Mortality associated with anesthesia in the developed world has been reduced significantly during the last 25 years. Although anesthetic-related deaths have declined overall, there has been an increase in the number of reports of mortality arising from institutional organizational problems such as poor supervision and rostering (1). Health care is continually changing and anesthesiologists also need to identify trends that could lead to reduced patient safety and develop corrective actions. For example, Cesarean delivery is now performed during epidural or spinal anesthesia so commonly that anesthesia trainees have few opportunities for supervised learning of general anesthesia for this procedure (2). Some specialist anesthesiologists may even become less skilled. There will, however, always be a need for general anesthesia for Cesarean delivery, and we believe new measures for training and maintaining the skills used in this situation need to be carefully considered.
Commercial aviation is an inherently risky industry that has been made safer through adherence to checklists and protocols. Aviation safety authorities mandate that commercial pilots use formal checklists to prepare for take-off and landing and this includes anticipation and management of untoward events that might suddenly occur. The checklist is to ensure that all procedures are followed rather than relying on human memory alone. An electronic checklist that can give voice prompts and highlight errors has been developed to assist one person checking in general aviation (3). This checklist system is simple to use and easily customizable.
We investigated whether preparation for a Cesarean delivery under general anesthesia could be helped by the use of a formal aviation-style checklist. To begin the study, we devised a checklist for use when preparing to administer general anesthesia for Cesarean delivery and then evaluated it using an anesthesia simulator.
Local Research Ethics Committee approval was obtained for this study. We used a modification of the Delphi technique (4) to generate a list of items that should be checked before the commencement of non-emergency Cesarean delivery under general anesthesia. We invited consultants with a special interest in obstetric anesthesia practicing across Australia to participate in this process. They were contacted through email and remained anonymous to the other participants. From the results of two questionnaires we constructed a list of items that should be checked before a Cesarean delivery under general anesthesia and divided it into equipment, drugs, monitoring, and patient checks.
The checklist was then programmed into the EC-TS electronic checklist (Aironautical Electronics Corporation Pty. Ltd.) (Fig. 1). The unit is a small, portable device that runs on an internal battery that can be charged from almost any external power source. The screen displays several lines of text and there is an optional speech facility using a voice synthesizer, which we used for our study. A button is pressed to confirm the item and progress to the next on the checklist.
Twenty experienced anesthetic registrars and consultants from the Flinders Medical Centre Clinical Anesthesia Department consented to participate in an evaluation of the electronic checklist. This was done using a high-fidelity anesthesia simulator in a realistic operating room setting, including an anesthetic machine, airway trolley, monitoring facilities and a full array of anesthetic and surgical equipment. The participants were first given the scenario of being asked to give a general anesthetic for a non-emergency Cesarean delivery. They were instructed to proceed as they normally would in this situation, starting with the checking of equipment. The “patient” (Laerdal SimMan®) was brought into the scenario at the appropriate time. The simulation was stopped immediately before induction of anesthesia. The participants were observed during their preparation by one of the investigators and items on the checklist were marked off as they were completed. At the end of the simulation, the participant and investigator went through the electronic checklist in verbal and written form. The investigator confirmed with the participant which of the items they had checked.
The participants were then asked to repeat the scenario using the checklist. The aim of this scenario was to give these anesthesiologists the opportunity to reflect on the clinical use of such a system. They were not timed during this process, although all completed the checks within a reasonable time period, and no one commented that the checklist unduly prolonged their usual checking procedure. They were asked to complete an anonymous feedback form which asked whether they found the checklist useful, whether they would like to use it for practicing scenarios, clinical situations, or both, whether they thought the checklist was too long, too short, or about right, whether they thought a checklist should be written or verbal, whether there were any other items which should be on the checklist, and for any comments.
The final checklist included 40 items that the expert panel believed to be important. The 20 participants had spent an average of 8.5 yr in anesthetics (range, 1–25 yr). The median number of items not checked was 13 (range, 7–23). The checklist and the number of participants who checked each item are listed in Table 1. All but one of the 20 participants (95%) reported that they felt the checklist was useful. Although 16 participants (80%) said they would like to use the checklist system for practicing simulated scenarios, only eight (40%) would like to use it in real life clinical situations. Most participants (90%) thought that the checklist was about the right length, but two (10%) said it was too long. Considering the way the checklist was presented, 12 (60%) would prefer a written checklist and eight (40%) preferred the verbal checklist.
Most participants felt that the checklist was complete but there were several suggestions for additional items including checking that the patient would have blood available (“grouped and saved”) and checking specific details in the preanesthetic assessment such as allergies and airway examination.
Anesthesiology is often considered to be a leading specialty with respect to medical risk management and patient safety (5). However, the discipline still lags far behind other industries such as aviation and nuclear power in terms of developing strategies to monitor and improve performance. Emergencies can never be totally eradicated from anesthesia because of the complexity and dynamism of our clinical environment (6), but they can be prepared for. Studies on clinical performance have shown a reduction in adverse events when using principles derived from aviation (7,8). Crisis management training has now been recommended for all anesthesiologists (1). It appears from our study, that a checklist can assist preparation for administering anesthesia and that many anesthesiologists would like to be able to practice using them.
Human memory is fallible, and acute stress can adversely affect performance and attention (9–11). The need for general anesthesia in obstetrics is unpredictable and often occurs under difficult and emotionally charged circumstances. There are several factors that make anesthesia for Cesarean delivery different from other general anesthetics; this, coupled with increasing unfamiliarity with the event, increases the risk of critical incidents occurring. Although these external circumstances cannot be controlled, an effective strategy for preventing problems is to make specific checks for triggering factors and correct these as necessary (6).
We chose to use the scenario of an elective Cesarean delivery using general anesthesia. This is becoming an uncommon event, and most obstetric general anesthetics are administered for emergency deliveries. In such situations, the life of the fetus or the mother is at risk. Preparation for such an event will depend on the time available as advised by the obstetric team, and so preoperative checks must be tailored to each individual situation. In dire emergencies there may only be time to check the most important things. For this reason, it was felt that an elective Cesarean delivery was the most suitable scenario for the evaluation of a checklist. Shorter checklists could be devised for use in emergency situations.
We devised a checklist using expert opinion in the field to study its use. A checklist such as this can easily be modified or developed for each hospital according to local policy, or even for individual anesthesiologists depending on their own preferences, particularly with regard to airway equipment. Aviation style checklists have been advocated to cover critical details in busy times (12). Devising one’s own checklist makes it useful (12) but a standardized system would be better for consistency in training.
On average, participants failed to check approximately one-third of the items on the list. Some check items were omitted because staff assumed they were present or had been checked by others. Such assumptions can have disastrous consequences. Obviously, some items are more vital than others, but most participants admitted that they did forget to check an item that they would deem as important and would routinely want to check. Although the simulator environment is useful, it has some obvious drawbacks in terms of realism, and participants may behave differently in this situation than they would in a clinical setting. In particular, the very large proportion of participants omitting to put the patient on left lateral tilt could be a result of the simulated environment. The simulated patient was already on the operating table, whereas transferring a patient onto the table often prompts this to be arranged. Also, the anesthetic machine used in our simulator lab is a basic machine, whereas anesthesia workstations are routinely used in our operating rooms. However, most of the participants had practiced scenarios in the simulation area in the past and so would have been familiar with its layout and the equipment.
Several participants had problems understanding some words on the verbal checklist because it uses a synthesized speech facility. This may account for many of the participants preferring a written checklist to a verbal checklist. The checklist also appeared in written form on the digital display screen of the device and so could be checked if there was any doubt about the item when spoken. The advantage of the verbal checklist is that the operator does not need to be near the device, or even looking at it, when the item is spoken. Another feature of the device is that a button must be pressed to progress on to the next item; this can be on the unit itself or by means of a remote trigger. With a written list, such as in paper form, it is easy to miss a line or read several items at once, and then forget one or more of them. Many new personal digital assistants and hand-held computers have facilities for recording. A verbal checklist could be read out, recorded and stored for later use. This also gives the advantage of having a spoken rather than synthesized voice. Personal digital assistants could be used to store written checklists if the individual prefers this format.
An aspect of checklists that is not often discussed in relation to anesthesia is that the task being checked becomes explicit and common knowledge. Thus, in a real-life setting similar to our simulation, if a checklist system is used, some of the steps could be efficiently and safely delegated to anesthesia assistants. Alternatively, if an institution has agreed protocols on checks to be performed by different personnel, then each may have their own checklist.
Only 40% of participants felt that this checklist would be useful in a clinical situation. Many expressed concerns about increased anxiety in the patient associated with its use. The speech facility of this checklist device may be disconcerting to an awake patient in what is often an already stressful situation for them. Of course, the opposite may be true, in the same way that knowing that everything has been checked and nothing has been ignored or overlooked by the flight crew comforts airline passengers. However, if this was felt to be the case, a checklist may still be useful for the equipment checks to be made before the patient is brought to the operating room.
The majority of participants said they would like to use the checklist for practicing patient care in simulations. The use of simulators is increasing in anesthesia (13), and their use for crisis management training is very much based on the way flight simulators are used for training in aviation (6). The use of a checklist could be studied as a potential training aid. The number of omissions may be reduced after using the checklist in real or simulated scenarios.
The time taken for the completion of the checklist was not measured in this study. Obviously, any unnecessary delays in inducing anesthesia must be avoided, particularly in emergency situations. Future studies could be designed to investigate this area further. As many participants expressed a preference for a written, rather than verbal, checklist, another area for further work would be a comparison of the two formats. The next stage in our studies would be to evaluate the use of the checklist in clinical situations.
In our simulation we observed that important checks were often forgotten when memory alone was relied on to prepare for a general anesthetic for Cesarean delivery and that the use of a checklist could improve this. Overall, participants’ perceptions of the checklist and the electronic device were very positive. We believe there is a role for the wider use of checklists in anesthesia training, and it also seems likely that checklists could be valuable in routine anesthetic practice to improve patient safety.
The EC-TS used in this study was loaned to us by Aironautical Electronics Corporation Pty. Ltd (www.airosafe.com.au). Dr. Kym Osborn provided valuable help with the Delphi rounds. We thank the staff of the Department of Anesthesia at Flinders Medical Centre who gave their time to help this research.
1. Mackay P. Safety of anaesthesia in Australia: a review of anaesthesia related mortality 1997–1999. Melbourne, Australia: Australian and New Zealand College of Anaesthetists, 2002.
2. Johnson RV, Lyons GR, Wilson RC, Robinson AP. Training in obstetric general anaesthesia: a vanishing art? Anaesthesia 2000;55:179–83.
3. Checklist Manual. Adeleide, Australia: Aironautical Electronics Corporation Pty. Ltd., 2002.
4. Bowles N. The Delphi technique. Nurs Stand 1999;13:32–6.
5. Pierce EC Jr. The development of anesthesia guidelines and standards. Qual Rev Bull 1990;16:61–4.
6. Gaba DM, Fish KJ, Howard SK. Theory of dynamic decision making and crisis management. In: Crisis management in anesthesiology. New York: Churchill Livingstone Inc., 1994:5–29.
7. Steiger HJ. Standards of neurosurgical procedures. Acta Neurochir Suppl 2001;78:89–92.
8. Hugh TB. New strategies to prevent laparoscopic bile duct injury: surgeons can learn from pilots. Surgery 2002;132:826–35.
9. Vedhara K, Hyde J, Gilchrist ID, et al. Acute stress, memory, attention and cortisol. Psychoneuroendocrinology 2000;25:535–49.
10. Roozendaal B. Stress and memory: opposing effects of glucocorticoids on memory consolidation and memory retrieval. Neurobiol Learn Mem 2002;78:578–95.
11. Leiberman HR, Tharion WJ, Shukitt-Hale B, et al. Effects of caffeine, sleep loss, and stress on cognitive performance and mood during U.S. Navy SEAL training. Psychopharmacology 2002;164:250–61.
12. Bauman LA. Checklists cover critical details. APSF Newsletter 2003;18:1–16.
13. Forrest FC, Taylor MA, Postlethwaite K, Aspinall R. Use of a high-fidelity simulator to develop testing of the technical performance of novice anaesthetists. Br J Anaesth 2002;88:338–44.