A pragmatic analysis of the association between center volume and outcomes following reoperative adult cardiac surgery

N. Cho¹, S. Mallick², S. Kim², J. Hadaya³, A. Vadlakonda⁴, Z. Tran⁵, P. Benharash^6
1David Geffen School of Medicine, Los Angeles, California ²UCLA, Los Angeles, California ³David Geffen School of Medicine at UCLA, Los Angeles, California ⁴David Geffen School of Medicine at UCLA, Saratoga, California ⁵University of California, Los Angeles, San Antonio, Texas ⁶UCLA Division of Cardiac Surgery, Los Angeles, California

Purpose: With improving surgical care and an aging population, a greater number of patients are expected to undergo reoperative cardiac surgery. However, there is a paucity of data evaluating the impact of center volume on outcomes following re-operation. We aimed to characterize clinical and financial endpoints associated with re-operative cardiac procedures.

Methods: All adult patients undergoing coronary artery bypass grafting or valve operations were identified using the 2016-2020 Nationwide Readmission Database. Diagnoses and procedures were tabulated using the International Classification of Diseases, Tenth Revision, codes. Patients undergoing ventricular assist device implantation, transplantation, or aortic operations were excluded. Major adverse event (MAE) was defined as a composite of in-hospital mortality and cardiovascular (cardiac arrest, myocardial infarction), respiratory (acute respiratory failure, pneumonia), thromboembolic (pulmonary embolism) and infectious (sepsis, surgical site infection) complications. A logistic regression to model hospital adverse events was developed, with the inclusion of annual cardiac surgery volume as restricted cubic splines. The volume (N=535) corresponding to the inflection point of the spline was used to designate institutions as low- (LVH) and high-volume hospitals (HVH). Multivariable-adjusted regression models were subsequently developed to examine the impact of HVH status on MAE, duration of hospitalization (LOS), hospitalization costs, 90-day readmission and non-home discharge.

Results: Of an estimated 1,765,372 patients undergoing cardiac procedures, 245,030 (13.9%) had a prior history of cardiac operation. The incidence of cardiac re-operation decreased over the study period (nptrend < 0.001). Relative to others, patients with a history of cardiac surgery were older (72 [63-79] vs 69 [60-76] years), more commonly male (74.3 vs 66.8%) and more frequently insured by Medicare (68.8 vs 61.8%) (All P< 0.001). While 28.8% of the centers were defined as HVH, 42.7% of patients undergoing cardiac procedures were managed at these centers. Among those undergoing re-operation, patients admitted to HVH less commonly had coronary artery bypass grafting (27.4 vs 65.9%, P< 0.001) as compared to those managed at LVH. On spline analysis, HVH demonstrated a significant decrease in the risk-adjusted rate of MAE compared to LVH (Figure). After adjustment, management at HVH was associated with lower odds of MAE (Adjusted Odds Ratio [AOR] 0.48, 95% Confidence Interval [95%CI] 0.46-0.51). Additionally, management at HVH was associated with lower incremental LOS by 3.7 days (95%CI 3.6-3.8 days) and reduced index hospitalization cost by $8,000 (95%CI $7,400-8,700). Patients undergoing re-operation at HVH had a lower risk of non-home discharge (AOR 0.38, 95%CI 0.36-0.40) and 90-day readmission (AOR 0.85, 95%CI 0.82-0.89).

Conclusion: A substantial proportion of patients undergoing cardiac cases appear to be reoperative. We noted significantly lower rates of perioperative complications at HVH performing redo cardiac operations with a threshold of ~500 annual cases. When feasible, preferential referral of such cases to high volume centers should be considered.

Identify the source of the funding for this research project: None