Author + information
- ↵⁎Reprint requests and correspondence:
Dr. John S. Rumsfeld, Cardiology (111B), VA Eastern Colorado Health Care System, 1055 Clermont Street, Denver, Colorado 80220
Because hospital readmissions are costly and potentially avoidable, reducing hospital readmission rates has been touted as a means to improve quality and reduce costs. Cardiovascular disease is a natural focus for this because of overall costs of treatment and frequency of readmission (e.g., nearly 1 in 4 Medicare patients is readmitted within 30 days after hospitalization for heart failure) (1). Since 2009, hospital-level 30-day risk-standardized readmission rates for acute myocardial infarction and heart failure have been publicly reported for Medicare patients (1). Starting in 2013, those outcomes measures will be linked to Medicare reimbursement. That has prompted an intense flurry of interest in reducing hospital readmission rates across all sectors of the U.S. healthcare system. Yet, hospital readmission is a complex, multifactorial outcome measure, and it remains unclear how hospital readmission rates can effectively and safely be lowered.
In this issue of JACC: Cardiovascular Interventions, Hannan et al. (2) provide important clarity about readmission after coronary artery bypass graft (CABG) surgery. Studying factors associated with readmission among >30,000 CABG surgery patients in New York State, the authors made several important findings. First, the most common cause of readmission within 30 days after CABG surgery was procedural complications. Post-operative infection and other complications together accounted for >25% of the readmissions. The next most common reason for readmission was heart failure, accounting for nearly 13%. Second, there was a 2.5-fold range of 30-day hospital readmission rates among hospitals. Finally, consistent with other literature, there was low correlation between 30-day readmission and 30-day mortality rates.
Why are these findings important? On the one hand, they reinforce that a principal focus on patient characteristics as a way of targeting interventions is unlikely to significantly lower hospital readmission rates. Patient characteristics may be associated with readmission, but they explain only a small proportion of the variance in the outcome and tend to be nonmodifiable (e.g., age). Risk models of hospital readmission have generally been shown to have poor discrimination, limiting their clinical utility (3).
On the other hand, finding that the most common cause of readmission is complications, and that there is a wide range of readmission rates across hospitals, argues strongly that the answer lies in improving institutional quality of care. Complication rates for CABG surgery are long-standing, validated quality measures, and the variability in readmission rates between institutions is a strong argument that some of these readmissions are preventable (4). At least some of the observed variation is a reflection of variation in quality of care. A number of interventions are known to reduce post-operative infections, arrhythmias, pneumonias, myocardial infarction, and venous thromboembolism—all of which were identified as common causes of readmission in New York State's post-CABG population.
Interestingly, recent data on cardiac surgery outcomes in the United Kingdom bolster the implications of the findings of the study by Hannan et al. (2). It has recently been shown that outcomes among cardiac surgery patients in the United Kingdom are better than in other parts of Europe; these results have been attributed to a commitment to quality measurement and improvement (5). Similarly, rehospitalization rates after cardiac surgery are significantly lower in the United Kingdom than in the United States. Although length of stay is similar to that in the United States, >75% of cardiac surgery patients in the United Kingdom receive cardiac rehabilitation after discharge, far outpacing the United States. It is not a stretch to postulate that more robust post-discharge care helps in earlier identification and treatment of complications that evolve after discharge, leading to fewer rehospitalizations.
The United Kingdom story thereby reinforces the message that a commitment to inpatient quality measures and improved transitions of care from the inpatient to the outpatient setting can translate into improved outcomes. Prior studies have identified gaps in transitions of care, such as a lack of early follow-up after heart failure hospitalization or medication errors, as factors associated with unnecessary readmission (6). In addition, interventions focused on improving transitions of care, such as transition coaches, have been shown to reduce readmission rates (7). This body of knowledge helps justify efforts, such as the American College of Cardiology/Institute for Healthcare Improvement Hospital to Home Quality Initiative, that are focused on care transitions in the effort to reduce unnecessary rehospitalization among cardiovascular patients. The current study (2) supports these efforts, suggesting that a strong focus on reducing periprocedural complications—in the hospital and in the transition period after discharge—should be a point of emphasis for such initiatives.
The frequency of readmission for heart failure after hospital discharge for CABG surgery in the study by Hannan et al. (2) also “fits” this picture. The finding may be somewhat surprising, given that successful revascularization will have been accomplished in most cases. Some may argue that heart failure after CABG surgery is another complication; others may assert it is endemic to chronic heart failure present in many CABG surgery patients. Either way, heart failure management in the post-procedure setting should be a modifiable target for quality improvement, and this again supports the need for optimal transitions of care from hospital to home.
Finally, the study by Hannan et al. (2) reinforces the difference between mortality and rehospitalization measures. While 30-day mortality rates have fallen in recent decades, readmission rates have remained steady. Intrahospital correlation between risk-adjusted mortality and readmission rates is poor. This lack of correlation between mortality and readmission rates has been seen among patients hospitalized for other conditions (8). The belief that outcomes measures should move in the same direction—namely, that mortality rates automatically reflect all domains of quality—is overly simple. Concordance will only be present if different outcomes measures capture largely overlapping domains of quality. There are reasons that the quality domains of 30-day mortality and readmission are distinct. For example, surgical experience with graft implantation and pre-operative selection/anesthesia are likely to largely improve survival; in contrast, transitions-of-care interventions and antiseptic sternal care are more likely to decrease readmissions. Therefore, a variety of outcomes measures are likely to be needed for each therapy to capture the range of relevant performance domains and improve the overall value of care as it matters to patients.
The real challenge is translating all of this into clinical practice. The current U.S. healthcare system still incentivizes readmission. In fact, diagnosis-related group payments have been linked to shorter length of stay with increases in readmission rates (8). The effectiveness, costs, implementation, and “ownership” of interventions that focus on the transition period between inpatient and outpatient care all remain to be adequately addressed. Nonetheless, healthcare reform promises to emphasize “episodes of care” that bridge inpatient and outpatient care. Starting in 2011, Medicare's value-based purchasing will provide differential reimbursement based on quality measures. Although the initial quality measures will be processes of care, outcomes measures including rehospitalization for cardiac conditions and procedures are set to follow in short order.
In an environment in which consumer groups, clinicians, hospitals, health systems, and payers are highly cognizant that readmission rates are being emphasized as measures of quality and will be tied to reimbursement, the study by Hannan et al. (2) provides a simple but powerful message: we must work for evidence-based reductions in unnecessary rehospitalization. As such, efforts to reduce procedural complication rates and improve transitions of care should be centric to quality improvement. If that happens, it may not only help the financial bottom line of hospitals and the country, but more important, it can help individual patients and their families.
Both authors have reported that they have no relationships to disclose.
↵⁎ Editorials published in JACC: Cardiovascular Interventions reflect the views of the authors and do not necessarily represent the views of JACC: Cardiovascular Interventions or the American College of Cardiology.
- American College of Cardiology Foundation
- Hannan E.J.,
- Zhong Ye,
- Lahey S.J.,
- et al.
- ↵National Quality Forum. National Voluntary Consensus Standards for Cardiac Surgery. 2004 National Quality Forum, Washington, DC; Available at: www.qualityforum.org. Accessed April 2, 2011.
- Bridgewater B.,
- Gummert J.,
- Kinsman R.,
- Walton P.