Pitfalls of Calculating Hospital Readmission Rates Based Solely on Nonvalidated Administrative DatasetsKeywords: surgery, recurrent disease, survey, cost, hospitalInteractive Manuscript
Ask Questions of this Manuscript:
What is the background behind your study?
Administrative databases are increasingly being used to establish benchmarks for quality of care and to compare performance across peer hospitals. As proposals for Accountable Care Organization are being developed, readmission rates will be increasingly scrutinized.
What is the purpose of your study?
The purpose of this study is to assess the accuracy of administrative datasets and identify independent predictors of readmission.
Describe your patient group.
Data for 5,854 consecutive spine admissions to UCSF Medical Center from July 2007- June 2011 was abstracted from the University Health-System Consortium (UHC) using the Clinical Database/Resource Manager. Of these admissions, 320 cases (5.8%) were rehospitalized within 30 days of the initial discharge date.
Describe what you did.
We performed an independent chart review to determine reasons for readmission and extracted hospital administrative data to calculate total and direct costs. Logistic regression analysis was used to test the odds of readmission on categorical variables. The two-sample t-test was used to test the difference of total and direct cost between readmission and non-readmission.
Describe your main findings.
The main reasons for readmission were infection (46.1%); planned, staged surgery (11.6%); and nonoperative management (9.8%). The UHC database overestimated the readmission rate. Based on our chart review, 50 cases (of the 320 total readmissions) were misclassified. Thirty-seven cases (11.6%) were planned, staged procedures and 13 cases (4.1%) were unrelated to the initial admission. When planned, staged readmission cases are excluded, the total cost of readmission is reduced by 18.2% (p=0.005). The cost variance is in excess of one million dollars.
Describe the main limitation of this study.
This is a retrospective study.
Describe your main conclusion.
Predictors of readmission were admission status (pOur findings uncover the potential pitfalls of calculating hospital readmission rates based solely on nonvalidated administrative datasets.
Describe the importance of your findings and how they can be used by others.
Benchmarking algorithms for defining a hospital’s readmission rate must take into account planned, staged surgery and eliminate unrelated reasons for readmission. Current tools overestimate the true readmission rate and cost.
Administrative databases are increasingly being used to establish benchmarks for quality of care and to compare performance across peer hospitals. As proposals for Accountable Care Organization are being developed, readmission rates will be increasingly scrutinized.
The purpose of this study is to assess the accuracy of administrative datasets and identify independent predictors of readmission.
Data for 5,854 consecutive spine admissions to UCSF Medical Center from July 2007- June 2011 was abstracted from the University Health-System Consortium (UHC) using the Clinical Database/Resource Manager. Of these admissions, 320 cases (5.8%) were rehospitalized within 30 days of the initial discharge date.
We performed an independent chart review to determine reasons for readmission and extracted hospital administrative data to calculate total and direct costs. Logistic regression analysis was used to test the odds of readmission on categorical variables. The two-sample t-test was used to test the difference of total and direct cost between readmission and non-readmission.
The main reasons for readmission were infection (46.1%); planned, staged surgery (11.6%); and nonoperative management (9.8%). The UHC database overestimated the readmission rate. Based on our chart review, 50 cases (of the 320 total readmissions) were misclassified. Thirty-seven cases (11.6%) were planned, staged procedures and 13 cases (4.1%) were unrelated to the initial admission. When planned, staged readmission cases are excluded, the total cost of readmission is reduced by 18.2% (p=0.005). The cost variance is in excess of one million dollars.
This is a retrospective study.
Predictors of readmission were admission status (pOur findings uncover the potential pitfalls of calculating hospital readmission rates based solely on nonvalidated administrative datasets.
Benchmarking algorithms for defining a hospital’s readmission rate must take into account planned, staged surgery and eliminate unrelated reasons for readmission. Current tools overestimate the true readmission rate and cost.
Project Roles:
