Unintended intraoperative awareness: An analysis of Perioperative Anesthetic Adverse Events in Thailand (PAAd Thai)

1. Introduction

Intraoperative recall awareness is the unpleasant perception of an event during general anesthesia, which can result in psychological distress with anxiety and flashbacks. Moreover, it can result in long lasting sequelae, and at worst can present itself as a post-traumatic stress disorder. In the past decade, recall awareness incidences have been continuously reported despite improvements in medical technology, anesthetic techniques, pharmacological agents and monitoring systems [1–3]. Between 2005 and 2008, the incidence of intraoperative recall awareness was found to be 1.0–1.4% in Spain and China [4,5]. Three years later, these incidences dropped to 0.82–0.91% and then decreased to 0.0005–0.25% during 2012 and 2014 [6–10].

In 2005, the Thai Anesthesia Incidents Study (THAI Study), which was hosted by the Royal College of Anesthesiologists of Thailand, reported an incidence of 0.08% [11,12]. In the following years, the Thai Anesthesia Incident Monitoring Study (Thai AIMS) found that anesthesia-related factors were the main cause of the incident [13,14]. After that, national guidelines, recommendations and educational improvement programs were launched [15]. However, a follow up national audit has not yet been undertaken. Data on the possible event-related factors, protective strategies, management schemes and clinical outcomes might differ from the preceding findings. This national multicenter study, supported by the Royal College of Anesthesiologists of Thailand, aims to examine the intraoperative recall awareness incidents that occurred in Thailand between January to December, 2015.

2. Materials and methods

This study was retrospectively registered in the Thai Clinical Trials Registry (TCTR) with identification number TCTR 20160523002. Ethical approval was obtained from each institutional ethics committee.

3. Results

From a total of 2000 incidents reported in this study, intraoperative recall awareness was recorded in nine patients. Of the nine patients, over half (55.6%) were females with an average age and body mass index of 44.5 ± 19.1 years and 23.4 ± 4.2 kg/m² respectively. The majority of the patients (66.7%) presented with a physical status class I–II according to the ASA. Eight patients (88.9%) received general anesthesia, balanced techniques with inhalation agents, opioids, and neuromuscular blocking agents and one patient (11.1%) received general anesthesia with a total intravenous anesthesia technique. The mean anesthetic duration was 224.0 ± 199.0 minutes. All patients were monitored for hemodynamic parameters but neither electroencephalography nor processed electroencephalography was used to guide doses of anesthetics. Two patients (22.2%) received midazolam as a premedication before surgery and one patient (11.1%) received nitrous oxide as an adjuvant.

Each anesthesia was performed and adjusted by anesthesia providers who had experience varying from 6 months to 35 years. The WHO surgical checklist was completed for three patients (33.3%). Flaws were detected during the period of “Sign in” for 44.4% of patients; during “Time out” for 55.5% of patients; and “Sign out” for 44.4% of patients. During the post-operative visit at 24 to 48 hours, intraoperative recall awareness information was given to certified nurse anesthetists by five patients (55.6%), to anesthesia residents by two patients (22.2%), to an anesthesiologist by one patient (11.1%) and to a registered ward nurse by one patient (11.1%). The average year of experience of these personnel was 9.2 ± 3.9 years. The intraoperative recall awareness results were as follows: experience of pain (38.1%), perception of sound (33.3%), perception of intubation (9.5%) and feeling of paralysis (14.3%). The clinical outcome, however, was good for all patients, with complete recovery and no significant morbidity or evidence of post-traumatic stress syndrome at hospital discharge. The characteristics of the recorded intraoperative recall awareness events are shown in Table 1.

The inter-observer reliability was evaluated among four independent assessors before the consensus assembly. The intraclass correlation (ICC) showed perfect agreement for the factors related to intraoperative awareness with an ICC coefficient of 0.91 (95% CI; 0.57 to 0.95) and good to excellent agreement for contributing factors, mitigating factors and suggested corrective strategies with an ICC coefficient of 0.72 (95% CI; 0.46 to 0.87), 0.78 (95% CI; 0.55 to 0.86) and 0.82 (95% CI; 0.45 to 0.89) respectively.

Figure 1 shows the observed factors which were considered to cause intraoperative awareness, i.e. anesthesia-related (100%), patient-related (55%), surgery- related (22.2%) and systematic process-related (22.2%).

OPEN IN VIEWER

Fig. 1.

Factors associated with intraoperative awareness.

The contributing factors, factors minimizing incident and suggested corrective strategies are listed in Table 2. The contributing factors were situational inexperience (77.8%), inappropriate patient evaluation (44.4%), and lack of necessary monitors (22.2%) while factors that could minimize the incident (incident-mitigating factors) were perceived clinical awareness of anesthetic performers (100%) and their experience (88.9%). The suggested corrective strategies were quality assurance activity (88.9%), improved supervision (44.4%) and quality control of medical equipment (33.3%).

This report was undertaken following a third nationwide surveillance of Perioperative Anesthetic Adverse Events in Thailand (PAAd Thai), and encouraged by the Royal College of Anesthesiologists of Thailand in order to determine contributing factors to such events and facilitate the creation of comprehensive strategies to reduce future occurrences in accordance with the quality assurance program regarding anesthesia safety in Thailand.

Recently, Mashour and Avidan found that the incidence of intraoperative recall awareness could vary and depended on the assessment method [17]. By using a structured interview (i.e. Brice Interview Questionnaire) the incidences could be 0.1 to 0.2% [18–20]. In contrast, studies using other methods such as a tool without specific questions, quality assurance data or spontaneous report, had lower incidences and varied from 0.0065% to 0.023% [21–23]. The estimated incidence of intraoperative recall awareness from our collective database was consistent with the incidence rates in studies which did not use the Brice Interview Questionnaire to detect the incident. Moreover, the period of our observation was too short (i.e. 24 to 48 hours after surgery) to detect the incident of intraoperative recall awareness and its psychological consequences such as post-traumatic stress disorder. Hence, the estimated incidence of the intraoperative recall awareness seemed to be underestimated. Despite this, we used the pertinent information to perform the critical incident analysis by four independent experts in order to reveal factors associated with, contributing factors to, factors for minimizing, and corrective strategies for the incident of intraoperative recall awareness. The results from our consensus meeting have been presented in the result section and are discussed here.

Regarding factors associated with intraoperative recall awareness, anesthesia-related factors were considered to be the most relevant. Inappropriate depth of anesthesia was the prominent factor in the anesthesia-related component followed by abandonment of premedication, monitoring ineffectiveness and improper evaluation. In our analysis we found that responses of hemodynamics were used only as clinical signs to guide doses of anesthetics in five out of nine cases (55.55%) while end tidal gas concentration combined with clinical signs were used in four out of nine cases (44.44%). The use of clinical signs alone could be unreliable to guide doses of anesthetics. For cases with intraoperative hemodynamic instability such as hypotension, anesthesia providers usually reduced doses of anesthetics in order to optimize hemodynamics. This could lead to anesthetic under dosage resulting in intraoperative awareness. From our database in the PAAd Thai study the end tidal anesthetic gas concentration was used in 32% [16], whereas processed EEG such as Bispectral index (BIS) was used in 0.5% of the reported cases. The effect of these monitors on the occurrence of intraoperative recall awareness could not be demonstrated in the PAAd Thai study due to the limited study design. However, the evidence supporting the effect of BIS-guided anesthesia as well as end tidal gas concentration-guided anesthesia on reducing the risk of intraoperative recall awareness in the surgical patients at a high risk for complications has been demonstrated [24–28]. In Thailand, the direct depth of anesthesia monitoring was not used routinely because it was recognized as a high cost issue coded by the Ministry of Finance and could not be reimbursed. The clinical awareness was used instead except in the high risk patients or high risk surgery, e.g. cardiac, thoracic surgery [26].

All the reported cases received neuromuscular blocking agents during the operation but only three of them (33.33%) felt paralysis during the operation. The risk of neuromuscular blocking agents on intraoperative awareness has been documented [18]. The neuromuscular blockade can impede the sign of muscular movement during light anesthesia.

Nitrous oxide was not used in eight of the nine reported cases (88.9%). It was demonstrated to be associated to decrease risk of intraoperative recall awareness in a previous study [13]. In contrast, the 5th National Audit Projects (NAP5) concluded that omission of nitrous oxide was not a risk factor [29]. Hence, this controversy needs to be elucidated.

Twelve percent of the reported patients did not receive proper preoperative evaluation. The preoperative evaluation is an important part to identify patients’ risk factors for intraoperative awareness. The proper preoperative evaluation should be done, e.g. looking for substance use or abuse, limited hemodynamic reserve, ASA physical status and previous history of intraoperative awareness. The known risk factors for intraoperative awareness could lead the anesthesia providers to be more concern about this problem and look for strategies to its ameliorate.

Regarding patient-related factors, hemodynamic instability was the most prominent factor (57%) followed by opioids dependency (29%) and previous personality (14%). The finding indicated that more than half of the reported patients were a high risk group with hemodynamic instability. In this situation, anesthesia providers tended to use low dosage of anesthetics. This could lead to a decreased depth of anesthesia resulting in the awareness. Opioid dependency was considered in nearly one-third of the reported cases. This group of patients had an increased likelihood of experiencing intraoperative awareness [30].

Three patients (33.33%) reported their awareness during the dynamic phases of anesthesia, i.e. two incidents during induction and intubation and two incidents during emergence. According to a previous article on accidental awareness during general anesthesia, 66.66% of the cases of accidental perioperative awareness could occur during dynamic phase of anesthesia, i.e. induction of and emergence from anesthesia [29]. Contributing factors during induction of anesthesia were use of thiopental, rapid sequence induction, obesity, difficult airway management, neuromuscular blocking agent, and interruptions of anesthetic delivery during movement from anesthetic room to theatre. During emergence, most patients perceived residual paralysis as accidental awareness.

Regarding surgical risk factors, high risk surgery was the prominent factor followed by surgical difficulty. This is determined by the type of surgery (for example, obstetric, cardiac and thoracic) in association with the patient’s characteristics, such as body mass index, hemodynamics, drugs taken prior to anesthesia and personality, and the anesthetic performer’s scientific and technical performance.

Attempts to minimize the factors which can be attributed to human error, such as the inexperience of anesthesia personnel, evaluation of errors and decision-making uncertainty from previous adverse incidents has been made [13]. Inadequate monitoring is important and must be emphasized for future incident prevention. In fact, the reduction of the incidence of adverse events by half during the past decade might be the consequence of the recognition of these factors and collaborative national efforts to prevent them [11,13,16]. The resulting national guidelines include the introduction of increased practice hours and greater clinical supervision, skill practice and knowledge-based training for anesthesia personnel. These deliberate measures to address intraoperative recall awareness have been successful. Education and training for anesthesia providers, such as rehearsal of specific tasks, practice under supervision and feedback provided by a mentor have had a positive effect [31–34]. The WHO checklist, initiated internationally, claimed to help reduce the level of human error in surgical anesthesia, however the findings from this study have shown that the checklist has, unfortunately, been ineffectual. Instead, a specific checklist to reduce anesthetic-relevant risk issues has been recommended [9].

Since 2005, The Royal College of Anesthesiologists of Thailand has recognized the occurrence of intraoperative recall awareness and has implemented effective strategies to reduce the incidence of these adverse events. In this study, the deliberate measures to prevent intraoperative awareness have been shown to be successful, and resulted in a decrease in occurrence of intraoperative recall awareness [11,16]. This is a positive result for patients and anesthetic staff.

The limitation of this study is the issue of the disclosure of intraoperative recall awareness. The events were mainly recorded from a volunteered narrative. However, the validity and reliability may be questionable. The Modified Brice Interview Questionnaire and Michigan Awareness Classification Instrument should be introduced for standardization of disclosure [35,36]. Another limitation is the process of data analysis which was the retrospective methods despite the authors designed the research methodology as the prospective. The possibility that the efficacy of the data collection was questionable must be noted and the reporting bias has perhaps remained. According to the anonymous basis, some inappropriate management or outcomes might be concealed. The quality control of data management and the rigorous action of individual site manager would be beneficial. Finally, the duration of intraoperative episodes was brief, so it is possible that patients may not have recognized the occurrence of intraoperative awareness or the exact period in which it occurred. Recall awareness is subjective and difficult to prove. This confusion may affect the data interpretation and therefore the results of this study.

5. Conclusion

The nine intraoperative recall awareness cases reported in this study were few, however the causes were preventable. Anesthesia-related factors seemed to be the most relevant causes. The result from our analysis has provided lessons for the creation of comprehensive strategies to ameliorate the incident in accordance with the quality assurance program regarding anesthesia safety. The revision of the anesthesia-specific WHO checklist related to our finding is recommended.