Child Psychiatry
August 10, 2016

“Staff enjoy this approach and have found through it a pleasant and positive therapeutic environment. There’s less tension in the air when you’re not getting into power struggles all day, and it’s much easier for patients to engage in their treatment.” – Elizabeth Reynolds, Ph.D.

Restraints and seclusion have generally been viewed by child psychiatry inpatient staff as unpleasant but sometimes necessary tools to manage aggressive behavior among hospital patients, reduce their agitation and prevent injuries to themselves and others. Professionals agree alternatives are needed, but there has been little research on specific, evidence-based behavioral interventions to reduce restrictive practices. Until now.

Modeling a proven prevention strategy that has been implemented in over 7,500 schools nationwide to reduce disciplinary actions and suspensions and improve academic performance—Positive Behavioral Interventions and Supports (PBIS)—Johns Hopkins child psychologist Elizabeth Reynolds and colleagues in a four-year prospective study were able to meaningfully reduce—from 543 events to 253 events—the use of restraints and seclusion on their inpatient unit. Also, the percentage of patients who were placed in seclusion or restraint markedly decreased, from 19.6 percent to 13.4 percent during the study period, and the mean duration of seclusion and restraint incidents decreased from 20.43 minutes to 8.18 minutes per episode (Psychiatric Services May 2016;67( 5): 570-573).

Given the concern that a reduction in seclusion and restraint can lead to an increase in administering medications pro re nata (PRN)—or as needed—to treat acute agitation or aggression, the researchers also monitored the use of PRN medications. After implementation of the PBIS model, there was a reduction in the use of PRNs from 1,705 to 1,014. The percentage of patients who received a PRN medication decreased from 41.6 percent to 29.4 percent.

“We were able to reduce the rates of seclusion and restraints, and the use of PRNs, significantly,” says Reynolds.

How? The proactive PBIS model, Reynolds explains, employs positively worded behavioral expectations for patients—“be safe,” “be responsible” and “be respectful”—and a reward system that reinforces appropriate behavior by patients. In this program, staff stamped patients’ passports with rewards that could be accumulated and reimbursed with physical products or privileges. Punitive actions were not included in the model.

“One of the biggest factors is changing the focus on behavior,” says Reynolds. “Rather than focusing on the behavior we don’t like, we focus on the behavior we want to see. When that behavior is demonstrated, we praise it and reward it.”

The model also incorporated targeted problem-solving conversations with patients who demonstrated problem behavior on the unit, and individualized behavior plans for patients who continued to have problematic behaviors after those problem-solving conversations. Key to success, Reynolds adds, is training for nursing staff that includes education and role-playing, and buy-in from staff significantly challenged in managing high-risk patients.

“Obviously this is a very difficult population to work with,” says Reynolds. “You see kids with significant trauma histories, poor home environments and lots of parental loss. It can be a really challenging place to work.”

But staff on the Johns Hopkins Child Psychiatry inpatient unit, Reynolds stresses, often look for ways to create a more therapeutic and caring environment, a place where patients feel more supported. They were ripe for a positive behavioral intervention, which turned out to reward staff as well as patients.

“Staff enjoy this approach and have found through it a pleasant and positive therapeutic environment,” says Reynolds. “There’s less tension in the air when you’re not getting into power struggles all day, and it’s much easier for patients to engage in their treatment. We really see it as a positive and preventive environment.”

The study did have its limitations, Reynolds notes, and was not able to identify the intervention components that were most instrumental in achieving outcomes. It also covered a relatively short duration of intervention, which raises questions about sustainability over time. The findings highlight the need for a future randomized clinical trial to more rigorously test the efficacy and effectiveness of the PBIS model in other inpatient settings. Next steps for Reynolds and her colleagues include a quality improvement initiative with ongoing monitoring of seclusion and restraints.

“We’re always trying to improve the quality of care we’re providing,” says Reynolds. “We’d like the rates of restraints and seclusion to be zero, to not be used at all.”