Evaluation of an alternative model for the management of clinical risk in an adult acute psychiatric inpatient unit.

Despite their widespread use, typical visual observation practices are not evidence-based and adverse events - such as self-harm and absconding - still occur even under the most intense forms of observation. This study aimed to (i) develop and implement an engagement-focused systematized model of clinical risk management in an adult acute psychiatric inpatient unit; and (ii) prospectively evaluate its effect on rates of violence, self-harm, absconding, sexually inappropriate behaviour, and seclusion. A new model of engagement-focused clinical risk management was developed using a participatory action research framework and implemented in an adult acute psychiatric inpatient unit. Using a mirror-image design, rates of violence/aggression, self-harm, absconding, sexually inappropriate behaviour, and seclusion were compared before and after implementation, and staff satisfaction levels were measured. The clinical engagement-based model was introduced, and 1087 admissions before implementation (24 months) were compared with 965 admissions post-implementation (18 months). The new model was associated with significantly reduced rates of absconding (pre: 10.5/1000 occupied bed days, 95% CI [9.0, 12.1] compared with post: 6.5/1000 occupied bed days [5.2, 8.1], P < 0.001) and seclusion (pre: 43.7/1000 occupied bed days, 95% CI [40.6, 46.9] compared with post: 30.9/1000 occupied bed days [27.9, 34.1], P < 0.0001). Rates of aggression, deliberate self-harm, and sexually inappropriate behaviour were non-significantly decreased. Findings suggest that this engagement-focused model of clinical risk management in an adult psychiatric inpatient unit significantly reduced adverse patient events and was preferred by staff over current practice. Other psychiatric inpatient facilities may see a reduction in adverse events following the introduction of this well-tolerated risk management model.