Author + information
- Viraj Bhise, MBBS, MPH,
- Pushkar Kanade, MD, MPH,
- Ghanshyam P.S. Shantha, MD, MPH,
- Prakash Balan, MD,
- Tom C. Nguyen, MD,
- Pranav Loyalka, MD,
- Biswajit Kar, MD,
- Anthony Estrera, MD,
- Richard W. Smalling, MD, PhD and
- Abhijeet Dhoble, MD, MPH∗ ()
- ↵∗Department of Cardiology, University of Texas Health Science Center, 6431 Fannin Street, MSB 1.224, Houston, Texas, 77030
Data on comparative outcomes and readmissions after transcatheter aortic valve replacement (TAVR) and surgical aortic valve replacement (SAVR) in patients with end-stage renal disease (ESRD) are scarce because these patients were excluded from major TAVR trials (1). We conducted retrospective analyses using the National Inpatient Sample (NIS) and the National Readmission Database (NRD), both of which provide de-identified information regarding hospitalizations in U.S. hospitals (2,3). In addition, the NRD also contains reliable verified patient linkage numbers to track patient admissions across hospitals within a state and identify 30-day readmissions (3). We followed recommendations from the Agency for Healthcare Research and Quality to use the NIS and the NRD for our analysis (2,3). We used the International Classification of Diseases-Ninth Revision, Clinical Modification (ICD-9-CM) to identify 2 cohorts of patients with ESRD (ICD-9-CM code 585.6) who underwent TAVR (ICD-9-CM codes 35.05 and 35.06) or SAVR (ICD-9 CM codes 35.21 and 35.22).
Primary outcomes of interest were in-hospital mortality, in-hospital complications (permanent pacemaker implantation and blood transfusion), post-procedure length of stay, discharge disposition, cost of hospitalization, and 30-day readmission rates. To ensure that the groups were comparable, we excluded patients younger than 50 years of age from our analysis. We also excluded hospitalizations with missing data, those with a same day transfer to home, to hospice, or to another hospital. We calculated the Charlson Comorbidity Index with Deyo Modification using the ICD-9 CM diagnoses available for every patient. We then used logistic regression based on demographic and risk factors to generate a propensity score for undergoing TAVR or SAVR for each patient, and 1:2 matches were identified by a match algorithm using the nearest available Mahalanobis metric (4). Outcomes were evaluated between the 2 propensity-matched groups (TAVR and SAVR) using the bootstrapped average treatment effect for the matched data, after ensuring that the balancing properties were satisfied. We also used patient- and hospital-level weights provided in both the NIS and NRD datasets to generate national estimates.
Over a 2-year period (2012 to 2013) in the NIS, a total of 704 patients with ESRD had an aortic valve replacement procedure; 158 underwent TAVR, and 546 underwent SAVR. The 2 cohorts were similar with respect to most comorbid conditions; however, patients who underwent TAVR were older (76.26 years vs. 67.08 years; p < 0.001). Of these, 93 (13.21%) died during the same hospitalization. The 2 propensity-matched groups consisted of 119 patients with TAVR and 244 patients with SAVR. The groups did not have the exact 1:2 ratios due to a tied Mahalanobis metric between some observations. We did not find significant differences between the in-hospital mortality and blood transfusion rates within the TAVR and SAVR groups (Table 1). However, post-procedure length of stay (8.3 days vs. 17.7 days; p < 0.001) and hospitalization costs ($66,672 vs. $99,676; p < 0.001) were lower in TAVR patients. The patients who underwent TAVR were also more likely to be discharged home compared with those who underwent SAVR (56.3 vs. 42.8%; p = 0.01). In a subgroup analysis, patients who underwent transfemoral TAVR had the lowest in-hospital mortality, post-procedure length of stay, need for blood transfusion, and cost of hospitalization, compared with transapical TAVR and SAVR.
Over the 11-month period in the NRD, a total of 446 patients with ESRD had undergone an aortic valve replacement; 118 had TAVR (73 transfemoral and 26 transapical), and 347 had SAVR. Of these, 104 (23.32%) patients had a readmission within a 30-day period after discharge. No differences between the 30-day readmission rates were observed after TAVR and SAVR. The median times to readmission, length of stay, and in-hospital mortality during the 30-day readmission were also similar for the patient cohorts.
Our findings suggested that the patients with ESRD are at higher risk of mortality (13.21%) and complications after aortic valve replacement procedures. Almost 1 of 4 patients were readmitted within 30 days. Despite selecting older ESRD patients, TAVR resulted in similar mortality and 30-day readmission rates when compared to SAVR, but had a shorter length of stay and lower hospitalization costs. The likelihood of discharge to home was higher after TAVR. Transfemoral TAVR appeared to be the safest and the most effective treatment for these patients, sustaining the least mortality.
Please note: Dr. Nguyen is a consultant for Edwards Lifesciences. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2017 American College of Cardiology Foundation
- Ohno Y.,
- Attizzani G.F.,
- Barbanti M.,
- et al.
- Houchens R.L.,
- Ross D.N.,
- Elixhauser A.,
- Jiang J.
- Healthcare Cost and Utilization Project
- Leuven E.,
- Sianesi B.