HLTH 8052 Week 7 DRG Payment Import Critique Example

Importing Case-Based Hospital Payment Into a System With Few Coders and Thin Cost Data: A Critique of the Diagnosis-Related Group Model's Dependencies and the Behavior It Can Produce When They Are Missing

Student Name

Walden University

HLTH 8052: Advanced Issues in Global Public Health

Instructor Name

Month Day, Year

Importing Case-Based Hospital Payment Into a System With Few Coders and Thin Cost Data: A Critique of the Diagnosis-Related Group Model's Dependencies and the Behavior It Can Produce When They Are Missing

Introduction

Payment by diagnosis-related group, in which hospitals receive a set amount for each case of a given type, arrived in many middle-income countries with a strong reputation. It is associated with shorter stays, more transparent hospital financing and an end to open-ended budgets. This critique tests that reputation against a composite middle-income country, called here the importing system, with partial electronic records, few trained clinical coders and reliable cost data from only a handful of hospitals. It treats the model not as a success story to be copied but as a design that depends on supports, and it asks what the model does to hospital behavior when those supports are thinner than its design assumed.

How the Model Works Where It Was Built

In its original form, case-based payment groups hospital stays into categories that are clinically similar and expected to use similar resources. Each group carries a weight reflecting its relative cost, and a hospital's payment for a case equals the weight multiplied by a base rate. Three supports make this work. Each discharge must be classified accurately, which requires complete clinical records and trained coders who translate diagnoses and procedures into standard codes. The weights must reflect real costs, which requires detailed cost accounting in a sample of hospitals. And claims must flow through an information system that can process them, audit them and pay on time. Case-based payment is therefore different from a global budget, which pays a hospital a fixed sum for a year regardless of cases, and from a fee schedule, which pays for each service delivered. Its incentive is to treat each case at a cost below its payment.

Dependency 1: Accurate Classification

The importing system's records are partly electronic and partly paper, and discharge summaries are often incomplete. It has few trained clinical coders, most of them in the capital's teaching hospitals. Reviewing case-based payment across lower-income and middle-income settings, Mathauer and Wittenbecher (2013) counted 12 countries had such systems and another 17 were piloting or exploring them, and reported that the requirements of coding standardization, data availability and information technology made implementation difficult. In the importing system, a hospital that cannot code accurately will be paid wrongly in both directions, sometimes too little for complex cases and sometimes too much for simple ones.

Dependency 2: Weights Based on Real Costs

Weights in the original setting come from cost data collected in many hospitals. The importing system has reliable cost accounting in only a few, mostly private or university hospitals whose case mix and costs differ from district facilities. The same review found that countries chose among imported and self-developed models and that most adapted them to local conditions (Mathauer & Wittenbecher, 2013). A country that imports weights from another system without local cost data will pay its hospitals according to another country's relative prices. Rural hospitals with higher transport and staffing costs may be paid below their costs for every case.

Dependency 3: An Information System That Can Pay and Audit

The model requires claims to be submitted, checked and paid on a predictable schedule. In the importing system, claims processing is manual in many regions and audit capacity is minimal. Without audit, there is no check on whether coded cases match the clinical record. Without timely payment, hospitals face cash shortages that can push them to delay purchases or ask patients to pay informally.

Behavioral Risks When the Supports Are Thin

Case-based payment changes hospital behavior, and some of the changes are documented responses rather than speculation. The first is upcoding: classifying cases into higher-paying groups than the clinical record supports. Even in a system with mature coding and audit, a study of Medicare hospitals found that between 1989 and 1996 the share of pneumonia and respiratory infection cases placed in the most generous group rose by 10 percentage points in not-for-profit hospitals, 23 points in for-profit hospitals and 37 points in hospitals converting to for-profit status (Silverman & Skinner, 2004). Where audit is weaker, the incentive is the same and the check is smaller. The second is early discharge, since a hospital paid per case gains by shortening each stay; without monitoring of readmissions, shorter stays can mean patients sent home before they are ready. The third is selection: hospitals may prefer cases whose payment exceeds their expected cost and avoid complex patients whose costs are likely to exceed the weight. In a system where public hospitals are also the provider of last resort, avoidance concentrates the most expensive patients in facilities least able to absorb the loss.

What the Evidence Can and Cannot Say

The review of low- and middle-income experience reported that all countries with case-based payment had set expenditure ceilings, that most piloted the system before national rollout and that most had managed to separate hospital financing from line-item public budgets; it also noted that more evidence on the impact of these systems was needed (Mathauer & Wittenbecher, 2013). The evidence therefore supports caution about sequence more than a verdict on the model. Busse and colleagues, describing the European experience, emphasized that case-based systems took years to mature and were repeatedly revised as data improved (Busse et al., 2011). Neither source shows that the model fails in a system like the importing one; both show that it depends on supports that take time to build.

Recommendation

The importing system should not adopt case-based payment as its main hospital payment method yet. It should adapt the model in stages. First, use the grouping logic for information only: classify discharges and report them back to hospitals, without linking payment, so coding skills and data quality improve. Second, develop local cost weights from a sample of public hospitals across regions, including district facilities. Third, link a small share of payment to case groups in pilot hospitals, under an expenditure ceiling and with audit of a random sample of coded records. Only when coding accuracy, cost data and claims processing reach defined standards should the share of payment based on case groups increase. During each stage the ministry should track four indicators: the share of a random sample of coded records whose code matches the clinical record on audit, average length of stay, readmissions within 30 days, and the share of complex cases treated at public hospitals. A rise in coding mismatches signals upcoding; falling stays with rising readmissions signal early discharge; a growing share of complex cases in public hospitals signals selection by others. Each indicator uses figures the ministry already gathers or could draw from a sample. This sequence keeps the model's main advantage, a payment that follows the patient's needs, while waiting for the supports that make it work as designed.

References

Busse, R., Geissler, A., Quentin, W., & Wiley, M. (Eds.). (2011). Diagnosis-related groups in Europe: Moving towards transparency, efficiency and quality in hospitals. Open University Press.

Mathauer, I., & Wittenbecher, F. (2013). Hospital payment systems based on diagnosis-related groups: Experiences in low- and middle-income countries. Bulletin of the World Health Organization, 91(10), 746-756A. https://doi.org/10.2471/BLT.12.115931

Silverman, E., & Skinner, J. (2004). Medicare upcoding and hospital ownership. Journal of Health Economics, 23(2), 369-389. https://doi.org/10.1016/j.jhealeco.2003.09.007