Interrupted time series: a comparison of spine surgery and hip replacement outcomes between a general hospital and a specialty orthopaedic center Open Access

Sun, Xin (2016)

Permanent URL:


Background: Whether or not a specialty medical center provides better patient outcomes and lower cost compared to a general hospital is controversial. In addition, few studies have focused on the specialty orthopaedic centers. The objective of our study is to evaluate if hospital length of stay (LOS) and cost were improved after the building of an orthopaedic and spine center.

Method: Emory University Orthopaedic & Spine Center (EUOSC) opened in September 2008. Medical records of patients who had spine surgery or hip replacement were collected from Emory University Hospital (EUH) during the period October 2006 to August 2008 and from EUOSC between October 2008 and September 2010. The primary outcome was average LOS and the secondary outcome was average medical cost per hospital day. Interrupted time series analysis with segmented regression model was used for outcomes of spine surgery and hip replacement, respectively. Autocorrelation of time series data was adjusted in the regression model.

Results: Average LOS of spine surgery patients sharply dropped by 0.8 day (change in level, p =.0108) immediately after the opening of EUOSC. In addition, LOS declined by 0.04 hospital day (i.e., one hour) over each month in EUOSC compared to baseline increasing trend in EUH (p=.0345). However, the LOS of patients with hip replacement was not significantly affected by EUOSC (change in level: p=.1448 and change in trend p=.1131). The mean cost per day of spine surgery patients increased by $25 (95%CI: 3.4-46.6) over each month in EUH and once surgeries were conducted at EUOSC this average increased by $1,094 (95%CI: 662.8-1,525.2). After Sept. 2008, the cost decreased by $41 every month in EUOSC compare to EUH (p=.0053). For hip replacements patients, cost had been continually increasing for the four year study period and was not impacted by EUOSC.

Conclusion: For spine surgery patients, LOS decreased immediately when treated at a specialty medical center, although the average cost were higher. For hip replacement patients, there was no evidence of a decreased LOS or increased cost impacted by the specialty orthopaedic and spine center.

Table of Contents

1 Introduction 1

1.1 Interrupted Time Series Review 3

1.2 Methods for Interrupted Time Series Analysis 5

1.3 Segmented Regression Model 7

2 Methods 9

2.1 Setting and Interventions 9

2.2 Study Sample 10

2.3 Outcomes and Demographic Variables 10

2.4 Statistical Analysis 11

3 Results. 15

3.1 Spine Surgery 15

3.1.1 Length of Stay (LOS) 16

3.1.2 Cost 17

3.2 Hip Replacement 18

3.2.1 LOS 19

3.2.2 Cost 20

4 Discussion 21

4.1 Tables and Figures 27

5 Appendix 33

5.1 Tables and Figures 33

5.2 SAS code 37

About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
Subfield / Discipline
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Partnering Agencies
Last modified

Primary PDF

Supplemental Files