| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;83:1412-1419
© 2007 The Society of Thoracic Surgeons
Division of Cardiovascular Surgery, Mayo Clinic, Rochester, Minnesota
Accepted for publication November 1, 2006.
* Address correspondence to Dr Wiegmann, Mayo Clinic, 200 First Street SW, Rochester, MN 55905 (Email: wiegmann.douglas{at}mayo.edu).
Presented at the Fifty-third Annual Meeting of the Southern Thoracic Surgical Association, Tucson, AZ, Nov 8–11, 2006.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionFootnotesReferences |
|---|
Methods: As a first step, we utilized a model previously employed in aviation to develop structured interviews of individuals in multiple roles (surgeons and allied health staff). To enhance relevance to the OR, Likert scale questions were formulated based on published sentinel event analyses and focus group studies in which specific factors found to be causally related to error in health care were described. Additional items from other high risk-consequence industries were generated to address theoretically important factors not highlighted previously.
Results: Application of the modified model to the interview responses allowed the identification of factors impacting performance in the OR and estimation of their relative importance. Analysis of correlations among responses were consistent with predictions of the model that the origin of errors can be traced to organizational influences that impact supervisory processes, which in turn establish preconditions predisposing to errors.
Conclusions: These data demonstrate a model of error causation derived from aviation can be modified and applied to the cardiac surgery OR. This tool may prove useful in identifying systemic factors impacting human performance and patient safety.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionFootnotesReferences |
|---|
Advances in technology will no doubt favorably impact the occurrence of errors to an extent. As demonstrated in commercial and military aviation, however, significant further reduction in accidents and error has been accomplished by an appreciation of the important role that human factors play in safety [3]. The same can be expected to be true in medicine.
We have instituted a program of human factors science within the Division of Cardiovascular Surgery at Mayo Clinic to systematically study the complex human factor issues that play a role in error causation and error management in the operating room (OR). As a first step, we utilized a model previously employed in aviation to develop structured interviews of individuals in multiple roles (surgeons and allied health staff). The purpose of this study was to determine the transferability of a human factors model developed in the aviation industry to the OR, and to receive feedback from OR staff about the relative importance of those human factors that have been highlighted previously in other domains. Results will serve to focus future empiric investigations into critical factors that impact patient safety in the OR, so that evidence-based interventions can be developed. This is the first step in a larger program whose goal is to systematically address those factors that impact patient safety in the OR, thereby providing a scientific foundation for addressing those factors that predispose a surgical team to error.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionFootnotesReferences |
|---|
|
|
Many of the questions generated for each category were based on published sentinel event analyses and focus group studies in which specific causes of medical and surgical errors were described. However, additional items were created to address theoretically important factors (eg, organizational climate) not highlighted previously in medicine.
Forty-six questions were administered by one experienced interviewer, and the respondents were asked to answer each question according to a five point Likert scale based on frequency of occurrence (always = 5, very often = 4, sometimes = 3, rarely = 2, never = 1). The length of each interview was approximately 30 minutes. Respondents were also allowed to comment on specific situations that applied to the questions being asked. The 46 questions were divided into four main categories and 17 subcategories according to HFACS (as shown in Table 1). Additionally, 11 open-ended questions were asked regarding the personality and leadership qualities of all members of the OR to evaluate any potential conflicts that may exist. These scores were compiled and the means were compared among different specialty groups, ages, and years of experience. To address the issue that members of the OR team would not be the best judges of their own performance [5], approximately 50% of all individuals in all roles involved with patient care in the OR were interviewed about the spectrum of human factor deficiencies in the cardiovascular surgery ORs at the Mayo Clinic (surgeons, surgical assistants [SA], certified surgical technologists [CST], residents-fellows, anesthesiologists-certified registered nurse anesthetists [CRNA], perfusionists, monitor technicians, and nurses [RN]). Approval for this protocol was obtained from the Institutional Review Board, and the requirement for written informed consent was waived due to the fact that participation was voluntary and no personal identifying information from the individual respondent was recorded.
Statistical Analysis
Composite scores were created for each of the 17 subcategories by averaging the respondents’ ratings to the individual items within each subcategory. Responses to the questions were initially examined and categorized by demographic and professional factors; however, because there are no statistical differences between any of these factors, the results are presented across all respondents. Pearson correlation coefficients were calculated to identify relationships between response subcategories. Given the large number of correlations, we controlled for type I error by considering significant R values of 0.33 or greater, thereby producing an r2 value of 0.10 or greater and a p value of 0.01 or less.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionFootnotesReferences |
|---|
The mean composite scores for the 68 participants are shown in Figure 2. Each of the four sections represent the four main human factor levels according to the HFACS model, and bars within each human factor level represent the subcategory (17 total) composite score. Analyses of the results are examined at each level beginning with unsafe acts and proceeding to the organizational influences level. For clarity purposes, results in each level will be summarized separately and correlations across adjacent subcategories will be presented.
|
Preconditions for Unsafe Acts
In section B of Figure 2, the responses are shown for the level of preconditions for unsafe acts. Preconditions to unsafe acts as a whole were not excessively frequent (overall composite score of 2.48); however, some specific preconditions were perceived as occurring often. Teamwork issues (crew resource management; CRM) (2.66) predominated in this category. Specific questions revealed that inappropriate communication was thought to occur most often (3.07); these issues mainly involved surgeon communication with other team members. Other issues revolved around the surgeon’s inability to effectively lead in the OR. When asked about surgeon’s communication effectiveness in the OR, 22 (32%) stated that communication is effective, 17 (25%) stated that communication is variable from surgeon to surgeon, ranging from good to effective to poor, 14 (21%) stated communication is poor, 7 (10%) that communication patterns between surgeons and other staff members is intimidating, and 4 (6%) refused to answer the question. When asked in an open-ended manner to characterize surgeon personality and its impact in the OR environment, 24 (35%) individuals stated that surgeons are arrogant or demeaning, negatively affecting team function, and 16 (24%) stated that surgeons have a strong or intense personality, also negatively affecting team function. Fifteen respondents (22%) felt that surgeon personality was variable among surgeons, ranging from arrogant or demeaning to strong-to-intense to lackadaisical. Eight individuals (12%) stated that surgeons behave in a professional manner positively affecting teamwork. Surgeons were characterized as intelligent and very skilled by 4 (6%) respondents and 1 (1%) described surgeons as considerate or caring.
As shown in Table 2 and Figure 3, preconditions for unsafe acts had a number of correlations with unsafe acts. As the physical-mental status of an individual deteriorates, the tendency to commit skill-based errors, decision-based errors, routine violations, and exceptional violations were thought to increase. Physiologic compromise of an individual by medical illness, medication side effects, or other means was thought to contribute to committing skill-based errors and decision-based errors. Compromise of teamwork was also thought to make individual members more prone to decision-based errors, skill-based errors, perceptual errors, and routine violations.
|
|
Interpersonal conflicts, which constituted the biggest problem area in inappropriate planning, occur when OR team members are broken off from the team with whom they regularly work. Conflicts originate when surgeons have expectations of these staff members regarding intricacies of OR decorum that vary from surgeon to surgeon. When conflicts do occur because of misunderstandings regarding surgeon preferences, frustration on behalf of the surgeon and OR team member occur. Additionally, occasions exist when poor chemistry occurs during the initial contact between a team member and a surgeon, tainting any further operative experience they may share together in the future.
As shown in Table 2 and Figure 3, statistical analysis revealed that deficiencies in the ability of management to perform adequate planning (management and assignment of work including risk management, crew pairing, and operational tempo) negatively affect physical-mental status, personal readiness, and teamwork. When there are deficiencies in problem correction (deficiencies among individuals, equipment, training, or other areas that are "known" to the supervisor yet are allowed to continue uncorrected) physical-mental status, personal readiness, and teamwork were perceived as being negatively effected.
Organizational Influences
In section D of Figure 2, organizational influences received an overall score of 2.43. Failures in resource management (2.55) and organizational process (2.49) occurred more frequently than organizational climate (1.94). Issues with resource management were thought to be due to staff turnover (as described previously), a lack of appropriate staff, and a lack of staff with proper education or training. Organizational process issues revolved around requirements of team members, OR times taking longer than expected, and the effect of this increase in operative time on OR teams.
As shown in Table 2 and Figure 3, statistical analysis revealed that the perception of deficiencies in organizational process (the process by which the vision of an organization is carried out including operations, procedures, and oversight) was found to negatively affect the ability of management to correct known problems (problem correction) and plan effectively (the ability to manage and assign work). Deficiencies in organizational resource management (how human, monetary, and equipment resources carry out the vision of the organization) were believed to negatively affect the ability of management to correct known problems (problem correction) and manage personnel and resources (supervision).
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionFootnotesReferences |
|---|
With regard to teamwork issues, 59% of respondents indicated that a surgeon’s personality negatively affects the way members of the OR team function. Furthermore, when teamwork issues were examined within the scope of human performance globally, these issues were associated with other human factors such as skill-based errors, decision-based errors, perceptual errors, and routine violations. Despite this acknowledgement of the important role that teamwork plays in the OR, the actual reported frequency of these failures was not great (2.66). These results indicate that although the frequency of failures may be relatively low, the impact these failures have are significant; thus, improvement in OR performance may be expected by addressing these issues. Many institutions have attempted to address this issue of teamwork by sending surgeons to charm school in order to improve their attitude in the OR; however, the exact role of teamwork in error causation is unclear, emphasizing the importance of additional research in this area before implementing interventions that have no scientific basis.
de Level and colleagues [6, 7] were the first to apply the human factors theory to the OR in their study of the arterial switch operation. Their study demonstrated that many events predisposing to error have their origin in organizational and supervisory structure (latent failures), and that investing in an understanding of the latent failures underlying active failures is the most cost efficient way to improve safety in health care. Our study is consistent and reaffirms the observations of de Leval and colleagues. Furthermore, this study has gone one step further by applying a methodology to examine the effect that these latent failures have in the OR environment. By understanding the associations that exist within the multiple levels affecting performance in the OR, we have completed the first crucial step in focusing on those aspects of an organization where applied interventions may yield the greatest reduction in surgical errors.
Prior work has demonstrated that seemingly trivial problems can accumulate to influence the outcome of a case, even those that go unnoticed [6]. Prior to this study, however, there have been few attempts to apply a structured method of analysis to the identification and classification of the myriad factors that affect performance in the OR. The results of our study demonstrate the applicability of HFACS to the OR environment and its role in understanding the effect of human factors on human performance.
While our use of a structured interview may be called into question by the previous findings of de Level and colleagues [6], who found brief postoperative questionnaires of little value in identifying errors in surgical cases, our interviews were broader and concerned with the dynamics of the OR in general, for the purpose of establishing the applicability of a human factors model constructed in another domain to surgical care. Of note, Carthey and colleagues [5] proposed that a panel of medical experts from different specialties should be involved in categorizing the active and latent failures. In attempting to adapt the HFACS to the OR, we therefore ensured interviews of a representative distribution of OR team members across all roles.
Limitations
The principle limitation of this study is its interview format rather than empiric observation of error. As such, the validity of associations identified is dependent on individual perception of respondents. This study does, however, provide a starting point for making such primary observations, and is a necessary first step in applying a model developed in aviation to medicine. Furthermore, interviews were conducted in only one institution and one specialty. The applicability and universality of these data to other surgical and medical fields can only be determined with further research. Despite the limitations of structured interviews (versus objective data recorded by trained observers), we found many associations that have been well-defined in other high risk domains. Li and Harris [8] recently used HFACS to analyze 523 aviation accidents that occurred in China during a 25 year period. Of the 26 statistically significant correlations we identified in this study, 12 were found to be identical to the correlations described by Li and Harris.
The findings here, and the consistency of our data with literature from both the medical and human factor fields, indicate that we successfully adapted and applied HFACS to the OR. These findings are promising because they provide a methodology to reduce errors in the OR by addressing issues that plague the surgical field. In order to truly understand the interplay between the human factors identified, we recommend objective assessment of OR performance by trained human factors observers using the model described here to quantify the nature and effect of error.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionFootnotesReferences |
|---|
I would like to also mention that Andrew ElBardissi is a medical student. His presentation is extraordinary for a person at that level of training, and what is even more extraordinary is the support that he has received from the surgeon mentors at the Mayo Clinic who have supported him in that. I would also like to call attention to the fact that they have the wisdom to include an extraordinary human factors scientist, Dr Wiegmann, who is before us now, as a member of their team.
We had a similar amazing experience when I was on the Dartmouth faculty at Concord, New Hampshire, of working very closely with Jeff Brown, a human factors scientist, and what he was able to teach us about what we take for granted, but shouldn’t, was extraordinary.
I had the privilege when I was a resident of having a lunch at the AATS with Harvey Bender, and all of us sat down with him and he was giving us pearls, and they consisted of basically two. One doesn’t really apply here but it was kind of fun anyway, which was that I wish the heck you guys would understand that air rises and blood sinks. That was the first thing. And the second one, though, that has stayed with me my entire career is that if you set things up so that they can happen, they eventually will. And that surgical principle that we all know as surgeons, that if we set things up in a certain way so that they can happen it will eventually happen to us, and the converse is that if we set things up in a different, safer way, we will get better outcomes is really what this paper is all about.
To illustrate this briefly, I would like each surgeon just to reflect on things that you know very well, two circumstances. One is a really great day in the OR and the other is a really bad day in the OR. Let’s think about that really great day. You come in, everybody knows exactly what is going to happen, the patient is there on time, lines and tubes just go right in, the case goes smoothly, your team knows exactly what you need, you hardly even need to talk, all the supplies that you need are right there, anesthesia is perfect, and the case is just fantastic and you are out of the room by 10:30. A really bad day is just the opposite. They have happened to not have the scrub nurse that you know, they have substituted somebody who came from OB/GYN, the anesthesiologist you have not ever seen before, the perfusionist is just new, rotated in from somewhere else, and you are struggling, and you finally get the case done but you go out of there just shaking your head. And the question that I have for all of us is, are our skills any different on that good day or the bad day, or is that good day or bad day because of something else? And what I would suggest to you is that human factors science and safety science suggests that our skills are pretty much the same on good days and bad days, that it is usually not the people, it is the context that we are working in. And traditionally we have not paid attention to our context, and what this group from the Mayo Clinic is doing is trying to provide very rigorous scientific evaluation of the context that we do work in.
And so I would like to just conclude with one comment and then two questions. The first is just to emphasize that what this study shows us is that rigorous study of how we give care is as important as study of what care to give. And the questions that I have for you, one is, what changes have you made in your operating room environment based on the studies that you have done so far? And the second is, what advice could you give the rest of us as we go about our journey of trying to step outside of our clinical practice culture and see it clearly afresh so that we can hopefully lead and transform it?
Thank you again for this outstanding presentation and the privilege of discussing it.
DR WIEGMANN: Thank you very much. Just to address a couple of questions that were raised, particularly the question regarding any changes that we have made. First, I would like to note that we have only just recently embarked on this endeavor. One of the things that a human factors professional will repeatedly emphasize is that in order for change to be effective, the fundamental nature of the problem must be understood and a clear set of goals needs to be established. We are still in the process of identifying what the issues are in a systematic way. We don’t want to simply implement "feel good" fixes that we hope will have an impact but we’re not really sure, because we haven’t taken the time to collect the data. We are optimistic about some of our ideas about how to make systemic changes in the OR. However, before we make any major changes we are definitely going to roll up our sleeves and do the hard work, which is collecting the data up front. We need to be able to prioritize and know we are addressing our most important problems. We also want to be able to validate whether or not our interventions have had any impact, or whether we need to do things differently.
In terms of advice for other people embarking in this area, I think that the comments that Dr Uhlig made were very insightful. We need to embrace the idea that skill doesn’t necessarily change on any given day. Rather, it is the contextual or systemic factors that very often impact a surgeon’s ability to maintain and achieve surgical excellence. We are not talking about turning bad surgeons into good surgeons. Rather, we are talking allowing great surgeons to achieve excellence. However, it can take considerable effort to refine your approach in the OR, or to do things a little bit differently with your surgical team. Nonetheless, these efforts will pay off.
DR ROBERT J. CERFOLIO (Birmingham, AL): First, I want to congratulate you on the excellent presentation and thank my good friends Joe Dearani and Thor Sundt and the Mayo Clinic for bringing this important topic to the STSA. I think some of you saw this paper and said "Oh, this is one of these non-scientific talks, I can skip this one and get a quick snack and take a break." Why did the Program Committee put this on? The reason we put it on this year’s program is because it represents the reality of the world we practice medicine in, and we need to be the ones who generate the data that affect our working ambiance. The reason we have a timeout before we make any incisions in our hospital, and probably everywhere else in the United States, is because of this type of data. I hear many of us complain that "we have to send somebody out of the operating room to go mark the next patient to make sure we open the correct side." Again this rule was promulgated from data generated from reports like this one. So I believe we need to be the ones generating this type of data. So my specific question to you is, since we are trying to study this exact issue critically at UAB—on the floors not in the OR—is what specific ways, what specific tools, not just surveys, can I implement to objectively study and generate data on this topic.
Thank you again.
DR WIEGMANN: I would agree. And just to mention that this is the first study in multiple studies that we have been conducting at Mayo Clinic in cardiac surgery. We started with this initial structured interview to get an idea of what some of the issues are and to see to what extent these human factors concepts from other domains would transfer into cardiac surgery. We have since done objective observations. We have observers in our operating rooms now documenting surgical flow disruptions and distractions that impact the team’s ability to concentrate fully on the case. In one operating room, we have also installed video cameras so that we can record our cases and systematically review them as an interdisciplinary team. There is also a lot of other patient and outcome data that we need to collect and correlate with the observations that we are making. Collecting and analyzing these additional empirical data is an important next step in our efforts to move forward.
DR WILLIAM A. GAY (St. Louis, MO): As I looked through the program and saw this paper on the program, I applauded in my reverie that I think it is outstanding to have a paper like this as a part of a scientific program. I just want to call your attention to that newspaper article that was the first slide in the presentation. There are a lot of people who get injured and hurt in hospitals. Hospitals are dangerous places to be. So what we are talking about here is not only true for the operating room, we are talking about safety measures for the hospital in general. And what you have heard today, if you take this message back home and implement some changes, that will probably save more lives than anything else you will hear at this meeting. I will guarantee you that there are more lives lost to medical mistakes and errors that occur in hospitals than there are to the diseases that we are treating. So just ponder that for a little while.
I would like to congratulate the authors. I think this was an outstanding and timely paper, and paying attention to things like this is going to become part of every physician’s armamentarium for the future. Thank you very much. I think this was very well done.
DR WIEGMANN: Thank you.
|
Footnotes |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionFootnotesReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionFootnotesReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. J. Pronovost, C. A. Goeschel, J. A. Marsteller, J. B. Sexton, J. C. Pham, and S. M. Berenholtz Framework for Patient Safety Research and Improvement Circulation, January 20, 2009; 119(2): 330 - 337. [Full Text] [PDF]
|
|
|
|
 |
|
 P. Barach, J. K. Johnson, A. Ahmad, C. Galvan, A. Bognar, R. Duncan, J. P. Starr, and E. A. Bacha A prospective observational study of human factors, adverse events, and patient outcomes in surgery for pediatric cardiac disease. J. Thorac. Cardiovasc. Surg., December 1, 2008; 136(6): 1422 - 1428. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Shuhaiber, K. Goldsmith, and S. A.M. Nashef Impact of Cardiothoracic Resident Turnover on Mortality After Cardiac Surgery: A Dynamic Human Factor Ann. Thorac. Surg., July 1, 2008; 86(1): 123 - 131. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
