DOES CASE-MIX EXPLAIN VARIATION IN LENGTH OF STAY ACROSS ROMANIAN HOSPITALS? AN ANALYSIS OF NATIONAL DRG DATA IN ROMANIA
Author(s)
András Attila Mezei, MSc1, Eniko Fulop, MSc1, Diána Elmer, MSc, PhD2, Edina Eross, MSc1, Tímea Csákvári, MSc, PhD3, Attila Dénes Orbán, MSc1, Zsuzsanna Kívés, MSc, PhD2, Imre Boncz, MSc, PhD, MD2.
1Doctoral School of Health Sciences, University of Pécs, Pécs, Hungary, 2Institute for Health Insurance, University of Pécs, Faculty of Health Sciences, Pécs, Hungary, 3Institute for Health Insurance, University of Pécs, Faculty of Health Sciences, Zalaegerszeg, Hungary.
1Doctoral School of Health Sciences, University of Pécs, Pécs, Hungary, 2Institute for Health Insurance, University of Pécs, Faculty of Health Sciences, Pécs, Hungary, 3Institute for Health Insurance, University of Pécs, Faculty of Health Sciences, Zalaegerszeg, Hungary.
OBJECTIVES: To examine the association between hospital case-mix and average length of stay (LOS) in Romanian hospitals, assess whether this relationship is linear or non-linear, compare patterns between acute and chronic care, and determine the extent of residual LOS variation after case-mix adjustment.
METHODS: We analyzed national Romanian diagnosis-related group (DRG) data for 2023 at the hospital level. The analysis included 403 hospitals with at least 100 acute cases. Hospital case-mix was measured using the acute case-mix index (CMI), derived from DRG weights and case volumes. Average LOS was calculated from total bed-days and total cases. Descriptive statistics, scatterplots, case-mix quintile comparisons, and linear and quadratic fitted models were used. Residual LOS was examined to identify hospital-level differences remaining after case-mix adjustment.
RESULTS: The median hospital acute CMI was 1.41 (IQR: 1.21-1.64; range: 0.36-6.43). Most hospitals were concentrated within a moderate case-mix range, with few high-complexity outliers. A weak negative correlation was observed between hospital case-mix and average acute LOS (Pearson’s r = −0.15, p = 0.003), while fitted curves suggested only modest non-linearity. Across case-mix quintiles, median acute LOS values were 5.10, 5.21, 5.54, 5.47, and 4.79 days for Q1-Q5, indicating limited variation attributable to case-mix alone. Acute and chronic care showed different patterns, with substantially greater heterogeneity in chronic care. After case-mix adjustment, residual hospital-level LOS differences remained pronounced, ranging from −4.81 to +15.58 days.
CONCLUSIONS: Case-mix contributes to variation in hospital LOS, but its explanatory power appears limited. Substantial hospital-level differences remained after case-mix adjustment, suggesting that organizational and practice-related factors also influence LOS beyond patient complexity. These findings support the use of case-mix-adjusted performance assessment in the Romanian DRG system and highlight the need for further investigation of institutional variation.
METHODS: We analyzed national Romanian diagnosis-related group (DRG) data for 2023 at the hospital level. The analysis included 403 hospitals with at least 100 acute cases. Hospital case-mix was measured using the acute case-mix index (CMI), derived from DRG weights and case volumes. Average LOS was calculated from total bed-days and total cases. Descriptive statistics, scatterplots, case-mix quintile comparisons, and linear and quadratic fitted models were used. Residual LOS was examined to identify hospital-level differences remaining after case-mix adjustment.
RESULTS: The median hospital acute CMI was 1.41 (IQR: 1.21-1.64; range: 0.36-6.43). Most hospitals were concentrated within a moderate case-mix range, with few high-complexity outliers. A weak negative correlation was observed between hospital case-mix and average acute LOS (Pearson’s r = −0.15, p = 0.003), while fitted curves suggested only modest non-linearity. Across case-mix quintiles, median acute LOS values were 5.10, 5.21, 5.54, 5.47, and 4.79 days for Q1-Q5, indicating limited variation attributable to case-mix alone. Acute and chronic care showed different patterns, with substantially greater heterogeneity in chronic care. After case-mix adjustment, residual hospital-level LOS differences remained pronounced, ranging from −4.81 to +15.58 days.
CONCLUSIONS: Case-mix contributes to variation in hospital LOS, but its explanatory power appears limited. Substantial hospital-level differences remained after case-mix adjustment, suggesting that organizational and practice-related factors also influence LOS beyond patient complexity. These findings support the use of case-mix-adjusted performance assessment in the Romanian DRG system and highlight the need for further investigation of institutional variation.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
HPR198
Topic
Epidemiology & Public Health, Health Policy & Regulatory
Topic Subcategory
Insurance Systems & National Health Care, Reimbursement & Access Policy
Disease
No Additional Disease & Conditions/Specialized Treatment Areas