The introduction of transcatheter aortic valve replacement (TAVR) has transformed the treatment landscape of aortic stenosis (AS), broadening the range of patients suitable for aortic valve replacement. Empirical evidence strongly backs the utilization of TAVR in a variety of clinical scenarios, with large analyses reporting minimal complication rates and decreased hospital readmissions post-TAVR [1, 2, 3, 4, 5, 6]. To further improve rates of morbidities and readmissions after cardiac procedures, home health care (HHC) has been proposed as a form of home-based care aimed to better facilitate patients’ postoperative transition to a more bolstering home environment [7, 8]. However, evidence regarding the associated benefits of HHC remains unclear. Broadly, HHC has been applied in the setting of mitral valve repair and heart failure optimization, with mainly sobering results of higher association with hospital readmissions in the short term [9, 10, 11]. Previous reports have shown similar findings in the post-procedural period following TAVR, with higher rates of heart failure-specific and all-cause hospital readmissions [12]. More comprehensive evidence regarding the utility of HHC following TAVR is lacking. This study aims to compare post-TAVR outcomes in patients receiving HHC compared to routine discharges, while also identifying predictors and clinical outcomes associated with post-TAVR hospital readmission.

In this study, we present a propensity-matched analysis of the postoperative clinical outcomes in the Nationwide Readmissions Database TAVR patients who underwent discharge via HHC vs routine to home. After 1:1 nearest-neighbor PSM, HHC TAVR patients experienced significantly higher rates of early mortality compared to routine discharge patients (0.47% vs. 0.20%, respectively). Moreover, rates of 30-day readmission, postoperative MI, arrhythmia, acute kidney injury, stroke/TIA, and sepsis were also significantly higher in HHC patients, while incidence of heart failure was higher in routine discharge patients. On multivariate logistic regression, HHC conferred a significantly increased risk of 30-day readmission following TAVR (HR: 1.34; 95% CI: 1.30–1.38). Though these findings should be bolstered by additional studies with longer clinical follow-up, the data here suggest that in TAVR patients, HHC is significantly associated with increased short-term morbidities and early mortality when compared to routine discharge patients.

Albeit limited in comparison to surgical aortic valve replacement, early complications following TAVR have been well-documented by recent randomized control trials and observational studies. Several major complications have been consistently identified in the periprocedural period following TAVR, namely moderate/severe paravalvular leakage (PVL), stroke, major vascular and bleeding complications, conduction abnormalities necessitating permanent pacemaker placement, and acute kidney injury [13, 14, 15, 16, 17, 18, 19]. Further expounding upon this evidence, studies have shown early moderate-to-severe PVL and major bleeding to be specifically associated with increased 1-year postprocedural mortality and decreased quality of life as measured by Kansas City Cardiomyopathy Questionnaire score [20]. Strategies have therefore recently been devised to limit such acute complications following TAVR, with institutions demonstrating the efficacy of various prophylactic and measurement devices such as stroke management protocols and aortic regurgitation indices for valvular dysfunction quantification [14, 21]. Nonetheless, as TAVR valves continue to evolve in morphology and function, similar to what has been seen with recent developments in cerebral embolic protection and anti-leak perivalvular occlusion, further reductions in initial postprocedural morbidities are expected [22, 23].

Comprised of home health assistants, nursing, and rehabilitation services, HHC is a system of care, similar to skilled nursing facilities and inpatient rehabilitation centers, that provides patients with multidisciplinary surveillance following a visit to an acute care institution [24]. HHC serves as an attractive paradigm to reduce the rate of early morbidities in the post-hospital setting as various analyses have shown HHC to be associated with advantageous outcomes such as lower 1-year mortality and readmission rates when compared to incomplete HHC [25, 26]. However, patients being managed for cardiac disease pathologies do not seem to experience similar benefits of HHC. HHC has been attributed to increased 30-day readmission rates following both HF admissions and transcatheter mitral valve repair, likely due to these patients having worse overall health status preceding TAVR [9, 10, 11]. This trend is consistent amongst TAVR patients as well, with a recent study using the NRD by Nazir et al. [12] showing that even after adjusting for covariates amongst TAVR patients, HHC utilization was correlated with longer median hospital length of stay, higher rates of 30-day all-cause readmissions, and 30-day heart failure readmissions compared to those who did not receive HHC. Outside of these few analyses, there is a general paucity of data comparing HHC vs routine discharge patients in a post-TAVR setting.

Here, we show that in propensity-matched TAVR patients, HHC is associated with a significantly higher risk of early mortality and 30-day hospital readmission compared to the routine discharge paradigm. Though many factors may explain the poorer outcomes associated with HHC in the studied TAVR cohort, it is plausible that the increased presence of trained healthcare professionals in the post-hospital setting may result in early detection of postprocedural complications, thereby resulting in increased hospital readmissions. Additionally, HHC programs may vary in structure on an institutional basis, such that inconsistencies in follow-up, care practices, and protocolization may affect post-procedural outcomes. Lastly, HHC utilization may also be associated with lower patient health literacy and socioeconomic status, both of which were unable to be accounted for in PSM, and are individual predictors for worsened outcomes after TAVR [27].

It is important to note that while differences in a variety of outcomes between HHC and routine discharge groups were found to be statistically significant, absolute effect sizes of these respective outcomes varied. For example, in the propensity-matched discharge groups, early mortality differed by 0.1% and early readmissions by 4.1%, with both outcomes found to be significantly higher in TAVR patients discharged to HHC. Nonetheless, we contend that even a 4.1% increase in readmissions after HHC is a clinically relevant outcome, especially given that typical 30-day all-cause readmission after TAVR occurs in roughly 15–17% of patients, suggesting nearly a 25% relative increase in readmissions [28]. Moreover, readmissions after TAVR have been shown to confer increased mortality risk, effectively augmenting the risk profile of the index procedure [27]. While the results of this retrospective analysis do not call for the direct rejection of HHC-guided postprocedural patient management, further comparative studies are required to justify the recommendation of such discharge practices in TAVR patients.

This study has several important limitations. There are inherent limitations in using observational data from large databases. Despite the nearest-neighbor propensity matching used here, this study displays a considerable degree of heterogeneity. Not only is the patient population diverse, but clinical decision-making and techniques can vary significantly between participating institutions. Patients referred for HHC after TAVR likely had a higher comorbidity burden than regularly discharged patients, a difference that could not be adequately adjusted for with PSM. Due to this inherent selection bias, the associations observed in this analysis may be derived from patient comorbidity burden rather than the properties of each respective care modality. Specifically, patient frailty metrics could not be included in our propensity score model due to the lack of these data in the NRD. Future studies of large clinical datasets integrating functional status may prove invaluable in better matching patients receiving HHC compared to regular discharge, further elucidating differences in outcomes between practices. Additionally, given that TAVR involves multidisciplinary efforts, the lack of granularity in such broad data repositories often fails to capture nuances in detailed variables that could explain differences in outcomes between HHC and routine discharge patients.

Additionally, large database studies with large sample sizes can detect small differences in data. Although statistically significant, these differences may not be clinically relevant. Therefore, to address such limitations, the results of large database studies should be further explored in tandem with single-institution analyses. Lastly, the NRD includes patients from 30 states in the United States. While comprehensive, the generalizability of this study’s findings may be limited as HHC practices may vary on a state-by-state or international basis.

After TAVR, the use of HHC services was correlated with higher in-hospital morbidities and an increased risk of 30-day readmission. Enhancing discharge procedures prior to HHC and revising HHC eligibility criteria could improve overall outcomes. For instance, frailty assessments, such as the 6-minute walk test, can be used as metrics to potentially predict non-home and HHC outcomes. Moreover, standardized implementation of such HHC eligibility criteria in the preoperative setting can allow for faster and more targeted use of adjunct rehabilitative services that benefit HHC TAVR patients with comorbidities likely driving higher hospital readmissions.

Data may be made available upon reasonable request to corresponding author.

EA, DA, IS contributed to study conception and design. EA, DA, YW contributed to data collection and analysis. SY, JAB, DSG, DK, CT, DW and AM contributed to the analysis and interpretation of data for the work. EA, DA, SY, JAB, YW, DSG, DK, CT, DW, AM, IS contributed to manuscript drafting and revision. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

The study was carried out in accordance with the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of the University of Pittsburgh approved on 4/17/2019 (STUDY18120143). Informed consent not applicable as data was obtained through the NRD.

This research received no external funding.

Dr. Sultan receives institutional research support from Abbott, Atricure, Artivion, WL Gore, Edwards, Medtronic, and Terumo Aortic. None of these are related to this manuscript. The remaining authors have no competing interests to declare.

Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.31083/HSF48321.