Share this article on:

A QUALITY IMPROVEMENT INITIATIVE: UCSF Neurologists Ratchet Down Readmissions

Butcher, Lola

doi: 10.1097/
Back to Top | Article Outline




The inpatient neurology department at the University of California, San Francisco reduced hospital readmissions by nearly half since the start of a new quality improvement initiative. The article offers an overview of the steps taken to achieve that goal.

When the University of California, San Francisco (UCSF) Medical Center began an intense focus on reducing avoidable hospital readmissions, neurologist S. Andrew Josephson, MD, was not convinced that the effort was worthwhile for neurology patients.

Four years later, Dr. Josephson, UCSF's vice chairman and medical director of inpatient neurology, reports 30-day readmission rates have fallen by more than half —from more than 20 percent to less than 10 percent. More significantly, neurology care has been dramatically improved by a series of interventions to reduce avoidable readmissions.

“While I'm certainly excited about numbers that look good, I am most interested in how this helps our patients because some of these unplanned readmissions suggest we have a gap in our health care delivery system that needs to be fixed,” he said. “The take-home point for us is that we can really make a difference in readmissions with the right resource commitment.”

Back to Top | Article Outline


Although neurology-specific readmission reduction initiatives are still rare, readmissions associated with some diagnoses — for example, congestive heart failure and pneumonia — have been targeted by quality improvement initiatives for nearly a decade. Dr. Josephson and his colleague, Anthony S. Kim, MD, medical director of the UCSF Stroke Center, started their work by borrowing from a successful model developed by the medical team that cares for patients with congestive heart failure.

“We thought some of that work was generalizable to neurology, so we started with the lessons learned and the infrastructure that had already been developed for that service,” Dr. Kim said.

Participating in a hospital-wide readmissions initiative has contributed to the neurology department's success. For one thing, the medical center's information technology support is key to tracking readmissions and process measures that help Drs. Kim and Josephson monitor their progress. For another, the opportunity to learn from others speeds everyone's learning.

Some of the lessons imparted are that readmissions quality improvement requires a multidisciplinary approach, constant data monitoring, and a willingness to change processes to improve outcomes. The neurology readmissions team of physicians, quality improvement analysts, nurses, social workers, case managers, and pharmacists meets at least monthly, but its members are in contact more frequently.

“There is no way to be able to analyze or make any difference with readmissions just from a physician standpoint, or just from a nurse standpoint; it really takes a team effort,” Dr. Josephson said.

Back to Top | Article Outline


The first goal was to learn why neurology patients are readmitted. Thus, each readmission is reviewed by the department's Morbidity and Mortality Committee to determine whether the readmission was potentially avoidable, what barriers prevented the patient's success after discharge, and what could be done to avoid a similar situation in the future.

Early on, this approach revealed that patients needed more effective education about their medication regimens and their discharge instructions. Now, a pharmacist meets with each patient before discharge to educate them about their medications, discuss how and when prescriptions will be filled, and identify and address any possible barriers to medication compliance.

Similarly, nurse educators meet with each patient to provide education about the patient's neurologic condition and the discharge instructions needed to successfully transition to home.

Both nurses and pharmacists use the “teach back” method to verify the patient and/or the family caregivers understand what they need to do. (See “Teach Back: A Quality Improvement Technique.”)

Back to Top | Article Outline


The neurology department has introduced several post-discharge interventions. For example, a nurse calls every patient within 72 hours of discharge to ask if the patient understands the medication regimen, has scheduled follow-up appointments, and is feeling well. Any patient with a concern or question is referred to a UCSF pharmacist or neurologist for a same-day phone call.

“It's really only about 10 percent of folks who have any questions,” Dr. Josephson said. “This approach seems to avoid a number of visits back to the emergency department as well as clinical deterioration.”

Similarly, a nurse calls every patient 30 days after discharge. Those conversations cover the same topics as the 72-hour calls, but they also gather information about whether a patient had an emergency department visit or hospitalization at another facility during the month. Without those calls, UCSF, a tertiary referral center, has no way of knowing its true readmission rate for patients who live a distance from San Francisco.



Those calls also provide a learning opportunity for the clinical staff. “The nursing staff learn a lot more about what patients are facing at home just by virtue of having made some of the calls to the patients at 72 hours and 30 days,” Dr. Kim said. “They incorporate that into how they think about patients when they are in the hospital and when they are preparing for discharge.”

Each patient is scheduled to visit the UCSF neurology discharge clinic within a week to 10 days to be examined by a neurohospitalist — typically the physician in charge of his or her inpatient care. This clinic, which takes a half-day each week, was created because patients frequently are unable to schedule an appointment with an outpatient neurologist for a few to several weeks after discharge.

“This allows us to follow up on tests that were planned, to make sure the patient is on the right medications, and to assess for clinical deterioration, therefore serving as sort of a bridge from the inpatient hospitalization to the outpatient clinic,” Dr. Josephson said.

Back to Top | Article Outline


In keeping with UCSF's hospital-wide practice, the neurology department's performance is tracked through a dashboard that displays metrics about patient care processes and readmissions. Creating the dashboard forces decisions about how success will be measured, Dr. Kim said.

“Without that process of thinking about what we are trying to change and why, we could end up with a change process that is not aligned with our goals,” he said. “We went through several iterations of the dashboard with input from everyone involved until we settled on what we cared about. That process is very important.”

Among other things, the dashboard shows:

  • Neurology readmission numbers and percentages on a monthly basis, with comparisons over time to monitor the effect of interventions.
  • The percentage of patients contacted within 72 hours and within 30 days of discharge.
  • The number of patients who were reviewed through the department's morbidity and mortality committee, and the number of patients that received medication education and nursing education at discharge.
  • The percentage of nurses that have completed formal training in “teach back” education.
Back to Top | Article Outline


Not all readmissions are bad, so determining an appropriate readmission rate presents a challenge. Some readmissions are planned, such as those for patients with two-step procedures. Some readmissions are unavoidable; for example, a stroke patient who later is admitted for an unrelated medical condition.

The most obvious strategy for reducing readmissions might be to lengthen the patient's stay, but that increases the chance of hospital-acquired infections and other problems — and potentially wastes financial resources.

The closer one examines readmissions, the more complex the topic becomes, Dr. Josephson said. For example, if a stroke patient is readmitted to the hospital for pneumonia, that readmission might reflect neurology's failure to take proper preventive action.

“Or what if somebody who was on the medicine service developed transient atrial fibrillation and was discharged without anticoagulation and comes back two weeks later with a stroke? That may be a problem that we can help with through discussions with the medicine quality improvement group,” he said.

For most neurologic conditions — and for patients with multiple medical conditions — national readmission rates are not available, so neurologists have no benchmarks against which to measure their performance.

In light of that, UCSF's approach is to track readmissions carefully to learn its baseline performance and how to improve from there. “We look at each one of these readmissions and figure out what happened,” Dr. Josephson said. “We want to understand these unplanned readmissions and figure out how to avoid them in the future, when possible.”

Back to Top | Article Outline


Teach Back is a quality improvement technique that involves asking the patient or family caregiver to restate in his or her own words what they thought they heard during their discharge instructions. Show Back, a variation of the technique, involves asking patients to demonstrate a skill such as putting a day's medications in a pill box organizer.

  • Teach Back is used to assess whether patients and their family members or other caregivers have the ability and confidence to follow discharge instructions regarding medication, nutrition, follow-up medical appointments and other topics. Steps include:
  • Explaining needed information the patient and other caregivers who need to know it.
  • Asking in a non-shaming way for the individual to explain the learning in his or her own words.
  • If a gap in understanding is identified, reiterate or expand on the original explanation and ask the patient to again “teach back” in their own words.
  • If the patient and/or caregiver cannot “teach back,” inform the patient's hospital care team so the discharge plan can be adjusted accordingly.

Source: How-to Guide: Improving Transitions from the Hospital to Post-Acute Care Settings to Reduce Avoidable Rehospitalizations, Institute for Healthcare Improvement, 2011.



This article is part of a continuing series of articles on quality improvement initiatives in neurology. The full archive of the series is available here:

Back to Top | Article Outline


•. Douglas VC, Scott BJ, Berg G, et al. Effect of a neurohospitalist service on outcomes at an academic medical center. Neurology 2012; 79(10):988–994.
    •. Cohen AB, Sanders AE, Swain-Eng RJ, et al. Quality measures for neurologists: Financial and practice implications. Neurol Clin Pract 2013:3(1):44–51.
      © 2013 American Academy of Neurology