National Trends in Surgical Readmissions IN THE ERA OF THE HOSPITAL - - PowerPoint PPT Presentation

National Trends in Surgical Readmissions

IN THE ERA OF THE HOSPITAL READMISSIONS REDUCTION PROGRAM

Winta T. Mehtsun MD MPH, Irene Papanicolas PhD, Jie Zheng PhD, E. John Orav PhD, Keith

• A. Lillemoe MD, Ashish K. Jha MD MPH

No Relevant Financial Disclosures

Reducing Readmissions is a Priority

• .
• However, readmissions are also a concern for surgical conditions.
• 1 in 7 surgical patients readmitted within 30 days.
• Readmissions have been at the center of policy efforts.
• Hospital Readmissions Reduction Program (HRRP)

Hospital Readmissions Reduction Program

• Part of the Affordable Care Act
• Passed in March 2010
• Financial penalties for higher-than-expected readmission rates
• Targeted Conditions
• AMI, heart failure, pneumonia (FY 2013)
• TKR/THR, COPD (FY 2015)
• CABG (FY 2017)
• Penalty size
• FY 2013: 1%
• FY 2015 to present: 3%

Medical Readmissions in the HRRP Era

Zuckerman et. al NEJM

Is HRRP Impacting Surgical Readmissions?

• Hypothesis 1:
• We may see reduced surgical readmission rates if hospitals implement

system-wide readmission reduction strategies.

• Hypothesis 2:
• We may not see reduced surgical readmission rates because the factors

that drive surgical readmissions (i.e. complications) are distinct from medical readmissions.

Research Questions

1. Have surgical readmission rates decreased over the past decade, especially in the HRRP era? 2. Were reductions, if any, more pronounced among surgical conditions that were targeted for future penalties ? 3. Did hospitals with the largest reductions in medical readmissions also have the largest reductions in surgical readmissions?

Study Cohort

• Medicare (FFS) beneficiaries 2005-2014
• 8 inpatient procedures
• Targeted:
• Coronary Artery Bypass Graft (CABG)
• Total hip replacement (THR)
• Total knee replacement (TKR)
• Non-targeted:
• Pulmonary Lobectomy
• Abdominal Aortic Aneurysm Repair (AAA)
• Colectomy
• Cholecystectomy
• Appendectomy

Primary Outcome

• 30-day all-cause readmission following

index procedure.

• Risk-adjusted:
• Sex
• Age
• Race
• Procedure type
• Elixihauser comorbidity
• DRG weight

Analysis

• Focused on three time periods:
• Time 1: Pre HRRP

Q1 2005 to Q1 2010

• Time 2: HRRP implementation

Q2 2010 to Q2 2012

• Time 3: Penalty phase

Q3 2012 to Q4 2014

• Unit analysis: Patient
• Linear interrupted time series approach

• 1. Have surgical readmission rates decreased?

Pre – ACA HRRP Implementation HRRP Penalty Period

12.1% 11.3% 10.7% 9.9% 8.7%

0% 2% 4% 6% 8% 10% 12% 14% 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

• 2. Were readmission reductions more pronounced

among surgical conditions targeted in the future?

HRRP Implementation Pre – HRRP

11.3% 9.8% 8.8% 8.4% 8.3% 7.1% 0% 2% 4% 6% 8% 10% 12%

Risk Adjusted Readmission Rate

HRRP Penalty Period Slope: -0.103% per quarter Slope: -0.129% Slope: -0.111%

Difference-in-Trends Penalty – Implementation: 0.02% (p=0.15) Difference-in-Trends: Implementation – Pre-ACA: 0.03%(p=0.02)

HRRP Implementation Pre – HRRP

14.0% 13.8% 13.7% 13.2% 12.7% 12.0% 0% 2% 4% 6% 8% 10% 12% 14% 16%

Risk Adjusted Readmission Rate

HRRP Penalty Period Slope: 0.023% per quarter Slope: -0.121% Slope: -0.090%

Difference-in-Trends Penalty – Implementation: 0.03% (p=0.18) Difference-in-Trends: Implementation – Pre-ACA:

• 0.14% (p<0.001)

• 3. Did hospitals with the largest reductions in medical

readmissions also have the largest decreases in surgical readmissions?

Medical Readmissions Performance Quartile Change in Medical Readmissions (annual slope) All Surgical Procedures Targeted Non- Targeted Top Performers

• 1.05%

2nd Quartile

• 0.46%

3rd Quartile

• 0.19%

Worst Performers 0.50%

• 0.57%*
• 0.37%
• 0.24%
• 0.03%*
• 0.70%*
• 0.43%
• 0.28%
• 0.07%*
• 0.34%*
• 0.25%
• 0.16%
• 0.02%*

_Trend across categories of medical readmission reductions * P < 0.05

Discussion

• Surgical readmissions have decreased.
• Largest reductions seen among non-targeted conditions.
• Anticipation vs. robust local quality improvements program
• Hospitals with largest drop in medical readmissions also have

largest drop in surgical readmissions.

Limitations

• Imperfect risk adjustment with claims data.
• No control conditions.
• Unmeasured confounders – i.e. change in surgical practice.

Policy Implications

Narrowly targeted program focused on a few conditions can have broad spillovers.

Special thanks to the Jha Lab

(where immigrant work ethic is welcomed)

…and PCORT for funding this work