Communication failures contributing to patient injury in anaesthesia malpractice claims☆

Abstract

Background

Communication amongst team members is critical to providing safe, effective medical care. We investigated the role of communication failures in patient injury using the Anesthesia Closed Claims Project database.

Methods

Claims associated with surgical/procedural and obstetric anaesthesia and postoperative pain management for adverse events from 2004 or later were included. Communication was defined as transfer of information between two or more parties. Failure was defined as communication that was incomplete, inaccurate, absent, or not timely. We classified root causes of failures as content, audience, purpose, or occasion with inter-rater reliability assessed by kappa. Claims with communication failures contributing to injury (injury-related communication failures; n=389) were compared with claims without any communication failures (n=521) using Fisher's exact test, t-test, or Mann–Whitney U-tests.

Results

At least one communication failure contributing to patient injury occurred in 43% (n=389) out of 910 claims (κ=0.885). Patients in claims with injury-related communication failures were similar to patients in claims without failures, except that failures were more common in outpatient settings (34% vs 26%; P=0.004). Fifty-two claims had multiple communication failures for a total of 446 injury-related failures, and 47% of failures occurred during surgery, 28% preoperatively, and 23% postoperatively. Content failures (insufficient, inaccurate, or no information transmitted) accounted for 60% of the 446 communication failures.

Conclusions

Communication failure contributed to patient injury in 43% of anaesthesia malpractice claims. Patient/case characteristics in claims with communication failures were similar to those without failures, except that failures were more common in outpatient settings.

Editor's key points.

• •The authors investigated the role of communication failures in patient injury using the Anesthesia Closed Claims Project database.
• •Failures in perioperative communication contributed to patient injury in 43% of 910 anaesthesia malpractice claims.
• •Patient characteristics in claims with communication failures were similar to those without failures.
• •The most common root cause of communication failure was insufficient or inaccurate information.

Communication between healthcare team members and with their patients is critical to providing safe and effective medical care.¹ Communication failures can jeopardise patient safety. Observational and video studies in surgery have found that communication failures occurred frequently during surgery: they were observed in 30% of procedurally relevant communications,² occurred every few minutes during the procedure,³ and contributed to episodes that compromised patient safety.⁴ These studies focused on surgery, rather than anaesthesia, and the direct contribution of communication failures to patient injury remains unexplored.

Malpractice claims represent a valuable collection of patient injury analyses used to identify safety problems in anaesthesia.⁵ Most of these analyses focused on clinical management rather than human factors or team performance. In surgery, analyses of 444 malpractice claims identified 60 claims (13.5%) with 81 communication breakdowns associated with patient injury.⁶ Analysis of the contribution of anaesthesia communication failures to patient injury is lacking.

We analysed communication failures contributing to patient injury in anaesthesia malpractice claims. We investigated the quantity of communication failures, the most common types and characteristics of failures, and how claims with communication failures differed from those without failures. We hypothesised that communication failures would be identified as contributing to patient injury or its severity in a substantial proportion of anaesthesia claims, and that patient and case characteristics would differ in claims with failures from those without failures.

Methods

The Anesthesia Closed Claims Project (CCP) database is a collection of closed anaesthesia malpractice claims previously described in detail.⁷ The University of Washington Institutional Review Board determined that this secondary analysis of de-identified CCP data does not involve human subjects as defined by federal and state regulations (study 00009904). Data were abstracted by board-certified practising anaesthetists from closed claims files at malpractice insurance companies across the USA. Data collection used an extensive standardised survey form, including surgery type, details of anaesthesia care, patient demographics and outcomes, and payments made. The anaesthetist–reviewer evaluated the type and severity of injury, cause of injury, and the appropriateness of anaesthesia care. The anaesthetist–reviewer also summarised the sequence of events and causes of injury in a narrative, which contained additional details. All data collection forms underwent rigorous quality review by the CCP oversight committee.

For this study, we used the CCP database of 10,546 claims, including events dating from 1970 to 2013. We classified the variables and planned the qualitative analysis before accessing the data. Inclusion criteria were claims associated with surgical/procedural/obstetric anaesthesia and postoperative (acute) pain management, where the adverse event occurred in the year 2004 or later (n=1132; Fig. 1). Claims associated with chronic pain medicine were not included. These 1132 claims were manually reviewed by the authors. Claims with insufficient information to assess whether a communication failure occurred (n=36) and claims in which a communication failure occurred but did not contribute to the patient injury (n=186) were excluded, leaving 910 claims for analysis. Evaluation of communication failures from claim narratives was independently conducted by the authors using a standardised survey form.

Fig 1.

Study inclusion. Injury-related failure claims: better communication would possibly have ameliorated the injury or the severity of the injury. The 389 claims with injury-related communication failures included a total of 446 communication failures (n=52 claims had >1 communication failure).

Definition of variables

Communication was defined as an attempt or the need to transfer information between two or more parties. A communication failure was defined as any such information transfer that was absent, incomplete, inaccurate, or untimely. Two authors (KBD and LSS) identified whether a communication failure occurred in each claim or not. Disagreements were adjudicated by a third author (KLP). Two anaesthetist-authors (KBD and JMD) then independently evaluated claims with failures to determine whether better communication might have potentially ameliorated the patient's injury or the severity of the injury. Disagreements were resolved through discussion. Claims in which these authors agreed that better communication might have ameliorated the patient's injury or its severity were classified as injury-related failures. Each failure was classified as to type, location, personnel, timing, topic, and communication method.

Two authors (KBD and LSS) classified the type of communication failure using definitions adapted from Lingard and colleagues² to identify four primary types of failures: occasion, audience, purpose, and content. Occasion failure was defined as a problem in the situation or context of a communication event² (e.g. failures in timing, such as delayed test results). Audience failure was defined as a gap in the composition of the group engaged in communication.² Purpose failure was defined as a communication event, in which the purpose was unclear, not achieved, or inappropriate² (e.g. disagreement between an anaesthetist and a surgeon regarding the appropriate clinical action). Content failure was defined as a situation, where insufficient, inaccurate, or no information was transmitted.² Modifications adapted the definitions of Lingard and colleagues² to this analysis of malpractice claims, which included wording changes and additional definitions beyond those described by Lingard and colleagues² (Supplementary Appendix).

Communication failure methods were classified as oral, written, both oral and written, test results, and paging. The topic of communication was classified as preoperative information, intraoperative events, postoperative care, patient status changes, and other. Personnel was defined as the anaesthesia team, surgical team, nursing, patient or patient's family, and other. Handover communication was defined as the transfer of information during transition of responsibility of patient care between healthcare professionals.⁸ For this analysis, we only examined anaesthesia handovers, defined as transfers of information between members of the anaesthesia team or between the anaesthesia team and other providers. We restricted the handover analysis to oral communication failures that occurred during the procedure (breaks or shift change), or postoperatively during transfer of patient care to the postoperative care team.

Severity of injury was collapsed into three categories: temporary/minor injuries, permanent disabling injuries, and death. Appropriateness of anaesthesia care was assessed as appropriate (based on reasonable or prudent practice at the time the event occurred), less than appropriate, or impossible to judge. Anaesthesia payments included payments made on behalf of anaesthetists and anaesthesia corporations.

Statistical analysis

Inter-rater reliability was determined for whether a communication failure occurred or not, the type of failure, and whether better communication might have ameliorated the injury or severity of injury using kappa scores. Anaesthesia payments were adjusted to 2019 dollar amounts with the Consumer Price Index,⁹ with median and inter-quartile range reported because payments were not normally distributed. Claims with injury-related communication failures were compared with claims without communication failures using Fisher's exact test, independent t-test for equality of means, or Mann–Whitney U-test with two-tailed tests and P<0.05 as the criterion for statistical significance. The sample size was based on available data; no a priori power analysis was conducted. All statistical analysis used SPSS 19 for Windows (International Business Machines Corporation, Armonk, NY, USA).

Results

Of the 910 eligible claims, 389 (43%) had at least one injury-related communication failure and 521 had no communication failures. Agreement on whether a communication failure occurred or not was excellent (κ=0.885). Most injury-related failure claims had only one communication failure (n=337; 87%), although 47 had two and five had three or more failures, resulting in a total of 446 failures. Patient injuries occurred between 2004 and 2013.

Patients in claims with injury-related communication failures were similar to patients in claims without communication failures (Table 1), although failures were more common in outpatient settings (34% vs 26%; P=0.004; Table 1). Although general anaesthesia was the most common technique, claims with failures were more likely to involve monitored anaesthesia care (21%) compared with other claims (8%; P<0.001; Table 1). There was no difference in the distribution of claims by category of anaesthesia care between those with failures and those without (Table 1).

Table 1.

Characteristics of claims with communication failures associated with injury and claims without communication failures. ∗Monitored anaesthesia care refers to sedation provided by an anaesthetist with preservation of spontaneous breathing and airway reflexes. N=910 claims unless otherwise stated. Claims with missing data excluded. Payment amounts in 2019 inflation adjusted dollars. P-values by Fisher's exact test, χ² test, t-test (age), and Mann–Whitney U-test (payment amount). Odds ratios for claims with one or more communication failures related to injury compared with claims with no communication failure. ASA, American Society of Anesthesiologists; NA, not applicable; sd, standard deviation.

	Claims with failures related to injury (n=389), n (%)	No communication failures (n=521), n (%)	Odds ratio (95% confidence interval)	P-value
Male	166 (43%)	228 (44%)	0.957 (0.734–1.247)	0.398
Age (yr), mean (sd) (n=904)	51 (19)	49 (18)	1.004 (0.997–1.012)	0.225
ASA physical status 3–5 (n=901)	209 (54%)	275 (53%)	1.041 (0.799–1.356)	0.410
Emergency (n=901)	80 (21%)	85 (16%)	1.330 (0.948–1.866)	0.059
Obese (n=703)	139 (49%)	190 (45%)	1.155 (0.854–1.563)	0.195
Outpatient (n=893)	131 (34%)	132 (26%)	1.489 (1.114–1.989)	0.004
Primary anaesthetic technique				<0.001
General	215 (55%)	354 (68%)	0.664 (0.425–1.038)
Regional	48 (12%)	80 (15%)	0.656 (0.380–1.133)
Monitored anaesthesia care∗	83 (21%)	40 (8%)	2.268 (1.296–3.970)
Other/none/unknown	43 (11%)	47 (9%)	Reference
Anaesthesia service				0.369
Surgical/procedural	297 (76%)	411 (79%)	1.032 (0.388–2.743)
Obstetric	41 (11%)	59 (11%)	0.993 (0.349–2.822)
Acute pain	44 (11%)	41 (8%)	1.533 (0.534–4.405)
Resuscitation/intubation	7 (2%)	10 (2%)	Reference
Severity of injury				0.083
Temporary or non-disabling	172 (44%)	236 (45%)	Reference
Permanent and disabling	77 (20%)	129 (25%)	0.819 (0.581–1.155)
Death	140 (36%)	156 (30%)	1.231 (0.911–1.664)
Care less than appropriate (n=824)	238 (68%)	145 (31%)	4.880 (3.622–6.757)	<0.001
Anaesthetist/corporation paid claim (n=906)	239 (62%)	221 (43%)	2.156 (1.648–2.820)	<0.001
Anaesthetist/corporation median payment (n=460)	$200 288	$297 025	NA	0.138
Anaesthetist/corporation payment range (n=460)	$63 765–$596 094	$71 286–$681 250

Severity of injury was similar between claims with and without communication failures (Table 1). Death occurred in 36% of claims with failures. The most common severe injuries were brain damage (n=34), nerve injury (n=15, with 10 being spinal cord injury), eye injury (n=12), airway injury (n=8), and burns (n=8). Claims with communication failures were more likely assessed as less than appropriate anaesthesia care (68% vs 31%; P<0.001) and were more likely to result in payment to the plaintiff (62% vs 43%; P<0.001; Table 1). There was no difference in payment amounts between claims with and without communication failures.

Types of communication failures

Typical examples of the types of communication failures are shown in Table 2. Content failures accounted for 268 (60%) of the 446 communication failures (κ=0.833; Table 3). In 31% of all failures, the content failure consisted of never providing information. Common failures of this type involved missing preoperative information or failure to communicate about developing intraoperative complications. In another 15% of all failures, correct information was provided, but not reviewed. These involved preoperative information and preoperative and postoperative test results.

Table 2.

Communication failure examples by type of failure.

Communication failure examples
Content failures
Medication management: a 70–80-yr-old ASA 3 man with multiple comorbidities, including hypertension, congestive heart failure, atrial fibrillation, and a prior stroke, was scheduled for elective orthopaedic surgery. The patient stopped warfarin 7 days before surgery, as instructed by the surgeon. The anaesthetist cancelled the case for further cardiac evaluation. Neither the anaesthetist nor the surgeon instructed the patient to restart warfarin. The patient was off warfarin for 17 days before the rescheduled surgery. He sustained an intraoperative stroke and died a few days later.
On-patient fire: a 55–65-yr-old ASA 2 woman presented for blepharoplasty under monitored anaesthesia care. There was no preoperative discussion between the anaesthetist, surgeon, scrub technician, and circulating nurse regarding fire prevention; 3 L min−1 of oxygen via nasal cannula was administered and propofol infusion started by the anaesthetist. The patient was prepped and draped. The surgeon applied the cautery to the patient's face and a spark occurred, igniting the drapes and melting the nasal prongs. The fire was extinguished. The patient sustained third-degree burns to her face, requiring skin grafting, but was left with visible scars.
Audience failures
Wrong site surgery: a 20–30-yr-old man presented for a right Achilles tendon repair under general anaesthesia. The correct surgical site was recorded in the pre-anaesthesia evaluation, the procedure schedule, and on the surgical consent form. The surgeon was talking on the phone during the preoperative ‘time out’ and not paying attention. The procedure was performed on the left ankle.
Purpose failures
Disagreement on treatment of intraoperative complication: a 50–60-yr-old ASA 4E man underwent emergency laparotomy for bowel perforation with ischaemic bowel. During the abdominal wall closure, the oxygen saturations dropped and high ventilation peak pressures occurred during attempted closure. The anaesthetist asked the surgeon to stop closing the abdomen because of concern of abdominal compartment syndrome. The surgeon closed anyway. Oxygen saturations remained very low through the end of the procedure and during the first hour in the ICU, when the patient had a cardiac arrest, was resuscitated, and then returned to the operating theatre. Oxygen saturations improved dramatically with surgical reopening of the abdomen. The patient sustained hypoxic brain damage.
Production pressure during line placement: difficult i.v. catheter placement in preoperative holding resulted in case delay. The surgeon arrived and threatened to cancel the case if not ready in 10 min. The anaesthetist took the surgeon aside to continue the discussion away from patient bedside. Later, central venous catheter placement in the operating theatre was rushed, resulting in a pneumothorax. The patient testified to conflict between surgeon and anaesthetist.
Timing/occasion failures
Delayed test results: a 25–35-yr-old ASA 2 woman presented for spine fusion. The hospital required preoperative pregnancy testing. The patient denied possibility of pregnancy during the pre-anaesthesia evaluation. The test results were not in the electronic hospital record before induction. After incision, the anaesthetist found the positive results. The procedure was aborted. The patient elected for termination and sued for emotional distress.

Table 3.

Types of communication failures related to patient injury (n=446). κ=0.833 for agreement on the four main types of failure.

Type of communication failure	N (%)
Content failure	268 (60)
Information was never provided	138 (31)
Correct information provided, but never reviewed	66 (15)
Insufficient, inaccurate, or ambiguous information was provided	47 (11)
Correct information provided, but misinterpreted	6 (1)
Other content failure	11 (2)
Audience failure	58 (13)
Fails to pass information to appropriate person	13 (3)
Recipient distracted when information provided	9 (2)
Healthcare provider failed to respond to page/call	8 (2)
Recipient of information not in attendance	5 (1)
Other audience failure	23 (5)
Purpose failure	67 (15)
Resolution to healthcare issue is inappropriate	29 (7)
Healthcare providers cannot agree on resolution to issue	25 (6)
Healthcare provider was rude or disrespectful of others	10 (2)
Other purpose failure	3 (1)
Timing/occasion failure	53 (12)
Information should have been provided sooner	53 (12)

Purpose failures accounted for 67 (15%) of the 446 communication failures. These were most commonly inappropriate resolution to a healthcare issue or failure to agree on a resolution (7% and 6%; Table 3). Audience failures accounted for another 58 (13%) of the 446 failures, including failing to pass along information to the appropriate person (3%), failing to respond to a page or call (2%), and distraction whilst information was being communicated (2%; Table 3). Timing/occasion failures accounted for 53 (12%) of the 446 communication failures, all with communication that occurred too late in clinical care to be effective.

Personnel, timing, topics, and communication method

A total of 1112 personnel were involved in 446 injury-related communication failures. An anaesthesia team member was involved in almost all (94%) failures (Table 4); 57% involved a surgical team member, 48% involved nursing (Table 4), and 18% involved the entire operating theatre team. Twenty-two percent of failures involved the patient or family members (Table 4). Half (47%) of communication failures occurred during the procedure, with the remainder divided between pre-procedure (28%) and post-procedure (22%) communication (Table 4).

Table 4.

Personnel, timing and location, topic, and method of communication failures related to patient injury (n=446). ∗Eighteen percent of all errors involved the whole operating theatre team consisting of at least one anaesthetist, surgeon, and nurse. ^†Subtotals sum to 102 rather than 99 because two claims had failures that occurred between PACU and ward, and one claim had communication failure between PACU and ICU.

	N (%)
Personnel involved in failure∗
Anaesthesiology team	420 (94)
Surgical team	253 (57)
Nursing	212 (48)
Patient/family	100 (22)
Other personnel: other physicians, pharmacy, blood bank, and laboratory)	127 (28)
Timing of communication failure
Pre-surgery or other procedure	125 (28)
During surgery or other procedure	210 (47)
Post-surgery or other procedure before discharge	99 (22)
Ward (n=48)†	48
PACU (n=36)†	36
ICU (n=18)†	18
After discharge	6 (1)
Other	6 (1)
Topic of communication failure
Intraoperative event (not related to operating theatre fire)	153 (34)
Preoperative information/medical history	120 (27)
Operating theatre fires related to oxygen and cautery or lasers	55 (12)
Postoperative care plan	54 (12)
Patient status changes post-surgery	47 (11)
Urgency of care needed	12 (3)
Other	5 (1)
Method of communication failure
Oral	276 (62)
Both oral and written	101 (23)
Written	47 (11)
Test results	11 (2)
Pager	11 (2)

The topics of communication in failures were most frequently intraoperative events (n=153; 34%) and preoperative information or medical history (n=120; 27%; Table 4). An additional 12% (n=55) involved operating theatre fire prevention and another 12% (n=54) involved postoperative care plans (Table 4). Communication failures about postoperative changes in patient status accounted for another 11% (n=47) of failures (Table 4).

Most of the 446 injury-related communication failures involved oral communication, either purely oral (n=276; 62%) or communication that was (or could have been accomplished by) oral or written communication (n=101; 23%; Table 4). Test results and pager problems represented relatively small proportions of communication failures (2% each; Table 4).

Table 5 illustrates the personnel pairs involved and type of communication failure by perioperative timing of the failure. The most common pair was anaesthesia team member(s) and the surgical team (with or without nursing; n=166; 40% of 412 failures analysed; Table 5). Most of these failures occurred during the procedure (n=118; 71%; Table 5). Of 67 purpose failures, 36 (54%) occurred between the anaesthesia and surgical teams, with failure to agree (n=20), inappropriate resolution (n=13), and rude behaviour (n=3). Most communication failures between the anaesthesia team and a patient or family occurred before surgery (n=36; 84%; Table 5). These were most often failures to either ask or record medical history pre-surgery, or failure of patient/family to provide relevant medical history.

Table 5.

Personnel involved and type of communication failure by timing of failure. Because of infrequent occurrence, six claims with communication failures post-discharge and six claims with failures at other times were excluded from this analysis. Communication failures that did not involve a member of the anaesthesia team were not included in the analysis of personnel. Each combination of personnel involves only those listed and no others (e.g. communication failures between members of the anaesthesia team plus patient or family exclude cases in which the communication failure also included members of the surgical team). χ2 P-value.

	Pre-surgery or other procedure, n (row %)	During surgery or other procedure, n (row %)	Post-surgery, before discharge, n (row %)	P-value
Personnel involved (n=412): anaesthesia team member(s) plus2				<0.001
No others (anaesthesia team only) (n=34, column % = 8%)	4 (12)	26 (76)	4 (12)
Surgical team (with or without nursing) (n=166; column %=40%)	21 (13)	118 (71)	27 (16)
Patient or family (n=43; column %=10%)	36 (84)	3 (7)	4 (9)
Other (n=169; column %=41%)	60 (36)	62 (37)	47 (28)
Type of communication failure (n=434)				0.002
Occasion (n=53)	12 (23)	26 (49)	15 (28)
Audience (n=53)	9 (17)	24 (45)	20 (38)
Purpose (n=43)	17 (27)	25 (40)	21 (33)
Content (n=169)	87 (33)	135 (51)	43 (16)

²Column % do not add to 100% due to rounding.

Out of 86 injury-related communication failures, in which an anaesthesia handover failure might have occurred, failures were identified in nine (10%). In five cases, the handover failure occurred between anaesthesia team members during a break or shift change; in four, it occurred during transition of care to the PACU or ICU care team.

Discussion

Perioperative communication failures contributed to patient injury in 43% of 910 anaesthesia malpractice claims studied, with multiple failures in 13% of claims. The most common communication failures were between an anaesthetist or anaesthesia team member and a surgeon or surgical team member. However, failures also occurred with the patient/family, with other healthcare providers, and within the anaesthesia team. Nearly half (47%) of the failures occurred during the procedure. Our results emphasise that perioperative communication failures can directly contribute to patient injury.

Prior studies of surgical malpractice claims^6,10 found a lower proportion (13.5–22%) of claims with communication failures associated with patient injury compared with our results (43%). A higher proportion of failures in anaesthesia claims occurred during the procedure (47% vs 30% in surgical claims⁶). We found fewer communication failures occurred during handovers (10% vs 43% in surgical claims⁶) and with transfers between locations (23% vs 39% in surgical claims⁶). The results may differ, as anaesthetists have a larger focus on operating theatre practice, compared with surgeons with clinic and postoperative care responsibilities.

Although most communication failures in both our anaesthesia claims (89%) and surgery claims⁶ (64%) occurred during elective surgery because of the overall higher numbers of elective cases, we found a higher proportion of failures resulting in patient injury occurred in outpatient settings (34% of claims with failures vs 26% of claims without failures; Table 1). Communication amongst all of the patient's care providers, including other physicians, may be more challenging in outpatient compared with inpatient settings, where most of the patient's healthcare providers are within the same hospital system. The higher proportion of monitored anaesthesia care in claims with communication failures reflects failed communication of intraoperative oxygen–cautery fire risk and prevention.

Content failures accounted for more than half (60%) of communication failures leading to patient injury in anaesthesia claims, with half (51%) occurring intraoperatively. The rest occurred preoperatively (e.g. medical history) or postoperatively (intraoperative concerns/patient management plans). Checklists, automated prompts, and communications built into patient monitors and electronic medical records have high potential to reduce content failures.

Purpose failures occurred in 15% of communication failures. Poor teamwork, including disrespect, rudeness, and disagreement on patient care, may contribute to patient harm by increasing team tension and reducing efficiency.² Although communication and teamwork amongst all team members are important, Cooper¹¹ pointed out that the surgeon–anaesthetist relationship has a critical role in patient safety. A well-functioning relationship is conducive to safe patient care. A dysfunctional relationship creates inadequate communication and unsafe working conditions, and contributes to adverse outcomes.¹² Negative communication behaviours by surgeons (e.g. ‘snaps at others when frustrated’ and ‘talks down’) and low scores on positive communication (e.g. considering suggestions, paying attention, and informing others) were associated with a higher risk of malpractice claims.¹³

Methods to improve perioperative communication

Perioperative checklists

Twelve percent of communication failures in our study involved failure to discuss perioperative fire-risk concerning use of supplemental oxygen by mask or nasal prongs in high-fire-risk surgical procedures utilising electrocautery. This communication failure should have been prevented by the surgical safety checklist. Surgical safety checklists have improved patient safety and reduced mortality^14,15; however, reductions in mortality and morbidity depend upon team checklist compliance.¹⁶ Despite widespread use of checklists to prevent surgical ‘never events’, such as wrong site surgery and fires, these never events still occur,^17,18 suggesting that system re-engineering fixes, such as automated reminders, are needed.

Handovers

Intraoperative handovers between anaesthesia providers and during transitions of care to the PACU and ICU also have high potential for communication failure^19,20 and for adverse patient consequences.21, 22, 23, 24, 25 In our study, handover communication failures only accounted for nine (10%) of the 86 claims, where a handover failure was possible. This low proportion most likely reflects the focus of malpractice claims on the clinical care by individual healthcare professionals, rather than system issues involving multiple healthcare team members. Handover issues may also be difficult to identify in a retrospective analysis. However, some studies indicate that intraoperative handovers are not associated with increased risks of adverse surgical outcomes.^26,27 Short breaks and new providers bring a fresh perspective.²⁸ Standardised handover training with complete and consistent teams using electronic tools facilitates improving the quality of the handover communication between anaesthesia teams and the PACU and ICU.^19,20,29,30

Team communication training

Methods to improve team communication based upon the aviation and military communities may improve team oral communication. Multiple communication strategies enhance medical communication,^31,32 including ‘flying by voice’ (announce what intend to do), avoiding mitigating or de-emphasising language (e.g. ‘we should think’ vs ‘do’), closed-loop communication (‘give epinephrine 1 mg and tell me when in’), and repeat back (‘epinephrine 1 mg given’).³² Graded assertiveness (escalation from hints to commands) and five-step advocacy strategies are used to improve assertiveness of team members.³² Use of situation, background, assessment, and recommendation also provides effective communication.³² Oral communication can be improved by use of structured callouts during emergencies (stop, notify, assessment, plan priorities, invite ideas)³³ and the two-challenge rule, which pairs advocacy (stating observation) and inquiry (request for the other's reasoning).³⁴ Speaking up with a safety concern is associated with better overall team performance in simulation studies.35, 36, 37, 38 Enhanced training of team communication techniques may help reduce patient injury.³⁶

Limitations

Malpractice claims represent an extreme case of patient injury. Therefore, our findings concerning the association of communication errors with patient injury may not be generalisable to all adverse patient outcomes. Analysis of closed malpractice claims has a number of other limitations.⁷ The CCP represents a large opportunity sample collected retrospectively and non-randomly. The database is derived from information collected by insurance companies for claims resolution. It is subject to several biases (selection, outcome, and hindsight bias) and geographic imbalance. Study personnel were not blinded to patient injury. However, there was no difference in severity of injury in the claims with and without communication failures.

We included communication failures from pagers, test results, and written communication alone (15% of failures). Communication studies often focus exclusively on oral communication. Given the year of the claims, it is possible that widespread use of electronic health records improved communication of test results and patient health information. However, electronic health records may also contribute to information overload and important test results and health information may be overlooked. The authors abstracted details about the errors from claim narratives written by anaesthetist claim reviewers based on the primary data sources at the insurance companies. This may result in missing information. The overall agreement was excellent (κ=0.885), lending confidence in these secondary assessments.

Conclusions

Failures in perioperative communication contributed to patient injury in a substantial proportion (43%) of anaesthesia malpractice claims. Patient/case characteristics in claims with communication failures were similar to those without failures, except that failures were more common in outpatient settings.

Authors' contributions

Study design: RND, LSS, KLP, KBD, SLM

Data acquisition: LSS, KLP, KBD, JMD, SLM

Data analysis: RND, LSS, KLP, KBD, JMD

Drafting: RND, LSS, ARB, KLP, KBD

Revision/final approval/accountability: all authors.

Handling editor: Hugh C Hemmings Jr

Declarations of interest

The authors declare that they have no conflicts of interest.

Funding

American Society of Anesthesiologists (ASA); Anesthesia Quality Institute through a grant to the University of Washington; institutional funding.

Appendix A. Supplementary data

The following is/are the supplementary data to this article: