Peri‐operative Morbidity and Mortality in Horses

The occurrences of adverse events have been acknowledged in equine anesthesia and identified as an important area for development (Hartnack et al. 2013; Johnston et al. 2002). The Confidential Enquiry into Perioperative Equine Fatalities (CEPEF), a prospective observational epidemiological multi‐center study, found a 7‐day post‐anesthesia mortality rate of 0.9% (out of 35 435 non‐colic surgeries, 95% confidence interval 0.8–1.0%) (Johnston et al. 2002). Although the incidence of adverse events in equine anesthesia (other than mortality) is unknown, Johnston et al. (2002) identified several factors in which error could play a role. These include time of day (odds ratio for midnight to 6 a.m. of 7.6; 95% confidence interval (CI) 2.2–26.7, odds ratio for 6 p.m. to midnight of 2.2; 95%CI 1.3–3.5) and day of the week (odds ratio of weekend of 1.5; 95%CI 1.0–2.3). The study suggested that fatigue and staffing levels could contribute to the risk associated with out‐of‐hours procedures (see the discussion on unsafe acts in the following text).

Person‐Based Approach

This is the traditional and current model in the majority of veterinary clinics and hospitals, in which the search for the cause of an adverse event begins and ends with the person closest to the event. This is rational in the sense that within complex systems, such as anesthesia, human error accounts for the majority (>70%) of adverse events or near misses (Arbous et al. 2001; Cooper et al. 1984, 2002). The usual outcomes of this approach are to “blame and train” (to blame the individual and enforce additional training) or to “blame and shame” (Pang et al. 2018; Wiegmann and Shappell 2005). This approach conveniently ignores that errors are often made by highly trained, well‐intentioned individuals, and fails to account for system and organizational influences that predispose the final act of human error (Allnutt 2002; Reason 2000, 2005, 2008; Wiegmann and Shappell 2005). As described by Ludders and Macmillan, the human is “the final common pathway for an error, thrust there by a flawed system” (Ludders and McMillan 2017).

Human error is unavoidable; it is therefore better to accept the fallibility of humans and act to minimize the factors that contribute to errors occurring, or the severity of their consequences (Allnutt 2002; Kohn et al. 2000; Reason 2000). Focusing on human error and blame has been likened to swatting mosquitoes: they can be swatted individually (and doing so may be satisfying) but is an ineffective long‐term strategy. To eliminate mosquitoes, their source, the swamp, must be drained. In this analogy, the swamp represents the conditions within a system (“latent conditions,” see below) that predispose it to error and draining it the act of identifying and removing these conditions (Reason 2005). Furthermore, allocating resources to control individuals (additional training, stricter rules and guidelines, etc.) is only successful when the individuals concerned are performing below the expected standard (those particularly error‐prone, inexperienced, unmotivated, or badly trained) (Reason 2005). In recognizing this, it is logical to allocate resources to improving the system, an approach that yields long‐term improvements.

Systems‐Based Approach

Developed and successfully implemented by the nuclear power and aviation industries, systems‐based approaches encompass humans (and their errors) alongside the systems in which they work (Allnutt 2002; Cooper et al. 1984, 2002; Kohn et al. 2000; Reason 2000). This holistic approach recognizes and accounts for human error but, more importantly, places these errors in the context of underlying systems factors, called latent conditions, that contribute to the frequency, likelihood, and consequences of error. Latent conditions comprise organizational influences, unsafe supervision, and preconditions for unsafe acts.

This is not to say that all responsibility for error lies with the “system,” absolving humans of all responsibility (Reason 2008; Wiegmann and Shappell 2005). Reviewing the processes by which errors occur reveals that personal responsibility and error wisdom have a role to play. A more nuanced position is to recognize that while humans are central to the commission of errors, they are also critical in the prevention of errors or recovery from errors (Reason 2008).

The ability of humans to react and adapt is a valued characteristic and a defining feature of high‐reliability organizations (HROs). This is exemplified by situations in which humans compensate successfully during complex technical procedures with a high risk of errors. Notably, study of such procedures, such as pediatric cardiac surgery, has also revealed that there is a limited ability to compensate for errors, even those that are minor i.e. resilience to error is limited (Carthey et al. 2003; De Leval et al. 2000; Reason 2000, 2008). Monitoring of human efforts is important, not only to identify unsafe acts and their underlying causes, but also to identify if humans are compensating unnecessarily for a flawed system.

Combining Person‐ and Systems‐Based Approaches

Current frameworks for the study of errors and accidents strive for a balance between systems and person models. In anesthesia, these frameworks are based on the understanding that unsafe acts come from people in contact with a patient, but that local and organizational factors create the conditions for errors.

James Reason, a psychologist whose work is synonymous with the study of organizational safety, developed the Swiss cheese model of accidents (Reason 2008). Through its various iterations, it has become the most influential accident model in use (Figure 15.1) (Reason 2000, 2005, 2008). Reason’s Swiss cheese model describes barriers, defenses, and safeguards through which an accident trajectory may penetrate to cause an adverse event. Each defensive layer represents a different factor within the system: organizational influences, unsafe supervision, preconditions for unsafe acts, and unsafe acts. Ideally, each layer would be solid, devoid of weaknesses; however, this does not reflect the reality of complex systems such as anesthesia. When enough weaknesses (holes) align, a series of events, which may be minor or innocuous when considered alone, culminate in an adverse event. These holes are usually dynamic; emerging, disappearing, and shifting as individual factors change, and are classified as latent conditions or active failures. Latent conditions are resident within a system, in place because of design or organizational decisions. They are usually in place for some time and can sometimes, but not always, be identified before they contribute to an adverse event. In contrast, active failures are unsafe acts, committed by a person and usually representing the tipping point that leads directly to an adverse event. Most adverse events result from a combination of latent conditions and active failures.

Human Factors Analysis and Classification System

Building on Reason’s Swiss cheese model, the Human Factors Analysis and Classification System (HFACS) was developed to define the holes within the model layers, with the goal of aiding accident investigation (Wiegmann and Shappell 2005). As such, the HFACS is a framework for analysis and investigation, comprising of four broad categories representing latent conditions and active failures (Figure 15.2) (Diller et al. 2014; Wiegmann and Shappell 2005). The HFACS should be viewed as an aid to facilitate a complete exploration of factors contributing to adverse events, rather than an exhaustive list. The HFACS is presented here in the order in which it would be applied during an adverse event investigation, i.e. beginning with the action most closely associated with the event, usually an unsafe act, and working through each level in turn, up to organizational influences.

Unsafe Acts

Unsafe acts have been variably categorized as slips, lapses, mistakes, errors (skill‐based, decision and perceptual), and violations (Diller et al. 2014; Elbardissi et al. 2007; Reason 2005, 2008; Wiegmann and Shappell 2005). This confusing nomenclature represents different systems for classifying overlapping behaviors. Using the terminology of HFACS, there are three basic error types: skill‐based, decision, and perceptual errors (Figure 15.2).

A skill‐based error results from a failure of attention or memory when performing a routine, highly automated task, often in familiar circumstances. Examples of skill‐based errors are common, e.g. not flushing an IV extension line before connection to an IV catheter, not giving antimicrobials (where indicated) before surgery begins, or placing oxygen tubing intranasally during recovery but failing to turn on the oxygen flowmeter. Skill‐based errors were identified in over 50% of adverse events in a hospital setting (Diller et al. 2014). Medication errors can often be categorized as skill‐based errors as they represent a routine task that is susceptible to attention or memory failure (Cooper et al. 1984; Mahajan 2010). They are often described as “wrong drug,” “wrong dose,” “wrong route,” “wrong time” (Kohn et al. 2000). Medication errors were identified as the most common error type in a veterinary study, contributing to 54% of 560 reported incidents, and “wrong dose” was the most common example (58%) of medication error (Wallis et al. 2019).

A decision error is when an action goes as planned but the plan is inadequate. A normally good plan is misapplied because of time pressure (procedural error), a bad plan may be applied through an error of judgment or training (poor choice error), or a faulty plan is applied in the face of a novel (or poorly understood) situation (problem‐solving error). Decision errors are extremely common, present in over 90% of adverse events (Diller et al. 2014). Not surprisingly, stage of training, experience, and supervision can play an important role in preventing or limiting these errors. Developing new plans in the face of novelty is prone to error because of limited mental resources, an incomplete or incorrect understanding of the situation and a tendency to fixate on a hypothesis while ignoring contradictory information. Unlike skill‐based errors, decision errors are associated with tasks involving conscious effort and are susceptible to cognitive bias.

A perceptual error occurs when sensory input is limited (e.g. alarm silenced) or incorrect (e.g. visual illusion) and an incorrect response leads to an error. Trying to insert a hypodermic needle into an injection port by feel so as to avoid displacing surgical drapes and potentially disrupt a procedure, rather than using a light or lifting the drapes, may result in a stick injury. Perceptual errors appear to be less commonly encountered in medicine than aviation, identified in 15% of adverse events (Diller et al. 2014).

In contrast to errors, violations are deliberate deviations from safe or standard practice. Routine violations (“bending the rules”) are habitual and often tolerated (e.g. driving a few miles or kilometers per hour in excess of the speed limit) (Wiegmann and Shappell 2005). They may be committed with the goal of improved efficiency, particularly where standard procedures are perceived to, or actually, inhibit efficient workflow (Diller et al. 2014). Examples could include using a smaller gauge needle to perform jugular IV injection and increasing the risk of intra‐arterial injection or tolerating senior staff walking in and out of the OR in outdoor footwear while everyone else wears shoe covers. When routine violations are identified, this should trigger a search for a cause, which may include tolerance by a supervisor (supervisory violation), an absence of, or inadequate, policy (organizational process), and/ or a workplace culture of bending the rules (organizational climate). Routine violations occur frequently, recorded in 80% of adverse events (Diller et al. 2014). In contrast, exceptional violations, which are dramatic departures from standard practice, occur much less frequently. In anesthesia, they are typically associated with intentional violations of well‐established standards of care (Diller et al. 2014). In general, the demographic of those most likely to violate are: young men, those with a high opinion of their skills (relative to others), those who are relatively experienced and not especially error‐prone, those with a history of errors and adverse events, and those less affected by how they are viewed by others (Reason 2008).

Preconditions for Unsafe Acts

While unsafe acts are present at the majority of adverse events (80% in aviation and >90% in medicine), stopping the investigation at this level returns to a person‐based system of blame (Diller et al. 2014; Wiegmann and Shappell 2005). Behind unsafe acts are predisposing preconditions as they apply to the operator (caregiver or anesthetist), personnel and environment (Figure 15.2).

Suboptimal operator performance can be created by an adverse mental or physiological state or physical/mental limitations. Mental fatigue (from sleep loss or other stressors) or distraction can easily predispose an individual to performing an unsafe act (Campbell et al. 2012). Distractions are commonplace, with one survey of 31 hours of anesthetic practice identifying a distracting event occurring on average of once every 4.5 minutes, and 22% of such events having a negative consequence for patient care (Campbell et al. 2012). Adverse physiological states represent illness, injury, or any other alteration (e.g. intoxication) that limits job performance. Adverse mental states are commonly reported (approximately 50% incidence) during adverse event investigations, whereas adverse physiological states are relatively rare (1%) (Diller et al. 2014). Physical/mental limitations refer to situations where the required task exceeds the ability of the individual. This may be as simple as having a hand that is too large to perform manual orotracheal intubation of a cow (leading to delayed intubation and potential aspiration of regurgitated fluid) or it may be a fundamental limitation. Occasionally, individuals that do not have an aptitude for anesthesia are encountered. While there may be organizational/institutional pressure to keep them in the service, this can put strain on other team members.

Personnel factors include team resource management and personal readiness (fitness for duty). Team resource management applies within and between teams. For example, a group of anesthesia providers may work well together, providing smooth, error‐free anesthesia, but this is of limited value if there is poor coordination with other teams so that cases frequently run early or late, cases are presented for anesthesia unexpectedly or critical information (e.g. suspected drug reaction) is not shared. Good communication and coordination are key to good team resource management. Ineffective communication between team members constituted 70% of personnel factors (n = 448), followed by lack of teamwork (7%), and failure of leadership (4%) (Diller et al. 2014). In a review of 444 surgical malpractice claims, of which 258 led to patient harm, 23% (n = 60) involved a breakdown of communication, almost all of which were verbal (Greenberg et al. 2007). Important contributing factors related to hierarchy (e.g. resident communication with more senior surgeon) and ambiguity regarding responsibility or leadership. In the majority of cases, information was either never transmitted (49%) or transmitted but inaccurately received (44%).

The frequency of communication failures in a study of six complex surgeries was high, with approximately one failure every eight minutes (13–48 failures per case) (Hu et al. 2012). Communication between disciplines failed twice as frequently as within a discipline and most failures were associated with absence of key individuals and failure to resolve an issue. Eighty‐one percent of failures resulted in a loss of efficiency. Interestingly, this study was performed in a hospital where the Surgical Safety Checklist (SSC) was performed for all cases. As the SSC has been shown to reduce communication failures, delays, and morbidity/mortality, these findings suggest that failures may have been even more frequent without its use (Haynes et al. 2009; Lingard et al. 2008; Nundy et al. 2008). In veterinary medicine, communication breakdown was the second most common error type (approx. 30%), after medication error (approx. 60%), in contributing to reported incidents (Wallis et al. 2019).

Fitness for duty (which overlaps with adverse physiological states) is the expectation that individuals are able to perform at the expected level. Failures could result from fatigue or failure to adequately prepare for a scheduled case. Fitness for duty is a contributing factor to adverse events in medicine was uncommon (3%) (Diller et al. 2014). This contrasts with the high incidence of contributions from team resource management. Recognition of physical and mental health limitations must be coupled with the freedom to self‐report without stigma or shame (Allnutt 2002).

Environmental factors comprise the physical and technological environment. In medicine and veterinary anesthesia, these apply to both personnel and patients. For example, noise during recovery from anesthesia could serve as a distraction to personnel and potentially affect the quality of recovery. Performing tasks in a darkened environment (e.g. during arthroscopy) can limit patient visibility, potentially leading to a perceptual error. Limited physical access to patients (e.g. surgery of the head/neck) can increase reliance on physiologic monitors and limit assessment of physical indicators of depth of anesthesia. Working in poorly designed environments (e.g. slippery flooring) or with poorly designed equipment (e.g. monitors that provide erroneous readings when heart rates are below 30 bpm) undoubtedly increase the risk of adverse events. Environmental factors were present in approximately 50% of human adverse events, with most instances related to a problem with equipment or environment design, though this may be higher in veterinary anesthesia (Diller et al. 2014).

Developing a Safety Climate

A safety climate may be defined as “the shared perceptions of practices, policies, procedures, and routines about safety in an organization” (Singer et al. 2010). In HROs, this climate permeates an organization at all levels, with a tangible effect on policy and practice. A general awareness of safety in human medicine was triggered by the 1999 Institute of Medicine report, “To Err Is Human.” This was preceded by early recognition in anesthesia practice of the importance of individual components that fall under the umbrella of safety climate (Allnutt 2002; Cooper et al. 1984, 2002; Cooper 1984; Gaba et al. 1987; Kohn et al. 2000; Reason 2005). While considerable progress has been made, most notably with the widespread adoption and implementation of the World Health Organization‘s (WHO) SSC, the differences in safety climate between anesthesia and aviation (a prototypical example of a HRO) remain striking (Bergs et al. 2014; Haynes et al. 2009; Singer et al. 2010). A survey of naval aviators (all personnel, including commanding officers) and healthcare workers (senior managers, physicians, and other workers) in the United States (n = 34 206 respondents) found that the safety climate was perceived as significantly safer by the naval aviators (Singer et al. 2010). The authors suggested that these differences resulted from multiple factors, including the use of standard operating procedures, the existence of mechanisms for reporting safety hazards and adverse events, managerial and administrative support for safety, adherence to guidelines and standards of care for fitness for duty, and continuous training and assessment.

The recognition of safety climate within veterinary anesthesia is in its infancy, and there are few studies of the application or outcome of promoting safe practice (Armitage‐Chan 2014; Hartnack et al. 2013; Hofmeister et al. 2014; McMillan 2014; Menoud et al. 2018). Hofmeister et al. (2014) reported promising results from a pre/ post‐intervention observational study to identify adverse events and near misses over an 11.5‐month period followed by the introduction of targeted interventions to reduce the recurrence of these incidents. Use of an anonymous reporting system identified 74 adverse events or near misses (3.6% of 2028 patients).

Checklists

The WHO SSC is the best known and most widely adopted peri‐anesthetic checklist in use. In a landmark observational study involving eight hospitals in eight countries across the world, prospective data collected before and after implementation of the SSC showed a reduction in mortality and surgical infection from 1.5% to 0.8% and 6.2% to 3.4%, respectively (Haynes et al. 2009). These findings have been repeated, with a recent meta‐analysis supportive of the benefits of using the WHO SSC (Bergs et al. 2014). In addition to reductions in morbidity and mortality, the WHO SSC has resulted in cost savings, improved communication, and improved safety climate (Haynes et al. 2009, 2011; Kearns et al. 2011; Semel et al. 2010).

The WHO SSC comprises three sections, each corresponding to a different phase in the surgical pathway (before induction of anesthesia, before skin incision, before patient leaves operating room). Each section has items that are service specific (e.g. “Is the anesthetic machine and medication check complete?”; “Is the [surgical] site marked?”) as well as those that ensure sharing of information and promote discussion (e.g. “What is the anticipated blood loss?”; “Are there any patient‐specific concerns?”). Common to these sections, and key to the development of checklists in veterinary anesthesia, is specifying the personnel who should be present, the role of a single person responsible for managing the checklist, the requirement for verbal acknowledgment and confirmation of checklist items, and the introduction of all personnel present (name and role, including trainees). These practices are believed to play a critical role in the successful or failed implementation of the WHO SSC in an individual institution and the overall impact on safety climate (Haynes et al. 2009).

There are several considerations key to successful implementation of a checklist system that is relevant to development and adoption for veterinary anesthesia and surgery (World Health Organization 2009; Alidina et al. 2018; Armitage‐Chan 2014). (i) There must be public support from the administration and department/service chiefs for prioritizing safety and use of an SSC. (ii) Forming a team from a core, multidisciplinary group of interested personnel to establish and promote adoption of the checklist. (iii) The checklist should be modifiable. This promotes adaptation to local conditions and encourages team member involvement and support. When modifying the checklist, the principles of checklist development should be applied: the checklist should be concise, focused, brief (less than one minute per section), contain actionable items, be performed verbally, developed collaboratively, tested in a small/limited setting, and integrated into existing processes. (iv) Start small. Implement the checklist and track its use and performance on a limited scale (e.g. single operating room [OR] or with a single surgical team) before expanding its use after any necessary modifications are made. (v) Track process and outcome measures to monitor performance and identify problems early. The tools used in clinical audit would be ideal for this (Rose and Pang 2021). (vi) The checklist coordinator should be carefully selected. They must be able to work with key team members as they have the responsibility for ensuring completion and adherence to the checklist, including the power to stop progress to the next phase of the procedure.

As with veterinary anesthesia safety climate, there are few studies on the application or outcome of promoting safe practice, such as checklists (Hofmeister et al. 2014; McMillan 2014; Menoud et al. 2018). Menoud et al. (2018) used a consensus discussion (Delphi method) among veterinary anesthetists to develop a peri‐anesthetic checklist based on the WHO SSC (Menoud et al. 2018). This work highlighted common obstacles encountered when attempting to introduce change associated with a checklist, including resistance to change, concerns regarding usefulness and relevance, and the time required to complete a checklist. Taken together, these underline the importance of garnering support, raising awareness and education of potential users. Such support is essential to achieving successful implementation of checklists and reaping the potential benefits in patient safety (Pickering et al. 2013).

Pre‐anesthesia Checkout Procedure

The pre‐anesthesia checkout (PAC) procedure comprises a protocol for checking components of anesthetic equipment paired with a checklist. In most instances, a physical checklist is used to prompt completion of all checks. Anesthetic equipment faults are relatively common, so a PAC is a simple, effective way to prevent a fault resulting in an adverse event (Barthram and Mcclymont 1992; Kendell and Barthram 1998). Anesthetic machine checks identify 30–60% of faults and approximately one‐fifth of these may be serious, posing a direct risk to the patient (Barthram and Mcclymont 1992; Kendell and Barthram 1998). Additionally, completion of a PAC is associated with a decreased risk of anesthetic morbidity and mortality (odds ratio 0.64, 95%CI 0.43–0.95) (Arbous et al. 2005). Using the imagery of the Swiss cheese model, performing a PAC may be visualized as closing a hole in a defensive layer (Figure 15.1). As anesthesia equipment becomes increasingly complex and equipment varies considerably between clinics, it is impossible to provide a comprehensive, universal PAC guide (Association of Anaesthetists of Great Britain and Ireland et al. 2012; Hartle 2013). Manufacturer guidelines should be followed for individual pieces of equipment. Table 15.2 presents an outline of key checks that should be completed, based on the most recent American Society of Anesthesiologists and Association of Anesthetists of Great Britain and Ireland guidelines (Anesthesiologists Society of Anesthesiologists 2008; Association of Anaesthetists of Great Britain and Ireland et al. 2012

Stay updated, free articles. Join our Telegram channel

Join