• Change in performance between hospitals in the intervention and hospitals in the control group on the proportion of prophylaxis administered within the recommended timeframes.
  

• Change in performance on appropriate selection of drug, appropriate duration, appropriate number of doses pre-op, appropriate use of beta-lactams in patients with allergies.
  

Numerous studies have shown that many surgical site infections (SSI) are preventable with appropriately timed antimicrobial prophylaxis. Patients receiving prophylaxis either well-before or well after surgery are up to 5 times more likely to develop an SSI than those receiving appropriate therapy (Classen et al, 1992). Unfortunately, errors in antimicrobial prophylaxis timing are extremely common with error rates typically reported to be between 35 and 40%. Given that errors in antimicrobial prophylaxis are common and the consequences of error so grave, identifying methods to assist hospitals in improving prophylaxis must be a high priority. TRAPE will evaluate a multifaceted, theory-based intervention to assist hospitals in progressing through stages of organizational change to improve the prophylaxis process. We will test the impact of the intervention using a rigorous group-randomized, nested, pretest-posttest design (Murray, 1998).

Our specific aims over the 4 year project period are:

1. determine the incidence of medication errors related to antimicrobial prophylaxis for cardiovascular surgery, joint replacement surgery, and hysterectomies in 44 hospitals recruited to participate in the study;
2. identify organizational and system factors associated with error rates;
3. randomize the 44 hospitals to evaluate the effectiveness of a multifaceted intervention.

The interventions consist of: a) the promotion of specific process changes likely to reduce error rates, b) a site-visit, c) customized process feedback, d) facilitated benchmarking, and e) peer consultation. The 22 intervention hospitals will be compared to 22 hospitals that receive written feedback of their error rates only. The study has 80% power to detect a 12-15% improvement in the timing of prophylaxis in the full intervention group compared to the written feedback only group. Data collection will be done at each participating hospital, and the medical records of 100 surgical patients before and after the intervention will be abstracted at each hospital to establish performance rates. Changes in the processes of care, and the evolution through stages of organizational change will also be assessed.