An analysis of unobserved selection in an inpatient diagnostic cost group model

Hongjun Kan, Dana Goldman, Emmett Keeler, Nasreen Dhanani, Glenn Melnick

Research output: Contribution to journalArticle

Abstract

The study assesses unobserved selection bias in an inpatient diagnostic cost group (DCG) model similar to Medicare's Principal Inpatient Diagnostic Cost Group (PIP-DCG) risk adjustment model using a unique data set that contains hospital discharge records for both FFS and HMO Medicare beneficiaries in California from 1994 to 1996. We use a simultaneous equations model that jointly estimates HMO enrollment and subsequent hospital use to test the existence of unobserved selection and estimate the true HMO effect. It is found that the inpatient DCG model does not adequately adjust for biased selection into Medicare HMOs. New HMO enrollees are healthier than FFS beneficiaries even after adjustment for the included PIP-DCG risk factors. A model developed over an FFS sample ignoring unobserved selection overestimates hospital use of new HMO enrollees by 28 percent compared to their use if they had remained in FFS. Models that better captures selection bias are needed to reduce overestimation of Medicare HMO enrollees' resource use.

Original languageEnglish (US)
Pages (from-to)71-91
Number of pages21
JournalHealth Services and Outcomes Research Methodology
Volume4
Issue number2
DOIs
StatePublished - Jun 2003
Externally publishedYes

Fingerprint

Health Maintenance Organizations
Inpatients
Costs and Cost Analysis
Medicare
Selection Bias
Risk Adjustment
Hospital Records

Keywords

  • Medicare HMOs
  • PIP-DCG
  • Risk adjustment

ASJC Scopus subject areas

  • Health Policy
  • Public Health, Environmental and Occupational Health

Cite this

An analysis of unobserved selection in an inpatient diagnostic cost group model. / Kan, Hongjun; Goldman, Dana; Keeler, Emmett; Dhanani, Nasreen; Melnick, Glenn.

In: Health Services and Outcomes Research Methodology, Vol. 4, No. 2, 06.2003, p. 71-91.

Research output: Contribution to journalArticle

Kan, Hongjun ; Goldman, Dana ; Keeler, Emmett ; Dhanani, Nasreen ; Melnick, Glenn. / An analysis of unobserved selection in an inpatient diagnostic cost group model. In: Health Services and Outcomes Research Methodology. 2003 ; Vol. 4, No. 2. pp. 71-91.
@article{2bfacaeddfc24eda98097a7e8965b175,
title = "An analysis of unobserved selection in an inpatient diagnostic cost group model",
abstract = "The study assesses unobserved selection bias in an inpatient diagnostic cost group (DCG) model similar to Medicare's Principal Inpatient Diagnostic Cost Group (PIP-DCG) risk adjustment model using a unique data set that contains hospital discharge records for both FFS and HMO Medicare beneficiaries in California from 1994 to 1996. We use a simultaneous equations model that jointly estimates HMO enrollment and subsequent hospital use to test the existence of unobserved selection and estimate the true HMO effect. It is found that the inpatient DCG model does not adequately adjust for biased selection into Medicare HMOs. New HMO enrollees are healthier than FFS beneficiaries even after adjustment for the included PIP-DCG risk factors. A model developed over an FFS sample ignoring unobserved selection overestimates hospital use of new HMO enrollees by 28 percent compared to their use if they had remained in FFS. Models that better captures selection bias are needed to reduce overestimation of Medicare HMO enrollees' resource use.",
keywords = "Medicare HMOs, PIP-DCG, Risk adjustment",
author = "Hongjun Kan and Dana Goldman and Emmett Keeler and Nasreen Dhanani and Glenn Melnick",
year = "2003",
month = "6",
doi = "10.1023/B:HSOR.0000027921.40194.9d",
language = "English (US)",
volume = "4",
pages = "71--91",
journal = "Health Services and Outcomes Research Methodology",
issn = "1387-3741",
publisher = "Springer Netherlands",
number = "2",

}

TY - JOUR

T1 - An analysis of unobserved selection in an inpatient diagnostic cost group model

AU - Kan, Hongjun

AU - Goldman, Dana

AU - Keeler, Emmett

AU - Dhanani, Nasreen

AU - Melnick, Glenn

PY - 2003/6

Y1 - 2003/6

N2 - The study assesses unobserved selection bias in an inpatient diagnostic cost group (DCG) model similar to Medicare's Principal Inpatient Diagnostic Cost Group (PIP-DCG) risk adjustment model using a unique data set that contains hospital discharge records for both FFS and HMO Medicare beneficiaries in California from 1994 to 1996. We use a simultaneous equations model that jointly estimates HMO enrollment and subsequent hospital use to test the existence of unobserved selection and estimate the true HMO effect. It is found that the inpatient DCG model does not adequately adjust for biased selection into Medicare HMOs. New HMO enrollees are healthier than FFS beneficiaries even after adjustment for the included PIP-DCG risk factors. A model developed over an FFS sample ignoring unobserved selection overestimates hospital use of new HMO enrollees by 28 percent compared to their use if they had remained in FFS. Models that better captures selection bias are needed to reduce overestimation of Medicare HMO enrollees' resource use.

AB - The study assesses unobserved selection bias in an inpatient diagnostic cost group (DCG) model similar to Medicare's Principal Inpatient Diagnostic Cost Group (PIP-DCG) risk adjustment model using a unique data set that contains hospital discharge records for both FFS and HMO Medicare beneficiaries in California from 1994 to 1996. We use a simultaneous equations model that jointly estimates HMO enrollment and subsequent hospital use to test the existence of unobserved selection and estimate the true HMO effect. It is found that the inpatient DCG model does not adequately adjust for biased selection into Medicare HMOs. New HMO enrollees are healthier than FFS beneficiaries even after adjustment for the included PIP-DCG risk factors. A model developed over an FFS sample ignoring unobserved selection overestimates hospital use of new HMO enrollees by 28 percent compared to their use if they had remained in FFS. Models that better captures selection bias are needed to reduce overestimation of Medicare HMO enrollees' resource use.

KW - Medicare HMOs

KW - PIP-DCG

KW - Risk adjustment

UR - http://www.scopus.com/inward/record.url?scp=4344607885&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=4344607885&partnerID=8YFLogxK

U2 - 10.1023/B:HSOR.0000027921.40194.9d

DO - 10.1023/B:HSOR.0000027921.40194.9d

M3 - Article

VL - 4

SP - 71

EP - 91

JO - Health Services and Outcomes Research Methodology

JF - Health Services and Outcomes Research Methodology

SN - 1387-3741

IS - 2

ER -