The purpose of the proposed study is to evaluate the impact of a behavioral intervention team (BIT) on 2 adult units: a general medical (8N); a cardiac/medical stepdown (8S) at Vanderbilt University Medical Center (VUMC) with a higher proportion of patients with behavioral health comorbidities.

The BIT will provide a proactive psychiatric team that will screen all patients on admission, provide consultation to those meeting established criteria, and recommend interventions to the primary care team for consideration. BIT will also serve to model therapeutics tactics in communication and goal setting to the clinical staff.

Research Questions for the proposed study:

Considering that behavioral health co-morbidity is common among hospitalized medical inpatients and is associated with higher costs of care and staff dissatisfaction, we will ascertain whether the addition of dedicated, trained behavioral intervention team, compared to nursing staff training on trauma informed care and de-escalation techniques provides:

1. meaningful, measureable improvement in the prevention and management of disruptive behavior in the healthcare setting at VUMC, and
2. Improvement in staff perceptions of their ability to manage patients exhibiting disruptive, threatening or acting out behavior

Specifically, using a pragmatic, cluster cross-over trial design where the BIT crosses between 8N and 8S, we will test the hypothesis that presence of the BIT results in:

1. Improvement in the prevention and management of patients exhibiting disruptive, threatening or acting out behavior
2. Improvement in staff perceptions of their ability to manage patients exhibiting disruptive, threatening or acting out behavior

Inclusion Criteria:

• All patients admitted to two adult medical surgical units (8N/8S) at Vanderbilt University Medical Center during the study period.
• All clinical nurses and staff on 8N/8S employed during the study period.

Exclusion Criteria:

• Patients admitted to the floor before the BIT team has begun the crossover period on the unit.

Individual participant data be available (including data dictionaries) will be shared, particularly individual participant data that underlie the results reported after deidentification.

Immediately following publication. No end date.

Proposals should be directed to the PI via the Learning Healthcare System platform secondary data use application