Author + information
- Avnish Tripathi1,
- Michael P. Flaherty1,
- Jinnette D. Abbot2,
- Gregg. C. Fonarow3,
- Abdur R. Khan1,
- Arti Saraswat4,
- Dhaval Kolte2,
- Srujal Patel1,
- Ajay J. Kirtane5 and
- Deepak L. Bhatt6
- 1Division of Cardiovascular Disease, University of Louisville School of Medicine, Louisville, KY
- 2Division of Cardiology, Warren Alpert Medical School, Brown University, Providence, RI
- 3Division of Cardiology, David Geffen School of Medicine, UCLA, Los Angeles, CA
- 4University of Louisville, Louisville, KY
- 5Division of Cardiology, Columbia University Medical Center, New York, NY
- 6Brigham and Women’s Hospital Heart & Vascular Center, Harvard Medical School, Boston, MA
In Medicare population, transcatheter aortic valve replacement (TAVR) was associated with higher cost and lower short-term readmissions compared to surgical aortic valve replacement (SAVR). However, these differences have not been evaluated in a national sample, including all payer sources.
We used Healthcare Cost and Utilization Project's National Readmission Database to identify TAVR and SAVR cases who survived index hospitalization from January through November, 2013. Weighted national estimates of 30-day readmission and associated cost were calculated. Propensity score matching was used to 1:1 match 2,448 TAVR patients with SAVR patients on discharge weights, demographics and comorbidities including Charlson's comorbidity score. Hierarchal multivariable regression models were used to examine adjusted differences in the matched groups.
In the US, an estimated 12,196 TAVR and 68,578 SAVR procedures were performed among patients with mean age of 82 (SE:0.21) and 69 (SE: 0.20) years respectively. Compared to SAVR, TAVR was associated with higher 30-day readmission rate (18.8% vs. 15.8%; P <0.001); higher 30-day mortality (1.26% vs. 0.58; P <0.001); and higher cumulative mean cost over 30-day follow up ($61,216 vs. $56,832; P <0.001).
TAVR cohort had higher rates of readmission due to heart failure (21% vs. 14%; P <0.001), septicemia/pneumonia (11% vs. 8%; P = 0.003), bleeding complications (6% vs. 4%; P = 0.014) and cerebrovascular events (4% vs. 2%; P = 0.004); whereas, SAVR cohort had higher rate of surgical or implant related complications (17% vs. 12%; P <0.001), and cardiac arrhythmia (13% vs. 7%; P <0.001) including atrial fibrillation (8% vs. 3%; P <0.001).
In multivariable analyses, compared to SAVR, TAVR was associated with 14.2% higher mean cost of index hospitalization (β: 0.142; 95% CI 0.123-0.161; P <0.001); and 13.5% higher 30-day cumulative cost (including cost for readmissions) (β: 0.135; 90% CI 0.115-0.155). However, no differences were found in likelihood of 30-day readmission (HR: 1.01; 95% CI 0.933-1.089; P = 0.863) and cost associated with re-hospitalization within 30 days (β: -0.009; 95% CI -0.137-0.118; P = 0.885).
In a national sample of patients, significant differences were observed in clinical outcomes and cost between SAVR and TAVR. Unlike some previous reports, risk of short-term readmission was not significantly different after accounting for individual and clinical characteristics.