Admission MedisGroups score and the cost of hospitalizations

UMMS Affiliation

Department of Quantitative Health Sciences

Publication Date


Document Type



Aged; *Costs and Cost Analysis; Diagnosis-Related Groups; Fees and Charges; Humans; Length of Stay; Massachusetts; Patient Admission; Probability; Sampling Studies; *Severity of Illness Index


Biostatistics | Epidemiology | Health Services Research


Concerns about the insensitivity of Medicare's diagnosis-related groups (DRGs) to illness severity heightened interest in the potential of alternative patient classification systems to improve the fairness of hospital reimbursement. This article examines the ability of admission MedisGroups score to explain the costs of hospital stays. The database contained 54,112 patients 65 years or older discharged in 28 high-frequency DRGs from 1984 to the middle of 1986 from 24 hospitals across the country. Admission MedisGroups score alone explained 3% of costs using trimmed data. Addition of admission MedisGroups score to DRGs modestly improved ability to predict differences in cost: for trimmed data, DRGs alone explained 52% of the variation in costs, compared with 55% when admission MedisGroups score was added. Within individual DRGs, the explanatory power of admission MedisGroups score ranged from 0% to 21%. The level of explanatory power was not related to the spread of cases across admission MedisGroups scores within DRG. No consistent clinical pattern explained these differences across DRGs.


Med Care. 1988 Nov;26(11):1068-80. Link to article on publisher's site

Journal/Book/Conference Title

Medical care

PubMed ID


Related Resources

Link to Article in PubMed