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 journalArticlepeer-review

2 Scopus citations


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
Issue number2
StatePublished - Jun 2003
Externally publishedYes


  • Medicare HMOs
  • Risk adjustment

ASJC Scopus subject areas

  • Health Policy
  • Public Health, Environmental and Occupational Health


Dive into the research topics of 'An analysis of unobserved selection in an inpatient diagnostic cost group model'. Together they form a unique fingerprint.

Cite this