Methods for understanding super-efficient data envelopment analysis results with an application to hospital inpatient surgery

Liam O'Neill, Franklin Dexter

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

We compare two techniques for increasing the transparency and face validity of Data Envelopment Analysis (DEA) results for managers at a single decision-making unit: multifactor efficiency (MFE) and non-radial super-efficiency (NRSE). Both methods incorporate the slack values from the super-efficient DEA model to provide a more robust performance measure than radial super-efficiency scores. MFE and NRSE are equivalent for unique optimal solutions and a single output. MFE incorporates the slack values from multiple output variables, whereas NRSE does not. MFE can be more transparent to managers since it involves no additional optimization steps beyond the DEA, whereas NRSE requires several. We compare results for operating room managers at an Iowa hospital evaluating its growth potential for multiple surgical specialties. In addition, we address the problem of upward bias of the slack values of the super-efficient DEA model.

Original languageEnglish
Pages (from-to)291-298
Number of pages8
JournalHealth Care Management Science
Volume8
Issue number4
DOIs
StatePublished - 1 Nov 2005

Fingerprint

Inpatients
Surgical Specialties
Operating Rooms
Reproducibility of Results
Decision Making
Data envelopment analysis
Surgery
Super-efficiency
Growth
Multi-factor
Managers

Cite this

@article{0260e42921df415cb901ed7e87197f24,
title = "Methods for understanding super-efficient data envelopment analysis results with an application to hospital inpatient surgery",
abstract = "We compare two techniques for increasing the transparency and face validity of Data Envelopment Analysis (DEA) results for managers at a single decision-making unit: multifactor efficiency (MFE) and non-radial super-efficiency (NRSE). Both methods incorporate the slack values from the super-efficient DEA model to provide a more robust performance measure than radial super-efficiency scores. MFE and NRSE are equivalent for unique optimal solutions and a single output. MFE incorporates the slack values from multiple output variables, whereas NRSE does not. MFE can be more transparent to managers since it involves no additional optimization steps beyond the DEA, whereas NRSE requires several. We compare results for operating room managers at an Iowa hospital evaluating its growth potential for multiple surgical specialties. In addition, we address the problem of upward bias of the slack values of the super-efficient DEA model.",
author = "Liam O'Neill and Franklin Dexter",
year = "2005",
month = "11",
day = "1",
doi = "10.1007/s10729-005-4139-3",
language = "English",
volume = "8",
pages = "291--298",
journal = "Health Care Management Science",
issn = "1386-9620",
publisher = "Kluwer Academic Publishers",
number = "4",

}

Methods for understanding super-efficient data envelopment analysis results with an application to hospital inpatient surgery. / O'Neill, Liam; Dexter, Franklin.

In: Health Care Management Science, Vol. 8, No. 4, 01.11.2005, p. 291-298.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Methods for understanding super-efficient data envelopment analysis results with an application to hospital inpatient surgery

AU - O'Neill, Liam

AU - Dexter, Franklin

PY - 2005/11/1

Y1 - 2005/11/1

N2 - We compare two techniques for increasing the transparency and face validity of Data Envelopment Analysis (DEA) results for managers at a single decision-making unit: multifactor efficiency (MFE) and non-radial super-efficiency (NRSE). Both methods incorporate the slack values from the super-efficient DEA model to provide a more robust performance measure than radial super-efficiency scores. MFE and NRSE are equivalent for unique optimal solutions and a single output. MFE incorporates the slack values from multiple output variables, whereas NRSE does not. MFE can be more transparent to managers since it involves no additional optimization steps beyond the DEA, whereas NRSE requires several. We compare results for operating room managers at an Iowa hospital evaluating its growth potential for multiple surgical specialties. In addition, we address the problem of upward bias of the slack values of the super-efficient DEA model.

AB - We compare two techniques for increasing the transparency and face validity of Data Envelopment Analysis (DEA) results for managers at a single decision-making unit: multifactor efficiency (MFE) and non-radial super-efficiency (NRSE). Both methods incorporate the slack values from the super-efficient DEA model to provide a more robust performance measure than radial super-efficiency scores. MFE and NRSE are equivalent for unique optimal solutions and a single output. MFE incorporates the slack values from multiple output variables, whereas NRSE does not. MFE can be more transparent to managers since it involves no additional optimization steps beyond the DEA, whereas NRSE requires several. We compare results for operating room managers at an Iowa hospital evaluating its growth potential for multiple surgical specialties. In addition, we address the problem of upward bias of the slack values of the super-efficient DEA model.

UR - http://www.scopus.com/inward/record.url?scp=29144484064&partnerID=8YFLogxK

U2 - 10.1007/s10729-005-4139-3

DO - 10.1007/s10729-005-4139-3

M3 - Article

C2 - 16379412

AN - SCOPUS:29144484064

VL - 8

SP - 291

EP - 298

JO - Health Care Management Science

JF - Health Care Management Science

SN - 1386-9620

IS - 4

ER -