Developing a Checklist: Consensus Via a Modified Delphi Technique.

OBJECTIVE: To create a universal checklist of key preparatory steps to aid anesthesiologists in patient separation from cardiopulmonary bypass. DESIGN: Multistep, iterative survey with statistically guided refinement of survey items using a modified Delphi technique. SETTING: Internet-based surveys. PARTICIPANTS: Ninety active members of the Society of Cardiovascular Anesthesiologists volunteered to participate, including geographically distributed private practice and academic physicians. INTERVENTIONS: A series of checklist items was created and distributed to 90 anesthesiologists, who assessed each item's importance in preparing for patient separation from cardiopulmonary bypass and added, deleted, or modified any items as they saw fit. Items meeting a threshold of greater than 90% group acceptance were carried forward to a second survey. These items then were evaluated using a 5-point Likert scale to grade relative importance and then compared with the group's responses, creating a third survey with refined checklist items. The results then were used to generate a final survey based on each item achieving certain predefined statistical criteria, which then were scored again by the participants, generating a final checklist via statistically guided consensus. MEASUREMENTS AND MAIN RESULTS: An initial checklist containing 28 possible items was proposed to the participants. After the iterative process was completed, a final checklist of 10 items deemed essential to prepare for bypass separation was created. CONCLUSIONS: A checklist to aid in bypass separation was created with key steps derived from a statistically driven Delphi process. This technique of iterative consensus building may be useful in developing additional safety checklists.

Checklist usage decreases critical task omissions when training residents to separate from simulated cardiopulmonary bypass.

OBJECTIVE: Separation from cardiopulmonary bypass (CPB) requires multiple preparatory steps, during which mistakes, omissions, and human errors may occur. Checklists have been used extensively in aviation to improve performance of complex, multistep tasks. The aim of this study was to (1) develop a checklist using a modified Delphi process to identify essential steps necessary to prepare for separation from CPB, and (2) compare the frequency of completed items with and without the use of a checklist in simulation. It was hypothesized that the use of a checklist would reduce the number of omissions. DESIGN: High-fidelity simulation study. SETTING: University-affiliated tertiary care facility. PARTICIPANTS: Seven cardiac anesthesiologists created a checklist using a modified Delphi process. Ten residents participated in 4 scenarios separating from CPB in simulation. INTERVENTIONS: Each scenario was performed first without a checklist and then again with a checklist. An observer graded participants' performance. MEASUREMENTS AND MAIN RESULTS: A pre-separation checklist containing 9 tasks was created using the Delphi process. Without using this checklist, 4 tasks were completed in at least 75% of scenarios, and 8 tasks were completed at least 75% of the time when using the checklist. There was a significant improvement in completion of 5 of the 9 items (p< 0.01). CONCLUSIONS: A modified Delphi process can be used to create a checklist of steps in preparing to separate from CPB. Using this checklist during simulation resulted in increased frequency of completing designated tasks in comparison to relying on memory alone. Checklists may reduce omission errors during complex periods of anesthesiologists' perioperative workflow.

Accuracy of noninvasive estimated continuous cardiac output (esCCO) compared to thermodilution cardiac output: a pilot study in cardiac patients.

OBJECTIVE: To compare the noninvasive estimated continuous cardiac output (esCCO), device-derived cardiac output (CO) to simultaneous pulmonary artery catheter (PAC) thermodilution (TD) CO. DESIGN: A prospective study comparing pulse wave transit time (estimated continuous cardiac output, esCCO; Nihon Kohden, Tokyo, Japan) to intermittent TD CO. SETTING: One academic hospital. PARTICIPANTS: Patients presenting for cardiac surgery. INTERVENTIONS: Intraoperative CO measurements at 4 distinct time points (after induction, after sternotomy, after cardiopulmonary bypass, and after chest closure). MEASUREMENTS AND MAIN RESULTS: The study population consisted of American Society of Anesthesiologists (ASA) IV subjects, 27 (77%) males and 8 (23%) females, with a mean age of 64.6 ± 12.2 years. Data points from esCCO and TD were collected simultaneously and means per time point compared using Bland-Altman, Pearson R coefficient, and percent error. Mean TD CO for the study was 5.4 L/min. The Pearson R coefficient, percent error, and bias in L/min were: 0.57, 44%, 0.66 (after induction); 0.54, 51%, 0.88 (after sternotomy); 0.60, 60%, 0.95 (after cardiopulmonary bypass); and 0.57, 60%, 0.75 (after chest closure) respectively. CONCLUSIONS: esCCO is easy to use and provides continuous CO measurements, but has wide limits of agreement and large percentage errors with a consistently positive bias in comparison to TD.

Evaluating the status of "translating research into practice" at a major academic healthcare system.

OBJECTIVES: The aim of this study was to assess the status of translating research findings into practice at a major academic healthcare system in Central Texas. METHODS: We conducted a cross-sectional survey addressing knowledge of and participation in translational research of physicians, residents, nurses and third- and fourth-year medical students in a major academic healthcare system in Central Texas. RESULTS: Out of 508 respondents, 428 (84.3 percent) completed all questions. A total of 68.9 percent of faculty reported having sufficient education and training to conduct research versus 44.4 percent of residents and 35.6 percent of nurses. Fifty-eight percent of faculty, 53 percent of residents and 9 percent of nurses reported current involvement in research activity. A total of 55.6 percent of residents reported that their departments provide them with protected time for research versus 18.4 percent of faculty and 10.3 percent of nurses. In addition, 33.9 percent of nurses reported interest in participating in research but do not know how to start. There were 86.4 percent of faculty, 77.8 percent of residents, and 58 percent of nurses who indicated they were familiar with translational research. However, only 42.7 percent of faculty, 46.7 percent of residents and 35.6 percent of nurses indicated they were aware of any changes in the delivery of care that resulted from research projects. CONCLUSIONS: The study results suggested failure to leverage members of the healthcare team in a systematic process to ensure translation of research findings into practice. Results highlighted the need to merge culture of safety and quality improvement with research while dealing with the daily pressures of patient care.

Using simulation to orient code blue teams to a new hospital facility.

OBJECTIVES: Prompt and successful cardiopulmonary resuscitation during a sudden cardiac arrest can be hindered by multiple variables, ie, ineffective communication, stress, lack of training, and an unfamiliar environment, such as a new hospital facility. The main objective of the study was to use high-fidelity simulations to orient Code Blue Teams (CBTs) to critical events in a new hospital facility. A secondary objective was to elucidate factors that may have contributed to responses by debriefing teams. METHODS: Mock Code Blue exercises using high-fidelity simulation were implemented in real workplace settings to orient CBTs to critical events. We measured arrival time of first responder, crash cart to code site, first six CBT responders, first chest compression, and first electrical shock. After each mock code, participants were debriefed to assess any barriers to effective response and decision making. RESULTS: Twelve mock codes were conducted at different locations of the new facility. Sixty-nine percent of the participants reported that the training was beneficial. The median time of arrival of the first responders was 42 seconds and the first CBT member was 66 seconds. The median time to initiation of chest compressions was 80 seconds, crash cart arrival was 68 seconds, and first electrical shock was 341 seconds. An additional outcome of the study was the identification of facility and systems issues that had the potential to impact patient safety. CONCLUSIONS: Clinical simulation can be effectively used to orient CBTs and identify critical safety issues in a newly constructed healthcare facility.