Diagnosis-Related Groups, Reimbursement, and the Quality Disconnect

Mini Abstract

Diagnosis Related Groups were established to reduce excess waste in the healthcare system. These classification systems provide standardized and scalable approaches to measuring severity of illness, assessment of treatment difficulty, and resource intensity. In this commentary, we discuss how the current DRG models in pediatrics—and for congenital heart surgery in particular—are not constructed to optimize quality.

Murin et al.¹ examined a cohort of infants who underwent open-heart procedures in Germany, and found that the case-mix index (CMI) used to determine reimbursement under the German Diagnosis Related Group (DRG) system was lower for patients extubated within the first 24 postoperative hours than for those extubated later, regardless of surgical complexity. They also found that the CMI was similar for patients undergoing low and high complexity surgeries within the early extubation cohort. This timely study succinctly illustrates the limitations of DRG systems, as they stand, in estimating reimbursement across the vast range of congenital cardiac care.

Diagnosis Related Groups were established to reduce excess waste in the healthcare system. These classification systems provide standardized and scalable approaches to measuring severity of illness, assessment of treatment difficulty, and resource intensity. Under DRGs, hospitals are reimbursed the same amount for clinically similar patients. In the development of DRGs, there was incentive to limit the number of different groupings, and hence more granular methods of case mix adjustment were not incorporated; given the small volume of pediatric patients in comparison to adults, lumping pediatric patients together had minimal effect on intra-DRG variability. While this might result in similar average per patient reimbursement across a hospital system, Murin et al.’s work highlights the fact that current DRG models in pediatrics—and for congenital heart surgery in particular—are not constructed to optimize quality.

The authors’ findings about case complexity are consistent with our previous investigations of the U.S. DRG system.² Patients with higher complexity disease, on average, require more resources than those with lower complexity disease. For neonates undergoing congenital heart surgery in the U.S., for example, all children are reimbursed under just four DRGs: neonates <1.5kg undergoing any major procedure, cardiac or non-cardiac, neonates 1.5-2.5kg undergoing any major procedure, cardiac or non-cardiac, neonates ≥2.5kg undergoing a major cardiovascular procedure, and neonates who undergo extracorporeal membrane oxygenation, with or without an associated cardiac procedure. These DRGs are divided into severity levels based on risk factors largely irrelevant to children, with almost all cases falling under the highest severity levels. As under the German DRG system, there are few considerations made in U.S. DRGs for cardiac anatomy or case complexity, despite vastly different resource requirements needed across these groups. This results in similar reimbursement for neonates undergoing Norwood procedures and those undergoing ventricular septal defect closures, so long as the children are of similar weight (Table 1). Given the known increase in expenditures necessary on average to care for high complexity children, this may strain centers caring for the most complex patients.

Table 1:

Average costs associated with hospitalization, by neonatal U.S. APR-DRG. Patients within each DRG are reimbursed the same amount for each hospital/payer, regardless of associated costs.

APR-DRG	DRG Description	Average Cost of Cardiac Cases (SD)	Average Cost of Non-Cardiac Cases (SD)
588 or 609	Neonates <2.5kg undergoing any major procedure—cardiovascular or not	$352K ($254K)	$277K ($238K)
583	Neonates undergoing extracorporeal membrane oxygenation—with or without a cardiovascular procedure	$418K ($285K)	$305K ($245K)
APR-DRG	DRG Description	Average Cost of RACHS-1, Highest Complexity Cases (SD)	Average Cost of RACHS-1, Lowest Complexity Cases (SD)
630	Neonates >2.5kg undergoing a major cardiovascular procedure	$260K ($193K)	$155K ($159K)

The authors’ conclusion about early extubation, while intuitively true, may be preemptive given the data presented. The authors note that their study investigates reimbursement, not profit. Reimbursement refers to payments received after medical services are rendered; profit refers to the difference between reimbursement and expenditures.³ Analyses of early use of positive pressure ventilation among preterm newborns in the U.S. support the authors’ conclusions.⁴ Yet multiple studies have shown that patients extubated early have shorter lengths of stay and reduced resource requirements following congenital heart surgery.^5–6 If these differences are greater than the reduction in reimbursement, profitability would increase with early extubation, despite reduced reimbursement, rather than decrease.

While even registry data used to study center quality do not contain the full granularity necessary to explain much of the observed variation in center-level outcomes, refining DRGs using existing knowledge of case complexity could improve sensitivity to variations in resource requirements. To start, DRGs for all neonates could differentiate those undergoing cardiac surgery from those undergoing non-cardiac surgery and could further differentiate cardiac surgical patients by surgical risk category.² Centers of Excellence designations could be refined, and base rates could potentially be augmented when centers demonstrate superior outcomes and complete repairs for the most complex patients. Uncoupling reimbursement from postoperative intubation duration, as the authors suggest, would further this concept. In addition, DRG modifiers could account for social determinants of health, which have increasingly shown associations with resource requirements.⁷ As clinicians and researchers, we must remain committed to optimizing patient care and outcomes, irrespective of potential implications on profit. Yet we should also advocate to incorporate changes to DRG systems that better reflect pediatric disease.