Improving clinical and cost outcomes in delirium: Use of practice guidelines and a delirium care team

Objective: To test the impact of two different prospective, practice guideline-driven interventions in improving clinical recognition, management, and outcomes in elderly patients with delirium. Setting: An urban, academic hospital with three geographically separate, 48-bed, general internal medicine floors. One floor was designated as a control site; two floors were designated for interventional procedures. Participants: Participants included: (1) 124 acutely confused older patients identified by the nursing staff; (2) house staff, attending faculty physicians, and medical nurses providing care for these patients; and (3) a geriatric delirium care team. Design, Measurements, and Interventions: A diagnostic and therapeutic geriatric practice guideline for patients with delirium was developed. In Phase I, this guideline was distributed to physicians providing care for patients who were judged to be possibly delirious. In Phase II, a geriatric physician/nurse delirium care team used the guidelines and consulted directly with the nurses and physicians providing care to patients diagnosed with delirium (by means of the Diagnostic and Statistical Manual of Mental Disorders, 4th edition, and the Confusion Assessment Method). Post-discharge chart review outcomes measured included documentation of delirium, length of stay, death, nursing home discharge, use of neuroleptics or restraints, and consultations with neurology and psychiatry departments. Main Results: Phase I: Distribution of guidelines, even when accompanied by group educational sessions, did not result in improved outcomes in the intervention units versus the control unit. Phase II: Recognition of documentation of delirium was significantly better in the intervention sites (93%) versus the control site (43%; P < .01). Average length of stay was shorter in the intervention sites versus the control site by 1.7 days (P = .03), and fewer consultations were requested in the intervention sites (P < .01). No significant differences between the intervention sites and the control site were found in number of deaths, use of restraints, use of neuroleptics, or nursing home placement (all were low frequency). Net cost 'savings' of $57,138 was identified over the six-week period of the trial. Conclusion: Passive use of a practice guideline did not improve clinical outcomes in patients with delirium. A brief, labor-intensive, guideline-based intervention using a geriatric physician/nurse team, focused on the professionals providing care for geriatric patients with delirium, produced improved outcomes and cost savings.