Implementing and Validating a Comprehensive Unit-Based Safety Program

JOBNAME: jops 1#1 2005 PAGE: 1 OUTPUT: Tue March 15 15:21:54 2005 ORIGINAL ARTICLE Implementing and Validating a Comprehensive Unit-Based Safety Program Peter Pronovost, MD, PhD,* Brad Weast, MHA, Beryl Rosenstein, MD, J. Bryan Sexton, PhD,* Christine G. Holzmueller, BLA,* Lori Paine, MSN, Richard Davis, PhD, and Haya R. Rubin, MD, PhD* Background: The IOM identified patient safety as a significant problem. This paper describes the implementation and validation of a comprehensive unit-based safety program (CUSP) in intensive care settings. Methods: An 8-step safety program was implemented in the Weinberg ICU, with a second control (SICU) subsequently receiving the intervention. Unit improvement teams (physician, nurse, administrator) were identified to champion efforts between staff and Safety Committee. CUSP steps: (1) culture of safety assessment; (2) sciences of safety education; (3) staff identification of safety concerns; (4) senior executives adopt a unit; (5) improvements implemented from safety concerns; (6) efforts documented/analyzed; (7) results shared; and (8) culture reassessment. Results: Safety culture improved post versus pre-intervention (35% to 52% in WICU and 35% to 67% in SICU). Senior executive adoption led to patient transport teams and pharmacy presence in ICUs. Interventions from safety assessment included: medication reconciliation, short-term goals sheet and relabeling epidural catheters. One-year post-cusp implementation, length of stay (LOS) decreased from 2 to 1 day in WICU and 3 to 2 days in SICU (P, 0.05 WICU and SICU). Medication errors in transfer orders were nearly eliminated, and nursing turnover decreased from 9% to 2% in WICU and 8% to 2% in SICU (neither statistically significant). Conclusions: CUSP successfully implemented in 2 ICUs. CUSP can improve patient safety and reduce medication errors, LOS, and potentially nursing turnover. Key Words: safety, ICU, culture, CUSP, error, patient safety (J Patient Saf 2005;1:33 40) The Institute of Medicine (IOM) report To Err Is Human identified patient safety as a significant nationwide problem and stated that efforts to improve safety must focus on systems, not providers. 1,2 These systems include technology, From *The Johns Hopkins University, Baltimore, Maryland; Johns Hopkins Medicine, Center for Innovations in Quality Patient Care, Baltimore, Maryland 21205; 319th Medical Group, Grand Forks AFB, North Dakota; and The Johns Hopkins Hospital, Baltimore, Maryland. Correspondence: Peter J. Pronovost, MD, PhD, 901 S. Bond Street, Suite 318, Baltimore, MD 21231 (e-mail: ppronovo@jhmi.edu). Copyright Ó 2005 by Lippincott Williams & Wilkins practices, procedures, policies, and more broadly the culture in organizations. Organizational culture is defined here as the collection of values, beliefs, and assumptions that guide members behaviors, 3 and is generally referred to as the way we do things around here. In a safe culture, employees are guided by an organizationwide commitment to safety, in which each member upholds their own safety norms and those of their coworkers. 4 Evidence from aviation supports an association between a culture of safety and better error management. 5 The IOM report To Err Is Human spurred healthcare organizations to implement initiatives that improve patient safety. 6 Understanding that culture changes incrementally and that all staff must live a culture of safety, we sequentially implemented a safety program. The comprehensive unit-based safety program (CUSP) targets the work unit level, to engage and empower staff to identify and eliminate patient safety hazards. The specific aims of this investigation were: (1) to implement and validate CUSP and demonstrate its validity through a variety of measures in the ICUs at The Johns Hopkins Hospital (JHH), and (2) to evaluate whether we could disseminate the CUSP to a second ICU. METHODS Study Design and Population We used a quasi-experimental design in which we implemented a safety program in the Weinburg ICU (WICU), while another surgical ICU (SICU) served as the control. Six months later, we implemented the intervention in the SICU. Thus, we had two 6-month measurement periods (pre and post) in the WICU, and three 6-month periods in the SICU (pre, control, and post). These units were selected because ICUs are high risk areas where mistakes are common and because one of the authors (PP) is an attending intensivist in both ICUs. 7,8 The WICU is a 14-bed oncology surgical ICU, and the SICU is a 15-bed surgical ICU that cares for general vascular surgery, trauma, and transplant patients. In both ICUs, all patients are co-managed by intensive care physicians and surgeons, and the nurse to patient ratio is either 1:1 or 1:2. Critical care fellows and surgery and anesthesiology residents are assigned to the ICUs and care for all patients. Development of the CUSP As a work in progress, CUSP evolved through literature review, discussion with experts, trial, and adaptation. Design J Patient Saf Volume 1, Number 1, March 2005 33

JOBNAME: jops 1#1 2005 PAGE: 2 OUTPUT: Tue March 15 15:21:56 2005 Pronovost et al J Patient Saf Volume 1, Number 1, March 2005 of the program was influenced by participation in the Institute for Healthcare Improvement s Quantum Leaps in Patient Safety. The goal of this program was to create a safety program that: (1) could be implemented sequentially in work units, (2) would improve the culture of safety, (3) would allow staff to focus safety efforts on unit-specific problems, (4) would help staff implement system wide safety initiatives, and (5) would include rigorous data collection (ie, be able to be published). The foundation for this program was the unit improvement team. Each team included a physician, nurse, and a senior executive, with pharmacists, respiratory therapists, and other staff encouraged to participate. After discussions with department chairs and nurse managers, it was decided that improvement team members would dedicate 4 to 8 hours per week to implement this improvement program. CUSP is an 8-step program designed to impact safety climate by empowering staff to assume responsibility for safety in their environment. This is achieved through education, awareness, access to organization resources, and a toolkit of interventions. CUSP is summarized below and outlined in Table 1. Step 1: Conduct Cultural Survey The first step involves assessing the culture of safety on the work unit. We used a medical derivative of aviation s Safety Climate Scale (SCS). 5,9 This 10-item survey is answered using a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree) (Appendix 1) and assesses the extent to which staff perceive a strong and proactive organizational commitment to patient safety. The SCS, adapted from the flight management attitudes questionnaire, measures attitudes toward stress, status hierarchies, leadership, and interpersonal skills. 10 Scales and individual items from the SCS questionnaires have demonstrated good reliability, 11 internal consistency, and replicable factor structure. 9,11 Moreover, it is predictive of pilot performance 5,12 and high speed rail incident rates, 13 is sensitive to training interventions, 14 17 and is used in medicine to better understand the environments in which care is delivered. 18 20 TABLE 1. Steps in the Comprehensive Unit-Based Safety Program (CUSP) Step Description 1. Measure safety climate Survey staff 2. Educate staff on the science of safety Talk presented to staff 3. Identify staff s safety concerns -Ask staff what is broken -Incident reporting 4. Senior executive adopt a unit Senior executive meets monthly with team 5. Implement improvements Select 3 interventions that do and do not require marginal resources 6. Document results Each project needs a metric 7. Share stories Web based safety tales 8. Re-measure safety culture Survey staff Step 2: Educate Staff on Sciences Related to Safety Once the cultural survey is complete, the unit is ready to learn the science related to safety. The CUSP champion (the physician co-chair of the patient safety committee (PCS) or the nurse safety coordinator) presents this concept to all unit staff, including physicians, nurses, pharmacists, respiratory therapists, aides, and clerks in a 40-minute presentation. This talk highlights lessons learned from To Err Is Human and explores the potential impact of systems on the risk of an adverse event. 2 The objectives of this session are to help staff understand the following: patient safety is a significant problem; efforts to improve safety should focus on improving systems rather than blaming caregivers; when harm occurs, it is preceded by a cascade of system breakdowns; interpersonal skills, such as speaking up when you have a concern, listening when others do, and acknowledging personal and organizational vulnerabilities, play a critical role in patient safety; and, finally, blame free does not mean responsibility free we all need to accept responsibility for the systems in which we work. The CUSP champion serves as a role model and remains the main point of contact for the CUSP. Step 3: Identify Staff s Safety Concerns After the science of safety presentation, work unit staff completes an open-ended safety assessment (Appendix 2) that asks 3 questions: (1) Tell us about the last patient who would have been harmed without your intervention. (2) How will the next patient be harmed? (3) How can we prevent this from happening by either preventing the mistake, making the mistake visible, or mitigating the harm should it occur? This information is collected and summarized by the PSC before the first unit meeting with the senior executive. Issues presenting significant patient safety risks are discussed immediately with the chairmen of the PSC. Summarized safety concerns and an action plan are sent in letter form to work unit participants and the senior executive (step 4) from the PSC. In addition, we used an incident reporting system to help identify patient safety hazards (http://icusrs.org, click Members Zone, then Training Form ). Step 4: Executive Adopts a Work Unit Next, members of the PSC meet with the senior executive who will adopt the unit, present the science of safety, and discuss the senior executive s role in the CUSP. From that point forward, the senior executive meets with unit staff once a month to discuss results from the staff safety assessment (step 3) and other staff concerns. These meetings are intended to remove barriers to system changes, provide resources for safety improvements, demonstrate the executive s commitment to patient safety, provide coaching for teams, and foster a trusting relationship between senior leadership and staff. The details of this step have been previously published. 21 The ability to provide rapid, useful, and accessible feedback is an important means for developing trust and creating a reporting culture. 22 34 q 2005 Lippincott Williams & Wilkins

JOBNAME: jops 1#1 2005 PAGE: 3 OUTPUT: Tue March 15 15:21:57 2005 J Patient Saf Volume 1, Number 1, March 2005 Implementing and Validating a Comprehensive Unit-Based Safety Program Step 5: Implement Improvements The next step is to implement improvements. First, work unit staff selects areas in which to focus improvement efforts. Areas are selected based on the institutional safety priorities, results of the safety assessment, discussions, and historical events. Staff are then instructed to prioritize improvement efforts based on the probability of the event occurring and the severity of harm should the event occur. We initially used a prioritization matrix adapted from the VA to help select improvement areas. 6 However, staff rarely used this form and it was discontinued. Instead, staff relied on their knowledge of the work environment to prioritize improvement efforts. Once a priority list is compiled, staff is directed to select 2 3 improvement efforts that require minimal resources (ie, funded out of their budget) and implement these immediately. They also select 2 3 improvement efforts that require substantial resources (ie, require funding from the hospital) and submit these to the PSC for review for possible funding. Staff is encouraged to discuss needed resources with their senior executive. When selecting safety interventions, staff is asked to consider 2 simple goals: (1) reduce complexity or the number of steps in a process; and (2) create independent redundancies or checks for key steps in a process, such as the use of evidence-based therapies. An example of an independent redundancy is to have several individuals, such as a physician, nurse, pharmacist, and patient or family member, check independent of one another to ensure that a patient s medications are appropriate or that barrier precautions are used when inserting central lines. Step 6: Document Results While each improvement effort has its own outcome measure, the primary outcome measure of the CUSP was the safety attitudes survey. Also, because improved culture may lead to reduced complications and, thus, reduced length of stay (LOS), and improved nursing satisfaction and, thus, reduced nursing turnover, we evaluated the median ICU LOS and nursing turnover rate as secondary outcomes. Documenting results provides validity and the potential for public dissemination. As such, teams are encouraged to maintain methodological rigor in their data collection and improvement efforts and to present data in annotated run charts. Many improvement efforts falter because they fail to collect measures and thus are unable to document their impact. 23 These types of efforts are often perceived by caregivers as unscientific efforts by hospital administrators to reduce hospital costs. To overcome these barriers, we focus on improving safety rather than reducing costs and create the expectation that improvement efforts should be rigorous enough for publication. Step 7: Share Stories Although a key element in organizational learning is to share stories, we generally do this poorly. To help foster organizational learning, we created a standard report called a Safety Tales Form. Staff uses this form to summarize and disseminate improvement efforts. Step 8: Repeat Cultural Survey Repeating the safety climate survey helps us evaluate the success of the CUSP through the eyes of frontline personnel. This is administered within 6 months of the baseline survey to compare changes. We administered the Safety Climate Survey (SCS) to all WICU staff in June 2001 (pre-cusp) and again in December 2001 (post-cusp). Because the physician staff overlaps between the WICU and SICU, we only surveyed nurses in the SICU. SICU nurses completed 3 surveys (pre intervention in June 2001, control period December 2001, and post intervention June 2002). The potential response categories were a 5-point Likert scale from 1 = strongly disagree to 5 = strongly agree. We scored responses as the percent agreement, that is, the percent of respondents who answered 4 or 5 to each question. We summarized the percent of respondents who scored 4 or 5 to all 10 questions as those reporting a positive safety climate. To evaluate improvements in safety climate, we only included respondents who completed both the pre and post intervention surveys. Interventions Implemented from CUSP Several interventions were implemented during step 5 (Implement Improvements) of CUSP in the WICU and SICU. One intervention was a short-term goals sheet (previously published 24 ). This form was used during daily rounds to help ensure transparency among the care team regarding the work needed to discharge the patient and to identify and mitigate safety concerns. The goal sheet encourages staff to explicitly state their care plan, tasks to be done, communication plan (with the team and family members), and safety risks. We also developed and implemented in both units a tool to reduce medication errors in transfer orders by providing an independent check for errors. This check involved the nurse reviewing discharge orders and medical record at the time of patient discharge from the ICU and answering 3 questions: (1) Do the medications listed in the discharge orders match what the patient is currently receiving? (2) Are allergies listed correctly in the discharge orders? (3) Did the patient start their home medications? If the answer to any question was no, the nurse was instructed to ask the patient s physician generally the chief resident on the surgical service if they intended to make this change. The nurse then asked the patient if the allergies and home medications were listed correctly. Our definition of a medication error was if, as a result of this process, the physician changed the discharge orders. Our experience with medication reconciliation has been published. 25,26 Measurement and Analyses All interventions were implemented during the intervention period. We defined nursing turnover as the annualized percent of nurse full-time equivalents (FTEs) who left their nursing unit, in this case the ICU. We measured the mean nursing turnover in the WICU for the 6-month preintervention period (January 1 through June 30, 2001) compared with the same 6 months after the intervention (January 1 through June 30, 2002) to account for seasonal changes in q 2005 Lippincott Williams & Wilkins 35

JOBNAME: jops 1#1 2005 PAGE: 4 OUTPUT: Tue March 15 15:21:57 2005 Pronovost et al J Patient Saf Volume 1, Number 1, March 2005 turnover rates. We measured the mean turnover in the SICU for the baseline period (July 1 through December 31, 2001), control period (January 1 through June 30, 2002), and postintervention period (July 1 through December 31, 2002). We measured the ICU LOS in the WICU for the 6- month pre-intervention period (January 1 through June 30, 2001) compared with the same 6 months after the intervention (July 1 through December 31, 2001). We measured the mean ICU LOS in the SICU for the baseline period (July 1 through December 31, 2001), control period (January 1 through June 30, 2002), and post-intervention period (July 1 through December 31, 2002). We used a x 2 test to evaluate differences in responses to the cultural survey and nursing turnover in the pre and post periods. To evaluate differences in LOS, we used a t test. A P value of less than 0.05 was considered significant. All analyses were done with STATA 8.0. RESULTS Two surgical ICUs (WICU and SICU) have completed all 8 steps of the CUSP. Described below are the results from the WICU and SICU. Cultural Survey In the WICU, we received completed surveys from 66 people (89% response rate) in the pre-intervention period and 64 people (86% response rate) in the post-intervention period. In the SICU, we received completed surveys from 23 nurses (89% response rate) in the baseline period, 23 (89% response rate) in the control period, and 21 (84% response rate) in the post intervention period. Table 2 describes the response rates by job category in the WICU and for nurses only in the SICU. Results of the cultural survey before and after CUSP implementation in the WICU and SICU are presented in Table 3. The percent of respondents who agreed (scores of 4 or 5 on the survey) increased for all 10 questions in the TABLE 2. Respondents to Safety Attitudes Survey in Pre- and Post-Intervention Periods Role Pre Intervention Period Number Surveyed Number Responding (%) Post Intervention Period Number Surveyed Number Responding (%) Physician 10 8 (80) 12 12 (100) Nurse in WICU 40 37 (93) 38 32 (84) Nursing technician 3 2 (67) 3 1 (33) Unit clerk 4 3 (75) 4 3 (75) Support associate 8 7 (88) 8 7 (88) Respiratory therapist 2 2 (100) 2 2 (100) Pharmacist 7 7 (100) 7 7 (100) Overall 74 66 (89) 74 64 (86) Nurse in SICU* 26 23 (89) 25 21 (84) *Data only available for nurses, 89% response rate for control period (23 responders). post-intervention survey relative to the pre-intervention survey, with the SICU increasing more than the WICU. In the WICU, the percent of respondents reporting a positive safety climate increased to 52% from pre- to post intervention, with 35% reporting a positive climate preintervention and 52% reporting a positive climate post intervention. In the SICU, the percent of respondents reporting a positive safety climate almost doubled, from 35% pre to 68% post intervention, with results from the control period remaining relatively similar to pre intervention (40%). These results are summarized in Figure 1. Interventions Implemented from CUSP The staff s safety concerns were similar in the WICU and SICU. Examples of issues were lack of trained patient transport teams, medication errors, and poor communication among ICU providers. Table 4 represents a summary of concerns collected from the safety assessments and discussions with senior executives. The lessons learned from our experience with senior executives adopting a unit have been previously published. 21 Senior executive engagement helped to improve patient safety locally. 27 Staff s safety concerns prompted funding of a patient transport team and a point of care pharmacist to facilitate medication ordering and distribution. Table 5 summarizes the staff s plans to improve patient safety with minimal and additional resources. A total of 5 safety concerns identified were acted upon. Results from the safety assessment indicated that staff perceived inadequate communication as a significant safety risk, particularly when patients are transferred out of the ICU. A short-term goals sheet was implemented to improve communication among ICU team members and with family members. The results of implementing the daily goals sheet are published elsewhere. 24 Medication reconciliation was associated with a significant reduction in medication errors. 25,26 In the 2 weeks prior to this intervention, 94% (31 of 33) of WICU orders contained an error. After making the medication reconciliation a part of the routine nursing discharge process, this type of error was eliminated. 25 In the SICU, approximately 40% of charts contained a medication error at transfer; these were also eliminated. 26 Nursing Turnover The CUSP was also associated with a reduction in nursing turnover. There were 28 full-time nurses in the WICU and 46 in the SICU during study periods. Nursing turnover rates decreased from 9% (pre-intervention) to 2% (post intervention) in the WICU and from 8% (pre-intervention) and 9% (control) to 2% (post intervention) in the SICU (Table 6). Due to small sample size, these reductions were not statistically significant. The overall hospital nursing turnover rate was 17% in 2001 and 15% in 2002. LOS The CUSP was associated with a reduction in LOS and nursing turnover in both the WICU and SICU. In the preintervention period, the mean ICU LOS was 2 days in the WICU and 3 days in the SICU. After CUSP implementation, 36 q 2005 Lippincott Williams & Wilkins

JOBNAME: jops 1#1 2005 PAGE: 5 OUTPUT: Tue March 15 15:21:59 2005 J Patient Saf Volume 1, Number 1, March 2005 Implementing and Validating a Comprehensive Unit-Based Safety Program TABLE 3. Percent of Respondents Who Agree With Questions in Safety Attitude Survey Overall % Agree Pre-safety Program is WICU (N = 66) Overall % Agree Post-safety Program in WICU (N = 64) Relative % Increase Pre- Versus Post-safety Program in WICU Overall % Agree Pre-safety Program in SICU (N = 23) Overall % Agree Post-safety Program in SICU (N = 21) Relative % Increase Pre- Versus Post-safety Program in SICU 1. The senior leaders in my hospital listen to me and care about my concerns. 43 58 35 52 67 29 2. The physician and nurse leaders in my area listen to me and care about my concerns. 50 71 42 43 95 121 3. My suggestions about safety would be acted upon if I expressed them to management. 40 61 53 57 86 51 4. Management/Leadership will never compromise safety concerns for productivity. 52 67 29 65 67 3 5. I am encouraged by my supervisors and coworkers to report any unsafe conditions I observe. 56 76 36 83 95 14 6. I know the proper channels to report my safety concerns. 56 83 48 74 90 22 7. I am satisfied with availability of clinical leadership (MD, RN, RPh). 47 76 64 83 95 14 8. Leadership is driving us to be a safety-centered institution. 46 72 57 52 81 56 9. I am aware that patient safety has become a major area for improvement in my institution. 63 89 41 65 95 46 10. I believe that most adverse events occur as a result of multiple system failures, and are not attributable to one individual s actions. 62 87 40 70 86 23 P, 0.05 using x 2 test for differences between period one and period two. the WICU LOS decreased to 1 day (P, 0.05) (Table 6). In the SICU, the LOS was 3 days in the pre-intervention and control periods and 2.3 days in the post-intervention period (P, 0.05) (Table 6). FIGURE 1. The percent of staff in Weinberg ICU and surgical ICU whose attitudes about safety (safety culture) were more positive in the post-cusp intervention survey compared with pre-intervention. DISCUSSION We described the implementation and validation of a comprehensive unit-based safety program designed to document measurable improvements in patient safety and the concomitant improvement in local perceptions of patient safety. The CUSP was associated with an improved safety climate and reduced LOS, medication errors, and potentially nursing turnover. Moreover, we were able to replicate these results in a second ICU. While much discussion has ensued regarding patient safety, there are few examples where measurable improvements have been documented. In this study, we present a novel safety program that can be implemented via a cascade-effect throughout an organization and is associated with significant improvements in patient safety. The CUSP s unit-based focus rather than organization-wide change is a more realistic and manageable approach when initiating cultural change in a large organization. The association between reduced nursing turnover in both ICUs and improved culture of safety suggests that the behaviors required to be a safe organization are similar to the behaviors needed to improve employee morale, further supporting the business case for safety. For both safety and q 2005 Lippincott Williams & Wilkins 37

JOBNAME: jops 1#1 2005 PAGE: 6 OUTPUT: Tue March 15 15:22:02 2005 Pronovost et al J Patient Saf Volume 1, Number 1, March 2005 TABLE 4. Staff Safety Concerns and Recommended Improvements Identified Through Safety Survey Identified Concern Recommended Improvements Lack of standardized concentrations of vasopressors among ICUs and operating room Need to standardize concentration and labels for vasopressor Lack of trained team to transport patients from ICUs to remote testing sites Use dedicated transport team Risk of medication errors Point of care pharmacist available on units Poor management of pain Create guideline or protocol for pain assessment and management Poor communication among ICU providers Create short-term goals sheet Poor communication during ICU discharge leading to medication errors in transfer orders Implement medication reconciliation process at ICU discharge Epidural and intravenous catheters too similar in appearance Make identification of catheters clear morale, staff needs to be part of a team in which they feel empowered and engaged. The CUSP evolved from our perception that many efforts to improve safety were fragmented, lacked data regarding local culture, were unable to document improvements, and did not engage work unit staff. We sought to develop a program that integrated many aspects of patient safety, could be implemented at the work unit level, had empirical results, and would produce a cascade-effect throughout the JHH. One advantage of CUSP is its empowerment of frontline staff to assume responsibility for patient safety by generating issues, prioritizing them, and implementing them according to local needs. In addition, this program provides simple tools to improve the culture of safety, a common metric to evaluate the culture of safety, a standard approach to improvement, and a system to disseminate results within the organization. CUSP evolved through the lessons we learned during these efforts. First, we learned to keep the tools simple. Second, we learned that ongoing discussions between senior executives and work unit staff is imperative. 27 Third, we learned that staff needs time to work on improvement efforts if the CUSP is to succeed. We ask each team member to devote between 4 to 8 hours per week to our program. We do not know why culture improvements in the SICU were greater than those in the WICU. It may be the result of a learning curve regarding how best to implement the program the SICU followed the WICU, or the early improvements in culture in the WICU influenced the SICU. We recognize several limitations to our program. First, we have only implemented the program in ICUs and need to evaluate the replicability of this program in other work units. Second, we implemented the program in an academic medical center. It is unclear how this program would work in community hospitals. Third, we do not know the full impact of the program yet. We do not have a wide sampling of the incidence of adverse events and have not evaluated the long-term impact of the program. Also, we cannot make causal statements about the relationship between improvements in safety climate and outcomes. Fourth, the instrument used to measure the culture of safety could be improved and expanded. Sexton and colleagues 11 have developed a survey instrument that evaluates job satisfaction, perceptions of management, teamwork climate, safety climate, stress recognition, and working conditions, thus providing keener insights into the organization s culture of safety. Despite these limitations, we believe that the safety improvements we have documented are significant and important to patients, families, and staff. To our knowledge, this is the first study to document that climate of safety can be improved in healthcare. 28 The safety climate at JHH compares favorably to aviation. 5 CONCLUSIONS We have validated CUSP in an ICU and replicated improvements in patient safety in a second ICU at an academic medical center, and demonstrated the validity through a variety of outcomes. As a result, we improved staff s perceptions about patient safety and reduced ICU LOS, medication errors, and potentially nurse turnover. These findings suggest that healthcare organizations can improve patient safety and provide evidence for the business case for safety. In closing, TABLE 5. Action Items That Have Been or are Being Implemented Required additional resources Identified Opportunity Dedicated patient transport team in order to improve safety of transports Point of care pharmacist: PharmD assigned to unit to facilitate medication ordering and distribution. Action Taken Team initiated for ICU patients, but not for all shifts Funded and implemented in ICUs Minimal additional resources Goal sheets: Clarify short-term goals for patients in ICUs Reconciliation: Reconciling pre-hospital, pre-operative, and ICU medications before transfer from ICU Labeling epidural catheters: Risk of medication errors if epidural catheter is connected to intravenous catheters Short-term goal sheet part of daily rounds Medication reconciliation part of routine nursing discharge; Audit of sample of charts is ongoing Anesthesiologists in OR place orange sticker on epidural catheters; ICU nurses also have these stickers 38 q 2005 Lippincott Williams & Wilkins

JOBNAME: jops 1#1 2005 PAGE: 7 OUTPUT: Tue March 15 15:22:04 2005 J Patient Saf Volume 1, Number 1, March 2005 Implementing and Validating a Comprehensive Unit-Based Safety Program TABLE 6. Impact of CUSP on ICU LOS and Nursing Turnover WICU SICU Pre Intervention Post Intervention Pre Intervention Control Post Intervention # of ICU admissions 674 641 868 749 751 ICU LOS (SD) 2.2 (5.1) 1.1 (4.1) 3.1 (5.0) 3.0 (5.2) 2.3 (3.6) Nursing turnover 9% 2% 8% 9% 2% P, 0.05 with t test comparing pre versus post groups. Changes in nursing turnover were not statistically significant. we recently truncated CUSP into six steps with the development of a new tool to investigate safety defects. We believe this shortened version will fit easily into staff s daily routines. ACKNOWLEDGMENTS Partial funding for research activities provided by the Agency for Healthcare Research and Quality (grant #U18HS11902-01). REFERENCES 1. Kohn L, Corrigan J, Donaldson M. Institute of Medicine Report. To Err Is Human: Building a Safer Health System. Washington, DC: National Academy Press; 2000. 2. Reason J. Human error: models and management. BMJ. 2000;320:786 770. 3. Schein E. Organizational culture. Am Psychol. 1990;45:109 119. 4. Helmreich RL, Merrrit AC. Culture at work in aviation and medicine: National, organizational, and professional influences. Aldershot, UK: Ashgate Publishing Limited; 1998; 176. 5. Sexton JB, Klinect JR. The link between safety attitudes and observed performance in flight operations. Proceedings of the Eleventh International Symposium on Aviation Psychology. Columbus, OH: The Ohio State University; 2001. 6. Bagian J, Lee C, Gosbee J, et al. Developing and deploying a patient safety program in a large health care delivery system: you can t fix what you donõt know about. Jt Comm J Qual Improv. 2001;27:522 532. 7. Donchin Y, Gopher D, Olin M, et al. A look into the nature and causes of human errors in the intensive care unit. Crit Care Med. 1995;23:294 300. 8. Andrews L, Stocking C, Krizek T, et al. An alternative strategy for studying adverse events in medical care. Lancet. 1997;349:309 313. 9. Sexton JB, Helmreich RL, Williams R, et al. The Flight Management Attitudes Safety Survey (FMASS). Research Project Technical Report 01-01. Austin, TX: The University of Texas; 2001. 10. Helmreich R. Cockpit management attitudes. Hum Factors. 1984;26:583 589. 11. Sexton J, Helmreich R, Rowan K, et al. The Safety Attitudes Questionnaire: A psychometric validation. Health Services Research. (under review) 12. Helmreich R, Foushee H, Benson R, et al. Cockpit resource management: exploring the attitude performance linkage. Aviat Space Environ Med. 1986;57:1198 1200. 13. Itoh K, Andersen HB. Motivation and morale of night train drivers correlated with accident rates. Proceedings of CAES: International Conference on Computer-Aided Ergonomics and Safety. Barcelona, Spain; 1999: 5-19-0099. 14. Irwin C. The impact of initial and recurrent cockpit resource management training on attitudes. Proceedings of the Sixth International Symposium on Aviation Psychology. Columbus, OH: The Ohio State University. 1991: 344 349. 15. Salas E, Fowlkes J, Stout R, et al. Does CRM training enhance teamwork skills in the cockpit?: Two evaluation studies. Hum Factors. 1999;41: 326 343. 16. Helmreich R, Wilhelm J. Outcomes of crew resource management training. Int J Aviat Psychol. 1991;1:287 300. 17. Gregorich S, Helmreich R, Wilhelm J. The structure of cockpit management attitudes. J Appl Psychol. 1990;75:682 690. 18. Sexton J, Helmreich R, Thomas E. Error, stress and teamwork in medicine and aviation: Cross sectional surveys. BMJ. 2000;320:745 749. 19. Thomas E, Sexton J, Helmreich R. Discrepant attitudes about teamwork among critical care nurses and physicians. Crit Care Med. 2003;32:956 959. 20. Helmreich R, Sexton J. Group interaction under threat and high workload. In: Dietrich R, Childress T, eds. Group Interaction in High Risk Environments. Aldershot, UK: Ashgate; 2004. 21. Pronovost P, Weast B, Bishop K, et al. The senior executive adopt a work unit. Jt Comm J Qual Saf. 2004;30:59 68. 22. Reason J. Managing the Risks of Organizational Accidents. Burlington, VT: Ashgate Publishing Company; 2000. 23. Ovretveit J, Bate P, Cleary P, et al. Quality collaboratives: lessons from research. Qual Saf Health Care. 2002;11:345 351. 24. Pronovost P, Berenholtz S, Dorman T, et al. Improving communication in the ICU using daily goals. J Crit Care. 2003;18:71 75. 25. Pronovost PJ, Weast B, Schwarz M, et al. Medication reconciliation: a practical tool to reduce the risk for medication errors. J Crit Care. 2003;18:201 205. 26. Pronovost P, Hobson D, Earsing K, et al. A practical tool to reduce medication errors during patient transfer from an intensive care unit. J Clin Outcomes Mgmt. 2004;11:1 6. 27. Pronovost P, Weast B, Bishop K, et al. Patient safety, senior executive adopt-a-work unit: a model for safety improvement. Jt Comm J Qual Saf. 2004;30:59 68. 28. Pronovost P, Weast B, Holzmueller C, et al. Evaluation of the culture of safety: survey of clinicians and managers in an academic medical center. Qual Saf Health Care. 2003;12:405 410. q 2005 Lippincott Williams & Wilkins 39

JOBNAME: jops 1#1 2005 PAGE: 8 OUTPUT: Tue March 15 15:22:05 2005 Pronovost et al J Patient Saf Volume 1, Number 1, March 2005 APPENDIX 1. Cultural Survey Questionnaire We are conducting a survey to evaluate the culture of safety in your unit. The survey will take approximately three minutes to complete. Please leave the completed survey in the survey mailbox or designated point of contact from your unit. Role (circle one): Attending/Fellow Physician / Resident Physician / Nurse / Respiratory Therapist / Support Associate / Other (please list): Unit (please write in title and/or location): Date: Please circle one answer per question 1. The senior leaders in my hospital listen to me and care about my concerns. 2. The physician and nurse leaders in my area listen to me and care about my concerns. 3. My suggestions about safety would be acted upon if I expressed them to management. 4. Management/Leadership will never compromise safety concerns for productivity. 5. I am encouraged by my supervisors and coworkers to report any unsafe conditions I observe. 6. I know the proper channels to report my safety concerns. 7. I am satisfied with availability of clinical leadership (MD, RN, RPh). 8. Leadership is driving us to be a safety-centered institution. 9. I am aware that patient safety has become a major area for improvement in my institution. 10. I believe that most adverse events occur as a result of multiple system failures, and are not attributable to one individual s actions. Thank you for engaging in patient safety! Quantum Leaps in Patient Safety Institute for Healthcare Improvement Modified from Brian Sexton/Robert Helmreich Aviation Cultural Survey APPENDIX 2. JHCUSP Staff Safety Survey Name: Role: Date: Unit: Please describe how you prevented a patient from being harmed. Please describe how the next patient in your work area will be harmed. Please describe how we can prevent this harm by the following: Preventing the mistake that lead to harm Making the mistake visible Reducing the harm should it occur 40 q 2005 Lippincott Williams & Wilkins