- Email: [email protected]
Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Veterinary Anaesthesia and Analgesia 2017, xxx, 1e10
http://dx.doi.org/10.1016/j.vaa.2017.06.007
RESEARCH PAPER
Systems analysis of voluntary reported anaesthetic safety incidents occurring in a Q9
university teaching hospital
Q8
Matthew W McMillan & Kristina S Lehnus Queen's Veterinary School Hospital, Department of Veterinary Medicine, University of Cambridge, Cambridge, UK Correspondence: Matthew McMillan, Queen's Veterinary School Hospital, Department of Veterinary Medicine, University of Cambridge, Madingley Road, Cambridge, CB3 0ES, UK. E-mail: [email protected]
Q1
Abstract Objective To identify factors contributing to the development of anaesthetic safety incidents. Study design Prospective, descriptive, voluntary reporting audit of safety incidents with subsequent systems analysis. Animals All animals anaesthetized in a multispecies veterinary teaching hospital from November 2014 to October 2016.
Q2
Methods Peri-anaesthetic incidents that risked or caused unnecessary harm to an animal were reported by anaesthetists alongside animal morbidity and mortality data. A modified systems analysis framework was used to identify contributing factors from the following categories: Animal and Owner, Task and Technology, Individual, Team, Work Environmental, and Organizational and Management. The outcome was graded using a simple descriptive scale. Data were analysed using Pearson's Chi-Square test for association and univariable and multivariable logistic regression analysis. Results Totally, 3379 anaesthetics were performed during the audit period. Of these, 174 incident reports were analysed, 163 of which impacted safe veterinary care and 26 incidents were considered to have had major or catastrophic outcomes. Incident outcome was believed to have been limited by anaesthetist intervention in 104 (63.8%) cases. Various factors were identified as: Individual in 123 (70.7%), Team in 108 (62.1%), Organizational and Management in 94 (54.0%), Task and Technology in 80 (46.0%), Work Environmental in 53 (30.5%) and Animal and Owner in 36 (20.7%) incidents.
Individual factors were rarely seen in isolation. Significant associations were identified between Experience and Supervision, X2 (1, n ¼ 174) ¼ 54177, p ¼ 0.001, Failure to follow a standard operating procedure and Task Management, X2 (2, n ¼ 174) ¼ 11318, p ¼ 0.001, and Staffing and Poor Scheduling, X2 (1, n ¼ 174) ¼ 36742, p ¼ 0.001. Animal Condition [odds ratio (OR) ¼ 16210, 95% confidence interval (CI) ¼ 5573e47147)] and anaesthetist Decision Making (OR ¼ 3437, 95% CI ¼ 1184e9974) were risk factors for catastrophic and major outcomes. Conclusions and clinical relevance Individual factors contribute to many safety incidents but tend to occur concurrently with other factors. Anaesthetist intervention limits the consequences of incidents for most animals. Keywords anaesthesia, human factors, incidents, patient safety, systems analysis. Introduction Patient safety and quality improvement are rapidly evolving disciplines in human medicine, involved in the reduction of harm caused to patients directly by the healthcare they are receiving. Voluntary reporting and analysis of safety incidents are considered as key elements of these disciplines (Vincent 2004; Leistikow et al. 2016). Incident reports can be viewed as a “window on the system” in which they arise, providing valuable insights into gaps and inadequacies in healthcare provision (Vincent 2004). Subsequent systems analysis, which aims to identify failures within a healthcare system and an organization as a whole rather than 1
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
Q3
focussing on individual failures, can highlight both current weaknesses and future problems, facilitating tailored interventions and improvements to healthcare provision (Vincent 2004). To date, this area of study has received little attention in veterinary medicine. Anaesthesia has always been a specialty at the forefront of patient safety (Gaba 2000). However, little is known about safety in veterinary anaesthesia beyond a number of studies on fatality rates which have disclosed limited numbers of risk factors primarily associated with animal health status, procedure and anaesthetic technique (Clarke & Hall 1990; Johnston et al. 2002; Bidwell et al. 2007; Brodbelt et al. 2008; Bille et al. 2012). To date, the only study to specifically assess safety incidents in veterinary anaesthesia by Hofmeister et al. (2014) documented them using an incident log consisting of a tally of specific incident types that occurred. This gave one of the first insights into the frequency of safety incidents, such as medication and equipment errors. The objective of this study was to collect, review and analyse short vignettes of safety incidents reported during the management of cases by the anaesthesia service of a university teaching hospital. Using a systems analysis modified for veterinary anaesthesia, we aimed to identify factors and causal themes that contributed to safety incidents. Materials and methods Study design A prospective, descriptive, voluntary reporting audit of safety incidents was performed as part of a selfgovernance initiative in a multidisciplinary university teaching hospital. Documents were managed in a confidential manner, with only the anaesthesia team aware of the audit being conducted. Information from individual reports was analysed and pooled prior to publication. Ethics and consent The study was an audit, as defined by the United Kingdom ethics committees (National Research Ethics Service 2009); therefore, ethical approval was waived by the local ethics and welfare committee. Consent was obtained from all anaesthetists to use the information in the reports and during morbidity and mortality rounds. 2
Setting and participants The hospital in which the audit was performed provides both first opinion and referral healthcare services for a range of small animals, equine and farm animals. The anaesthesia service is responsible for approximately 1500e1700 anaesthetics per annum, the majority being in dogs (approximately 1000 cases), followed by cats, horses, farm animals and others (approximately 600, 50, 25 and 10 cases, respectively). Anaesthesia is performed in several discrete locations in the hospital: a small animal surgical unit; small animal endoscopy, dental and electro-diagnostics suite; two small animal radiography suites; a magnetic resonance imaging suite; a computed tomography suite; a radiotherapy unit and a large animal surgical unit. During the period of the audit, anaesthetics were performed by three European College Veterinary Anaesthesia and Analgesia Diplomates, one post-residency clinical anaesthetist, three anaesthesia residents, one registered veterinary nurse and a number of rotating interns. Final year veterinary students were involved with many of the cases, rotating through the anaesthesia service for approximately 80% of the time the audit was conducted. Data sources and collection The audit was performed over a period of 23 months between November 2014 and October 2016. The number of anaesthetics performed during this period was determined using electronic hospital records. Based on the World Health Organization's International Classification for Patient Safety, “Safety Incidents” were defined as any event or circumstance which could have resulted, or did result, in unnecessary harm to an animal or member of staff (Runciman et al. 2009). All the staff members performing anaesthesia were asked to report any such incidents occurring within the peri-anaesthetic period of an animal in a handwritten “Safety Incident Diary” held centrally in the anaesthesia induction suite. Whether to report an incident was left to the individual anaesthetist's discretion, but it was encouraged that any incident, irrespective of severity, be reported. Fatalities and significant anaesthetic related morbidities were also reported. In this study, the peri-anaesthetic period was defined as any time point in which an anaesthetist could be in contact with an animal between pre-anaesthetic assessment and discharge.
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
Limited guidance was given regarding the format of incident reporting. The anaesthetist was asked to provide a short narrative without giving reference to individual animals or staff; the use of broader terms such as species, surgeon, clinician, resident, intern, student, nurse or assistant staff was encouraged. These notes then formed the basis for subsequent anaesthesia mortality and morbidity rounds, planned for every 4e8 weeks. Meetings were conducted in a “no-blame” fashion, with anaesthetists encouraged to look beyond individual culpability at the underlying factors that may have contributed to the incident as well as its impact. This allowed the team to offer their opinion on any contributing, causative or ameliorating factors specifically anaesthetist intervention. Ways in which similar incidents could be avoided or better managed in the future were also discussed. To encourage an open forum, incidents involving senior staff were discussed at the start of the first few meetings. Thereafter, incidents were discussed in chronological order. Neither transcripts nor recordings of the meetings were made, but brief notes documenting the timeline of the incident and contributing factors were added to the existing records. Analysis At the end of the audit period, all reports and accompanying notes from meetings were evaluated. Further analysis was performed when clear problems with the delivery of care could have, or did result, in harm to an animal or staff member. Additionally, all reported fatalities and significant post-anaesthetic morbidities were included for analysis, irrespective of the presence of an obvious problem in the delivery of care. Incident inclusion and exclusion is outlined in Fig. 1. Incidents which impacted scheduling or staffing but not safe veterinary care were excluded, while those prolonging the procedure or general anaesthetic time were included. The aim of the analysis was to identify contributing factors and themes implicated in the incident reports and subsequent discussions. A simplified “Systems” analysis was performed using a framework of contributing factors modified for use in veterinary anaesthesia (Vincent et al. 2000; Woloshynowych et al. 2005). Contributing factors were divided into six main categories: 1) Animal and Owner; 2) Task and Technology; 3) Individual; 4) Team; 5) Work Environmental; and 6) Organizational and Management. A comprehensive list of contributing factors
and examples is provided in Appendices S1eS6. Every report was assessed systematically by a single assessor (main author) to identify contributing factors from each level of this framework (Fig. 2). Due to the nature and details of the reports and notes taken subsequently during the morbidity and mortality rounds, independent analysis was not possible. In order to allow comparison to previous data produced by Hofmeister et al. (2014), Medication Errors and Equipment Failures were also identified. These represented incidents primarily associated with the administration of drugs, or errors such as leaving adjustable pressure limiting (APL) valves closed. Anaesthetist Non-Technical Skills, comprising Decision Making, Situation Awareness, Teamwork and Task Management (Flin et al. 2010), were grouped and compared with the traditional competencies of Knowledge, Experience and Technical Skills. The estimated impact of the incident on the animal was assessed using a simple descriptive scale based upon similar scales used in human medicine (Table 1) (Cook et al. 2009). Statistical methods Descriptive statistics and qualitative data are reported. Normality of data was assessed using the ShapiroeWilk test where necessary. Non-normally distributed data are described using the median (interquartile range). Associations were assessed between factors which were identified together on 10 occasions using Pearson's Chi-square test for association. Data is reported as X2 (degrees of freedom, sample size) Pearson’s Chi-square value, p value. Statistical significance was set at p ¼ 0.05 before Bonferroni correction. Outcome analysis was conducted using logistic regression analysis, considering all cases that experienced major or catastrophic outcomes as a result of a safety incident. Only those contributing factors with n > 10 animals and p < 0.1 were included for multivariable analysis with forwards stepwise elimination. Final logistic regression fit was assessed using receiver-operating characteristic curves and area under the curve calculations, reported with 95% confidence intervals (CI), with the significance level set at p < 0.05. Results are reported as an odds ratio (OR) with respective 95% CI for factors with p < 0.1 and p < 0.05 for univariable and multivariable analysis. All statistics were performed by using SPSS version 23(IBM).
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
3
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
Q4
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
Total number of anaesthetics (n = 3379)
Safety diary entries (n = 204) Excluded entries not impacting animal or staff safety (n = 30)
Reported peri-anaesthetic fatalities and morbidities (n = 14)
Reported safety incidents (n = 160)
Cases included for further analysis (n = 174) Excluded cases with environmental impact only (n = 11) Cases included for logistic regression analysis (n = 163) Figure 1 Inclusion and exclusion flow chart for the handling and further analysis of safety incidents.
Results
Q5
During the audit period, 3379 anaesthetics were performed and 204 diary entries were made. The audit identified 160 safety incidents and a further 14 cases with significant morbidity or mortality. Therefore, 174 (prevalence 5.1%) cases were included for further analysis. Fourteen morbidity and mortality rounds were performed within the audit period with a gap of 4e14 weeks between meetings. Total monthly reported incidents ranged between 0 and 20, with a median of 7 (3e5). There was no discernible pattern to the number of reports made per month; reporting rates did not appear to trend upward or downward over time or be linked with case number. Of the total incidents, 159 directly involved individual animals; these included 17 Medication Errors (prevalence 0.5%), 11 Equipment Failures (prevalence 0.3%) and 4 incidents of APL valves being left closed 4
(prevalence 0.1%). Fifteen incidents were associated with generalized problems: four incidents involved pipeline oxygen pressure dropping because of back up cylinder failure requiring the use of machine-mounted cylinders; in 11 cases, environmental contamination occurred with anaesthetic gases and vapours. A total of 677 contributing factors were identified across all incidents. A median of four (3e5) factors were identified per reported incident, with three (2e4) out of the six framework categories present. Individual factors were identified in 123 (70.7%), Team factors in 108 (62.1%), Organizational and Management factors in 94 (54.0%), Task and Technology factors in 80 (46.0%), Work Environment factors in 53 (30.5%) and Animal and Owner factors in 36 (20.7%) incidents (Table 2). Where factors belonging to the Individual category were identified (n ¼ 123), factors from further down
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
Incident selected for analysis
Animal and Owner factors considered
Animal Condition, Temperament, Financial constraint, Communication with owner, Owner attitude
Report and supplementary notes evaluated
Task and Technology factors considered
Incident timeline developed
Individual factors considered
Experience, Knowledge, Technical skills, Decision making, Task management, Situation awareness, Team working, Health, stress or fatigue
Problems with delivery of care identified
Team factors considered
Supervision, Written communication, Verbal communication, Attitude, Seeking help, Leadership
Contributing factor framework applied
Work Environmental factors considered
Distraction, Layout and ergonomics, Social atmosphere
Impact severity scale applied
Organizational and Management factors considered
Staffing level, Poor scheduling, Culture and priorities, Training, Equipment levels, Unrealistic expectations, Not responding to staff concerns, Planning and preparation
Failure to follow Standard operating procedures, Equipment checks, Equipment failure, Over-reliance on technology, Equipment maintenance, Equipment availability, Drug labelling
Figure 2 Flow chart demonstrating the framework used for the systematic analysis of anaesthetic safety incidents.
the framework in the Team, Work Environmental and Organizational and Management categories occurred concurrently in 89.4% (n ¼ 110) incidents. In the remaining incidents (n ¼ 13, 10.6%) the contributing factor of Health, Stress and Fatigue was identified alongside other factors from the Individual category. Analysis of sub-groups within the Individual category indicated that Anaesthetist NonTechnical Skills (n ¼ 101, 58.0%), comprising Decision Making, Situation Awareness, Teamwork and Task Management contributed more to incidents than the “traditional competencies” such as Knowledge, Experience or Technical Skills (n ¼ 65, 37.4%). Pearson’s Chi-square test for associations was performed for 39 factor combinations with n 10.
Bonferroni correction set statistical significance at p < 0.0013. Significant associations were identified between Experience and Supervision, X2 (1, n ¼ 174) ¼ 54177, p ¼ 0.001, Failure to follow a standard operating procedure (SOP) and Task Management, X2 (2, n ¼ 174) ¼ 11318, p ¼ 0.001, and Staffing and Poor Scheduling, X2 (1, n ¼ 174) ¼ 36742, p ¼ 0.001. In total, 163 (93.7%) incidents impacted the animal. Of these, the consequences of 17 (9.8%) were classified as catastrophic, 9 (5.2%) as major, 24 (13.8%) as moderate, 77 (44.2%) as minor and 36 (20.7%) as insignificant. The impact of the incident on the animal was believed to have been limited by anaesthetist intervention in 104 (63.8%) of these
Table 1 Simple descriptive scale to grade the severity of the incident impact on the animal Severity
Description
Insignificant Minor Moderate Major Catastrophic
Incident Incident Incident Incident Incident
had little or no effect on the animal led to an increased level of care or caused minor short-term harm only led to short-term but significant loss or impairment of physiological function led to a significant loss or impairment of physiological function led to death or injury that led directly to euthanasia
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
5
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus Table 2 Overview of the prevalence of contributing factors identified during the analysis of the reported safety incidents Framework category
n (%)
Sub-categories
Sub-total n (%)
Patient and Owner
38 (5.6)
Task and Technology
100 (14.8)
Individual
194 (28.7)
Team
148 (21.9)
Work Environmental
54 (7.9)
Organizational and Management
143 (21.1)
Animal condition Temperament Financial constraint Communication with owner Owner attitude Failure to follow SOP Equipment checks Equipment failure Over-reliance on technology Equipment maintenance Equipment availability Drug labelling Decision making Experience Health, stress, fatigue Task management Knowledge Situation awareness Technical skills Team working Supervision Written communication Verbal communication Attitude Seeking help Leadership Distraction Layout and ergonomics Social atmosphere Staffing level Poor scheduling Culture and priorities Training Equipment levels Unrealistic expectations Not responding to staff concerns Planning and preparation Lack of SOP Failure of SOP
30 (78.9) 5 (13.3) 1 (2.6) 1 (2.6) 1 (2.6) 51 (51.0) 28 (28.0) 11 (11.0) 4 (4.0) 2 (2.0) 2 (2.0) 2 (2.0) 65 (33.5) 44 (22.7) 28 (14.4) 24 (12.4) 16 (8.2) 11 (5.7) 5 (2.6) 1 (0.5) 48 (32.4) 38 (25.7) 33 (22.3) 23 (15.5) 4 (2.7) 2 (1.4) 42 (77.8) 8 (14.8) 4 (7.4) 53 (37.0) 32 (22.4) 28 (19.6) 12 (8.4) 7 (4.9) 3 (2.1) 3 (2.1) 3 (2.1) 1 (0.7) 1 (0.7)
SOP, standard operating procedure.
163 incidents. Of the catastrophic consequences, three (17.6%) were considered potentially preventable, whereas the remainder (82.4%) were associated with severe disease, injury or surgical catastrophe. Of the 174 reported incidents, 11 (6.3%) did not result in harm to the animal but were associated with environmental contamination and therefore posed a risk to staff. All the 163 cases with consequences for the animal were considered for logistic regression analysis to assess the relationship between contributing factors and a combined major and catastrophic outcome (n ¼ 26, 16.0%). Following multivariable analysis, 6
Animal Condition (p < 0.001) and anaesthetist Decision Making (p ¼ 0.023) were significantly associated with developing a major or catastrophic outcome as a result of the safety incident. Results of the univariable and multivariable analyses are shown in Tables 3 and 4, respectively. Discussion Key results Our audit demonstrated that safety incidents occurred at a rate of approximately 1 in every 20 anaesthetics during the study period. The impact of
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus Table 3 Results of univariable logistic regression for contributing factors with p < 0.1 when assessing incidents with major or catastrophic impact for the animal. Percentages are in terms of the total number of contributing factors identified (n ¼ 677) Factor
n (%)
p
OR (95% CI)
Animal condition Task and technology Failure to follow SOP Experience Decision Making Situation awareness Work Environmental Equipment levels
30 (4.4) 100 (14.8) 51 (7.5) 44 (6.5) 65 (9.6) 11 (1.6) 54 (8.0) 7 (1.0)
< 0.001 0.035 0.031 0.001 0.017 0.068 0.095 0.065
18017 (6698e48468) 0.352 (0.133e0.932) 0.195 (0.044e0.864) 4345 (1814e10404) 2873 (1211e6817) 3377 (0.912e12501) 0.384 (0.125e1183) 4337 (0.910e20660)
CI, confidence interval; OR, odds ratio; SOP, standard operating procedure.
Table 4 Results of multivariable logistic regression for contributing factors with p < 0.05 when assessing incidents with major or catastrophic impact on the animal. Percentages are given in terms of the total number of cases that experienced adverse events (n ¼ 163) Factor
n (%)
p
OR (95% CI)
AUC (95% CI)
Animal condition Decision Making
30 (18.4%) 65 (40.0%)
< 0.001 0.023
16210 (5573e47147) 3437 (1184e9974)
0.779 (0.666e0.893) 0.629 (0.511e0.747)
AUC, area under the curve; CI, confidence interval; OR, odds ratio.
the majority of incidents was considered to be insignificant, minor or moderate and therefore was not felt to have caused any significant or lasting harm. However, it should be noted that in almost two-thirds of cases, consequences were considered to have been limited by the intervention of an anaesthetist. Although some problems in delivery of care were identified when analysing morbidity and mortality data, these were not thought to have adversely affected outcome as the disease process or injuries suffered by the animal were untreatable in the majority of cases. Nevertheless, these findings could highlight areas where veterinary care may be improved. Overall, the most commonly identified contributing factors belonged to the Individual and Team categories. Within the Individual category, anaesthetist Non-Technical Skills was identified more often than Knowledge, Experience or Technical Skills. This highlights the importance of skills, such as Decision Making, Situation Awareness, Team Working and Task Management, in the provision of safe anaesthesia. Contributing factors from the Individual category were either identified with other factor categories further down the framework or together with the Individual contributing factor Health, Stress and Fatigue. These findings highlight the importance of looking beyond individual culpability when
analysing the causality of incidents, i.e. other external factors may be negating influencing the performance of the individual, and suggest staff health and welfare is also an area worthy of consideration. Although multiple contributing factors were identified for all incidents reported, only few statistically significant associations were noted. Perhaps unsurprisingly, these were found between Anaesthetist Experience and Supervision as well as between Staffing Levels and Scheduling and between Failure to follow an SOP and Task management. It would appear logical that less experienced staff members without ready access to appropriate supervision may be involved in more incidents, and also that Staff numbers and Scheduling, and Failure to follow an SOP and Task management are intimately related. Outcome analysis indicated that most catastrophic events were strongly linked to the animal's underlying condition. This may indicate that animals with serious underlying conditions are more physiologically unstable and therefore are less able to tolerate the insult of an incident. This perhaps leads to the anaesthetist having less time to intervene and avert serious consequences. It is of note that problems with Decision Making were also identified in many of these cases which could indicate that Decision Making during critical time and stressful situations is an area
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
7
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
that warrants further investment in terms of education and training. It may also reflect an overly critical retrospective analysis of an individual’s performance during crisis situations. Limitations Voluntary reporting systems are likely to underreport safety incidents. Reasons cited for this include time constraints, lack of or poor access to an appropriate reporting system, uncertainty of what to report and the feeling that nothing will be done to counter contributing factors (Garrouste-Orgeas et al. 2012). Furthermore, fear of repercussions, such as blame, litigation or disciplinary action following reporting an incident is another significant barrier (Heard et al. 2012). Even if the incident is reported, individuals may still underplay their involvement, feel embarrassed, shame or guilt regarding the event. These feelings may have been exacerbated in this study by anaesthetists knowing the incidents would be discussed later in morbidity and mortality rounds (Heard et al. 2012) even if assurances were given about confidentiality and a “no-blame” culture. In addition, the reporting method permitted only a relatively superficial analysis of contributing factors. Details were often not elucidated until the incident was discussed at morbidity and mortality meetings, and although these discussions allowed a more thorough narrative of the incident and surrounding circumstances to be developed, this retrospective assessment may have left out significant contributing factors and potentially introduced hindsight bias. Furthermore, only a single person with intimate knowledge of the working environment and circumstances during the audit period was able to perform the systems analysis, adding a certain degree of observer bias. Nonetheless, we believe that the data represents a fair and balanced assessment of the contributing factors that were encountered in the hospital during the timeframe of the audit. Interpretation In light of these limitations, the findings of this audit must be considered in context. Although the analysis used was designed to provide a framework for both, quick and easy incident analysis as well as more indepth investigations, brief incident reports such as those collected during this study can only allow a superficial analysis and are unlikely to fully illuminate all the factors contributing to an incident. In 8
truth, a more in-depth analysis of a few keys incidents is likely to elucidate more specific information about the failings of a system and allow more targeted action to be taken than superficial analysis of a large number of incidents (Vincent et al. 1999a). Nonetheless, identifying even a few themes and factors should help focus attention to problem areas and thus may present opportunities to improve animal safety. In this study, information on incidents was gathered through written reports and via morbidity and mortality round discussions. Despite being a traditional practice and being embedded in the curriculum of veterinary and medical education, the structure and function of morbidity and mortality rounds appears poorly defined (Kuper et al. 2010; Higginson et al. 2012). Often morbidity and mortality rounds are used to discuss clinical conundrums and correlations between clinical and pathological findings and are explicitly aimed at improving medical knowledge (Kuper et al. 2010). Indeed, this was the case in our hospital prior to this investigation. However, through the performance of this study, we have found that the discussion of reported safety incidents during morbidity and mortality rounds provided an open forum to specifically address safety issues and identify potential solutions. Establishing blame-free group discussions had the additional benefit of allowing individual anaesthetists to reflect on their role in the incident and consider factors contributing to the incident both inside and outside of their control. This positive approach may have helped encourage continued reporting as demonstrated by a consistent reporting rate. Generalizability The ability to apply these findings to other clinical settings may be limited by the unique working environments and circumstances of different hospitals. However, a proportion of contributing factors and themes identified in this audit may well be universal. Differences in contributing factors may account for variations in incident rates observed between different hospitals. The only previous study to investigate safety incidents in veterinary anaesthesia found a greater frequency of medication error and closed APL valves than this study. Prior to making a number of safety interventions, Hofmeister et al. (2014) reported 25 medication errors and 20 incidents of APL valves being left closed in 2028 cases (prevalence, 1.2% and 1.0%, respectively). This compares to 17 medication errors and 4 incidents of APL valves being left closed in 3379 animals in this
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
study (0.5% and 0.11%, respectively). The reasons for these differences are unclear but likely reflect differences in individual working practices and protocols. It also suggests that a set of safety interventions suited to one clinical environment may not have the same impact in another with a different environment and working conditions. For Hofmeister et al. (2014), the instigation of SOP when administering medications to animals and a check box on the anaesthetic record regarding checking the APL valve reduced the frequency of medication errors to 14 and that of closed APL valves to 5 incidents in 2112 animals (0.7% and 0.2%, respectively). These are similar levels to those seen in our audit despite no specific safety measures being in place to avoid these incidents in our hospital. Future studies In order to perform more in-depth analyses, the quality and quantity of information obtained from incident reports must be improved. A better timeline that includes the sequence of events and the individual’s thought processes and feelings during the incident should be generated. This can be achieved in a way by encouraging a more information-rich narrative. Electronic reports are used in most safetyconscious industries, including human medicine, in order to facilitate this. The Australian Incident Monitoring System, the Anesthesia Incident Reporting System and the National Reporting and Learning System are three electronic systems available to report anaesthetic safety incidents (Woloshynowych et al. 2005). Such systems could be shared between institutions allowing information to be pooled and shared. In addition to this, transcripts or recordings of meetings and semi-structured interviews with individual anaesthetists could be utilized to improve the quality of safety data. Finally, reporting and analysing safety incidents alone cannot improve safety in isolation. After significant contributing factors are identified, interventions should be made to counter these factors, either by making attempts to remove them or putting barriers in place to avoid the potential consequences. The effects these interventions and subsequent modifications to operating processes have on safety should then be assessed through recurrent clinical audit cycles. This feedback loop has the added advantage of staff seeing reports actioned, which should in turn help encourage continued reporting (Vincent et al. 1999b). In this study, Communication
and Scheduling were two readily actionable areas highlighted as potentially contributing to a high number of incidents, and this has led to changing practices in these areas. The impact of these changes is yet to be explored. Methods to support anaesthetist Decision Making especially during crisis situations are also being developed. Conclusion Safety incidents occur at a measurable rate in veterinary anaesthetic practice, and simple methods can be used to gather data on their nature and frequency. Analytical frameworks developed for the use in human medicine can be used with minimal modification to identify potential contributing factors in a veterinary setting. Although Individual factors contribute to a significant proportion of incidents, they occur concurrently with other factors in a majority of cases. It is also significant that interventions by anaesthetists appear to limit the consequences of incidents for the majority of animals, highlighting the importance of resilience and non-technical skills such as situation awareness in anaesthesia. Such investigations can potentially be used to improve anaesthetic safety through the development of targeted interventions. Authors' contributions MWM: study concept and design, data acquisition, analysis of reports, manuscript authorship; KSL: data acquisition and sorting, statistical analysis, interpretation of data, tables and figures, editing and major revisions of manuscript. Funding This study received no financial backing Conflict of interest statement The authors declare no conflict of interest. References National Research Ethics Service: Defining Research NRES guidance to help you decide if your project requires review by a Research Ethics Committee, 2009. http://www.nres.npsa.nhs.uk/applications/is-yourproject-research/. Last accessed 20 October 2014. Bidwell LA, Bramlage LR, Rood WA (2007) Equine perioperative fatalities associated with general anaesthesia at a private practiceea retrospective case series. Vet Anaesth Analg 34, 23e30. Bille C, Auvigne V, Libermann S et al. (2012) Risk of anaesthetic mortality in dogs and cats: an observational
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
9
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
Q6
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Systems analysis of safety incidents MW McMillan and KS Lehnus cohort study of 3546 cases. Vet Anaesth Analg 39, 59e68. Brodbelt DC, Blissitt KJ, Hammond RA et al. (2008) The risk of death: the confidential enquiry into perioperative small animal fatalities. Vet Anaesth Analg 35, 365e373. Clarke KW, Hall LW (1990) A survey of anaesthesia in small animal practice: AVA/BSAVA report. Vet Anaesth Analg 17, 4e10. Cook TM, Bland L, Mihai R, Scott S (2009) Litigation related to anaesthesia: an analysis of claims against the NHS in England 1995-2007. Anaesthesia 64, 706e718. Flin R, Patey R, Glavin R, Maran N (2010) Anaesthetists' non-technical skills. Br J Anaesth 105, 38e44. Gaba DM (2000) Anaesthesiology as a model for Patient safety in health care. BMJ 320, 785e788. Garrouste-Orgeas, Philipart F, Bruel C et al. (2012) Overview of medical errors and adverse events. Ann Intensive Care 2, 2. Heard GC, Sanderson PM, Thomas RD (2012) Barriers to adverse event and error reporting in anaesthesia. Anesth Analg 114, 604e614. Higginson J, Walters R, Fulop N (2012) Mortality and morbidity meetings: an untapped resource for improving the governance of patient safety. BMJ Qual Saf 21, 576e585. Hofmeister EH, Quandt J, Braun C, Shepherd M (2014) Development, implementation and impact of simple patient safety interventions in a university teaching hospital. Vet Anaesth Analg 41, 243e248. Johnston GM, Eastment JK, Wood JLN, Taylor PM (2002) The confidential enquiry into perioperative equine fatalities (CEPEF): mortality results of Phases 1 and 2. Vet Anaesth Analg 29, 159e170. Kuper A, Zur Nedden N, Etchells E et al. (2010) Teaching and learning in morbidity and mortality rounds: an ethnographic study. Med Educ 44, 559e569.
10
Leistikow I, Mulder S, Vesseur J, Robben P (2016) Learning from incidents in healthcare: the journey, not the arrival, matters. BMJ Qual Saf. Published Online First: 01 April 2016. Q7 Runciman W, Hibbert P, Thomson R, Van Der Schaaf T, Sherman H, Lewalle P (2009) Towards an International Classification for Patient Safety: key concepts and terms. Int J Qual Health Care 21, 18e26. Vincent CA, Taylor-Adams S, Chapman EJ, Hewett D et al. (1999a) A protocol for the investigation and analysis of clinical incidents. University College London/Association of Litigation and Risk Management, London. Vincent C, Stanhope N, Crowley-Murphy M (1999b) Reasons for not reporting adverse incidents: an empirical study. J Eval Clin Pract 5, 13e21. Vincent C, Taylor-Adams S, Chapman EJ et al. (2000) How to investigate and analyse clinical incidents: Clinical Risk Unit and Association of Litigation and Risk Management protocol. BMJ 320, 777e781. Vincent CA (2004) Analysis of clinical incidents: a window on the system not a search for root causes. Qual Saf Health Care 13, 242e243. Woloshynowych M, Rogers S, Taylor-Adams S, Vincent C (2005) The investigation and analysis of critical incidents and adverse events in healthcare. Health Technol Assess 9. Received 6 January 2017; accepted 7 June 2017. Available online xxx
Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.vaa.2017.06. 007.
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64
Veterinary Anaesthesia and Analgesia 2017, xxx, 1e10
http://dx.doi.org/10.1016/j.vaa.2017.06.007
RESEARCH PAPER
Systems analysis of voluntary reported anaesthetic safety incidents occurring in a Q9
university teaching hospital
Q8
Matthew W McMillan & Kristina S Lehnus Queen's Veterinary School Hospital, Department of Veterinary Medicine, University of Cambridge, Cambridge, UK Correspondence: Matthew McMillan, Queen's Veterinary School Hospital, Department of Veterinary Medicine, University of Cambridge, Madingley Road, Cambridge, CB3 0ES, UK. E-mail: [email protected]
Q1
Abstract Objective To identify factors contributing to the development of anaesthetic safety incidents. Study design Prospective, descriptive, voluntary reporting audit of safety incidents with subsequent systems analysis. Animals All animals anaesthetized in a multispecies veterinary teaching hospital from November 2014 to October 2016.
Q2
Methods Peri-anaesthetic incidents that risked or caused unnecessary harm to an animal were reported by anaesthetists alongside animal morbidity and mortality data. A modified systems analysis framework was used to identify contributing factors from the following categories: Animal and Owner, Task and Technology, Individual, Team, Work Environmental, and Organizational and Management. The outcome was graded using a simple descriptive scale. Data were analysed using Pearson's Chi-Square test for association and univariable and multivariable logistic regression analysis. Results Totally, 3379 anaesthetics were performed during the audit period. Of these, 174 incident reports were analysed, 163 of which impacted safe veterinary care and 26 incidents were considered to have had major or catastrophic outcomes. Incident outcome was believed to have been limited by anaesthetist intervention in 104 (63.8%) cases. Various factors were identified as: Individual in 123 (70.7%), Team in 108 (62.1%), Organizational and Management in 94 (54.0%), Task and Technology in 80 (46.0%), Work Environmental in 53 (30.5%) and Animal and Owner in 36 (20.7%) incidents.
Individual factors were rarely seen in isolation. Significant associations were identified between Experience and Supervision, X2 (1, n ¼ 174) ¼ 54177, p ¼ 0.001, Failure to follow a standard operating procedure and Task Management, X2 (2, n ¼ 174) ¼ 11318, p ¼ 0.001, and Staffing and Poor Scheduling, X2 (1, n ¼ 174) ¼ 36742, p ¼ 0.001. Animal Condition [odds ratio (OR) ¼ 16210, 95% confidence interval (CI) ¼ 5573e47147)] and anaesthetist Decision Making (OR ¼ 3437, 95% CI ¼ 1184e9974) were risk factors for catastrophic and major outcomes. Conclusions and clinical relevance Individual factors contribute to many safety incidents but tend to occur concurrently with other factors. Anaesthetist intervention limits the consequences of incidents for most animals. Keywords anaesthesia, human factors, incidents, patient safety, systems analysis. Introduction Patient safety and quality improvement are rapidly evolving disciplines in human medicine, involved in the reduction of harm caused to patients directly by the healthcare they are receiving. Voluntary reporting and analysis of safety incidents are considered as key elements of these disciplines (Vincent 2004; Leistikow et al. 2016). Incident reports can be viewed as a “window on the system” in which they arise, providing valuable insights into gaps and inadequacies in healthcare provision (Vincent 2004). Subsequent systems analysis, which aims to identify failures within a healthcare system and an organization as a whole rather than 1
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
Q3
focussing on individual failures, can highlight both current weaknesses and future problems, facilitating tailored interventions and improvements to healthcare provision (Vincent 2004). To date, this area of study has received little attention in veterinary medicine. Anaesthesia has always been a specialty at the forefront of patient safety (Gaba 2000). However, little is known about safety in veterinary anaesthesia beyond a number of studies on fatality rates which have disclosed limited numbers of risk factors primarily associated with animal health status, procedure and anaesthetic technique (Clarke & Hall 1990; Johnston et al. 2002; Bidwell et al. 2007; Brodbelt et al. 2008; Bille et al. 2012). To date, the only study to specifically assess safety incidents in veterinary anaesthesia by Hofmeister et al. (2014) documented them using an incident log consisting of a tally of specific incident types that occurred. This gave one of the first insights into the frequency of safety incidents, such as medication and equipment errors. The objective of this study was to collect, review and analyse short vignettes of safety incidents reported during the management of cases by the anaesthesia service of a university teaching hospital. Using a systems analysis modified for veterinary anaesthesia, we aimed to identify factors and causal themes that contributed to safety incidents. Materials and methods Study design A prospective, descriptive, voluntary reporting audit of safety incidents was performed as part of a selfgovernance initiative in a multidisciplinary university teaching hospital. Documents were managed in a confidential manner, with only the anaesthesia team aware of the audit being conducted. Information from individual reports was analysed and pooled prior to publication. Ethics and consent The study was an audit, as defined by the United Kingdom ethics committees (National Research Ethics Service 2009); therefore, ethical approval was waived by the local ethics and welfare committee. Consent was obtained from all anaesthetists to use the information in the reports and during morbidity and mortality rounds. 2
Setting and participants The hospital in which the audit was performed provides both first opinion and referral healthcare services for a range of small animals, equine and farm animals. The anaesthesia service is responsible for approximately 1500e1700 anaesthetics per annum, the majority being in dogs (approximately 1000 cases), followed by cats, horses, farm animals and others (approximately 600, 50, 25 and 10 cases, respectively). Anaesthesia is performed in several discrete locations in the hospital: a small animal surgical unit; small animal endoscopy, dental and electro-diagnostics suite; two small animal radiography suites; a magnetic resonance imaging suite; a computed tomography suite; a radiotherapy unit and a large animal surgical unit. During the period of the audit, anaesthetics were performed by three European College Veterinary Anaesthesia and Analgesia Diplomates, one post-residency clinical anaesthetist, three anaesthesia residents, one registered veterinary nurse and a number of rotating interns. Final year veterinary students were involved with many of the cases, rotating through the anaesthesia service for approximately 80% of the time the audit was conducted. Data sources and collection The audit was performed over a period of 23 months between November 2014 and October 2016. The number of anaesthetics performed during this period was determined using electronic hospital records. Based on the World Health Organization's International Classification for Patient Safety, “Safety Incidents” were defined as any event or circumstance which could have resulted, or did result, in unnecessary harm to an animal or member of staff (Runciman et al. 2009). All the staff members performing anaesthesia were asked to report any such incidents occurring within the peri-anaesthetic period of an animal in a handwritten “Safety Incident Diary” held centrally in the anaesthesia induction suite. Whether to report an incident was left to the individual anaesthetist's discretion, but it was encouraged that any incident, irrespective of severity, be reported. Fatalities and significant anaesthetic related morbidities were also reported. In this study, the peri-anaesthetic period was defined as any time point in which an anaesthetist could be in contact with an animal between pre-anaesthetic assessment and discharge.
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
Limited guidance was given regarding the format of incident reporting. The anaesthetist was asked to provide a short narrative without giving reference to individual animals or staff; the use of broader terms such as species, surgeon, clinician, resident, intern, student, nurse or assistant staff was encouraged. These notes then formed the basis for subsequent anaesthesia mortality and morbidity rounds, planned for every 4e8 weeks. Meetings were conducted in a “no-blame” fashion, with anaesthetists encouraged to look beyond individual culpability at the underlying factors that may have contributed to the incident as well as its impact. This allowed the team to offer their opinion on any contributing, causative or ameliorating factors specifically anaesthetist intervention. Ways in which similar incidents could be avoided or better managed in the future were also discussed. To encourage an open forum, incidents involving senior staff were discussed at the start of the first few meetings. Thereafter, incidents were discussed in chronological order. Neither transcripts nor recordings of the meetings were made, but brief notes documenting the timeline of the incident and contributing factors were added to the existing records. Analysis At the end of the audit period, all reports and accompanying notes from meetings were evaluated. Further analysis was performed when clear problems with the delivery of care could have, or did result, in harm to an animal or staff member. Additionally, all reported fatalities and significant post-anaesthetic morbidities were included for analysis, irrespective of the presence of an obvious problem in the delivery of care. Incident inclusion and exclusion is outlined in Fig. 1. Incidents which impacted scheduling or staffing but not safe veterinary care were excluded, while those prolonging the procedure or general anaesthetic time were included. The aim of the analysis was to identify contributing factors and themes implicated in the incident reports and subsequent discussions. A simplified “Systems” analysis was performed using a framework of contributing factors modified for use in veterinary anaesthesia (Vincent et al. 2000; Woloshynowych et al. 2005). Contributing factors were divided into six main categories: 1) Animal and Owner; 2) Task and Technology; 3) Individual; 4) Team; 5) Work Environmental; and 6) Organizational and Management. A comprehensive list of contributing factors
and examples is provided in Appendices S1eS6. Every report was assessed systematically by a single assessor (main author) to identify contributing factors from each level of this framework (Fig. 2). Due to the nature and details of the reports and notes taken subsequently during the morbidity and mortality rounds, independent analysis was not possible. In order to allow comparison to previous data produced by Hofmeister et al. (2014), Medication Errors and Equipment Failures were also identified. These represented incidents primarily associated with the administration of drugs, or errors such as leaving adjustable pressure limiting (APL) valves closed. Anaesthetist Non-Technical Skills, comprising Decision Making, Situation Awareness, Teamwork and Task Management (Flin et al. 2010), were grouped and compared with the traditional competencies of Knowledge, Experience and Technical Skills. The estimated impact of the incident on the animal was assessed using a simple descriptive scale based upon similar scales used in human medicine (Table 1) (Cook et al. 2009). Statistical methods Descriptive statistics and qualitative data are reported. Normality of data was assessed using the ShapiroeWilk test where necessary. Non-normally distributed data are described using the median (interquartile range). Associations were assessed between factors which were identified together on 10 occasions using Pearson's Chi-square test for association. Data is reported as X2 (degrees of freedom, sample size) Pearson’s Chi-square value, p value. Statistical significance was set at p ¼ 0.05 before Bonferroni correction. Outcome analysis was conducted using logistic regression analysis, considering all cases that experienced major or catastrophic outcomes as a result of a safety incident. Only those contributing factors with n > 10 animals and p < 0.1 were included for multivariable analysis with forwards stepwise elimination. Final logistic regression fit was assessed using receiver-operating characteristic curves and area under the curve calculations, reported with 95% confidence intervals (CI), with the significance level set at p < 0.05. Results are reported as an odds ratio (OR) with respective 95% CI for factors with p < 0.1 and p < 0.05 for univariable and multivariable analysis. All statistics were performed by using SPSS version 23(IBM).
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
3
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
Q4
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
Total number of anaesthetics (n = 3379)
Safety diary entries (n = 204) Excluded entries not impacting animal or staff safety (n = 30)
Reported peri-anaesthetic fatalities and morbidities (n = 14)
Reported safety incidents (n = 160)
Cases included for further analysis (n = 174) Excluded cases with environmental impact only (n = 11) Cases included for logistic regression analysis (n = 163) Figure 1 Inclusion and exclusion flow chart for the handling and further analysis of safety incidents.
Results
Q5
During the audit period, 3379 anaesthetics were performed and 204 diary entries were made. The audit identified 160 safety incidents and a further 14 cases with significant morbidity or mortality. Therefore, 174 (prevalence 5.1%) cases were included for further analysis. Fourteen morbidity and mortality rounds were performed within the audit period with a gap of 4e14 weeks between meetings. Total monthly reported incidents ranged between 0 and 20, with a median of 7 (3e5). There was no discernible pattern to the number of reports made per month; reporting rates did not appear to trend upward or downward over time or be linked with case number. Of the total incidents, 159 directly involved individual animals; these included 17 Medication Errors (prevalence 0.5%), 11 Equipment Failures (prevalence 0.3%) and 4 incidents of APL valves being left closed 4
(prevalence 0.1%). Fifteen incidents were associated with generalized problems: four incidents involved pipeline oxygen pressure dropping because of back up cylinder failure requiring the use of machine-mounted cylinders; in 11 cases, environmental contamination occurred with anaesthetic gases and vapours. A total of 677 contributing factors were identified across all incidents. A median of four (3e5) factors were identified per reported incident, with three (2e4) out of the six framework categories present. Individual factors were identified in 123 (70.7%), Team factors in 108 (62.1%), Organizational and Management factors in 94 (54.0%), Task and Technology factors in 80 (46.0%), Work Environment factors in 53 (30.5%) and Animal and Owner factors in 36 (20.7%) incidents (Table 2). Where factors belonging to the Individual category were identified (n ¼ 123), factors from further down
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
Incident selected for analysis
Animal and Owner factors considered
Animal Condition, Temperament, Financial constraint, Communication with owner, Owner attitude
Report and supplementary notes evaluated
Task and Technology factors considered
Incident timeline developed
Individual factors considered
Experience, Knowledge, Technical skills, Decision making, Task management, Situation awareness, Team working, Health, stress or fatigue
Problems with delivery of care identified
Team factors considered
Supervision, Written communication, Verbal communication, Attitude, Seeking help, Leadership
Contributing factor framework applied
Work Environmental factors considered
Distraction, Layout and ergonomics, Social atmosphere
Impact severity scale applied
Organizational and Management factors considered
Staffing level, Poor scheduling, Culture and priorities, Training, Equipment levels, Unrealistic expectations, Not responding to staff concerns, Planning and preparation
Failure to follow Standard operating procedures, Equipment checks, Equipment failure, Over-reliance on technology, Equipment maintenance, Equipment availability, Drug labelling
Figure 2 Flow chart demonstrating the framework used for the systematic analysis of anaesthetic safety incidents.
the framework in the Team, Work Environmental and Organizational and Management categories occurred concurrently in 89.4% (n ¼ 110) incidents. In the remaining incidents (n ¼ 13, 10.6%) the contributing factor of Health, Stress and Fatigue was identified alongside other factors from the Individual category. Analysis of sub-groups within the Individual category indicated that Anaesthetist NonTechnical Skills (n ¼ 101, 58.0%), comprising Decision Making, Situation Awareness, Teamwork and Task Management contributed more to incidents than the “traditional competencies” such as Knowledge, Experience or Technical Skills (n ¼ 65, 37.4%). Pearson’s Chi-square test for associations was performed for 39 factor combinations with n 10.
Bonferroni correction set statistical significance at p < 0.0013. Significant associations were identified between Experience and Supervision, X2 (1, n ¼ 174) ¼ 54177, p ¼ 0.001, Failure to follow a standard operating procedure (SOP) and Task Management, X2 (2, n ¼ 174) ¼ 11318, p ¼ 0.001, and Staffing and Poor Scheduling, X2 (1, n ¼ 174) ¼ 36742, p ¼ 0.001. In total, 163 (93.7%) incidents impacted the animal. Of these, the consequences of 17 (9.8%) were classified as catastrophic, 9 (5.2%) as major, 24 (13.8%) as moderate, 77 (44.2%) as minor and 36 (20.7%) as insignificant. The impact of the incident on the animal was believed to have been limited by anaesthetist intervention in 104 (63.8%) of these
Table 1 Simple descriptive scale to grade the severity of the incident impact on the animal Severity
Description
Insignificant Minor Moderate Major Catastrophic
Incident Incident Incident Incident Incident
had little or no effect on the animal led to an increased level of care or caused minor short-term harm only led to short-term but significant loss or impairment of physiological function led to a significant loss or impairment of physiological function led to death or injury that led directly to euthanasia
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
5
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus Table 2 Overview of the prevalence of contributing factors identified during the analysis of the reported safety incidents Framework category
n (%)
Sub-categories
Sub-total n (%)
Patient and Owner
38 (5.6)
Task and Technology
100 (14.8)
Individual
194 (28.7)
Team
148 (21.9)
Work Environmental
54 (7.9)
Organizational and Management
143 (21.1)
Animal condition Temperament Financial constraint Communication with owner Owner attitude Failure to follow SOP Equipment checks Equipment failure Over-reliance on technology Equipment maintenance Equipment availability Drug labelling Decision making Experience Health, stress, fatigue Task management Knowledge Situation awareness Technical skills Team working Supervision Written communication Verbal communication Attitude Seeking help Leadership Distraction Layout and ergonomics Social atmosphere Staffing level Poor scheduling Culture and priorities Training Equipment levels Unrealistic expectations Not responding to staff concerns Planning and preparation Lack of SOP Failure of SOP
30 (78.9) 5 (13.3) 1 (2.6) 1 (2.6) 1 (2.6) 51 (51.0) 28 (28.0) 11 (11.0) 4 (4.0) 2 (2.0) 2 (2.0) 2 (2.0) 65 (33.5) 44 (22.7) 28 (14.4) 24 (12.4) 16 (8.2) 11 (5.7) 5 (2.6) 1 (0.5) 48 (32.4) 38 (25.7) 33 (22.3) 23 (15.5) 4 (2.7) 2 (1.4) 42 (77.8) 8 (14.8) 4 (7.4) 53 (37.0) 32 (22.4) 28 (19.6) 12 (8.4) 7 (4.9) 3 (2.1) 3 (2.1) 3 (2.1) 1 (0.7) 1 (0.7)
SOP, standard operating procedure.
163 incidents. Of the catastrophic consequences, three (17.6%) were considered potentially preventable, whereas the remainder (82.4%) were associated with severe disease, injury or surgical catastrophe. Of the 174 reported incidents, 11 (6.3%) did not result in harm to the animal but were associated with environmental contamination and therefore posed a risk to staff. All the 163 cases with consequences for the animal were considered for logistic regression analysis to assess the relationship between contributing factors and a combined major and catastrophic outcome (n ¼ 26, 16.0%). Following multivariable analysis, 6
Animal Condition (p < 0.001) and anaesthetist Decision Making (p ¼ 0.023) were significantly associated with developing a major or catastrophic outcome as a result of the safety incident. Results of the univariable and multivariable analyses are shown in Tables 3 and 4, respectively. Discussion Key results Our audit demonstrated that safety incidents occurred at a rate of approximately 1 in every 20 anaesthetics during the study period. The impact of
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus Table 3 Results of univariable logistic regression for contributing factors with p < 0.1 when assessing incidents with major or catastrophic impact for the animal. Percentages are in terms of the total number of contributing factors identified (n ¼ 677) Factor
n (%)
p
OR (95% CI)
Animal condition Task and technology Failure to follow SOP Experience Decision Making Situation awareness Work Environmental Equipment levels
30 (4.4) 100 (14.8) 51 (7.5) 44 (6.5) 65 (9.6) 11 (1.6) 54 (8.0) 7 (1.0)
< 0.001 0.035 0.031 0.001 0.017 0.068 0.095 0.065
18017 (6698e48468) 0.352 (0.133e0.932) 0.195 (0.044e0.864) 4345 (1814e10404) 2873 (1211e6817) 3377 (0.912e12501) 0.384 (0.125e1183) 4337 (0.910e20660)
CI, confidence interval; OR, odds ratio; SOP, standard operating procedure.
Table 4 Results of multivariable logistic regression for contributing factors with p < 0.05 when assessing incidents with major or catastrophic impact on the animal. Percentages are given in terms of the total number of cases that experienced adverse events (n ¼ 163) Factor
n (%)
p
OR (95% CI)
AUC (95% CI)
Animal condition Decision Making
30 (18.4%) 65 (40.0%)
< 0.001 0.023
16210 (5573e47147) 3437 (1184e9974)
0.779 (0.666e0.893) 0.629 (0.511e0.747)
AUC, area under the curve; CI, confidence interval; OR, odds ratio.
the majority of incidents was considered to be insignificant, minor or moderate and therefore was not felt to have caused any significant or lasting harm. However, it should be noted that in almost two-thirds of cases, consequences were considered to have been limited by the intervention of an anaesthetist. Although some problems in delivery of care were identified when analysing morbidity and mortality data, these were not thought to have adversely affected outcome as the disease process or injuries suffered by the animal were untreatable in the majority of cases. Nevertheless, these findings could highlight areas where veterinary care may be improved. Overall, the most commonly identified contributing factors belonged to the Individual and Team categories. Within the Individual category, anaesthetist Non-Technical Skills was identified more often than Knowledge, Experience or Technical Skills. This highlights the importance of skills, such as Decision Making, Situation Awareness, Team Working and Task Management, in the provision of safe anaesthesia. Contributing factors from the Individual category were either identified with other factor categories further down the framework or together with the Individual contributing factor Health, Stress and Fatigue. These findings highlight the importance of looking beyond individual culpability when
analysing the causality of incidents, i.e. other external factors may be negating influencing the performance of the individual, and suggest staff health and welfare is also an area worthy of consideration. Although multiple contributing factors were identified for all incidents reported, only few statistically significant associations were noted. Perhaps unsurprisingly, these were found between Anaesthetist Experience and Supervision as well as between Staffing Levels and Scheduling and between Failure to follow an SOP and Task management. It would appear logical that less experienced staff members without ready access to appropriate supervision may be involved in more incidents, and also that Staff numbers and Scheduling, and Failure to follow an SOP and Task management are intimately related. Outcome analysis indicated that most catastrophic events were strongly linked to the animal's underlying condition. This may indicate that animals with serious underlying conditions are more physiologically unstable and therefore are less able to tolerate the insult of an incident. This perhaps leads to the anaesthetist having less time to intervene and avert serious consequences. It is of note that problems with Decision Making were also identified in many of these cases which could indicate that Decision Making during critical time and stressful situations is an area
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
7
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
that warrants further investment in terms of education and training. It may also reflect an overly critical retrospective analysis of an individual’s performance during crisis situations. Limitations Voluntary reporting systems are likely to underreport safety incidents. Reasons cited for this include time constraints, lack of or poor access to an appropriate reporting system, uncertainty of what to report and the feeling that nothing will be done to counter contributing factors (Garrouste-Orgeas et al. 2012). Furthermore, fear of repercussions, such as blame, litigation or disciplinary action following reporting an incident is another significant barrier (Heard et al. 2012). Even if the incident is reported, individuals may still underplay their involvement, feel embarrassed, shame or guilt regarding the event. These feelings may have been exacerbated in this study by anaesthetists knowing the incidents would be discussed later in morbidity and mortality rounds (Heard et al. 2012) even if assurances were given about confidentiality and a “no-blame” culture. In addition, the reporting method permitted only a relatively superficial analysis of contributing factors. Details were often not elucidated until the incident was discussed at morbidity and mortality meetings, and although these discussions allowed a more thorough narrative of the incident and surrounding circumstances to be developed, this retrospective assessment may have left out significant contributing factors and potentially introduced hindsight bias. Furthermore, only a single person with intimate knowledge of the working environment and circumstances during the audit period was able to perform the systems analysis, adding a certain degree of observer bias. Nonetheless, we believe that the data represents a fair and balanced assessment of the contributing factors that were encountered in the hospital during the timeframe of the audit. Interpretation In light of these limitations, the findings of this audit must be considered in context. Although the analysis used was designed to provide a framework for both, quick and easy incident analysis as well as more indepth investigations, brief incident reports such as those collected during this study can only allow a superficial analysis and are unlikely to fully illuminate all the factors contributing to an incident. In 8
truth, a more in-depth analysis of a few keys incidents is likely to elucidate more specific information about the failings of a system and allow more targeted action to be taken than superficial analysis of a large number of incidents (Vincent et al. 1999a). Nonetheless, identifying even a few themes and factors should help focus attention to problem areas and thus may present opportunities to improve animal safety. In this study, information on incidents was gathered through written reports and via morbidity and mortality round discussions. Despite being a traditional practice and being embedded in the curriculum of veterinary and medical education, the structure and function of morbidity and mortality rounds appears poorly defined (Kuper et al. 2010; Higginson et al. 2012). Often morbidity and mortality rounds are used to discuss clinical conundrums and correlations between clinical and pathological findings and are explicitly aimed at improving medical knowledge (Kuper et al. 2010). Indeed, this was the case in our hospital prior to this investigation. However, through the performance of this study, we have found that the discussion of reported safety incidents during morbidity and mortality rounds provided an open forum to specifically address safety issues and identify potential solutions. Establishing blame-free group discussions had the additional benefit of allowing individual anaesthetists to reflect on their role in the incident and consider factors contributing to the incident both inside and outside of their control. This positive approach may have helped encourage continued reporting as demonstrated by a consistent reporting rate. Generalizability The ability to apply these findings to other clinical settings may be limited by the unique working environments and circumstances of different hospitals. However, a proportion of contributing factors and themes identified in this audit may well be universal. Differences in contributing factors may account for variations in incident rates observed between different hospitals. The only previous study to investigate safety incidents in veterinary anaesthesia found a greater frequency of medication error and closed APL valves than this study. Prior to making a number of safety interventions, Hofmeister et al. (2014) reported 25 medication errors and 20 incidents of APL valves being left closed in 2028 cases (prevalence, 1.2% and 1.0%, respectively). This compares to 17 medication errors and 4 incidents of APL valves being left closed in 3379 animals in this
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Systems analysis of safety incidents MW McMillan and KS Lehnus
study (0.5% and 0.11%, respectively). The reasons for these differences are unclear but likely reflect differences in individual working practices and protocols. It also suggests that a set of safety interventions suited to one clinical environment may not have the same impact in another with a different environment and working conditions. For Hofmeister et al. (2014), the instigation of SOP when administering medications to animals and a check box on the anaesthetic record regarding checking the APL valve reduced the frequency of medication errors to 14 and that of closed APL valves to 5 incidents in 2112 animals (0.7% and 0.2%, respectively). These are similar levels to those seen in our audit despite no specific safety measures being in place to avoid these incidents in our hospital. Future studies In order to perform more in-depth analyses, the quality and quantity of information obtained from incident reports must be improved. A better timeline that includes the sequence of events and the individual’s thought processes and feelings during the incident should be generated. This can be achieved in a way by encouraging a more information-rich narrative. Electronic reports are used in most safetyconscious industries, including human medicine, in order to facilitate this. The Australian Incident Monitoring System, the Anesthesia Incident Reporting System and the National Reporting and Learning System are three electronic systems available to report anaesthetic safety incidents (Woloshynowych et al. 2005). Such systems could be shared between institutions allowing information to be pooled and shared. In addition to this, transcripts or recordings of meetings and semi-structured interviews with individual anaesthetists could be utilized to improve the quality of safety data. Finally, reporting and analysing safety incidents alone cannot improve safety in isolation. After significant contributing factors are identified, interventions should be made to counter these factors, either by making attempts to remove them or putting barriers in place to avoid the potential consequences. The effects these interventions and subsequent modifications to operating processes have on safety should then be assessed through recurrent clinical audit cycles. This feedback loop has the added advantage of staff seeing reports actioned, which should in turn help encourage continued reporting (Vincent et al. 1999b). In this study, Communication
and Scheduling were two readily actionable areas highlighted as potentially contributing to a high number of incidents, and this has led to changing practices in these areas. The impact of these changes is yet to be explored. Methods to support anaesthetist Decision Making especially during crisis situations are also being developed. Conclusion Safety incidents occur at a measurable rate in veterinary anaesthetic practice, and simple methods can be used to gather data on their nature and frequency. Analytical frameworks developed for the use in human medicine can be used with minimal modification to identify potential contributing factors in a veterinary setting. Although Individual factors contribute to a significant proportion of incidents, they occur concurrently with other factors in a majority of cases. It is also significant that interventions by anaesthetists appear to limit the consequences of incidents for the majority of animals, highlighting the importance of resilience and non-technical skills such as situation awareness in anaesthesia. Such investigations can potentially be used to improve anaesthetic safety through the development of targeted interventions. Authors' contributions MWM: study concept and design, data acquisition, analysis of reports, manuscript authorship; KSL: data acquisition and sorting, statistical analysis, interpretation of data, tables and figures, editing and major revisions of manuscript. Funding This study received no financial backing Conflict of interest statement The authors declare no conflict of interest. References National Research Ethics Service: Defining Research NRES guidance to help you decide if your project requires review by a Research Ethics Committee, 2009. http://www.nres.npsa.nhs.uk/applications/is-yourproject-research/. Last accessed 20 October 2014. Bidwell LA, Bramlage LR, Rood WA (2007) Equine perioperative fatalities associated with general anaesthesia at a private practiceea retrospective case series. Vet Anaesth Analg 34, 23e30. Bille C, Auvigne V, Libermann S et al. (2012) Risk of anaesthetic mortality in dogs and cats: an observational
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
9
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
Q6
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Systems analysis of safety incidents MW McMillan and KS Lehnus cohort study of 3546 cases. Vet Anaesth Analg 39, 59e68. Brodbelt DC, Blissitt KJ, Hammond RA et al. (2008) The risk of death: the confidential enquiry into perioperative small animal fatalities. Vet Anaesth Analg 35, 365e373. Clarke KW, Hall LW (1990) A survey of anaesthesia in small animal practice: AVA/BSAVA report. Vet Anaesth Analg 17, 4e10. Cook TM, Bland L, Mihai R, Scott S (2009) Litigation related to anaesthesia: an analysis of claims against the NHS in England 1995-2007. Anaesthesia 64, 706e718. Flin R, Patey R, Glavin R, Maran N (2010) Anaesthetists' non-technical skills. Br J Anaesth 105, 38e44. Gaba DM (2000) Anaesthesiology as a model for Patient safety in health care. BMJ 320, 785e788. Garrouste-Orgeas, Philipart F, Bruel C et al. (2012) Overview of medical errors and adverse events. Ann Intensive Care 2, 2. Heard GC, Sanderson PM, Thomas RD (2012) Barriers to adverse event and error reporting in anaesthesia. Anesth Analg 114, 604e614. Higginson J, Walters R, Fulop N (2012) Mortality and morbidity meetings: an untapped resource for improving the governance of patient safety. BMJ Qual Saf 21, 576e585. Hofmeister EH, Quandt J, Braun C, Shepherd M (2014) Development, implementation and impact of simple patient safety interventions in a university teaching hospital. Vet Anaesth Analg 41, 243e248. Johnston GM, Eastment JK, Wood JLN, Taylor PM (2002) The confidential enquiry into perioperative equine fatalities (CEPEF): mortality results of Phases 1 and 2. Vet Anaesth Analg 29, 159e170. Kuper A, Zur Nedden N, Etchells E et al. (2010) Teaching and learning in morbidity and mortality rounds: an ethnographic study. Med Educ 44, 559e569.
10
Leistikow I, Mulder S, Vesseur J, Robben P (2016) Learning from incidents in healthcare: the journey, not the arrival, matters. BMJ Qual Saf. Published Online First: 01 April 2016. Q7 Runciman W, Hibbert P, Thomson R, Van Der Schaaf T, Sherman H, Lewalle P (2009) Towards an International Classification for Patient Safety: key concepts and terms. Int J Qual Health Care 21, 18e26. Vincent CA, Taylor-Adams S, Chapman EJ, Hewett D et al. (1999a) A protocol for the investigation and analysis of clinical incidents. University College London/Association of Litigation and Risk Management, London. Vincent C, Stanhope N, Crowley-Murphy M (1999b) Reasons for not reporting adverse incidents: an empirical study. J Eval Clin Pract 5, 13e21. Vincent C, Taylor-Adams S, Chapman EJ et al. (2000) How to investigate and analyse clinical incidents: Clinical Risk Unit and Association of Litigation and Risk Management protocol. BMJ 320, 777e781. Vincent CA (2004) Analysis of clinical incidents: a window on the system not a search for root causes. Qual Saf Health Care 13, 242e243. Woloshynowych M, Rogers S, Taylor-Adams S, Vincent C (2005) The investigation and analysis of critical incidents and adverse events in healthcare. Health Technol Assess 9. Received 6 January 2017; accepted 7 June 2017. Available online xxx
Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.vaa.2017.06. 007.
© 2017 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by Elsevier Ltd. All rights reserved., ▪, 1e10
Please cite this article in press as: McMillan MW, Lehnus KS Systems analysis of voluntary reported anaesthetic safety incidents occurring in a university teaching hospital, Veterinary Anaesthesia and Analgesia (2017), http://dx.doi.org/ 10.1016/j.vaa.2017.06.007
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64