2000-2001 Grant Awards
James S. Todd Memorial Award for Patient Safety Research
Diagnostic Errors in Internal Medicine
Mark Graber, MD (PI), Nancy Franklin, PhD, Gail Holtz, MS, RN, and Stefanie Marcin
This project sought to better understand the root causes of diagnostic errors encountered in internal medicine. To conduct the study, one hundred cases of diagnostic error were identified from autopsy discrepancies, quality assurance activities, and voluntary reports. Contributing factors were identified by root cause analysis with the help of a cognitive psychologist. We modified several existing taxonomies to develop one that was useful in categorizing the contributing factors, grouped under System-Related, Cognitive, or No Fault categories.
Seven cases exclusively involved “no fault” errors. Of the other 93 cases, 548 different system- or cognition-related errors were found, with an average of 6 system or cognition related errors per case. System errors contributed to the diagnostic error in 65% of the 100 cases; cognitive errors contributed in 74% of the errors. The most common system-related errors involved problems with policies and procedures, inefficient processes, teamwork, and communication. The most common cognitive problems involved faulty synthesis of the available information, most often related to premature closure, context errors, anchoring, and other issues. Faulty or inadequate knowledge was the least common error.
The study contributed one of the first good working definitions of diagnostic error and a useful taxonomy. The finding that most errors involve both system and cognitive elements was somewhat of a surprise, and points out the need for healthcare institutions to focus on this problem as a systems issue. The detailed analysis of each case, including interviews with the responsible staff, also offered unique insights into the origins of
these errors, details that are typically lacking in many retrospective error reviews.
This work, completed in 2004, and further discussions of diagnostic error have also served to bring together a growing number of clinicians and other stakeholders concerned about this problem. An outgrowth of this has been an annual international meeting dedicated to the topic of diagnostic error, beginning in 2008. Details about the upcoming and past meetings can be found at http://www.smdm.org/diagnostic_errors.shtml.
The Impact of Dedicated Medication Nurses Upon the Rate of Medication Administration Errors: A Randomized, Controlled Trial
Nancy Lynn Greengold, MD, MBA (PI), Rita Shane, PharmD, Linda Burnes-Bolton, DrPH, RN, and Philip Schneider, RPh, MS
The goal of this study was to determine whether the provision of a dedicated, specially trained medication nurse would reduce the number of medication administration errors in the hospital setting. The study was performed at two large hospitals. At both institutions combined, medication nurses had a 15.7% error rate compared to 14.9% for general nurses (P
Dr. Greengold is now heading up a commercial business unit called the Zynx Health Device Network, which has just launched a new patient-safety solution called Prime-A-Pump. This customizable Web-based program offers clinical decision support to help hospitals build the “drug libraries” that are used to configure “smart” infusion pumps. Zynx provides the evidence-based and expert-based recommendations for how to set the soft and hard minimum and maximum limits/alerts in these pumps to try to prevent under- and over-dosing errors at the nurse administration stage of medication delivery.
Understanding Errors in Emergency Departments: A Convergence Approach
Robert Wears, MD, MS (PI), Kathleen M. Sutcliffe, PhD, and Shawna J. Perry, MD.
This study sought to develop an understanding of the nature and mechanisms of errors in Emergency Departments (ED). Building on previous research that addressed the type and scope of errors in ED, this research team looked to answer “why” and “how” those failures occur. The study was carried out in two academic emergency department sites at the investigators´ home institutions, the University of Florida (UF) and the University of Michigan (UM).
Four different methods of data collection were used: direct ethnographic observations of both routine ED work and critical incidents or adverse events; voluntary, anonymous reporting of incidents and unsafe conditions; semi-structured interviews of ED staff at all levels; and a survey of safety-related attitudes. The final analysis was undertaken after all the observational efforts were completed. This involved both qualitative and quantitative analytic methods. This study has led to several publications as well as several national and international presentations. Through this work, the research team determined that implications for patient safety in ED care can be expressed as a series of dilemmas, or tradeoffs that must be negotiated. The abundant negotiation at many levels gives an opportunity to develop methods of inter-relating more heedfully—or it could result in a normalization of deviance, as parties implicitly agree to ignore aberrancies and drift toward the limits of safe practice. Similarly, enriching communication channels could improve the transfer of information and the development of shared mental models— or it could add to the information overload already experienced by caregivers. Finally, the demand/resource mismatch seems to be an opportune target for improvement—but risks becoming the main focus of attention, or a self-serving suboptimization.
Perhaps the most important outcome from the project has been the improvement in the team’s ability to attract additional research funding to continue investigation in this area. The information gained through this process will improve basic understanding of how success and failure occur in the ED. An immediately applicable outcome will be an extension of existing frameworks for analysis of critical events.
Our initial NPSF project, aimed at understanding how clinicians make sense of and manage risk in the emergency department, was a transformational experience: it introduced our team to issues of complexity, sense-making, resilience, and above all the situatedness of work that we had previously been unaware of, and changed the direction of all of our subsequent research efforts. It has led to further funded work on communication, turnovers, problem recognition, and resilience in complex, uncertain, demanding, high-risk settings.
The Surgeons Checklist
R. Scott Jones, MD, FACS (PI), J. Forrest Calland, MD, Reid B. Adams, MD, Don Detmer, MD, Viktor Bovbjerg, PhD, Stephanie Guerlain, PhD, and Owen Seely
Based on the successful use of checklists developed in aviation and anesthesia practices, the investigators of this study developed and evaluated an intra-operative safety checklist to reduce the frequency of procedural variance, adverse events, and errors in the operating room.
The use of the checklist led to a significant improvement in the frequency with which surgical teams conducted a pre-procedure briefing, but decreased the team members’ satisfaction with communication. Participant questionnaires revealed that group situational awareness was not significantly better among checklist participants. Furthermore, there was not a significant difference between intervention and control group surgeons’ technical proficiency. This study brings into question the utility of checklists without supplemental communications training or
debriefing of the operative team.
