Evidence of the linkage between hospital-based care coordination strategies and excess 30-day readmission ratios

Objective: In the new era of value-based payment models and pay for performance, hospitals are in search of the silver bullet strategy or bundle of strategies capable of reducing their 30-day readmission rates. Our objective was to determine if hospitals that self-report consistent use of care coordination strategies have less Medicare readmissions.

Data: We merged hospital level data from the 2015 AHA Annual Survey, 2015 AHA Survey of Care Systems and Payment Survey, 2015 and 2016 CMS Hospital Compare, FY 2018 CMS Hospital Readmission Reduction Program (HRRP), and 2015 Area Health Resource File.

Methods: We operationalized excess 30-day readmission ratios from the HRRP as binary measures comparing hospitals with ratios of 1 or less to hospitals with ratios greater than 1 and used multivariate logistic regression. We ran separate models for each of the 6 condition-specific ratios: chronic obstructive pulmonary disease (COPD), heart failure, pneumonia, myocardial infarction, coronary artery bypass grafting (CABG), and total knee or hip arthroscopy (TKA/THA). To evaluate the association between care coordination strategies (e.g., medication reconciliation, follow-up calls, visit summaries) and readmissions more globally, we also ran models with performance on the HCAHPS (patient satisfaction survey) care transition composite measure and overall hospital (star) ratings as an outcome variable, comparing hospitals rated 4 or 5 to hospitals rated 3 or less. We ran separate models for each of the 12 strategies included in the AHA survey. We also ran models with the number of strategies implemented and a corresponding quadratic term to determine the benefit of adding additional strategies and if there are diminishing returns.

Results: Of the 3305 HRRP hospitals, 1115 hospitals participated in the AHA 2015 Survey of Care Systems and Payment Survey. 696 hospitals answered all twelve care coordination survey questions and were linked with the CMS HRRP program. The hospitals responding to all 12 questions were more likely to be large, academic, and non-profit, less likely to be in rural or high-poverty communities, and less likely to have excess CABG, COPD, and heart failure readmissions.

We found that increasing the number of strategies is associated with lower excess readmissions for heart failure (AOR=1.29, se=0.14, p<0.05), COPD (AOR=1.28, se=0.14, p<0.05), pneumonia (AOR=1.34, se=0.16, p<0.05), AMI (AOR=1.24, se=0.13, p<0.05), higher care transition satisfaction (AOR=1.57, se=0.18, p<0.001), and higher overall star ratings (AOR=1.43, se=0.18, p<0.01). However, we found that odds of success had a slight downward trend for AMI readmissions (AOR=0.98, se=0.0, p<0.05), care transition satisfaction (AOR=0.97, se=0.18, p<0.001), o. We found that hospitals with consistent medication reconciliation, visit summaries, follow-up calls, disease management programs, transitional care programs, and case manager follow-up have significantly higher odds of top star-ratings. None of the strategies were positively associated with lower readmissions for TKA/THA or CABG.

Conclusions: Based on this preliminary analysis, it appears that care coordination strategies typically used by hospitals may be effective at reducing readmissions for patients initially hospitalized for medical conditions, but less effective for surgical admissions. There also appears to be diminishing returns when implementing more than seven strategies.