per case, enabling module-level plausibility checks.
- Control totals: audited-account reconciliation totals and key mapping tables (cost center mapping, feeder-to-case linkage rates).
3.3 Step 3 - Implement hospital PLICS “ledger → cost pools → patient activity → case cost”
A national standard should require hospitals to implement a repeatable pipeline:
(a) Cost quantum (GL → costing ledger):
- Map the general ledger to costing cost pools/cost centers, excluding out-of-scope costs per national rules. England’s CP1 emphasizes “correct cost quantum” and how the General Ledger (GL) is used for costing.
- Reconcile to audited accounts (see Controls, Section 4).
(b) Direct attribution (“trace where you can”):
- Drugs, implants/prostheses, blood products, some external services: attribute via patient-level feeder systems where available (AHPCS guidance repeatedly specifies feeder file fields and QA expectations).
(c) Activity-based allocation (for indirect and overhead):
- Allocate indirect costs to patient activity using causal drivers (ward days, ICU hours, OR minutes, imaging RVUs/points, lab workload measures), consistent with the “method of causation” principle described in the German scheme.
(d) Produce case-cost outputs in a standardized module structure:
- Ensure every case produces a complete module vector (even if some modules are zero), enabling national plausibility checks and trimming.
3.4 Step 4 - National validation, plausibility checks, and outlier handling
To produce robust weights, the national authority should:
- Run technical and formal checks: schema validation, identifier integrity, period alignment.
- Run plausibility checks: distribution corridors by module/DRG, feeder completeness expectations, reconciliation checks. The InEK process is explicitly described as using plausibility and conformity checks and may exclude cases from calculation if costs fall outside expected corridors.
- Define inlier/outlier logic: trimming rules for extremely short/long stays and high-cost cases (method must be published and stable across years).
3.5 Step 5 - Construct relative weights and normalize
A defensible national methodology typically:
- Computes mean (or trimmed mean) cost per DRG from validated (reliable) cases.
- Divides by the overall mean cost (or a reference group mean) to create relative weights.
- Applies a normalization factor so that the weighted average is stable year-to-year (CMS publishes normalization-related supplemental files; Australia publishes national price weights and associated adjustments).
Critical governance point: publish the verified and validated calculation methodology, version history, and summary statistics (case counts, exclusion rates) so the system is auditable and contestable.
4. Implementing a practical controls checklist for auditability and trust
Controls must connect three assurance layers:
- Financial integrity (General Ledger (GL) reconciliation),
- Data integrity (feed completeness + linkage), and
- Method integrity (allocation rule governance + coding QA/QC).
England’s costing guidance explicitly structures the process around cost quantum (CP1), cost identification (CP2), and cost allocation (CP3). Germany’s InEK handbook includes explicit reconciliation expectations (“Abgleich”) between audited annual accounts and the costing base and provides chapter checklists. Australia’s AHPCS guidance embeds feeder QA checks as a recurring requirement.
Note: There are five costing process standards (NHS, UK):
- CP1: Ensuring the correct cost quantum: set outs how the general ledger is used for costing and highlights the areas that require review to support accurate costing.
- CP2: Clearly identifying costs: ensures costs are in the correct starting position and correctly labelled for costing.
- CP3: Allocating costs to activities: ensures the correct quantum of costs is allocated to the correct activity using the most appropriate costing allocation method.
- CP4: Matching costed activities to patients: ensures consistency in assigning costed activities to the correct patient event.
- CP5: Reconciliation: sets out the process for reconciling costs and income to organisation accounts and reconciling the activity counts reported by the organisation.
4.1 Controls checklist (implementation-ready)
Below is a practical checklist format suitable for internal audit and external assurance. “Evidence” should be retained in a controlled repository per cycle.
| Control domain | Control objective | Minimum control(s) | Evidence (examples) | Frequency / owner |
|---|---|---|---|---|
| GL reconciliation (CP1 / Abgleich) | Cost quantum complete and consistent with audited accounts | (1) Reconcile total in-scope costs to audited trial balance; (2) Reconcile cost-type to cost-center ledger; (3) Document exclusions and adjustments | Signed reconciliation workbook; mapping table; adjustment log; approval trail | Monthly close + annual submission / Finance controller + Costing lead |
| Feed completeness | Patient activity feeds are complete and linkable to cases | (1) Interface monitoring (record counts vs source); (2) Missingness thresholds for key fields (patient ID, episode ID, date/time); (3) Linkage rate KPIs by feeder | Automated completeness reports; exception queue; root-cause tickets | Daily/weekly ops + monthly certification / Data engineering + Costing |
| Allocation rule governance (CP2–CP3) | Allocation methods are causal, versioned, and approved | (1) Cost driver dictionary with rationale; (2) Version control and change approval; (3) Materiality thresholds; (4) Segregation of duties (build vs approve) | Allocation rule register; change requests; sign-off minutes | Quarterly + on change / Costing governance committee |
| Feeder QA (AHPCS “perform QA checks”) | Feeder files are reliable for costing | (1) Validate date ranges; (2) Remove error dates; (3) Duplicate detection; (4) Unit consistency; (5) Price field validity where relevant | QA scripts output; exception logs; corrected extracts | Every costing run / Costing analysts |
| Coding QA/QC | DRG assignment inputs are accurate and compliant | (1) Routine coding audits; (2) Targeted audits for high-impact DRGs and CC/MCC capture; (3) Education feedback loop; (4) Clinical documentation improvement linkage | Audit sampling plan; audit results; coder education records | Monthly/quarterly / HIM + CDI leadership |
| Grouper/version control | DRG logic consistent over time and reproducible | (1) Lock grouper version per period; (2) Regression testing on reference cases; (3) Document mapping changes year-to-year | Grouper version manifest; test suite results | On release / HIM + Analytics |
| Submission certification (national weights) | National dataset is explainable and defensible | (1) Hospital CEO/CFO attestation; (2) Disclosure of deviations from costing standard; (3) Exclusion rate and linkage KPI disclosure | Signed attestation; deviation register | Annual / CFO + National authority |
This checklist is intentionally aligned to published costing process standards and reconciliation expectations in England and Germany, and to feeder QA expectations in Australia.
5. Implementing patient-based costing for internal hospital case-mix and pathway management
National reimbursement weights answer “what should we be paid (on average)?” Internal management needs a different lens: “why does this pathway cost what it costs here, and what can we change without harming outcomes?”
5.1 Core internal outputs
Using the same PLICS pipeline, hospitals can produce:
- Patient-level pathway costs from referral/admission to discharge, identifying cost concentration points (“where on the pathway costs are incurred”).
- Service-line profitability under the prevailing tariff, decomposed into volume, case-mix (CMI), and cost per case.
- Variance analytics: compare observed cost vs expected cost for the same DRG (or DRG+severity subclass, where used). For APR-DRG, the methodology explicitly incorporates severity of illness (SOI) and risk of mortality (ROM) subclasses (often used for risk adjustment and profiling).
5.2 Operationalizing pathway improvement (PLICS + TDABC targeting)
A practical approach is hybrid:
- Use PLICS to identify high-variance, high-volume, high-cost DRGs/pathways.
- For those pathways, deploy TDABC (process-time mapping) to identify operational drivers of cost and capacity use, consistent with TDABC’s intent to connect resource capacity and time to cost.
This avoids the common failure mode where hospitals attempt TDABC everywhere (too resource intensive) or rely purely on top-down allocations (too insensitive to process redesign).
5.3 Case-mix stewardship and clinical engagement
The National Institute for Health and Care Research (NIHR) PLICS study emphasizes that patient-level costs can strengthen benchmarking and clinical ownership - provided clinicians actually use the information to change practice, not just to identify “unprofitable” areas. Therefore, implementation should include:
- Clinical dashboards showing pathway cost components (ward, ICU, OR, drugs, imaging, labs) and key outcome/quality indicators (to prevent cost-only optimization).
- Standard work for monthly service-line reviews: cost outliers, length-of-stay drivers, implant/drug utilization, readmissions.
- Documentation and coding integrity as a quality issue, not merely revenue protection, using structured audit programs.
6. Discussion: risks and safeguards
Even rigorous patient-level costing does not automatically produce fair reimbursement weights. Known risks include:
- Participation and representativeness bias in national samples; Germany’s recalculation historically relied on a subset of hospitals (e.g., Vogl reports 263 hospitals in 2009), requiring careful plausibility checks and policy adjustment.
- Geographic price variation and labor market differences: national weights may need separation of relative weights from base-rate adjustments (wage indices/market-force factors) to avoid embedding regional costs into “clinical” weights. (CMS publishes wage index and standardized amount tables separately from DRG weights, illustrating this separation of mechanisms.)
- Coding response to incentives (upcoding/documentation intensity and cherry picking/lemon dropping): this is why coding QA/QC and clinical validation loops are not optional.
- Allocation distortions where feeders are weak; national standards (AHPCS/England/InEK) explicitly push feeder QA and causal drivers, but enforcement and assurance capacity must match ambition.
7. Conclusion
To produce reliable and robust national DRG reimbursement weights, patient-based costing must be treated as national infrastructure: standardized costing rules, consistent grouper logic, disciplined data submissions, rigorous plausibility checks, and transparent statistical weight construction. To be auditable, the system requires a controls checklist that ties GL reconciliation, feed completeness, allocation governance, and coding QA/QC into a single assurance narrative with retained evidence. Finally, the same patient-level costing foundation enables internal hospital case-mix and pathway management when coupled with targeted TDABC on high-variance pathways and sustained clinical engagement.
Bibliography
Australian Independent Hospital Pricing Authority (IHACPA). Australian Hospital Patient Costing Standards (AHPCS), Version 4.2 and Costing Guidelines.
Australian Independent Hospital Pricing Authority (IHACPA). National Efficient Price Determination 2025–26 (price weights and adjustments).
CMS (Centers for Medicare & Medicaid Services). FY 2026 IPPS Final Rule home page and supporting data files (relative weights inputs, CCRs/HCRIS, standardizing file, AOR/BOR).
CMS. MS-DRG classifications and grouper software; MS-DRG definitions manual.
CDC/NCHS. ICD-10-CM Official Guidelines for Coding and Reporting. (FY 2026)
CMS. ICD-10-PCS Official Guidelines for Coding and Reporting. (FY 2026)
England (NHS England). Approved Costing Guidance (including quick guide CP1–CP3) and National Cost Collection overview. (2025)
NHS Digital / NHS England. PLICS information and PLICS collections (patient-level activity and costing).
InEK / German hospital self-administration partners. Handbook for calculating treatment costs (Kalkulationshandbuch), Version 4.0 (2016).