Background: Hospital readmission rate is viewed as an indirect indicator of quality of care provided by a hospital. We completed a small pilot study looking at a multidisciplinary bedside huddle in the 24 – 48 hours before discharge which decreased readmission rates in high risk patients from 36% to 20% in a small pilot study. We are revisiting this data after a maintenance phase and comparing with a similar nursing unit that did not do routine huddles on high risk patients.

Methods: Our hospital is a 550-bed academic medical center with close to 150 medicine patient readmissions each month. We initiated a “Bedside Huddle” to be done within 24 – 48 hours of anticipated discharge.
The bedside huddle was convened for patients at high risk for readmission based on any one of the following: HAARD score (a locally devised and validated tool based on age, comorbidities, number of medications on admission and ED visits or hospitalization in the prior six months), length of stay > 7 days, or admission in the prior 30 days.

The bedside huddle consisted of a multidisciplinary team of provider (physician or midlevel provider), Registered Nurse (RN), pharmacist, case manager RN or social worker. Physical and occupational therapy were involved when applicable. The huddle is scheduled for a time when family or relevant caregivers can participate. The process starts with a standardized questionnaire which assesses patient and family’s concerns about the discharge process. The huddle focused on addressing patient concerns and providing education about disease severity and prognosis, expected symptoms and management, medication management, hospital follow up and the role of family and caregivers.

During the maintenance phase of the huddle project, one of the medicine units began using a survey completed by the patient to ascertain patient’s readiness for discharge. Patients were allowed to opt out of the huddle on this unit. The comparative nursing unit continued to perform huddles based on the prior criteria.

Results: A total of 43 huddles were performed over 5 months on both units. The 30-day readmission rate was significantly lower among high risk patients who received a bedside huddle (20.9%) as compared to high risk patients who did not receive a huddle (41%), χ2(1)=5.135, p=0.023, (V=0.20). Fewer huddles were held on the unit allowing patients to opt out of the huddle based upon a survey (4% versus 89% of eligible patients). More high-risk patients were readmitted on the opt-out unit (38.7% versus 28.3%).

Conclusions: There is limited high quality evidence as to how to best achieve readmission reduction, as suggested by only poor to modest performance of most readmit-risk prediction models and interventions.
In our institution we have found that a standardized bedside huddle can be helpful in decreasing readmission rates in high risk patients. The number of bedside huddles decreased and the readmission rate increased when patients were allowed to opt out of the huddles.