You are here

Article Text

Article menu
Valvular heart disease
Original research
Hospital variation of 30-day readmission rate following transcatheter aortic valve implantation
  1. Tomo Ando1,
  2. Said Ashraf2,
  3. Toshiki Kuno3,
  4. Alexandros Briasoulis4,
  5. Hisato Takagi5,
  6. Cindy Grines6,
  7. Aaqib Malik7
  1. 1 Division of Cardiology, Kawasaki Saiwai Hospital, Kawasaki, Japan
  2. 2 Division of Interventional Cardiology, New York University Langone Medical Center, New York City, New York, USA
  3. 3 Department of Internal Medicine, Mount Sinai Beth Israel Hospital, New York City, New York, USA
  4. 4 Division of Cardiology, University of Iowa, Iowa City, Iowa, USA
  5. 5 Division of Cardiovascular Surgery, Shizuoka Medical Center, Shizuoka, Japan
  6. 6 Division of Interventional Cardiology, Northside Hospital Cardiovascular Institute, Atlanta, Georgia, USA
  7. 7 Division of Cardiology, Westchester Medical Center and New York Medical College, Valhalla, New York, USA
  1. Correspondence to Dr Tomo Ando, Cardiology, Kawasaki Saiwai Hospital, Kawasaki 212-0014, Japan; andotomo{at}hotmail.co.jp

Abstract

Objectives Thirty-day readmission rate is one of the hospital quality metrics. Outcomes of transcatheter aortic valve implantation (TAVI) have improved significantly, but it remains unclear whether hospital-level variance in 30-day readmission rate exists in the contemporary TAVI era.

Methods From the 2017 US Nationwide Readmission Database, endovascular TAVI were identified. The unadjusted 30-day readmission rate and 30-day risk-standardised readmission rate (RSRR) were calculated and we then conducted model testing to determine the relative contribution of hospital characteristics, patient-level covariates and economic status to the variation in readmission rates observed between the hospitals.

Results A total of 44 899 TAVI from 338 hospitals were identified. The range of unadjusted 30-day readmission rate and 30-day RSRR was 2.0%–33.3% and 9.4%–15.3%, respectively. Median 30-day RSRR was 11.8% and there was a significant hospital-level variation (median OR 1.22, 95% CI 1.16 to 1.32, p<0.01) and this was similar in both readmissions caused due to major cardiac and non-cardiac conditions. Patient, hospital and economic factors accounted for 9.6%, 1.9% and 3.8% of the variability in hospital readmission rate, respectively.

Conclusions There was significant hospital-level variation in 30-day RSRR following TAVI. Further measures are required to mitigate this variance in the readmission rate.

  • transcatheter aortic valve replacement
  • quality of health care
  • aortic valve stenosis

Data availability statement

No data are available.

https://doi.org/10.1136/heartjnl-2020-318583

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Transcatheter aortic valve implantation (TAVI) is an alternative to surgical aortic valve replacement for patients with severe aortic stenosis at intermediate and high surgical risk and has tremendously increased in volume over the last few years.1 2 In addition, it has become a much safer procedure with smaller calibre sheaths and accumulation of experience in both operators and centres.3 4

    With the increase in TAVI cases, TAVI has become more accessible but it remains unknown whether the same level of care is provided across the hospitals. Thirty-day readmission rates are considered as one of the qualities of care metrics in TAVI5 6 and could incur hospitals with financial penalties under the Affordable Care Act if the risk-adjusted readmission rates are higher than the national benchmarks.7 The Centers for Medicare & Medicaid Services (CMS) publicly report the hospital’s 30-day risk-standardised readmission rates (RSRR) on six specific conditions/procedures.8 Even though TAVI is less invasive compared with surgical aortic valve replacement, the cost of care and the frequency of readmission during follow-up was higher in TAVI compared with surgical aortic valve replacement.9 10 The higher readmission rate of TAVI compared with surgical aortic valve replacement is likely multifactorial but one of the reasons may be the difference in the provided quality of care represented by variance in 30-day RSRR among hospitals performing TAVI. There is a paucity of data on contemporary, national perspectives on hospital variation in 30-day RSRR following TAVI.

    Therefore, the primary aim of our study was to assess the unplanned 30-day RSRR following TAVI and whether there is the hospital-level variance of 30-day RSRR among hospitals after standardisation for hospital case mix from the US Nationwide Readmission Database (NRD) 2017.

    Methods

    Study population

    The study was performed following the Strengthening the Reporting of Observational Studies in Epidemiology guidelines (https://www.equator-network.org/reporting-guidelines/strobe/). There were no patients or public involved in the design of the study. All admissions who underwent non-transapical TAVI above the age of 50 years and those who survived to discharge were identified from the US NRD 2017 through the International Classification of Diseases, Tenth Revision, Clinical Modification/Procedure Coding System (ICD-10-CM/PCS). Online supplemental table 1 provides ICD-10-CM/PCS codes that were used for this study. Admissions with both TAVI and surgical aortic valve replacements codes were excluded. Cases performed in December each year were also excluded because it does not allow follow-up for 30-day readmission. Non-transapical TAVI was identified with ICD-10-CM/PCS of 02RF3JZ, 02RF3KZ, 02RF38Z and 02RF37Z. NRD is developed by the Agency for Healthcare Research and Quality’s Healthcare Cost and Utilisation Project (www.hcup-us.ahrq.gov/nrdoverview.jsp). Briefly, the NRD is the largest publicly available all-payer in-patient readmission database in the USA and includes all discharge records except long-term acute care facilities. National estimates (weighted) are available through a variable ‘discwt’ and the NRD contains data of 36 million weighted discharges from 27 geographically dispersed states accounting for roughly 50% of the total US population and hospitalisations. We have used the weighted data.

    Supplemental material

    Clinical end points

    The primary outcome was hospital-level variance in all-cause unplanned 30-day RSRR. Range (lowest and highest) and IQR of 30-day RSRR for all-cause as well as for major cardiac and non-cardiac causes were reported. Any readmission with the status of ‘non-elective’ was considered unplanned readmission. Major cardiac conditions for readmission were congestive heart failure, arrhythmia, hypertension, conduction disorders, ischaemic heart disease, acute myocardial infarction, bioprosthetic failure, aortic re-intervention and endocarditis. Major non-cardiac causes included gastrointestinal disorders, respiratory, genitourinary, stroke, acute kidney injury, peripheral vascular disease, bleeding, skin disorders and cancer.

    Statistical analysis and calculation of 30-day risk-standardised readmission rate

    Baseline demographics, comorbidities, hospital characteristics and the clinical outcomes mentioned above were compared between the two groups, using Pearson’s χ2 test for categorical variables and one-way analysis of variance for continuous variables. Categorical and continuous variables were reported as percentages and mean±SD, respectively.

    We used a similar procedure employed by CMS to estimate the 30-day RSRR by using a hierarchical logistic regression model with a hospital-specific intercept to account for patient clustering. We adjusted for case mix by including patient age, sex, insurance status, length of stay, 29 Elixhauser comorbidities, median household income by ZIP code, hospital procedural volume and hospital characteristics, including hospital ownership status, teaching status and bed size. This approach models data at patient and hospital levels to account for the variance in patient outcomes within and between hospitals. At the patient level, it models the log-odds of hospital readmission within 30 days of discharge using age, selected clinical covariates as identified in table 1 and a hospital-specific effect. At the hospital level, the approach models the hospital-specific effects as arising from a normal distribution. The hospital-specific effect represents the underlying risk of readmission at the hospital, after accounting for patient risk. The RSRR is calculated as the ratio of the number of ‘predicted’ readmissions to the number of ‘expected’ readmissions at a given hospital, multiplied by the national observed readmission rate. This approach is widely used and accepted.11 A p value <0.05 was considered statistically significant. We later calculated the median OR that is defined as the median value of the OR between the patient readmissions at highest risk and the patient readmissions at the lowest risk when randomly picking out two hospitals. It can be understood as the increased risk (in median) that a patient would have of readmissions if moving from one hospital to a different hospital with a higher risk. We finally identified the tertiles of hospitals that were reclassified based on readmission rates after risk standardisation.

    View this table:
    Table 1

    Baseline characteristics of TAVI by tertile of observed 30-day readmission rate of Nationwide Readmission Database in 2017

    For each TAVI performed, we then conducted model testing to determine the relative contribution of hospital characteristics, patient-level covariates and economic status to the variation in readmission rates observed between the hospitals. For that purpose, we first calculated the variance attributed to hospital-level random effects using an empty model. We also repeated these models with patient-level covariates first and then sequentially added economic status, and then hospital characteristics, each time getting a newer model. The relative decrease in variance attributed to hospital-level random effects was then calculated to determine the impact of each parameter on observed variance. There were no missing data. The variance unexplained by this process is deemed unknown. All data management and statistical analyses were performed using Stata V.16.1 (StataCorp, College Station, Texas, USA).

    Results

    Patient-level and hospital-level demographics

    Hospitals were divided into three groups (low, medium and high) based on tertiles of observed 30-day readmission rate. A total of weighted 44 899 (low, medium and high tertile, 9246, 19 224 and 16 429, respectively) TAVI admissions from 338 hospitals were identified. The admission selection flow chart is summarised in online supplemental figure 1. A significant difference was found in mean age, the percentage of the male sex, prior myocardial infarction and procedural volume but other patient-level and hospital-level characteristics were similar among low, medium and highest tertile groups (table 1). New pacemaker implantation and non-routine discharge occurred more frequently in the highest tertile group. Other complication rates were similar across different groups (table 2).

    Supplemental material

    View this table:
    Table 2

    Clinical outcomes of TAVI at index admission

    30-Day readmission rate

    The median unadjusted 30-day readmission rate in the low, medium and high tertile group was 6.1%, 10.9% and 16.3% (p<0.01), respectively. Median unadjusted readmission rate was 13.7% (IQR 9.7%–17.1%) with a range of (2.0%–33.3%). The RSRR in the low, medium and high tertile group was 10.6%, 11.6% and 13.1% (p<0.01), respectively (table 1). Distribution and range of unadjusted 30-day readmission rate and RSRR are presented in figures 1A,B and 2. The median 30-day all-cause RSRR was 11.8% (IQR 11.0–12.9) with a range of 9.4%–15.3%. The median OR was 1.22 (95% CI 1.16 to 1.32, p<0.01). This corresponds to a risk difference of 22% for 30-day readmission between two randomly selected hospitals for the same case. The median 30-day RSRR for major cardiac causes was 4.5% (IQR 4.4–4.7) with a range of 4.0%–5.3%. Similarly, the median 30-day RSRR for major non-cardiac causes was 6.3% (IQR 6.2–6.4) with a range of 6.0%–6.5% (online supplemental figure 2A, B). The median OR of major cardiac and non-cardiac readmission causes were 1.16 (95% CI 1.06 to 1.43, p<0.01) and 1.11 (1.02 to 1.63, p<0.01), respectively (table 3). Hospital procedure volume was not a significant predictor of outcomes after adjustment in the case mix (p=0.38). A total of 186 (55%) hospitals were reclassified after adjustments. In the lowest and highest tertile hospitals, 43.3% (49/113) and 60.7% (68/112) hospitals were up-classified and down-classified, respectively. In the medium tertile group, 15.0% (17/113) and 46.0% (52/113) hospitals were up-classified and down-classified, respectively (figure 3, online supplemental figure 3).

    Supplemental material

    Supplemental material

    Supplemental material

    Figure 1
    Figure 1

    Distribution of unadjusted (A) and adjusted (B) all-cause 30-day readmission rate for individual hospitals. X-axis shows readmission rates by percentage and y-axis demonstrates number of hospitals corresponding to each readmission rate.

    Figure 2
    Figure 2

    Plot of unadjusted (solid lines) and adjusted (dashed lines) 30-day readmission rate with 95% CIs for individual hospitals. X-axis shows number of hospitals included in the study and y-axis shows risk standardised readmission rates by percentage. Note that the y-axis does not start from zero.

    Figure 3
    Figure 3

    Change in classification in the hospital category for 30-day risk standardisation readmission rate after adjustments. Y-axis shows tertile in unadjusted readmission. TAVI, transcatheter aortic valve implantation.

    View this table:
    Table 3

    Thirty-day RSRR for all-cause, major cardiac and major non-cardiac readmission rate

    Causes of 30-day readmission rate variance

    Admission-level (comorbidities, age and gender), hospital (hospital size, teaching or non-teaching, ownership, geographical location) and economic factors (ie, income level) accounted for 9.6%, 1.9% and 3.8% of the variability in hospital readmission rate, respectively. The majority of the variability was not explained by these factors.

    Discussion

    From this 2017 nationwide claims readmission database, the major findings were as follows: (i) there was a significant hospital-level variation in 30-day RSRR following TAVI performed in 2017 in the USA; (ii) more than half of the hospitals were reclassified when readmission rate was adjusted for admission-level and hospital-level characteristics; (iii) only a minority of hospital variance in 30-day RSRR was explained by the admission and hospital factors.

    The 30-day readmission rate is an important metric as it is considered a quality measure for hospitals. The presence of hospital-level variation of 30-day RSRR is clinically important as it raises a target for quality improvement. The presence of hospital-level variation in 30-day RSRR has been reported for several conditions such as congestive heart failure, myocardial infarction, pneumonia and peripheral arterial disease,11 12 providing areas of improvement in these conditions. There are currently limited reports on hospital variation in 30-day RSRR post-TAVI. Ten hospitals in Canada did not demonstrate significant hospital-level difference in 30-day readmission rate post-TAVI and differed from our results.13 Conversely, there was a significant difference of 41% (95% CI 37% to 44%) in 30-day readmission risk between two random hospitals for the same post-TAVI from the Medicare fee-for-service beneficiaries aged over 65 years between 2011 and 2013 in the USA.14 Our study has unique differences compared with these studies. We included a higher number and diverse hospital characteristics (ownership, bed size, procedure volume and teaching status) and all-payers across the USA. Another important distinction of our study is that we only included non-transapical TAVI that were performed in 2017 in the USA, reflecting more contemporary practice of TAVI (ie, new-generation TAVI devices, transfemoral first approach). Our study found a significant variation in 30-day RSRR post-TAVI and suggests that despite improvement in TAVI procedural practice and postprocedural care over the last several years, there still exists significant variation in these metrics across hospitals. This may not be that surprising as even centres participating in the major randomised clinical trials demonstrated significant variability in TAVI outcomes.15 However, it is reassuring to observe the narrowing of variance compared with the previous study, which included TAVI from the early period from 2011 to 2013 (22% vs 41%). This is likely as a result of continuous improvement in perioperative complications related to readmission.2 An additional contribution of our study is that we also demonstrated that there was hospital variation in readmission rate for both major cardiac and non-cardiac causes. Our study does not provide information on the causes of readmission that significantly accounted for the variation in 30-day RSRR. However, considering the gap in evidence, this proposed metric is important. Previous national data showed that readmission rates were 39.3% and 61.7% for cardiac and non-cardiac causes, respectively. Furthermore, respiratory diseases were the most common causes of non-cardiac readmission and heart failure was the most common cause of cardiac readmission following TAVI.16 Our study suggested that large proportion of source of variance in 30-day RSRR remains unknown but implementation of disease-specific measures to reduce common causes of hospital readmissions post-TAVI (heart failure and respiratory disease)17 18 as well as other general measures to reduce hospital readmission (patient education, medication reconciliation, follow-up telephone calls, postdischarge home visits, etc)19 should be recommended to hospitals with higher 30-day RSRR compared with the national benchmark.

    Hospitals that were categorised as medium or high for observed 30-day RSRR had higher percentage of high volume hospitals. A previous study demonstrated that there was an inverse relationship between hospital TAVI volume and the 30-day readmission rate.16 Although we did not specifically examine the TAVI volume and readmission rate of the hospitals, we have included this important variable to calculate the 30-day RSRR. Expansion of TAVI into lower risk patients, wide use of newer-generation device and standardisation of implantation technique may have lowered effect on volume outcome relationship in TAVI. This requires future study.

    Roughly half of the hospitals were reclassified into different tertile categories after adjustments. Our results are in agreement with previous studies that a fair percentage of hospitals were re-categorised into different class after adjustments.20 21 This highlights the importance of using RSRR for benchmarks of hospital performance instead of the unadjusted readmission rates. Another important finding from our study was that admission and hospital factors accounted only for a minority of hospital-level variance of 30-day RSRR post-TAVI. Admission characteristics accounted for the largest whereas hospital characteristics contributed the smallest portion for hospital variation, which is in agreement with the findings of a previous study.22 We speculate that one of the reasons for this is that 30-day readmission is a heterogeneous outcome because the causes of readmission are diverse. Both the US and Canadian consensus statement for TAVI includes 30-day readmission rate as one of the quality indicators for TAVI,5 6 but our results suggest that a large portion of hospital-level variance is of unclear source aside from patient or hospital characteristics. Therefore, it may not be a suitable quality measure index and other direct outcomes such as in-hospital mortality or stroke may be preferred.

    Our study has several major limitations. First, there is currently no reliable predicted model to calculate the 30-day readmission rate post-TAVI using the ICD-10-CM/PCS codes and therefore, the calculation of RSRR may differ in a different database. Second, the NRD is an administrative database where comorbidities, procedures and outcomes are identified through ICD-10-CM/PCS. The codes are not 100% accurate and subject to miscoding. However, the NRD is a well-developed database and one of only a few databases that can be used to examine TAVI outcomes at a national level. In addition, the codes are entered through specially trained coordinators. Third, because of the nature of our study, our results contain all the biases related to observational studies and unmeasured confounders. In addition, cases performed in December had to be excluded as 30-day readmission rate could not be calculated and this may have introduced selection bias. Fourth, we only included TAVIs that were performed through non-transapical access as this is the standard approach for TAVI in the contemporary era. Fifth, the case-mix adjustment method used by CMS for estimation of 30-day RSRRs for hospital uses 12 months of claims data before the admission, which is not available in this dataset. However, our work is consistent with prior work as mentioned previously. Sixth, NRD includes hospitalizations from community hospitals and lacks all the Veterans Affairs health systems. Similarly, readmissions from different states other than index admissions are not captured as it is a compilation of state inpatient databases, which might have led to a small underestimation of readmission rates. Lastly, the NRD does not include granular data related to TAVI such as details of access site, echocardiographic data and valve type or size.

    In conclusion, there was a significant hospital-level variation in 30-day RSRR following TAVI. Further measures are required to mitigate this variance in the readmission rate.

    Key messages

    What is already known on this subject?

    • Thirty-day readmission rate is one of the quality measure metrics of transcatheter aortic valve implantation (TAVI).

    What might this study add?

    • There was significant hospital-level variation in risk-standardised 30-day readmission rate in contemporary TAVI in the USA.

    • Patient-level and hospital-level characteristics only accounted for minority of the source in variation.

    How might this impact on clinical practice?

    • Further study is required to identify the source of variation and decrease the hospital variation in 30-day readmission rate.

    Supplemental material

    Data availability statement

    No data are available.

    Ethics statements

    Patient consent for publication

    Ethics approval

    The study was considered exempted from approval from the Institutional Board of Review as the NRD is publicly available data and cases are appropriately de-identified.

    References

    1. ↵
      2. Nishimura RA ,
      3. Otto CM ,
      4. Bonow RO , et al
      . 2017 AHA/ACC Focused Update of the 2014 AHA/ACC Guideline for the Management of Patients With Valvular Heart Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol 2017;70:252–89.doi:10.1016/j.jacc.2017.03.011 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28315732
      OpenUrlFREE Full Text
    2. ↵
      2. Akintoye E ,
      3. Ando T ,
      4. Sandio A , et al
      . Aortic valve replacement for severe aortic stenosis before and during the era of transcatheter aortic valve implantation. Am J Cardiol 2020;126:73–81.doi:10.1016/j.amjcard.2020.03.038 pmid:http://www.ncbi.nlm.nih.gov/pubmed/32336533
      OpenUrlCrossRefPubMed
    3. ↵
      2. Ando T ,
      3. Villablanca PA ,
      4. Takagi H , et al
      . Meta-Analysis of Hospital-Volume Relationship in Transcatheter Aortic Valve Implantation. Heart Lung Circ 2020;29:31519–7.doi:10.1016/j.hlc.2019.10.016 pmid:http://www.ncbi.nlm.nih.gov/pubmed/32089491
      OpenUrlPubMed
    4. ↵
      2. Russo MJ ,
      3. McCabe JM ,
      4. Thourani VH , et al
      . Case volume and outcomes after TAVR with Balloon-Expandable prostheses: insights from TVT registry. J Am Coll Cardiol 2019;73:427–40.doi:10.1016/j.jacc.2018.11.031 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30704575
      OpenUrlFREE Full Text
    5. ↵
      2. Asgar AW ,
      3. Ouzounian M ,
      4. Adams C , et al
      . 2019 Canadian cardiovascular Society position statement for transcatheter aortic valve implantation. Can J Cardiol 2019;35:1437–48.doi:10.1016/j.cjca.2019.08.011 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31679616
      OpenUrlCrossRefPubMed
    6. ↵
      2. Bavaria JE ,
      3. Tommaso CL ,
      4. Brindis RG , et al
      . 2018 AATS/ACC/SCAI/STS expert consensus systems of care document: operator and institutional recommendations and requirements for transcatheter aortic valve replacement: a joint report of the American association for thoracic surgery, American College of cardiology, Society for cardiovascular angiography and interventions, and society of thoracic surgeons. J Am Coll Cardiol 2019;73:340–74.doi:10.1016/j.jacc.2018.07.002 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30031107
      OpenUrlFREE Full Text
    7. ↵
      Patient protection and Affordable care act, public law No. 111-148 2010.
    8. ↵
    9. ↵
      2. Brescia AA ,
      3. Syrjamaki JD ,
      4. Regenbogen SE , et al
      . Transcatheter versus surgical aortic valve replacement episode payments and relationship to case volume. Ann Thorac Surg 2018;106:1735–41.doi:10.1016/j.athoracsur.2018.07.017 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30179625
      OpenUrlCrossRefPubMed
    10. ↵
      2. Goldsweig AM ,
      3. Tak HJ ,
      4. Chen L-W , et al
      . Relative costs of surgical and transcatheter aortic valve replacement and medical therapy. Circ Cardiovasc Interv 2020;13:e008681. doi:10.1161/CIRCINTERVENTIONS.119.008681 pmid:http://www.ncbi.nlm.nih.gov/pubmed/32406261
      OpenUrlPubMed
    11. ↵
      2. Sukul D ,
      3. Sinha SS ,
      4. Ryan AM , et al
      . Patterns of readmissions for three common conditions among younger us adults. Am J Med 2017;130:1220.e1–1220.e16.doi:10.1016/j.amjmed.2017.05.025 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28606799
      OpenUrlCrossRefPubMed
    12. ↵
      2. Secemsky EA ,
      3. Schermerhorn M ,
      4. Carroll BJ , et al
      . Readmissions after revascularization procedures for peripheral arterial disease: a nationwide cohort study. Ann Intern Med 2018;168:93–9.doi:10.7326/M17-1058 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29204656
      OpenUrlCrossRefPubMed
    13. ↵
      2. Elbaz-Greener G ,
      3. Qiu F ,
      4. Webb JG , et al
      . Profiling Hospital performance on the basis of readmission after transcatheter aortic valve replacement in Ontario, Canada. J Am Heart Assoc 2019;8:e012355. doi:10.1161/JAHA.119.012355 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31165666
      OpenUrlPubMed
    14. ↵
      2. Murugiah K ,
      3. Wang Y ,
      4. Desai NR , et al
      . Hospital variation in outcomes for transcatheter aortic valve replacement among Medicare beneficiaries, 2011 to 2013. J Am Coll Cardiol 2015;66:2678–9.doi:10.1016/j.jacc.2015.10.008 pmid:http://www.ncbi.nlm.nih.gov/pubmed/26670069
      OpenUrlFREE Full Text
    15. ↵
      2. Greason KL ,
      3. Blackstone EH ,
      4. Rajeswaran J , et al
      . Inter- and intrasite variability of mortality and stroke for sites performing both surgical and transcatheter aortic valve replacement for aortic valve stenosis in intermediate-risk patients. J Thorac Cardiovasc Surg 2020;159:1233–44.doi:10.1016/j.jtcvs.2019.04.112 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31350027
      OpenUrlCrossRefPubMed
    16. ↵
      2. Khera S ,
      3. Kolte D ,
      4. Gupta T , et al
      . Association between hospital volume and 30-day readmissions following transcatheter aortic valve replacement. JAMA Cardiol 2017;2:732–41.doi:10.1001/jamacardio.2017.1630 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28494061
      OpenUrlPubMed
    17. ↵
      2. Hale TM ,
      3. Jethwani K ,
      4. Kandola MS , et al
      . A remote medication monitoring system for chronic heart failure patients to reduce readmissions: a Two-Arm randomized pilot study. J Med Internet Res 2016;18:e91. doi:10.2196/jmir.5256 pmid:http://www.ncbi.nlm.nih.gov/pubmed/27154462
      OpenUrlCrossRefPubMed
    18. ↵
      2. Keshishian A ,
      3. Xie L ,
      4. Dembek C , et al
      . Reduction in hospital readmission rates among Medicare beneficiaries with chronic obstructive pulmonary disease: a real-world outcomes study of nebulized bronchodilators. Clin Ther 2019;41:2283–96.doi:10.1016/j.clinthera.2019.09.001 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31630815
      OpenUrlCrossRefPubMed
    19. ↵
      2. Hansen LO ,
      3. Young RS ,
      4. Hinami K , et al
      . Interventions to reduce 30-day rehospitalization: a systematic review. Ann Intern Med 2011;155:520–8.doi:10.7326/0003-4819-155-8-201110180-00008 pmid:http://www.ncbi.nlm.nih.gov/pubmed/22007045
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      2. Hawkins BM ,
      3. Kennedy KF ,
      4. Aronow HD , et al
      . Hospital variation in carotid stenting outcomes. JACC Cardiovasc Interv 2015;8:858–63.doi:10.1016/j.jcin.2015.01.026 pmid:http://www.ncbi.nlm.nih.gov/pubmed/25999111
      OpenUrlAbstract/FREE Full Text
    21. ↵
      2. Pandey A ,
      3. Keshvani N ,
      4. Vaughan-Sarrazin MS , et al
      . Evaluation of risk-adjusted home time after acute myocardial infarction as a novel hospital-level performance metric for Medicare beneficiaries. Circulation 2020;142:29–39.doi:10.1161/CIRCULATIONAHA.119.044765 pmid:http://www.ncbi.nlm.nih.gov/pubmed/32408764
      OpenUrlCrossRefPubMed
    22. ↵
      2. Singh S ,
      3. Lin Y-L ,
      4. Kuo Y-F , et al
      . Variation in the risk of readmission among hospitals: the relative contribution of patient, hospital and inpatient provider characteristics. J Gen Intern Med 2014;29:572–8.doi:10.1007/s11606-013-2723-7 pmid:http://www.ncbi.nlm.nih.gov/pubmed/24307260
      OpenUrlPubMed

    Supplementary materials

    Footnotes

    • Contributors Writing manuscript, design of the study and responsible for the overall content of the manuscript: TA. Writing manuscript and design of the study: SA. Design of the study and responsible for the overall content of the manuscript: TK. Writing manuscript, design of the study and responsible for the overall content of the manuscript: HB. Design of the study and writing manuscript: HT. Responsible for the overall content of the manuscript: CG. Writing manuscript, design of the study, performed statistical analysis and responsible for the overall content of the manuscript: AM.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests None declared.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.