Recovery room incidents: a review of 419 reports from the Anaesthetic Incident Monitoring Study (AIMS)


Correspondence to: Dr M. T. Kluger E-mail:


Summary Four hundred and nineteen incidents that occurred in the recovery room were extracted from the Anaesthetic Incident Monitoring Study database, representing 5% of the total database of 8372 reports. Incidents were reported mainly in daylight hours, with over 50% occurring in ASA 1–2 patients. The most common presenting problems related to respiratory/airway issues (183; 43%), cardiovascular problems (99; 24%) and drug errors (44; 11%). One hundred and twenty-two events (29%) led to a major physiological disturbance and required management in the High Dependency Unit or Intensive Care Unit. Contributing factors cited included error of judgement (77; 18%), communication failure (57; 14%) and inadequate pre-operative preparation (29; 7%), whilst factors minimising the incident included previous experience (97; 23%), detection by monitoring (72; 17%) and skilled assistance (54; 13%). Staffing and infrastructure of the recovery room needs to be supported, with ongoing education and quality assurance programmes developed to ensure that such events can be reduced in the future.

Both the Closed Claims Study in the United States [1] and the Australian Incident Monitoring Study (AIMS) [2] have highlighted the importance of recovery areas and the potential for serious adverse events that may occur there. The role of the recovery area has evolved from one of passive observation to an important determinant of speed of recovery and discharge of patients. It is therefore important to identify the types of problems encountered in this area and their mode of presentation and, from these, to suggest improvements. Previous studies have identified the overall incidence of adverse events occurring in recovery areas as 5–30%[3–6]. Minor complications appear to present much more frequently than major events. In a study of 37 079 patients in the recovery area, Bothner et al. identified minor complications in 22.1% and major complications in 0.2%[6].

The AIMS project has been used to investigate several aspects of anaesthetic practice, including crisis management [7], obstetric practice [8], aspiration pneumonia [9], pre-operative preparation [10] and the role of the anaesthetic assistant [11]. Recovery room incidents were described in the 1993 symposium edition of the original AIMS study, which included a total of 2000 reports [2]. The aim of this study was to report on a larger number of incidents and identify problems and incidents that occurred in the recovery area, and to determine any areas in which further improvements in care might lead to a reduction in adverse events.


The AIMS project is a voluntary, self-reporting audit of actual or potential incidents that occur during anaesthesia [12]. Incidents are reported on a standard form, using a combination of tick boxes that relate to patients' demographics, contributing and alleviating factors, and suggested corrective strategies along with outcome of the incident. In addition, respondents complete a free narrative relating to the incident. The reports are coded from the key words and free text, and added to the database. The ongoing project has been approved by the Royal Adelaide Hospital Ethics Committee and by the Commonwealth of Australia Department of Health and Aged Care.

Incidents that occurred in the recovery area were identified by searching the database for ‘Recovery area’ within the ‘Where and when it happened’ section of the AIMS form. As the interpretation of the data differed from that in the original study of recovery room incidents reported in 1993 [2], the complete database was reviewed and recoded. Therefore, incidents reported in the 1993 symposium issue are included in this report. The relevant information was entered into a Microsoft Excel spreadsheet (Microsoft Corporation, Seattle, WA, USA). Data were entered as originally recorded on the individual AIMS reports and were analysed with respect to the following fields: nature of surgery, surgical category, ASA status, outcome, presenting problem, contributing factors and suggested corrective strategies. The presenting problem categories were categorised to allow further investigation of their aetiology. Some categories had degrees of overlap; in these cases, the presenting problem that appeared to be the most important in development of the incident was designated as the main problem.


Of the 8372 reports in the AIMS database, 478 were extracted in which the incident occurred in the recovery area. Of these, 59 were discarded because of inadequate data, or because the incident did not actually occur in the recovery area or no significant incident occurred. This left 419 reports, or 5% of the entire database, for subsequent evaluation.

The majority of incidents (384; 92%) involved adults (> 14 years); 24 (6%) involved children (1–14 years) and seven (2%) involved infants (< 1 year). Most (377; 90%) occurred during daytime hours (08:01–22:00); 17 (4%) occurred at night (22:01–08:00) and time was not recorded in the remaining 6% of reports. All surgical types were represented with general surgical and orthopaedic patients making up the majority of reports (Table 1) Most of the reported cases involved patients with mild or moderate coexisting disease (Fig. 1).

Table 1.  Type of procedure involved with 419 recovery room incidents reported to AIMS.
Surgical typeNo. (proportion)
General surgery118 (28%)
Orthopaedics 72 (17%)
Ear, nose and throat 42 (10%)
Ophthalmology 33 (8%)
Gynaecology 30 (7%)
Urology 28 (2%)
Vascular 25 (6%)
Dental and maxillofacial 14 (3%)
Plastic 12 (3%)
Cardiothoracic 12 (3%)
Obstetric 10 (2%)
Electroconvulsive therapy  6 (1%)
Neurosurgery  5 (1%)
Pain block  2 (0.5%)
Radiology  1 (0.2%)
Vascular access  1 (0.2%)
Figure 1.

ASA status in 419 recovery room incidents reported to AIMS.

The most common primary presenting problems were cardiovascular, respiratory, airway related or due to drug error (Table 2). Cardiovascular problems are shown in more detail in Tables 3 and 4. Pulmonary oedema appeared to be secondary to airway obstruction (11; 3%), fluid overload (7; 2%), ischaemic heart disease (6; 1%), drugs (3; 1%), sepsis (1; 0.2%) or with no discernible cause (1; 0.2%). Hypotension was caused by a wide variety of conditions including drugs, anaesthetic techniques, allergy, bleeding, sepsis and myocardial ischaemia. In the 20 patients who suffered a cardiac arrest in the recovery area, seven were orthopaedic patients and 17 had ASA grades between 1 and 3.

Table 2.  Primary presenting problem in 419 recovery room incidents reported to AIMS.
Primary presenting problemNo. (proportion)
Cardiovascular99 (24%)
Respiratory97 (23%)
Airway86 (21%)
Drug error44 (11%)
Central nervous system32 (8%)
Equipment27 (6%)
Communication problems 7 (2%)
Hypothermia 6 (1%)
Regional block problems 4 (1%)
Inadequate documentation 4 (1%)
Hyperthermia 3 (1%)
Trauma 3 (1%)
Dental problems 2 (0.5%)
Renal 1 (0.2%)
Skin 1 (0.2%)
Blood transfusion 1 (0.2%)
Facility limitations 1 (0.2%)
Gastrointestinal problems 1 (0.2%)
Table 3.  Presenting problem in 99 cardiovascular incidents out of 419 recovery room incidents reported to AIMS.
cardiovascular problem
No. (proportion of total incidents)
Pulmonary oedema29 (7%)
Hypotension26 (6%)
Cardiac arrest20 (5%)
Bradycardia 7 (2%)
Myocardial ischaemia 7 (2%)
Tachycardia 4 (1%)
Bleeding 2 (0.5%)
Hypertension 2 (0.5%)
Allergy 1 (0.2%)
Radial artery ischaemia 1 (0.2%)
Table 4.  Details of 20 patients who suffered a cardiac arrest within the recovery area, out of 419 recovery room incidents reported to AIMS.
ProcedureASAComorbiditiesInitial rhythm*OutcomeComments
  • *

    VF: ventricular fibrillation; VT: ventricular tachycardia; PEA: pulseless electrical activity.

Insertion of peritoneal  dialysis catheter4Obesity, ischaemic heart  disase, chronic renal failureAsystoleSurvived 
Total knee joint  replacement2 AsystoleSurvivedSpinal anaesthetic
Adrenalectomy2GlaucomaAsystoleSurvivedPossible primary respiratory arrest  from inadequate reversal  of neuromuscular blockade
Femoral nailing3Congestive cardiac failure,  metastatic cancerAsystoleDiedMetastatic cancer with possible  hypovolaemia
Tonsillectomy1NilNot connected  to ECGSurvivedSevere laryngospasm, hypoxia  and cardiac arrest
Eye surgery3 VFSurvivedUneventful general anaesthetic;  given metoclopramide 20 mg,  followed by VF
Unknown  Not statedSurvived 
Forefoot amputation3Insulin-dependent diabetes  mellitus, ischaemic heart Not stated
 disease, chronic  renal failure
SurvivedInitial bleed in recovery room  followed by cardiac arrest; possible  hypovolaemia
Radical prostatectomy2 Not statedSurvived 
Transurethral resection  of prostate2HypertensionVT, VFDied 
Hip replacement2 Not statedDiedMassive myocardial infarct
Hip replacement2Supraventricular tachycardia,  non-insulin-dependent  diabetes mellitus,  hypertensionVFSurvivedSciatic nerve block plus epidural;  total of 480 mg lidocaine and  150 mg bupivacaine
Colonoscopy2 Not statedSurvived 
Vascular3Atrial fibrillationVFSurvived 
Hemicolectomy3 AsystoleSurvivedPossible respiratory arrest following  opioids
Nailing of femur1NilVFSurvived 
Insertion of dialysis  catheter4Insulin dependent diabetes  mellitus, chronic renal  failure, hypertensionNot statedDied 
Repair of fractured hip3 PEASurvivedPossible opioid overdose and  hypovolaemia
Knee replacement2 Not statedSurvivedPossible vagal response to acute  bleed
Eye surgery3Insulin dependent diabetes  mellitusNot statedSurvived 

Respiratory complications are detailed in Table 5. Inadequate reversal of neuromuscular blockade was the most commonly cited reason for respiratory failure (29; 7%), followed by sick, debilitated patient (10; 2%), lung collapse/consolidation (9; 2%), opioid overdose (6; 1%), obesity-related factors (5; 1%) and unknown aetiology (11; 3%). Problems with the airway and drug related errors are detailed in Tables 6 and 7, respectively.

Table 5.  Presenting problem in 97 respiratory incidents out of 419 recovery room incidents reported to AIMS.
respiratory problem
No. (proportion of
total incidents)
  • *

    including a range of conditions that led to inadequate oxygenation and/or ventilation.

Respiratory failure*74 (18%)
Aspiration 7 (2%)
Respiratory arrest 6 (1%)
Bronchospasm 5 (1%)
Pneumothorax 5 (1%)
Table 6.  Presenting problem in 87 airway incidents out of 419 recovery room incidents reported to AIMS.
airway problem
No. (proportion of
total incidents)
  • *

    11 adult; 7 child.

Airway obstruction59 (14%)
Laryngospasm18 (4%)*
Jaw dislocation 2 (0.5%)
Allergy 2 (0.5%)
Foreign body (throat pack) 2 (0.5%)
Failed extubation 1 (0.2%)
Failed intubation 1 (0.2%)
Endobronchial intubation 1 (0.2%)
Table 7.  Details of 44 drug error-related incidents out of 419 recovery room incidents reported to AIMS.
Type of drug errorNo. (proportion of total incidents)
Inappropriate drug17 (4%)
Overdosage14 (3%)
Inadequate drug/dosage 6 (1%)
Misconnection/wrong route 3 (1%)
Side-effect 2 (0.5%)
Withdrawal 1 (0.2%)
Allergy 1 (0.2%)

Outcomes for patients who suffered an adverse event in the recovery area are shown in Fig. 2. The reasons for transfer to specialist units included provision of ventilatory support (71; 17%), inotropic support (21; 5%), observation of the airway (17; 4%), provision of sedation (7; 2%) and monitoring (6; 1%). Nine patients died (2%), of whom two were emergency cases and seven were elective cases. The aetiological factors for these deaths were primarily cardiovascular (seven) or respiratory (one).

Figure 2.

Outcome in 419 recovery room incidents reported to AIMS. Each report is classified into only one category. Major and Minor outcomes are self-reported subjective outcomes nominated by the reporter. ICU/HDU: admission to Intensive Care Unit, High Dependency Unit or Coronary Care Unit.

Contributing factors involved in the generation of the incidents are shown in Table 8, whilst factors cited as helping to minimise the incident are shown in Table 9.

Table 8.  Contributing factors in 419 recovery room incidents reported to AIMS.
Contributing factorNo. (proportion)
Error of judgement77 (18%)
Communication57 (14%)
Fault of technique29 (7%)
Inadequate patient assessment29 (7%)
Sick patient14 (3%)
Inattention11 (3%)
Other equipment11 (3%)
Surgical team10 (2%)
Lack of equipment10 (2%)
Inexperience 8 (2%)
Haste 8 (2%)
Problem with monitor 4 (1%)
Problem with equipment 3 (1%)
Fatigue 3 (1%)
Lack of facility (bed shortage) 3 (1%)
Drug labelling 2 (0.5%)
Inadequate assistance 2 (0.5%)
Unfamiliar environment 2 (0.5%)
Distraction 1 (0.2%)
Lack of monitor 1 (0.2%)
Pressure to proceed 1 (0.2%)
Relief anaesthetist 1 (0.2%)
Stress 1 (0.2%)
Unfamiliar equipment 1 (0.2%)
Table 9.  Factors minimising the severity of 419 recovery room incidents reported to AIMS.
Factors minimising incidentNo. (proportion)
Prior experience or training97 (23%)
Detection by monitoring device72 (17%)
Skilled assistance54 (13%)
Supervision17 (4%)
Healthy patient14 (3%)
Relief anaesthetist10 (2%)
Improved communication 1 (0.2%)
Quality assurance activity 1 (0.2%)


Incidents that occur in the recovery area are frequently associated with major physiological changes and transfer of care to an Intensive Care Unit (ICU)/High Dependency Unit (HDU) setting. Overall, recovery incidents accounted for 1 in 20 of all reports submitted to the AIMS database. This figure is similar to the analysis of the first 2000 AIMS reports (6%) [2] and the North American Closed Claims Study that looked at recovery room mishaps (7.1%) [1].

It is important to note that complications occurred predominantly in general surgical and orthopaedic patients. Orthopaedic and abdominal operations have been shown to have a higher complication rate and duration of stay in recovery than other procedures [4, 13]. This has important institutional implications. Experience from our own hospital suggests that patents recovering from major orthopaedic surgery have undergone long, often complex surgery and utilise recovery area resources to a higher degree (nursing time, equipment, other procedures) than other groups. Often these patients are not accepted for HDU/ICU care, but are also not stable enough to return to the general ward. Beds in HDU/ICU are in high demand, are costly and occupy the services of HDU/ICU nursing staff, already in short supply. Ward staff are keen to continue to nurse their ‘own’ patients, in order to maintain skills, yet wards are frequently short staffed and ill equipped. A common scenario is for patients to remain in the recovery area under the care of the anaesthetic department. This too has resource implications both for utilisation of recovery area space and from an anaesthetic and nursing staffing perspective.

Complications may occur in patients with little or no coexisting disease, and in some cases this can lead to serious sequelae, including cardiac arrest. This finding is consistent with Hines et al.'s prospective survey of over 38 000 patients, which found that over three-quarters of adverse events occurred in patients who were of ASA status 1–2 [4].

Many ‘minor’ recovery area adverse events have an important impact on utilisation. Bothner et al. found that even ‘minor’ events were considered significant predictors of recovery area utilisation, increasing duration of stay by up to 26%[6]. In the present study, almost one-third of patients suffered a significant physiological disturbance that warranted transfer to HDU/ICU. This figure may reflect certain geographical trends and institutional limitations. In some countries, patients might be automatically transferred to such units, bypassing the recovery area. A study of 15 000 cases by Rose et al. noted that the admission rate to ICU for ventilation was 2.2%, and of these cases, only 9% were unplanned [14]. The present study identified 71 cases in which ventilatory support was required, all of which were unplanned. A survey from the UK noted that admission to an ICU for purely anaesthetic-related reasons was 0.05%[15] whilst the incidence from Rose et al.'s Canadian study looking at all reasons for admission was 0.3%[14].

Although much time and effort goes into the management of ischaemic heart disease in the peri-operative period, the most common cardiovascular presentation was pulmonary oedema, precipitated in the majority of cases by airway obstruction. This specific complication is being increasingly recognised in the immediate postoperative period [16]. Twenty patients suffered a cardiac arrest whilst in the recovery area, with the majority surviving the incident. This compares to a single arrest noted in the first 2000 reports [2]. The reasons for the arrest varied, and half of the patients were categorised as ASA 1–2, illustrating again that serious complications can follow surgical procedures in previously well patients. Data are limited in evaluating cardiac arrest in the recovery area. The study by Hines et al. noted a major cardiovascular event rate (pulmonary oedema or myocardial ischaemia) of 0.1%; however, no patients suffered a cardiac arrest [4]. A French group reviewed cardiac arrest associated with anaesthesia and noted that the overall frequency was 1.1 per 10 000 procedures, with ASA status 3–5 and age > 84 years identified as risk factors [17]. Of particular note, the cited causes for the 11 arrests described included anaesthetic overdose, hypovolaemia and hypoxaemia. At least one human error was implicated with each arrest and inadequate pre-operative evaluation and preparation were considered important in 70% of arrests. The data from the present study reflect similar factors.

Respiratory incidents account for a significant proportion of adverse events in the recovery area, ranging from 9% to 55%[14, 18]. The classification used in the present study separates airway problems from other respiratory complications. In a prospective study of 24 000 patients, Rose et al. noted that critical respiratory events occurred in 1.3% of all admissions to the recovery area [19]. Most were treated easily with insertion of oral/nasal airways or simple manipulation of the airway, and in only 0.1% was tracheal intubation required. We used respiratory failure as the term used to indicate inadequate oxygenation and/or ventilation. Despite the advent of newer neuromuscular blocking agents, residual paralysis was cited as the most common reason for respiratory failure. This may reflect the widespread use of shorter acting drugs and a lack or routine objective evaluation of neuromuscular blockade. It was not uncommon for reports to indicate that either a nerve stimulator was available but not used, or that reversal had not taken place, despite administration of a non-depolarising agent. The recommendation for objective assessment of all patients who have received neuromuscular blockade is strongly supported not only for prevention of inadequate ventilation but also for prevention of awareness [20]. Although considerable effort has gone into the prevention of aspiration in anaesthetic practice, research and recommendations tend to focus on the pre-operative period. Aspiration can occur in the postoperative phase, and this complication needs to be considered for all patients. Simple measures such as transporting unconscious patients on their side and awake extubation in patients at high risk need to be re-emphasised. In one incident, tracheal extubation was performed in the supine position despite the patient's being high-risk, because the orthopaedic surgeon was concerned about dislocating a new hip replacement [9].

Despite considerable information regarding the importance of drug errors in medical practice, and the significance placed on this from previous AIMS data, it is disappointing that drug errors continue to be highly represented in the present AIMS database. Whilst inadequate dosage, overdosage, side-effects and allergy can be obviated to a large extent by education and increased awareness, inappropriate drug delivery and misconnection should be preventable. Indeed, new automated drug delivery systems are being developed that dispense medications through a locked cupboard, via a dedicated system that makes inadvertent swaps such as water instead of lidocaine less likely. A recent system has taken this further and allows bar coding of syringes with auditory and visual feedback of the drug about to be given [21]. Whilst originally intended for use in the operating room, this system has applications for the whole institution and given the range and type of medication used in the recovery area, would seem to be a useful adjunct to improving patients' safety in this environment.

The pattern of contributing factors for incidents in the recovery area is consistent with other areas of anaesthetic practice. In a technical, procedure-orientated area, error of judgement and fault of technique are amongst the most frequent precipitating factors. Of note is the recurring problem of failure of communication. This factor was also deemed to be of significance with the AIMS data reflecting inadequate patient pre-operative preparation [10]. Improvements in written, verbal and electronic transmission of information need to be addressed as a matter of urgency.

Conclusions drawn from incident reporting studies are frequently cited as being inappropriate, invalid and non-representative. All of these criticisms can be argued. What is indisputable is that the incidents provide a wealth of clinical information from which we can suggest possible corrective strategies and conclusions. First, recovery is an extremely important area that requires considerable support from an institutional viewpoint. Dealing with a wide cross-section of surgical procedures and other activities, it has a crucial position in the safe and efficient running of an operating suite. Second, adequate staff numbers and competencies are required for this area. This is reflected in the factors minimising the incidents (Table 9). In a similar context to anaesthetic assistants, recovery area staff are frequently relied upon to provide an immediate extension of the anaesthetist's role, and must be skilled in airway management, cardiovascular manipulation and drug delivery. This comes about by ensuring comprehensive training specific to the area and encouraging ongoing education, supported strongly by their anaesthetic colleagues. Thirdly, drug errors should be addressed by the use of automated drug delivery systems with feedback. In addition, manufacturers of equipment need to take responsibility to ensure that safety is prioritised (e.g. connections for intravenous lines are not compatible with epidural infusion equipment). Fourth, the role of the recovery area as a medium-term recovery unit needs to be addressed. Just under one-third of all patients were transferred to an HDU/ICU environment. Although not formally reviewed in this study, there is evidence that recovery areas will come under increasing pressure to act as intermediate care facilities. Increasing pressure on ICUs along with procedures being undertaken on older, sicker patients with multiple comorbidities mean that patients may be returned to a ward setting with inadequate facilities. Certainly from a local perspective, shortages in trained nursing and medical staff and lack of monitoring equipment place inordinate pressure on general wards. This is against an ever-increasing backdrop of litigation and expectations from families and the general public. Whilst fast-tracking of some procedures may allow patients to bypass the recovery area altogether, increasing duration of stay for other groups may negate any potential savings.


The support of the Australian Patient Safety is greatly appreciated, along with the contribution of the anaesthetists who continue to complete the AIMS forms.