Can use of an administrative database improve accuracy of hospital-reported readmission rates?

James R. Edgerton, Morley A. Herbert, Baron L. Hamman, W. Steves Ring

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Objectives: Readmission rates after cardiac surgery are being used as a quality indicator; they are also being collected by Medicare and are tied to reimbursement. Accurate knowledge of readmission rates may be difficult to achieve because patients may be readmitted to different hospitals. In our area, 81 hospitals share administrative claims data; 28 of these hospitals (from 5 different hospital systems) do cardiac surgery and share Society of Thoracic Surgeons (STS) clinical data. We used these 2 sources to compare the readmissions data for accuracy. Methods: A total of 45,539 STS records from January 2008 to December 2016 were matched with the hospital billing data records. Using the index visit as the start date, the billing records were queried for any subsequent in-patient visits for that patient. The billing records included date of readmission and hospital of readmission data and were compared with the data captured in the STS record. Results: We found 1153 (2.5%) patients who had STS records that were marked "No" or "missing," but there were billing records that showed a readmission. The reported STS readmission rate of 4796 (10.5%) underreported the readmission rate by 2.5 actual percentage points. The true rate should have been 13.0%. Actual readmission rate was 23.8% higher than reported by the clinical database. Approximately 36% of readmissions were to a hospital that was a part of a different hospital system. Conclusions: It is important to know accurate readmission rates for quality improvement processes and institutional financial planning. Matching patient records to an administrative database showed that the clinical database may fail to capture many readmissions. Combining data with an administrative database can enhance accuracy of reporting.

Original languageEnglish (US)
JournalJournal of Thoracic and Cardiovascular Surgery
DOIs
StateAccepted/In press - Jan 1 2018

Fingerprint

Patient Readmission
Databases
Thorax
Thoracic Surgery
Medicare
Quality Improvement
Surgeons

Keywords

  • Accuracy of readmission
  • Administrative database
  • Clinical database
  • Database
  • Hospital reported
  • Readmission
  • Readmission after cardiac surgery

ASJC Scopus subject areas

  • Surgery
  • Pulmonary and Respiratory Medicine
  • Cardiology and Cardiovascular Medicine

Cite this

Can use of an administrative database improve accuracy of hospital-reported readmission rates? / Edgerton, James R.; Herbert, Morley A.; Hamman, Baron L.; Ring, W. Steves.

In: Journal of Thoracic and Cardiovascular Surgery, 01.01.2018.

Research output: Contribution to journalArticle

@article{86051310856a4da492c18359947b017e,
title = "Can use of an administrative database improve accuracy of hospital-reported readmission rates?",
abstract = "Objectives: Readmission rates after cardiac surgery are being used as a quality indicator; they are also being collected by Medicare and are tied to reimbursement. Accurate knowledge of readmission rates may be difficult to achieve because patients may be readmitted to different hospitals. In our area, 81 hospitals share administrative claims data; 28 of these hospitals (from 5 different hospital systems) do cardiac surgery and share Society of Thoracic Surgeons (STS) clinical data. We used these 2 sources to compare the readmissions data for accuracy. Methods: A total of 45,539 STS records from January 2008 to December 2016 were matched with the hospital billing data records. Using the index visit as the start date, the billing records were queried for any subsequent in-patient visits for that patient. The billing records included date of readmission and hospital of readmission data and were compared with the data captured in the STS record. Results: We found 1153 (2.5{\%}) patients who had STS records that were marked {"}No{"} or {"}missing,{"} but there were billing records that showed a readmission. The reported STS readmission rate of 4796 (10.5{\%}) underreported the readmission rate by 2.5 actual percentage points. The true rate should have been 13.0{\%}. Actual readmission rate was 23.8{\%} higher than reported by the clinical database. Approximately 36{\%} of readmissions were to a hospital that was a part of a different hospital system. Conclusions: It is important to know accurate readmission rates for quality improvement processes and institutional financial planning. Matching patient records to an administrative database showed that the clinical database may fail to capture many readmissions. Combining data with an administrative database can enhance accuracy of reporting.",
keywords = "Accuracy of readmission, Administrative database, Clinical database, Database, Hospital reported, Readmission, Readmission after cardiac surgery",
author = "Edgerton, {James R.} and Herbert, {Morley A.} and Hamman, {Baron L.} and Ring, {W. Steves}",
year = "2018",
month = "1",
day = "1",
doi = "10.1016/j.jtcvs.2017.11.071",
language = "English (US)",
journal = "Journal of Thoracic and Cardiovascular Surgery",
issn = "0022-5223",
publisher = "Mosby Inc.",

}

TY - JOUR

T1 - Can use of an administrative database improve accuracy of hospital-reported readmission rates?

AU - Edgerton, James R.

AU - Herbert, Morley A.

AU - Hamman, Baron L.

AU - Ring, W. Steves

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Objectives: Readmission rates after cardiac surgery are being used as a quality indicator; they are also being collected by Medicare and are tied to reimbursement. Accurate knowledge of readmission rates may be difficult to achieve because patients may be readmitted to different hospitals. In our area, 81 hospitals share administrative claims data; 28 of these hospitals (from 5 different hospital systems) do cardiac surgery and share Society of Thoracic Surgeons (STS) clinical data. We used these 2 sources to compare the readmissions data for accuracy. Methods: A total of 45,539 STS records from January 2008 to December 2016 were matched with the hospital billing data records. Using the index visit as the start date, the billing records were queried for any subsequent in-patient visits for that patient. The billing records included date of readmission and hospital of readmission data and were compared with the data captured in the STS record. Results: We found 1153 (2.5%) patients who had STS records that were marked "No" or "missing," but there were billing records that showed a readmission. The reported STS readmission rate of 4796 (10.5%) underreported the readmission rate by 2.5 actual percentage points. The true rate should have been 13.0%. Actual readmission rate was 23.8% higher than reported by the clinical database. Approximately 36% of readmissions were to a hospital that was a part of a different hospital system. Conclusions: It is important to know accurate readmission rates for quality improvement processes and institutional financial planning. Matching patient records to an administrative database showed that the clinical database may fail to capture many readmissions. Combining data with an administrative database can enhance accuracy of reporting.

AB - Objectives: Readmission rates after cardiac surgery are being used as a quality indicator; they are also being collected by Medicare and are tied to reimbursement. Accurate knowledge of readmission rates may be difficult to achieve because patients may be readmitted to different hospitals. In our area, 81 hospitals share administrative claims data; 28 of these hospitals (from 5 different hospital systems) do cardiac surgery and share Society of Thoracic Surgeons (STS) clinical data. We used these 2 sources to compare the readmissions data for accuracy. Methods: A total of 45,539 STS records from January 2008 to December 2016 were matched with the hospital billing data records. Using the index visit as the start date, the billing records were queried for any subsequent in-patient visits for that patient. The billing records included date of readmission and hospital of readmission data and were compared with the data captured in the STS record. Results: We found 1153 (2.5%) patients who had STS records that were marked "No" or "missing," but there were billing records that showed a readmission. The reported STS readmission rate of 4796 (10.5%) underreported the readmission rate by 2.5 actual percentage points. The true rate should have been 13.0%. Actual readmission rate was 23.8% higher than reported by the clinical database. Approximately 36% of readmissions were to a hospital that was a part of a different hospital system. Conclusions: It is important to know accurate readmission rates for quality improvement processes and institutional financial planning. Matching patient records to an administrative database showed that the clinical database may fail to capture many readmissions. Combining data with an administrative database can enhance accuracy of reporting.

KW - Accuracy of readmission

KW - Administrative database

KW - Clinical database

KW - Database

KW - Hospital reported

KW - Readmission

KW - Readmission after cardiac surgery

UR - http://www.scopus.com/inward/record.url?scp=85040354530&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85040354530&partnerID=8YFLogxK

U2 - 10.1016/j.jtcvs.2017.11.071

DO - 10.1016/j.jtcvs.2017.11.071

M3 - Article

JO - Journal of Thoracic and Cardiovascular Surgery

JF - Journal of Thoracic and Cardiovascular Surgery

SN - 0022-5223

ER -