Thirty day readmission rates are a growing clinical area of scrutiny in a changing healthcare climate. Early hospital readmissions give insight into opportunities to implement changes in care during the index hospitalization and early after implantation to avoid utilization of resources that could potentially be spared. Hospitals and insurance carriers are increasingly viewing 30 day readmissions as avoidable hospitalizations. We sought to evaluate the national rate and cause of 30 day readmission in newly implanted ventricular assist devices (VAD) by analyzing one of the largest publicly available all-payer inpatient databases.
Heart failure is the leading cause of hospitalization among patients over the age of 65 years and it accounts for nearly 30 billion dollars in healthcare expenditures, with both numbers predicted to continue climbing over the next decade.1 A growing number of these patients are now being treated with VAD, which are used to mechanically assist and support circulation in patients with advanced heart failure refractory to medical therapies. Although device-related complications and overall survival have improved with the development of newer generation VADs,2 the device is still not entirely without complications, especially early postimplantation. Ventricular assist devices have also had an impact on healthcare expenditures. The average cost of left ventricular assist device (LVAD) implantation is around $175,420 and average readmission costs increased from $12,377 before LVAD to $19,465 after LVAD. The total lifetime costs of patients with LVADs were estimated to be around $726,200, with a significant portion resulting from hospital readmissions.3 While there have been single-center studies evaluating 30 day readmissions after VAD implantation,4–6 we wanted to examine multicenter national rates and etiologies of 30 day readmissions after index hospitalization for implant.
Materials and Methods
We used one of the largest publicly available all-payer inpatient databases: The Nationwide Readmissions Database (NRD), a subset of the Healthcare Cost and Utilization Project (HCUP) sponsored by the Agency for Healthcare Research and Quality (AHRQ), and evaluated data from 2014, the last full year of available data at time of publication. Data were derived from 22 geographically dispersed states and 15 million (weighted 35 million) discharges per year. The data account for 51.2% of the total US resident population and 49.3% of all US hospitalizations. We utilized Implantable Cardioverter Defibrillator (ICD) 9 code 37.66 (insertion of implantable heart assist system) to identify index hospital admissions for VAD therapy. Subjects who were <18 years old, those with missing data (age, sex, or mortality), and/or new primary December admissions (lack of 30 day follow-up data) were excluded. Comorbidities were identified by analytical code “CM_” variables provided by NRD and ICD-9 codes. Primary outcome was 30 day readmission and secondary outcomes were the leading etiologies of readmission. Logistic regression was utilized to assess for statistical significance.
We studied 1,481 patients who underwent implantable heat assist system placement during primary admission. Mean age was 56.6 years (63.8% >55 years of age), 78% were male, length of stay (LOS) averaged 39.1 ± 1.6 days, and 57% had Medicare/Medicaid as the primary payer. Of the studied patients, 1,315 survived to hospital discharge during primary admission (mortality rate of 11.2%), and 60.6% were discharged to a nursing facility (Table 1). One hundred thirty-one patients (10%) were readmitted within 30 days of their primary hospitalization. The leading etiologies for 30 day readmission were bleeding (24%), the majority of which were from gastrointestinal source (90.6% of patients with bleeding), heart failure (18%), device complications (14%), and coagulopathy (8%; Table 2). The mean LOS during readmission was 13.8 days with a mortality rate of 2.1%. The overall trend of 30 day readmissions was notable for 50% of patients being readmitted between days 22 and 30 and nearly 80% of readmissions occurred after day 15 (Figure 1). Predictors of 30 day readmission, including sex, age, socioeconomic class, discharge location, and presence of comorbidities such as hypertension, diabetes mellitus, and kidney disease were analyzed and were not statistically significant.
While mechanical circulatory assist device therapy has been shown to improve quality of life and survival in the appropriate patient population with advanced heart failure, its use has been associated with important complications, including thrombosis, cerebrovascular accidents, bleeding (most commonly gastrointestinal bleeding), acute and chronic right ventricular heart failure, and device-related infections.7Table 2 shows bleeding as the leading cause for 30 day readmission for patients receiving VAD therapy in this study. As observed in prior studies,8 the majority of these bleeding complications were gastrointestinal in etiology, likely as a result of anticoagulation in the setting of acquired von Willebrand disease which is known to develop after VAD implantation.9 Based on our findings which show that bleeding is the most frequent etiology of readmission after device implant, particular attention to gastrointestinal risk factors/medications and vigilant observation and titration of anticoagulation may help reduce 30 day readmissions, and is an important area of ongoing study in large multicenter databases.
In our study, heart failure accounted for the second highest number of 30 day readmissions. As shown in Figure 1, the majority of patients readmitted in the first 30 days after discharge presented in the latter half of that time period, suggesting a gradual onset of heart failure. Based on this finding, one could consider implementing early LVAD speed optimization studies with aggressive diuretic and heart failure medication titrations to prevent early hospital readmissions. Identification of risk factors for the development of right ventricular failure (RVF) postoperatively is also paramount in the prevention of early hospitalizations. A RVF risk score has been developed to help identify patients who may benefit from early right ventricular assist device (RVAD) implantation or augmented medical therapy.10
Recent studies examining single-center cohorts of patients receiving LVADs identified device-related infections as the leading cause for early readmission.4,5 Interestingly, in our analysis, device complications, which includes device-related infections, were found to be the third leading cause of readmission in the studied population. The difference in frequency among the studies may be related to fewer VAD driveline infections secondary to standardization of driveline care and judicious antibiotic choices in the postoperative period.11 With the development of new VAD technology, including reduced shear stress forces on blood and with a programmed artificial pulse, there is hope that the rates of device-related complications including hemolysis and pump thrombosis will continue to decline.12
Rehospitalization after LVAD implant occurs in approximately 58% of patients in the first 6 months after implant as seen in Figure 2,13 and prior studies have shown that early readmission was associated with increased mortality.6 However, these studies have generally evaluated smaller, single-center cohorts. A separate study assessed the cost-effectiveness of LVAD therapy, which demonstrated an incremental cost-effectiveness ratio of $209,400 per quality-adjusted life-year in patients and also predicted a greater than 50% drop if hospital readmission rates could be reduced by half.3 This is likely because healthcare costs associated with readmissions in patients with VADs appear to be more complicated, and therefore more costly, than hospitalizations in patients without devices. This is corroborated by the fact that patients with LVADs have higher medical costs yet, according to our study, lower hospital readmission rates than those previously reported for patients with heart failure on medical therapy alone.14 Our evaluation of the NRD allows for important socioeconomic and cost analyses which are not currently publicly available in the INTERMACS registry including median household income, primary payer, and discharge disposition, which are of epidemiologic importance in a complex health care environment, and important areas of future research.
There are some limitations to our study which should be considered, including the fact that the use of ICD-9 codes does not fully identify the type of device implanted, which would be an important area for future evaluation. We also did not identify the reason for primary hospitalization, which may be of interest as it may identify additional risk factors for future complications, nor take into account complex comorbidities which could factor into the index hospitalization including bleeding episodes, need for RV support, etc. It is also important to recognize that in our evaluation, device-related complications included many factors including pump thrombosis and infections, although this inclusive rate of device complications was surprising lower than prior studies have shown. Our evaluation of index hospitalization patient characteristics as seen in Table 1 revealed no statistically significant predictors of rehospitalization frequency or etiology.
A surprising trend in our evaluation was the very high percentage of discharges to a nursing facility after index hospitalization. We identified 60% of patients being discharged to nursing facilities, which is higher than previously published reports which showed rates of 29% and 38% being discharged to postacute care facilities.15,16 Our higher rate may be related to the fact that the study by Dunlay et al.16 was limited to evaluation in privately insured patients. It may also be because they reviewed patients from 2006 to 2013, and rates may be increasing as postacute care facilities are becoming more comfortable carding for VAD patients with time and education. The study by Lander et al.15 evaluated a smaller number of patients and was limited to one center, which may account for the difference in our percentages. It is also important to note that the NRD only evaluates approximately half of all US hospitalizations, and may not capture the full geographic trend in use of these facilities. The use of postacute care facilities and their influence on outcomes and readmissions will be an important area of continued investigation.
We have identified gastrointestinal bleeding, heart failure, and device complications as the leading etiologies of 30 day readmission post-VAD implantation in one of the largest available inpatient databases. By understanding the etiologies and trends of these hospitalizations, we hope to offer insight into continued strategies to prevent early hospital readmission, significantly decrease cost of care after VAD implantation, and further improve the quality of life of patients.
1. Heidenreich PA, Albert NM, Allen LA, et al. Forecasting the impact of heart failure
in the United States: A policy statement from the American Heart Association. Circ Heart Fail 2013.6: 606–619.
2. Slaughter MS, Rogers JG, Milano CA, et al.; HeartMate II Investigators: Advanced heart failure
treated with continuous-flow left ventricular assist device
. N Engl J Med 2009.361: 2241–2251.
3. Baras Shreibati J, Goldhaber-Fiebert JD, Banerjee D, Owens DK, Hlatky MA. Cost-effectiveness of left ventricular assist devices in ambulatory patients with advanced heart failure
. JACC Heart Fail 2017.5: 110–119.
4. Da Silva M, MacIver J, Rodger M, et al. Readmissions following implantation of a continuous-flow left ventricular assist device
. J Card Surg 2016.31: 361–364.
5. Kimura M, Nawata K, Kinoshita O, et al. Readmissions after continuous flow left ventricular assist device
implantation. J Artif Organs 2017.20: 311–317.
6. Gupta S, Roy S, Cogswell R, Thenappan T, Liao K, John R. Readmission within 30 days after left ventricular assist device
implantation is associated with increased long-term mortality. J Heart Lung Transplant 2016.35: S259–S260.
7. Potapov EV, Stepanenko A, Krabatsch T, Hetzer R. Managing long-term complications of left ventricular assist device
therapy. Curr Opin Cardiol 2011.26: 237–244.
8. Demirozu ZT, Radovancevic R, Hochman LF, et al. Arteriovenous malformation and gastrointestinal bleeding in patients with the HeartMate II left ventricular assist device
. J Heart Lung Transplant 2011.30: 849–853.
9. Nascimbene A, Neelamegham S, Frazier OH, Moake JL, Dong JF. Acquired von Willebrand syndrome associated with left ventricular assist device
. Blood 2016.127: 3133–3141.
10. Drakos SG, Janicki L, Horne BD, et al. Risk factors predictive of right ventricular failure after left ventricular assist device
implantation. Am J Cardiol 2010.105: 1030–1035.
11. Cagliostro B, Levin AP, Fried J, et al. Continuous-flow left ventricular assist devices and usefulness of a standardized strategy to reduce drive-line infections. J Heart Lung Transplant 2016.35: 108–114.
12. Mehra MR, Naka Y, Uriel N, et al. A fully magnetically levitated circulatory pump for advanced heart failure
. N Engl J Med 2017.376: 440–450.
13. Kirklin JK, Naftel DC, Pagani FD, et al. Seventh INTERMACS annual report: 15,000 patients and counting. J Heart Lung Transplant 2015.34: 1495–1504.
14. Dharmarajan K, Hsieh AF, Lin Z, et al. Diagnoses and timing of 30-day readmissions after hospitalization for heart failure
, acute myocardial infarction, or pneumonia. JAMA 2013.309: 355–363.
15. Lander BS, Patel K, Blackstone EH, Nordseth T, Starling RC, Gorodeski EZ. Post-acute care trajectories in the first year following hospital discharge after left ventricular assist device
implantation. J Am Med Dir Assoc 2016.17: 908–912.
16. Dunlay SM, Haas LR, Herrin J, et al. Use of post-acute care services and readmissions after left ventricular assist device
implantation in privately insured patients. J Card Fail 2015.21: 816–823.