Reducing Hospital Readmissions

A Major Opportunity for Rapid Savings

One of the best ways for communities to reduce healthcare costs quickly and improve patient care in the process is to implement initiatives to reduce hospital readmissions. Research studies and quality-reporting initiatives around the country show that 15-25% of people who are discharged from the hospital will be readmitted to the hospital within 30 days or less, and that many of these readmissions are preventable. The patients certainly would not mind having fewer hospitalizations, and billions of dollars in spending on hospital stays could be saved if these hospitalizations could be avoided. In other words, reducing readmissions is a win-win for both cost and quality, without a hint of rationing. Moroever, savings can be achieved rapidly, since the principal focus is on a short-term outcome: readmissions within 30 days.

Although there are many efforts around the country today to reduce readmissions, most have one or more of three key weaknesses:

• They aren't data driven, either in understanding where readmission problems exist or in determining how well interventions are working;
• They're narrowly focused on hospital-based changes or "transitions," which don't naturally extend to avoiding readmissions after 30 days or to helping chronic disease patients avoid being hospitalized in the first place; and
• They aren't supported by payment reforms that give physicians, hospitals, and other providers the flexibility they need to redesign care in order to reduce readmissions as well as accountability for succeeding.

One of the best examples of how to reduce readmissions through data-driven changes across the full continuum of care is a project CHQPR Executive Director Harold Miller designed and coordinated for the Pittsburgh Regional Health Initiative, a multi-stakeholder Regional Health Improvement Collaborative that has been a national leader in healthcare quality improvement for over a decade. The PRHI project to reduce readmissions achieved dramatic results quickly with only a small investment of upfront resources. You can watch a presentation describing this project here. Many of the lessons below are drawn from PRHI's work.

Step 1: Analyze Data to Identify Types and Causes of Readmissions

It is critical to first determine which types of readmissions to focus on, since the causes of the readmissions and the mechanisms for preventing them can differ dramatically. Preventable readmissions can be divided into three broad categories:

• Readmissions for complications or infections arising directly from the initial hospital stay, e.g., if a surgery patient develops a surgical site infection or other complication and has to return after discharge.
• Readmissions because of poorly managed transitions during discharge, e.g., if a patient or a caregiver does not receive clear instructions from the hospital about the types of medications to take or what to do or not do during recuperation.
• Readmissions because of a recurrence of a chronic condition that led to the initial hospitalization, e.g., an exacerbation of asthma, congestive heart failure, or chronic obstructive pulmonary disease.

Readmissions in the first category and many readmissions in the second category can be viewed as primarily the responsibility of the hospital, and most of these readmissions occur quickly; typically within 15-30 days. But data analyses show that the largest volume of readmissions occurs among patients with chronic disease. Many of these patients are being admitted and readmitted not because of something that does or does not happen to them in the hospital, but because they are not receiving good primary care support in the community, and that is not likely to change simply by changing the way hospital care is delivered or penalizing hospitals for these readmissions. Moreover, a 30-day window of time is arbitrary; more than half of the readmissions among chronic disease patients occur after 30 days, so efforts focused solely on transitions after discharge from the hospital may only address a subset of the reasons for readmission, and projects which focus on transitions cannot be readily extended to preventing the initial hospitalizations before they occur.

Two examples of the kinds of analyses to guide readmission reduction efforts that can be done using multi-payer claims databases are illustrated in the Readmission Briefs issued by the Pittsburgh Regional Health Initiative:

• Analysis of Readmissions for Six Chronic Diseases (asthma, chronic obstructive pulmonary disease, diabetes, depression, heart attack, and heart failure)

• Patterns of Hospital Admission and Readmission Among HIV-Positive Patients in Southwestern Pennsylvania

These analyses showed, for example, that:

• Heart Failure and COPD alone account for 9% of all medical admissions, and have some of the highest readmission rates (23%-26%) of any type of patient;
• Among patients with chronic diseases who are readmitted to the hospital, fewer than half of the readmissions occur within the first 30 days of discharge. Over half of patients with COPD and heart failure are readmitted to the hospital within 12 months.
• The majority of patients with COPD and heart failure are readmitted to the hospital for the same diagnosis or a closely-related condition, suggesting that better management of the patient's condition could prevent both the initial hospitalization as well as readmissions.

Step 2: Reinvent Care Delivery Across the Continuum

The majority of readmission reduction initiatives across the country have focused narrowly on improving the hospital discharge process or the transition from hospital to the home or other post-acute care setting. However, if the largest categories of readmissions are chronic disease patients and their readmissions occur over a multi-month period, then purely hospital-based initiatives or a focus on the time immediately around the hospital discharge may well fail to address the true root causes of the readmissions. Moreover, hospital-based readmission reduction initiatives cannot easily be expanded to try and prevent initial chronic disease admissions (since if the patient never goes to the hospital at all, they could not access services that presume a hospitalization has occurred!). It is likely that to achieve the maximum reduction on hospital readmissions and initial admissions, there will need to be a strong focus on improving primary care/outpatient management of patients with chronic disease.

As described in detail in its Readmission Reduction Guide, the Pittsburgh Regional Health Initiative (PRHI) sought to reinvent care for chronic disease patients across the entire continuum of care, both during a patient's inpatient stay and also in primary care practices. Because of its emphasis on enhancing primary care rather than merely improving hospital discharges and transitions, the approach PRHI used can be used to reduce the number of initial hospitalizations for patients with chronic diseases as well as readmissions.

In addition to the PRHI Readmission Reduction Guide, PRHI developed a video illustrating the work done at one pilot site in Pittsburgh which was able to reduce readmission rates by 44% in the first year of implementing changes designed using the techniques described in the Readmission Reduction Guide.

Step 3: Create Payment Systems That Make Care Changes Sustainable

Many projects across the country and around the world have shown that the rate of hospital readmissions and initial admissions can be reduced dramatically through very simple and low-cost changes in the way care is delivered to patients. The major reason that these projects have not been replicated broadly is the barriers posed by current healthcare payment systems. Under Medicare, Medicaid, and most commercial health plans, primary care practices cannot be reimbursed for providing care management services to patients (or even for answering a patient's phone call when they are experiencing health problems that could lead to a hospitalization). In addition, under current payment systems, hospitals lose a significant amount of revenue when readmissions are reduced, even though they still have to cover the fixed costs of having beds and staff available for those patients who do need to be hospitalized.

The approach being taken by Medicare and many private health plans is to reduce or eliminate payments to hospitals for readmissions. While this may create a greater "incentive" for hospitals to reduce readmissions, it assumes readmissions can be prevented by the hospital, and it does nothing to provide more resources to primary care practices to better manage the care of chronic disease patients so they do not need to be hospitalized or readmitted as frequently.

Is there a better way to pay to support and encourage reductions in readmissions? Yes:

• Hospitals could be paid using episode-of-care payments, with a built-in limited "warranty" for treating readmissions due to infections, complications, etc. that are caused by the hospital. This would reward them for reducing readmissions, rather than penalizing them for doing so.

• Primary care physicians could be paid (through their physician groups or health systems) using comprehensive care payments. This would give the physicians flexibility to provide better care management services to their patients, but also give them accountability for the rates at which their patients are hospitalized and rehospitalized for chronic disease exacerbations and other conditions that are best managed outside of the hospital.

CHQPR has developed a presentation showing how different payment approaches support or fail to appropriately support efforts to reduce readmissions.