Existing literature on aggression within psychiatric hospitals suggests that treating an aggressive patient’s symptoms could be complemented by (a) milieu environments that mitigate violence and (b) hospital-wide policies and procedures that focus on creating a safe environment. Described as an ecological approach, examples of how this broader, situational approach can reduce inpatient violence in psychiatric settings are provided throughout. The authors identify potential barriers to focusing on wards and institutional rules as well as patient treatment. Last, details of how this ecological approach has been implemented at one state hospital in California are provided.

Objectives: A mechanism to increase the influence of Health Technology Assessments (HTAs) on hospital policy decisions was developed.

Methods: We describe the process and results of an experiment in which a local in-hospital HTA unit was created to provide sound evidence on technology acquisition issues, and to formulate locally appropriate policy recommendations. The Unit consists of a small technical staff that accesses and synthesizes the evidence incorporating local health and economic data, and a Policy Committee that develops policy recommendations based on this evidence. It represents administration, health-care professionals, patients, and representatives of the clinical disciplines affected by each issue. The level of success of the Unit was independently evaluated.

Results: To date, 16 reports have been completed, each within 2–4 months. Five recommended unrestricted use, seven recommended rejection, and four recommended very limited use of the technology in question. All have been incorporated into hospital policy. Budget impact is estimated at approximately $3 million of savings per year.

Conclusions: This local in-house HTA agency has had a major impact on the adoption of new technology. Probable reasons for success are (i) relevance (selection of topics by administration with on-site production of HTAs allowing them to incorporate local data and reflect local needs), (ii) timeliness, and (iii) formulation of policy reflecting community values by a local representative committee. Because over one third of all health-care costs are incurred in the hospital, diffusion of this model could have a significant effect on the quantity and quality of health-care spending.