30 Day Readmission Rates After Cardiac Surgery, a Single-Center QI Study

Abstract

Research Question: Are there significant demographic differences between patients aged 18-90 who were readmitted within 30 days after discharge from cardiac surgery, and are there any specific systemic changes that can be made to potentially decrease rates of readmission? We predict that follow up with a provider within 30 days of discharge will have a significant negative correlation with readmission.

Background, Significance, and Rationale for the Question: Readmission after cardiac surgery is common across the nation and associated with significant morbidity and healthcare-associated cost to the patient and the hospital. We explored the potential risk factors associated with increased likelihood of readmission in a single-center review with the hope of modifying current hospital practices to decrease readmission rates in the future.

Materials and Methods: Using administrative data, we identified patients readmitted to the same institution within 30 days of cardiac surgery over a 36-month period (n = 61). Time-matched patients meeting the same inclusion criteria were the control group (n = 487). Cardiac surgery included valve replacement surgery, aortic surgery, coronary artery bypass grafting (CABG), or any combination of these procedures. We performed a standardized review of readmitted patients’ medical records to evaluate timing and potential risk factors for readmission including comorbidities, emergent status, type of procedure, and whether they had followed up with a physician prior to readmission. We evaluated timing of readmission by procedure and tested for univariate associations between characteristics of readmitted patients and non-readmitted patients in our clinical registry. Patient demographics and perioperative comorbidities were evaluated by univariate analyses. Logistic regression analysis identified independent risk factors for readmission.

Results: Following 548 hospitalizations for cardiac surgery, 61 patients were readmitted to the index hospital within 30 days for a readmission rate of 11.13%. Median time to readmission was 11.5 days. There was no significant difference in readmission rates between different types of surgery (p = 0.26) or emergent status (p = 0.402). However, follow up with a physician after discharge was negatively correlated with readmission (p = <0.001). There were no statistically significant effects of comorbidities on readmission status in this population. Discussion/

Conclusions: Using univariate analysis and logistic regression, there were no significant differences between the readmission and nonreadmission groups except for follow up status (p <0.001). This confirms our hypothesis of early follow up being the major predictor of readmission within 30 days.