Improving hospital discharge
By Christopher Guadagnino, Ph.D.
As physicians, hospitals and payors continue to ramp up their efforts to reduce preventable adverse medical events, relatively little attention has been paid to a moment in the care process when the patient is particularly vulnerable: the hospital discharge.
Nearly 18 percent of Medicare patients are readmitted to a hospital within 30 days of discharge, and patients with multiple chronic conditions are readmitted at rates as high as 25 percent, according to Medicare Payment Advisory Commission (MedPAC) estimates, accounting for $15 billion in spending. Research on care transitions suggests that as many as 20 to 30 percent of adverse events following discharge are preventable, and another 30 percent are ameliorable – i.e., their severity could be reduced if corrective measures were instituted earlier and more effectively.
The hospital discharge is a “prototypical condition” for the patient safety movement: it is common and risky; it is nonstandardized from patient to patient and hospital to hospital; responsibility for its implementation is fragmented among many hospital staff; and adverse events occur in approximately one in five discharges that may lead to preventable hospital use, according to a June 2007 study published in the Journal of Patient Safety.
Decentralized responsibility is a central defect of the discharge process, says Judith Black, M.D., MHA, medical director of senior products at Highmark Blue Cross Blue Shield. After a total hip replacement, for example, an orthopedic surgeon writes orders, a primary care physician writes orders, and a case manager issues instructions. The current system does not adequately ensure that patients and their care partners reconcile and understand this information, that it is transmitted timely across care settings, or that patients promptly follow up with their primary care physicians, she notes.
Accreditation entities such as the Joint Commission and the National Committee for Quality Assurance monitor that communication occurs between the hospital and the receiving parties (the patient and their caregiver, for discharge to the home), but they are limited in their ability to monitor the actual effectiveness of the process. The Joint Commission looks at whether the reasons for discharge are based on the patient’s assessed needs; whether discharged patients are given information about continued care and treatment; and whether the patient and appropriate practitioners, staff and family members are involved. But feedback from, and oversight of discharge effectiveness is difficult, as many settings to which patients are discharged are not accredited (most obviously, the home) says Robert Wise, M.D., the Joint Commission’s vice president, Division of Standards and Survey Methods.
The situation is ripe for change, and emerging data suggest that interventions before, during and after discharge can reduce the number of post-discharge adverse events and prevent rehospitalizations. Pilot projects are now addressing documented failures of hospital discharge practices – including incomplete or inaccurate reconciliation of information from multiple clinicians, poor comprehension by patients and lack of appropriate follow-up care.
Those interventions include standardized discharge toolkits, and transitional care nurses using proven, patient-focused protocols who interact with patients more closely for a period after hospital discharge. Systemic and financial hurdles may stymie widespread dissemination of those interventions, although support by government, private foundations and some health insurers is giving those projects momentum and may push them into the mainstream.
A reliable theme in medical research literature is that physicians overestimate patients’ comprehension of medical instructions, including those given at discharge. Patients have difficulty remembering their care plans, even better-educated patients do not adequately understand and remember instructions, and physicians underutilize after-care providers – instead relying too heavily on patients’ information retention, says Jeffrey Greenwald, M.D., associate professor of medicine at Boston Medical Center and member of the Society of Hospital Medicine’s Quality and Patient Safety Committee.
As co-investigator in a study called Project RED (Re-engineered Hospital Discharge), which is funded by the Agency for Healthcare Research and Quality, Greenwald says that half of patients who get traditional discharge instructions can’t name key information, such as why they were in the hospital, a list of their medications, and their follow-up plans. Median age is in the 40s for patients in the study.
“This problem is rampant and underappreciated: the perception is that we do a good job at it,” says Greenwald. Contributing to the problem, he says, is the limited amount of time dedicated to the discharge process in an era of shorter hospital stays, inadequate communication with after-care providers, and inadequate systems of after-care. “The level of change required is not insubstantial, and requires multidisciplinary coordination, as well as buy-in from the ‘C-suite’ of hospital leadership,” he adds.
“Hospitals can only control what’s within their walls, and not a patient’s access to transportation for follow-up care, or their ability to cover their medication costs, or the social ties of an isolated elderly person living in a high-rise,” says Aline Holmes, RN, senior vice president of clinical affairs for the New Jersey Hospital Association.
Other recent research corroborates the problem of discharge failure. Many patients have poor comprehension of various aspects of their care, according to a study of four categories of emergency department discharge information: diagnosis and course, ED care, post-ED care, and return instructions, published online in July by the Annals of Emergency Medicine. Researchers wrote that “the chaotic nature of the environment and transient interactions pose significant challenges to communication” during the discharge process, noting that three-quarters of the patients studied demonstrated deficient comprehension in at least one of the four categories – most commonly post-ED care (medications, ancillary measures, and follow-up), while half of patients demonstrated deficient comprehension in two or more categories. Perhaps more disturbing, most patients weren’t even aware of their lack of understanding, and perceived their difficulty with comprehension only 20 percent of the time. Clinicians, therefore, cannot simply ask patients to identify their areas of confusion, the researchers noted.
Care transitions are especially important for elderly patients and other high-risk patients who have multiple comorbidities. More than half of patients over age 70 years responding to a posthospitalization telephone survey did not recall anyone talking with them about how to care for themselves after hospitalization, according to a study published last September in the Journal of Hospital Medicine.
There are significant gaps in quality and safety when patients transition from the hospital to an outpatient setting, and patient safety problems are exceedingly common in the early discharge period, writes Alan Forster, M.D., MSc, Assistant Professor at the University of Ottawa and Scientist, Clinical Epidemiology, Ottawa Health Research Institute – who Greenwald says is a seminal researcher on hospital discharge failure and improvement. Several studies performed about five years ago in the U.S. and Canada by Forster and colleagues demonstrate that approximately one in five medical patients experience an adverse event during the first several weeks after hospital discharge, while one-third of those events are associated with disability and half of them are associated with use of additional health services.
Close to two-thirds of post-discharge adverse events are preventable or ameliorable, such as a patient who develops C. difficile diarrhea complicated by severe dehydration or sepsis following discharge, Forster writes in a December 2004 commentary published in Hospital Medicine. The most prevalent type is an adverse drug event, such as a side effect that harms the patient, while other types include procedural complications or hospital-acquired infections that become clinically apparent only after patients go home – a phenomenon exacerbated by hospitals shortening patient length of stays, according to Forster. Diagnostic and therapeutic errors account for approximately 10 percent of post-discharge adverse events, a frequency that Forster believes may be underestimated, as patients in post-discharge research studies are generally followed for one month, at most, which may be too short a follow-up duration to identify poor outcomes related to such errors.
Deficiencies of the discharge process itself lead to adverse events, writes Forster. Patients who are unable to remember a discussion with their care provider about the side effects of their medication are at a three-fold greater risk of experiencing an adverse event than patients who can recall such information. Patients see multiple providers before, during and after a hospital encounter who are often practicing in different locations. Discharge summaries often lack important information describing the most responsible diagnosis, the results of important tests, the medications prescribed at discharge, or the follow-up plans; and are often not handed off in a timely manner, if at all, to multiple follow-up physicians. There is a lack of infrastructure to adequately monitor patients’ conditions or test results after they get home, Forster writes, and patients often have trouble getting in contact with physicians who cared for them while hospitalized to discuss new symptoms, side effects of medications, or follow-up that is not proceeding as planned.
Medical Home Within the Home
Physicians are witnessing a culture change whereby blaming preventable rehospitalizations on noncompliant patients is giving way to a paradigm of clinician- and system-level improvement to better prepare patients for their self-management role in their next care setting following hospital discharge, says Greenwald.
Those improvements are gaining momentum and showing promising preliminary results: increasing patients’ understanding of their follow-up plans, encouraging them to recognize and act on important new health problems and complications after discharge, and coordinating their ability to do so.
A common element among various intervention models is the use of a dedicated discharge advocate who plays a greatly enhanced care coordination role than traditional discharge planners – analogous to the coordination function of the medical home model of primary care – tying together fragmented medical records and care management advice; ramping up the use of proven patient-centered techniques; encouraging better involvement and follow-through by patients and their caregivers; and in some cases directly facilitating that follow-through.
Greenwald’s Project RED attempts to sort through a maze of discharge procedures by identifying clear roles and responsibilities of physicians, nurses and others involved in the discharge process. Using root cause and qualitative analyses to study systems and processes related to patients who are frequently admitted to the hospital, Greenwald and his colleagues identified specific failures of the hospital discharge system to inform a reengineered discharge process. Components of Project RED, which Greenwald says are being tested in ongoing randomized controlled trials, include:
· A dedicated discharge advocate appointed to each patient to coordinate activities during admission and prior to discharge, and who serves as the patient’s key contact person after discharge. The advocate prepares patient education throughout the hospital stay – not just at discharge, oversees medication education and reconciliation, arranges medication pickups and durable medical equipment delivery post-discharge, and ensures that patients have scheduled transportation.
· A comprehensive after-hospital care plan document, including photos of the primary treating physician and the discharge advocate; explicit instructions for the first days and weeks after discharge; follow-up appointment details; pending test results; a list of medications, medical indications and diagnoses, and doses and schedule, with photos of each pill; and diet, exercise and lifestyle recommendations.
· A phone call from the hospital pharmacist within two to three days after discharge to identify medication issues or concerns, check on the patient’s clinical condition, and remind them about follow-up appointments.
Patients in preliminary survey data indicate feeling better-prepared; having their questions answered better; and having a greater understanding of appointments, medications and diagnoses.
Many components of Project RED were incorporated into a quality standard on hospital discharge endorsed in late 2006 by the National Quality Forum, providing a clear road map for hospitals to follow, says Greenwald.
Greenwald is also co-investigator of a workflow redesign trial that attempts to standardize the hospital discharge – with local institutions adapting implementation based on their resources and workflow processes. The initiative, called Project BOOST (Better Outcomes for Older adults through Safe Transitions), was launched this summer by the Society of Hospital Medicine (SHM) at six pilot sites, with support by a grant from the John A. Hartford Foundation.
The initiative features a hospitalist physician as the leader of discharge quality improvement, in collaboration with other care providers, and incorporates best practices endorsed by care transitions researchers, hospital medicine physicians and pharmacists, and specialists in process improvement, health quality and patient safety. The initiative is led by a national advisory board that includes representatives from the Agency for Healthcare Research and Quality, The Joint Commission, the Centers for Medicare & Medicaid (CMS), and the Blue Cross and Blue Shield Association.
The project’s centerpiece is a discharge planning toolkit that includes:
· A risk stratification process and universal patient discharge checklist addressing key transition-readiness issues.
· A sheet of written discharge instructions (separate from the traditional discharge summary, and written in lay terminology) listing reason for the hospitalization; medications (ideally, distinguishing which medications were old, new, or changed and which medications the patient was on prior to admission that he or she should no longer take) with name, dose, route and frequency; what types of complications may occur and what to do if they happen; list of follow-up appointments for tests and clinical visits, with their dates, times and locations; and list of relevant contact information (e.g., principal care providers, the pharmacy, and inpatient physicians).
· A teach-back guide for assessing a patient’s understanding of a concept, i.e., asking the patient open-ended questions to encourage them to explain their understanding of the issue at hand, rather than telling the patient the information and asking for yes/no confirmation of comprehension.
· A risk-specific intervention plan, ensuring that the patient understands how to manage predictable events after discharge and why, when and how to access medical attention.
· A tracking mechanism to confirm that the patient’s principal outpatient provider receives the discharge summary, and encouragement of phone communication between inpatient and outpatient providers to address identified risk factors, primary issues of the hospitalization (including therapies initiated and discontinued), and outstanding issues, tests, appointments and follow-up plans for the patient.
· A commitment to arrange telephone contact (by either the hospital-based team or the outpatient care provider) to the patient or their caregiver within 48-96 hours of discharge in order to assess the patient’s condition and adherence and to reinforce follow-up.
The year-long Project BOOST trial will measure the interventions’ impact on length of stay, readmission rate and user (patient and provider) perception of effectiveness, says Greenwald. One of the six pilot sites is the Hospital of the University of Pennsylvania, and enrollment of an additional 24 sites began last month.
Another initiative which focuses on elderly patients – the Care Transitions Intervention – promotes the principle that patients and their caregivers need to become more active participants in their medical care. Spearheaded by Eric Coleman, M.D., MPH, director of the Care Transitions Program at the University of Colorado, the program features a personal health record and an advanced practice nurse “transition coach” who initiates a home visit and three telephone calls during a 28-day posthospitalization discharge period to encourage the patient and caregiver to assert a more active role during care transitions, and to ensure that the patient’s needs are being met.
The transition coach focuses on four conceptual areas or “pillars” derived from Coleman’s earlier investigations of what patients and their caregivers said would be most valuable to them during care transitions: assistance with medication self-management, a personal health record owned and maintained by the patient to facilitate cross-site information transfer, timely follow-up with primary or specialty caregivers, and a list of red flags that indicate a worsening condition with instructions on how to respond to them.
Goals of the home visit include reconciling the patient’s medication regimens; role-playing effective communication strategies so that the patient is prepared to clearly articulate his or her care needs during subsequent physician encounters; and reviewing red flags for worsening conditions, what initial management steps to take, and when to contact the appropriate health care professional.
Coleman’s intervention was able to reduce rehospitalization rates at 30 and 90 days, including rehospitalization for the same diagnosis, and decrease hospitalization costs at 180 days, even in a heavily penetrated Medicare Advantage market which has for many years attempted to reduce hospital use, according to results published in Sept. 2006 in the Archives of Internal Medicine.
The Care Transition Intervention model shifts the advance practice nurse’s role from doing things for the patient to the supportive role of facilitator who encourages the patient to do as much as possible independently. That less intense role also allows transition coaches to manage more patients – 24 to 28 patients per coach at any given time – and avoid redundancy with existing health care practitioners such as home health nurses and case managers, according to Coleman.
Another approach, the Transitional Care Model (TCM), deliberately assigns advance practice nurses a more intense role – not just as a coach to promote a good handoff, but as a caregiver who actively delivers both acute and community-based services, monitoring and managing in-hospital planning and home follow-up for chronically ill high-risk older adults. A demonstration project using the model is being implemented in the Philadelphia region by its principal architect, Mary D. Naylor, Ph.D., RN, Marian S. Ware Professor in Gerontology, University of Pennsylvania School of Nursing.
Unlike traditional case management, the purpose of the model is not to provide ongoing care to patients but to optimize patient outcomes specifically following a defined acute episode of illness. The model features transitional care nurses, each managing an active caseload ranging from 15 to 20 patients, who follow them from the hospital and make regular home visits, and who offer ongoing telephone support seven days per week through an average of two months post-discharge, says Naylor. After the initial visit, a minimum of one home visit per week during the first month is made, followed by semi-monthly visits until discharge from the program. The transitional care nurse prepares a transition summary – which includes patient’s goals, progress in meeting them, and ongoing or unresolved issues with the plan of care – for patients and primary care providers who assume responsibility for continuing care.
While the model is nurse-led, Naylor notes that it is a multidisciplinary approach that includes communication to, between and among the patient, family and informal caregivers, physicians, nurses, social workers, discharge planners and pharmacists, with a focus on increasing patients’ and caregivers’ ability to manage their care. The nurse accompanies the patient on their first post-discharge visit to their primary physician (and on subsequent visits, if needed), assisting them in generating a list of questions to ask, and assisting them in understanding the physician’s instructions immediately after the visit.
Randomized controlled trials have demonstrated that the TCM model has resulted in fewer rehospitalizations for patients’ primary illnesses and coexisting conditions, says Naylor. Among patients who required rehospitalizations, the time between their primary discharge and readmission was longer and the number of inpatient days was generally shorter than expected. The program has led to improved health outcomes, greater satisfaction of care and reduced all-cause hospital readmissions – at a savings of about $5,000 per patient, says Naylor.
Overcoming Implementation Barriers
The current reimbursement system offers two challenges that transition care improvement initiatives must overcome: incentive to improve and wherewithal to improve.
“As long as hospitals are paid for rehospitalizations, it is difficult to see the incentive for them to prevent them,” except in capitated systems, says Greenwald. “Incentives are still not sufficiently aligned to prevent rehospitalization. Some hospitals may lose revenue if they lower readmission rates. But, I say this unapologetically: it’s the right thing to do,” Greenwald says.
Reducing payments to hospitals with excessive readmission rates is one approach that was recently recommended by the Medicare Payment Advisory Commission (MedPAC). But physicians and insurers interviewed for this article regard that as using “a hatchet instead of a scalpel,” given the complexity of factors that contribute to hospital readmissions.
“We must ensure that we are incenting behaviors we are truly seeking, and avoid unintended consequences,” says Don Liss, M.D., Aetna’s medical director for the mid-Atlantic region. “Readmission is too broad a measure in itself, and one must consider several other variables such as patient severity and risk,” he adds.
The reputational impact on hospitals of transparency initiatives by CMS and other entities – which publicize hospitals’ readmission rates as a quality outcome measure – injects a level of accountability to spur improvement, says Liss.
Absence of a reimbursement mechanism for care across care settings remains another key barrier to widespread adoption of care transition initiatives, which require additional money and resources. Government and private foundation grants have funded demonstration pilots, but sustainable payor models have yet to be worked out. Proponents of transitional care models are beginning to build relationships with commercial insurers to fund the projects, and CMS is also beginning to take an interest in them.
Highmark is implementing Coleman’s model in several venues, with a focus on empowering patients and caregivers to accept a more active role in their transition, says Black. In January, the insurer plans to roll out a pilot project using a dedicated care transition coach with a community hospital. This summer, Highmark launched a care transition collaborative project with Quality Insights of Pennsylvania and six hospitals primarily in the Westmoreland Co. region, focusing on Medicare fee-for-service patients.
Highmark has also incorporated Coleman’s principles in an ongoing Medicare Advantage pay-for-performance pilot with two hospitals to improve transitions of care for patients discharged from a hospital to a skilled nursing facility, and anticipates eventual cost savings associated with reduced readmissions, Black notes. Highmark plans to share the results from the pilot program at local, state and national forums to stimulate further programs in transitional care improvement, adds Black.
A transition coaching model similar to Coleman’s is being implemented at ten hospitals in New Jersey, focusing on African-American and Hispanic/Latino patients diagnosed with heart failure, and funded by the New Jersey Health Initiatives – a statewide grant-making program of the Robert Wood Johnson Foundation, says Holmes. “Expecting Success: Excellence in Cardiac Care” features an advanced practice nurse who maintains contact with heart failure patients, primarily through telephone calls, to coach them on diet and nutrition; offer medication and medical management; advise them on when to contact a physician for a worsening condition; and ensure that they keep follow-up appointments with their primary care physician, she notes. The program began in early 2007 and runs through June 2009.
Health insurers are also beefing up their case management programs and focusing them more heavily on care coordination during the discharge and outpatient periods for high-risk patients. Independence Blue Cross (IBC) over the past couple of years has doubled the number of nurses involved in case management programs, and is adopting such a focus, says Richard Snyder, M.D., senior vice president of health services. In its Healthy Lifestyles: Keys to Wellness program, IBC nurse coordinators make telephone contact with patients to monitor gaps in care, medications and follow-up appointments, he notes.
Two months ago, New Jersey’s Horizon Blue Cross Blue Shield compeleted a year-long pilot project for which it contracted with a company to make a post-discharge phone call to every Horizon patient within 24 hours (except those who have home care services scheduled) to review their status, ask whether they are adhering to follow-up care instructions and whether they are encountering any problems, and refer them to Horizon’s clinical staff or to their physician to arrange for follow-up care if necessary, says Richard Popiel, M.D., Horizon’s vice president and chief medical officer. Horizon is examining the results of the pilot to see its impact on readmission rates, he notes.
Aetna supported a demonstration project using Naylor’s Transitional Care Model for about 200 of its Medicare managed care members in the mid-Atlantic region, which was completed six months ago, says Naylor. The success of that project has led the University of Pennsylvania Health System (UPHS) to adopt the model two months ago.
Independence Blue Cross has signed on as the first insurer to reimburse for Naylor’s program through UPHS, focusing initially on heart disease and diabetes, says Snyder. IBC’s case managers will also be involved in exchanging information with the transitional care nurses. “It is a fully integrated and coordinated program – a safety net during a particularly vulnerable time for patients,” says Snyder. “We’ll be interested in demonstrating better clinical and financial outcomes in all phases of care,” adds Snyder, who says it is preposterous to pour large sums of money into high-tech inpatient procedures, only to discharge patients to an environment with little coordinated support. Snyder says IBC is talking to other hospital systems about the Transitional Care Model. “We want to make this a mainstream-type program.”
This past August, CMS invited Naylor, Coleman and others to discuss their models with Quality Improvement Organizations from 14 states, with plans to award contracts for care transitions improvement programs, says Naylor. “We hope CMS will eventually offer a benefit for patients to access evidence-based transitional care services. That is our ultimate goal,” she adds.