Skip to main content

Table 2 Fifteen examples of SA errors

From: Situation awareness errors in anesthesia and critical care in 200 cases of a critical incident reporting system

Case number Case description Analysis from the SA perspective SA level
1 An anesthesiologist took over a patient who had undergone massive transfusion including catecholamine therapy. He reports to have received a “detailed handover” and that his job was to finish the procedure and to transport the patient to ICU. Just before leaving the OR he replaced an empty infusion bag with a new one in order to continue volume replacement. Immediately afterwards, the patient’s suffered from ventricular arrhythmia and the systolic blood pressure increased to 250 mmHg. “During check of the i.v.-lines I noticed that the adrenaline syringe pump had been connected to the central line by two extension lines type Heidelberger. Obviously they had filled with highly concentrated adrenaline which was administered unintentionally during volume resuscitation.” The anesthesiologist was not aware about a significant amount of adrenaline in the lines. Possibly, the hand-over, which he felt to be “detailed”, did not include information about this fact (SA-I). Alternatively, he may have forgotten this information in face of a complex situation where gaining complete SA in short time is challenging for someone who had not been involved until this moment. SA I
data not available or memory loss
2 “The code blue physician does not hear the beeper. The beeper turns off after a certain time. The causes are a significant noise exposure on the ICU and the high frequency of phone calls.” The code blue physician did not perceive the alarm (SA-I). The reporting individual mentions acoustic barriers on one hand and high workload on the other hand as causes. SA I
hard to detect
3 “For economic reasons, sometimes, nurses program the syringe pumps. In this case a syringe pump programmed for propofol ran with remifentanil and, accordingly, it ran too slow. […] The only striking point was that we had propofol in the remifentanil line repeatedly and despite high infusion rates, we still had the first syringe of remifentanil after hours. Having a closer look we were able to recognize propofol in small fonts on the display whereas remifentanil was indicated on the syringe label.” It largely remains unclear why the nurse allocated the drugs incorrectly. Assumingly some information (syringe content or pump program) has been forgotten. However, the reporting individual clearly states that important information was displayed in small fonts hindering a fast and quick recognition of the content of syringe pumps (SA-I). SA I
data hard to discriminate
4 “After putting the drapes, the access to both peripheral iv lines was hampered. During team-time-out one of the surgeon leant against the arm compressing the iv lines while the anesthesiologist paged through the patient’s health record […] so that the anesthetics entered the infusion bag of the crystalloids. During skin incision the patient showed increase of heart rate and moved the arms. Then, we switched the administration of anesthetics to the other iv access.” Important visual information from iv lines (obstruction) was not perceived due to a visual barrier (drapes). Furthermore, the visual attention was directed to the patient’s health record during team time out. It remains speculative why a non-return valve had not been used and whether the use of such a valve had resulted e.g., in high-pressure alarms in the syringe pumps (SA-I). SA I
hard to detect and failure to observe
5 After uneventful anaesthesia the patient was transferred to another location. There, the first systolic blood pressure assessed was 60 mmHg. “In this OR a transport monitor does not exist. The short transfer regularly is done without monitoring. Every time a monitor is required, we have to get it from elsewhere which is time-intensive.” The case reveals structural problems as a monitoring device is not easily available and the anesthesiologists avoid time delays in face of assumingly uncomplicated cases. As a result, important information is missed (SA-I). SA I
failure to monitor
6 “To keep open an arterial line, HES [hydroxyethyl starch] was used instead of saline. Both look similar but HES is an emergency substance so that it should be stored in a different place.” As both infusions look similar (look-alike problem), the information was correct but obviously misperceived (SA-I). SA I
7 “A patient is transferred to ICU with several syringe pumps including a pump for TIVA [total intravenous anaesthesia] that had been equipped with a catecholamine. The ICU personnel are not familiar with that type of pumps. […] Unintentionally, the patient got a high bolus.” A health care provider works with a syringe pump he is not familiar with. Although all the dynamic information is present (rates, drugs, indication), the individual applies an incorrect mental model of the pump’s operating mode and thus, he lacks of comprehension (SA-II). SA II
use of incorrect mental model
8 “two oral drugs […] had been given via the central venous line instead of the gastric tube.” Assumingly, all the relevant information (e.g., package insert, drug orders) was present, but the individual lacks of a mental model with respect to how these drugs are administered (SA-II). As a result he does not comprehend that these drugs have to be administered in another way. SA II
use of incorrect mental model
9 […] On the third postoperative day […] the epidural was stopped. On the next morning, the anesthetist cannot visit the patients due to concurrent obligations. During the evening visit, the anesthetist noticed that ropivacaine was re-started but that it was connected to a peripheral venous line. The infusion was stopped immediately.” Assumingly, all the relevant basic information was present: drug, patient and indication (SA-I). But the information was not properly integrated, due to missing knowledge or the use of missing or an incorrect mental model (SA-II). If someone is confronted with a set of information he can’t process due to missing contextual contents in the long-term memory, he will probably ask for assistance. If an incorrect model is used, he won’t recognize the error as long as there is no additional information such as visible adverse effects. SA II
use of incorrect mental model
10 "During TIVA a change of the syringe (remifentanil) was pending. The syringe had been prepared by the nurse (50 ml, clear solution). The label “remifentanil” and the ampoule lied besides the syringe. The nurse told to the anesthesiologist that the remifentanil syringe was prepared. The anesthesiologist changed the syringe; in the following minutes, the patient shows tachycardia and high blood pressure, deepening anaesthesia is without success. When the nurse came back, she asked if the anesthesiologist had added the remifentanil to the prepared syringe. As it turned out, the communication […] was unclear and stated a potential danger for the patient.” The anesthesiologist incorrectly assumed a syringe to be correctly prepared (SA-II). Visible information (the ampoule next to the syringe) was not perceived or not integrated in order to come to the conclusion that the syringe contained purely saline. Additionally, the reporting individual identified a lack of information resulting from unclear communication as the cause. SA II
over-reliance on default values
11 “A critically ill patient with complex pains, who was visited by pain physicians for 4-fold analgetic medication. During change of syringe pump, ketamine is administered in wrong dosage, 50 mg/ml instead of 1 mg/ml is administered, as it is usual for sedation. During shift change the error is recognized. […] The patient was awake throughout the case […] but suffered from headache.” The nurse who changed the syringe prepared the dosage as usual (assuming standard values), despite differing information from the medication order as indicated through the fact that this was recognized during shift change. This may have happened through an over-reliance on default values (SA-II) although additional information was available that would have resulted in a different action (preparing the correct dosage). SA II
over-reliance on default values
12 “During thoracic surgery (VATS lobectomy) the suction catheter was introduced too deep in the tracheal part of the double-lumen tube. […] Lobectomy is performed using a stapler. The suction catheter could not be removed for checking for leakiness […]. As a cause, the stapler had fixed the suction catheter. An anterior thoracotomy was performed […] and the suction catheter was removed successfully.” The anesthesiologist, assumingly, was aware about the surgical procedure to be performed (use of stapler). Additionally he had the information about the suction catheter as he himself had inserted it. This information has not been integrated properly as he relied on his experience from prior situations where removing the device was always without problems and long-term memory content such as a mental model or prototypical situations suited to successfully integrate the basic data was not used or not present. As a result, also a problem on the level of projection emerges as an anterior thoracotomy had to be performed unexpectedly. SA II
lack of or incomplete mental model
13 “A surgeon indicated emergency surgery. There is no written information about patient history and it is impossible to get the information orally [from the patient]. The patient is assessed clinically, an old scar from tracheostomy is visible which indicates possible intubation problems. The anesthesiologist put himself under pressure and induces anaesthesia without investigating the background or consulting the admitting hospital. A rapid sequence induction is performed. Intubation with a 8.0 size tube is not possible, bag mask ventilation works, a 7.0 mm is not introducible as well, and a laryngeal mask (4 and 5) is not tight so that adequate ventilation is impossible. Finally, another physician successfully intubates.” Unexpectedly, the anesthesiologist ran into intubation difficulties, indicating an error on the SA level of projection (SA-III). This is supported by the retrospective statement that he worked under avoidable time pressure and that, as a consequence, search for additional information was omitted (SA-I). Regardless of the fact whether the simple presence of a scar from tracheostomy should prompt the preparation for difficult airway management, a mental model that integrates the basic data (tracheostomy in the past) to SA on the level of projection “expected difficult intubation” was absent (SA-III). SA III
lack of or incomplete mental model
14 A patient is scheduled for hip replacement. […] Until the use of palacos bone cement everything went fine. […] Immediately after inserting palacos bone cement, end-tidal CO 2 drops from 37 to 13 mmHg. Oxygen saturation does not provide values. At the beginning, a sinus tachycardia of 140 bpm is noticed, quickly followed by deformed QRS complexes. Heart rate drops to 20 bpm. Cardiopulmonary resuscitation is initiated immediately. The working hypotheses are air embolism, fat embolism and allergic reaction. An unexpected deterioration due to the use of palacos bone cement is described (SA-III). A dramatic change of vital parameters is the basic information (SA-I) that results in a re-evaluation of the situation. As a consequence, the anesthesiologist comprehends that cardiopulmonary resuscitation is required (SA-II). Additionally, based on basic information, possible causes are discussed. SA III
over-projection of current trends
15 A geriatric patient with dementia is transported to the emergency department. He has a visible laceration on the head after having fallen out of the bed. The laceration was sutured and a CT scan ordered in face of increasing somnolence. “A medical student saw that nobody had placed a cervical collar and that the patient complaint about pain when the head was positioned for suturing. He did not communicate his observation […]. The scan showed a facture of atlas and axis.” There are relevant cues that indicate the possibility of a lesion of the cervical spine (fall, laceration on head, increasing somnolence, pain during movement of the head). The reporting individual emphasizes that the team did not comprehend the possibility of a spine lesion that is, they either did not possess over the mental model that allowed for meaningful integration of the information mentioned above (SA-II) or they simply did not perceive some piece of information, e.g., pain during movement of the neck (SA-I). SA I
failure to observe
Another point refers to a lack of communication as the medical student did not speak up (Team SA). Communication can refer to the SA level of comprehension (e.g., “we cannot rule out a spinal lesion, therefore cervical collar makes sense”) or to the level of perception (e.g., “every time the patients head/spine is moved, the patient complaints about pain”). missing mental model
  1. Fifteen cases during which SA errors led to errors or near misses. SA-I refers to the level of perception, SA-II to the level of comprehension, SA-III to the level of projection, respectively