Organisational failures in urgent and emergency surgeryA potential peri-operative risk factor
Abstract
Medical error is an important cause of morbidity and mortality. Organisational failure in the pre-operative period has been associated with catastrophic outcome. Little information is available regarding peri-operative organisational problems. The incidence and nature of organisational failure before urgent and emergency surgery in a district general hospital was studied prospectively in 159 cases over a 30-day period. Organisational failure affected more than half of the cases overall, but varied in both its incidence and its complexity between surgical disciplines. Various causative factors were identified, e.g. 8% of cases were subject to delay due to clinical emergencies. The median [range] time required to rectify the problems was 115 [5−750] min. A consultant anaesthetist and surgeon were present in 30 and 20% of cases, respectively. Difficulty with the preparation of patients for emergency surgery is an important but underevaluated cause of medical error that may put patients at risk.
‘It may seem a strange principle to enunciate as the very first requirement in a hospital that it should do the sick no harm’ (Florence Nightingale, 1859) [1]. The concept that Florence Nightingale expounded cannot have been new even in her day, yet injury and death continue to occur as a result of medical error. Recent figures from the USA suggest that medical errors result in 1000 000 excess injuries each year [2]. The Department of Health estimates the figure in the UK to be 850 000 a year despite a smaller population [3]. The cost of the resulting litigation to the National Health Service (NHS) may be as high as £400 million annually, a figure which continues to rise [3, 4].
The death of a 12-year-old child in 1997 following an intrathecal injection of vincristine under general anaesthesia is of particular interest. Manslaughter charges against doctors caring for him were dropped when an expert witness concluded that hospital failures had a greater responsibility for the death than was first thought. Important organisational failures included the failure to fast the child for an earlier operating theatre list and admission to an inappropriate ward [5]. The more recent death of a patient following the accidental removal of a healthy kidney again illustrates the consequences of inadequate pre-operative preparation [6].
The preparation of cases for surgery is a complex process that may involve more than 20 members of staff. Common organisational difficulties include missing investigation results, inadequate fasting and staff shortages. These problems are latent conditions within the system of pre-operative care [7]. Latent conditions are passive faults in procedure that in isolation may have no adverse consequences, but can promote the occurrence of active errors [7]. Many major disasters resulting from human error could have been prevented had known latent conditions been corrected [8].
Anecdotal evidence suggests that the preparation of cases for emergency surgery is particularly prone to organisational failure. The short time scale from admission to surgery leaves little time for errors to be identified. The urgent nature of many cases puts pressure on the operating team to justify time spent preparing the patient. The importance of pre-operative resuscitation has been emphasised in the National Confidential Enquiry into Peri-Operative Deaths (NCEPOD) reports [9]. Unfortunately, other aspects of pre-operative care have not been subject to the same level of scrutiny. A Medline search of the literature back to 1966 failed to reveal any studies assessing the incidence of pre-operative organisational problems in either emergency or elective surgery. This study was conducted to assess the incidence and nature of organisational problems before urgent and emergency surgery in a district general hospital.
Methods
This prospective study was conducted in a typical acute district general hospital with ≈ 180 surgical beds, serving a population of some 300 000 people in south-east London. Approximately 7500 inpatient surgical procedures are performed each year. Tertiary referral care is provided for ear, nose and throat, vascular and paediatric surgery. The main operating suite consists of five operating theatres with separate obstetric, paediatric and day surgery facilities. During the study period, ≈ 30% of surgical cases were performed on an elective inpatient basis, 35% as emergency inpatients and 35% in the day surgical unit. A 1-week pilot study led to the following patient selection and definitions. The study received approval by the local research ethics committee.
The study was conducted over a 30-day period (without Bank holidays) during which there were 21 scheduled weekday theatre sessions for trauma surgery (weekday trauma list) and 21 scheduled weekday theatre sessions for other surgical emergencies (weekday emergency list). All patients undergoing surgery on either the daytime trauma or emergency lists or the out of hours emergency list were eligible for inclusion. Emergency obstetric patients and the majority of paediatric cases were not studied, as they were performed in separate operating theatre suites. A record of all emergency and trauma activity was kept, allowing an assessment of response rate. A data sheet, which included the patient's age, ASA grade and operation, was completed by the anaesthetist in charge of each case. Further records were kept at the operating theatre reception area to allow verification of data.
The anaesthetist recorded the earliest time at which anaesthesia could have commenced had it not been for the difficulties encountered. The difference between this time and the actual time anaesthesia was started allowed estimation of the number of minutes spent correcting any problems. Problems were classified as ‘organisational’ or as ‘clinical emergencies’. Organisational problems included anaesthetist or surgeon unavailability within office hours, nursing problems, portering delays, missing investigation results, missing patient consent forms and failure to fast patients despite adequate time to do so. Cardiac arrests, trauma calls and obstetric emergencies were classified as emergencies only when these occurred outside office hours. Delays due to other emergency cases and time spent on patient resuscitation before surgery were classified as emergencies at all times of day. Any case delayed until the following day was classified as postponed. Where cases had been postponed from previous days, only that delay incurred on the day of surgery was recorded.
The data were analysed using Microsoft excel. Differences in the incidence of organisational failure between groups were analysed using the chi-squared test. Differences in the length of time spent correcting organisational problems were analysed using the Wilcoxon rank sum test. Significance was defined as p < 0.05.
Results
One hundred and fifty-nine cases were included in the study (93% of those operated on in the study period), of which 93 were female (58%). The mean [range] age of the patients was 46 [7–92] years. Only two cases were delayed for resuscitation. Thirty-three cases were booked for gynaecological surgery, 58 for general and vascular surgery and 61 for orthopaedic surgery. Seven cases were booked by other specialities. Seventy-seven of the procedures were classed as minor, 51 as intermediate and 31 as major.
Eighty-six cases were subject to some form of organisational problem (54%). Thirteen cases encountered delay due to clinical emergencies (8%), whereas 23 cases (14%) experienced multiple problems. The median [range] time to rectify each problem was 115 [5−750] min (Fig. 1). Sixteen cases were postponed by 1 or more days (10%); these often experienced further delay on the day that they were actually performed.
Number of cases categorised by time taken to rectify problems. Delays due to clinical emergencies have been included. The number of undisrupted cases is given for comparison (time = 0 min).
The most common organisational problems encountered were the presence of a non-urgent prior case, e.g. elective list overrunning into the scheduled emergency list, and the lack of results of investigations (Table 1). Lack of operating theatre staff was seldom a problem, but lead to long delays when it did occur. Being booked for a daytime trauma or emergency list did not guarantee that the patient would have their surgery in these slots. On average, a third of such patients were eventually operated upon out of hours (Table 2), yet an average of only 1.7 and 2.4 cases were booked on each weekday trauma and emergency list, respectively, despite the fact that many of these operations were relatively short surgical procedures.
| Organisational problem | No. of cases | Median [range] time taken to rectify problem; min |
|---|---|---|
| Non-urgent prior case | 21 | 195 [25−635] |
| Investigation results unavailable | 16 | 150 [20–450] |
| Surgeon unavailable | 14 | 135 [10–420] |
| Clinical emergency | 13 | 180 [10–750] |
| Patient not prepared by ward | 13 | 90 [45–300] |
| Portering delay | 10 | 38 [5−330] |
| Patient not fasted | 8 | 150 [30–380] |
| Theatre staff unavailable | 7 | 165 [90–635] |
| No consent | 7 | 115 [20–300] |
| Anaesthetist unavailable | 7 | 30 [10–280] |
| Overall | 99 | 115 [5−750] |
| Weekday trauma | Weekday emergency | Out of hours | Overall | |
|---|---|---|---|---|
| No. of cases | 35 | 50 | 74 | 159 |
| No. affected by organisational problems | 19 (54) | 35 (70) | 32 (43) | 86 (54) |
| No. affected by clinical emergencies | 1 (3) | 3 (6) | 9 (12) | 13 (8) |
| No. performed out of hours | 12 (34) | 16 (32) | 72 (97) | 100 (63) |
| Time taken to rectify problem; min | 45 [5−360] | 118 [10–635] | 150 [10–750] | 115 [5−750] |
Seventy per cent of gynaecology cases experienced organisational problems; these required a median time of 120 min to rectify. Forty per cent of all the remaining cases suffered significant organisational delays, taking a median of 105 min to rectify (Table 3). The figures suggest greater direct consultant supervision within the operating theatre by anaesthetists than surgeons (Table 4); no data are available regarding cases supervised at a distance from within the operating theatre suite or within the hospital. The direct involvement of a consultant anaesthetist had no effect on whether a case was likely to encounter organisational failure (trainees = 63%, consultants = 60%, p = 0.54). In addition, the difference in the length of time required to rectify organisational problems, between trainees (median [range]= 120 [5–525] min) and consultants (median = 70 [10–750] min) was not significant (p = 1.0).
| Orthopaedics | General and vascular | Gynaecology | Other | Overall | |
|---|---|---|---|---|---|
| Number of cases | 61 | 58 | 33 | 7 | 159 |
| Number affected by organisational problems | 32 (52) | 27 (47) | 23 (70) | 4 (57) | 86 (54) |
| Number affected by clinical emergencies | 4 (7) | 8 (15) | 1 (3) | 0 (0) | 13 (8) |
| Number postponed | 10 (16) | 5 (9) | 1 (3) | 0 (0) | 16 (10) |
| Time taken to rectify problem; min | 90 [5−360] | 150 [10–750] | 120 [15–635] | 75 [10–165] | 115 [5−750] |
| Weekday trauma | Weekday emergency | Out of hours | Overall | |
|---|---|---|---|---|
| Consultant anaesthetist | 16 (46) | 20 (40) | 12 (16) | 48 (30) |
| Consultant surgeon | 10 (29) | 14 (28) | 8 (11) | 32 (20) |
The difference in the incidence of organisational failure in those cases performed by a consultant surgeon and those performed by a trainee surgeon failed to reach statistical significance (trainees = 65%, consultants = 50%, p = 0.06). The difference in the length of time required to rectify organisational problems between surgical trainees (median [range] = 115 [5−635] min) and consultants (median [range] = 90 [25–750] min) was not significant (p = 0.38).
One hundred and fifteen patients (72%) were classified as ASA I or II and the remaining 44 as ASA III or IV (28%). The ASA grade had no effect on the incidence of organisational delay (ASA I and II = 60%, ASA III and IV = 68%, p = 0.27). The duration of delay was not significantly affected by ASA grade (ASA I and II median [range] = 120 [5−750] min, ASA III and IV median [range] = 85 [20–420] min, p = 1.0).
Fourteen cases were delayed by the unavailability of a surgeon due to other commitments, e.g. outpatient clinic or ward round; on only one occasion was this because of an emergency commitment. An anaesthetist was unavailable in seven instances, four of these being due to clinical emergencies elsewhere.
Table 5 describes four vignettes in which organisational problems put patients at risk of injury.
| 1. | A patient with an ectopic pregnancy booked for theatre was found to have no full blood count or group and save results available. It is understood that the laboratory did not receive any blood samples. |
| 2. | Two elderly females were admitted on a Friday evening, both having sustained a fractured neck of femur. Neither patient was clerked by the admitting doctor and no investigations were ordered. In the ensuing confusion, one patient did not arrive in theatre until 36 h after admission and the other until 64 h after admission. |
| 3. | A haemophilia carrier with reduced Factor VIII levels presented with an ectopic pregnancy. No group and save sample was taken. Pre-operative advice from a consultant haematologist was requested. However, the suggested treatment regimen was not prescribed or administered. The necessary therapy was eventually arranged by the anaesthetist who discovered the omission. |
| 4. | A patient attended the Accident and Emergency department with appendicitis. Poor communication between medical staff resulted in no surgeon assessing the patient until 10 h after arrival. |
Discussion
This study revealed an unexpectedly high frequency of organisational failures before urgent and emergency surgery, affecting over half of the cases examined. Even more revealing was the time taken to correct such failures, e.g. a median time of 115 min was required to replace a missing consent form. Several cases illustrated the causative link between organisational failure and medical error (Table 5), which could have affected operative safety. While we are unaware of any additional morbidity or mortality due to these organisational delays, we did not specifically look for them. Data were provided principally by the anaesthetist in charge of each case and then verified with records kept in the operating theatre reception. By applying a rigid definition of delay and providing a well-structured data sheet, observer bias was kept to a minimum. This study has been discussed at length within our institution; changes to care pathways and clinical governance may have an impact on the problems discussed. Survival rates at University Hospital Lewisham are well above the national average for both emergency and elective surgery [10].
Organisational failures appeared to be less common when fewer staff were involved; patients cared for by a small team out of hours had fewer problems than those scheduled on the weekday emergency list (Table 2). Similarly, orthopaedic trainees, working a traditional on call rota, experienced fewer difficulties than gynaecology trainees working partial shifts (Table 3). This may reflect better communication and fewer links in the chain.
The study assessed the hospital procedures for emergency surgery, but no attempt was made to assess other aspects of care. As 48% of the procedures were classified as ‘minor’, there may not have been a great sense of urgency in getting some patients adequately prepared for the operating theatre. However, delayed non-elective surgery is unlikely to benefit the patient. Furthermore, minor procedures are often performed on patients with significant comorbidity. Where multiple delays occurred, it was not possible to distinguish between the time spent correcting each individual problem. The amount of time spent rectifying organisational problems may therefore be overestimated in 14% of cases. The range of times required to correct failures included brief intervals of < 10 min, suggesting that, in some cases, rapid correction of the problem was possible. However, it is worrying that, for the majority of patients, organisational failings took a much longer period to correct.
A literature review failed to find any other studies detailing the incidence or nature of organisational problems before surgery. A recent study quantified delays in urgent trauma surgery but made little attempt to explain why they occurred [11]. Weekday operating lists for emergency surgery were introduced to improve supervision of trainees and to reduce the number of cases starting late at night [12]. It is therefore frustrating to confirm previous data showing poor utilisation of these lists [13]. The failure to perform cases during scheduled weekday emergency and trauma sessions was an important reason for the poor level of consultant supervision for cases booked on these lists (Table 4). However, more detailed guidance is needed on exactly what level of supervision is appropriate for the development of surgeons and anaesthetists.
Two important causes of disruption were over-running elective lists and surgeon unavailability (due to simultaneous responsibilities in several areas). Whilst NHS targets continue to emphasise elective surgery and outpatient commitments over emergency work, this situation is unlikely to change. However, better use of weekday emergency lists would reduce out-of-hours operating theatre expense, shorten inpatient hospital stay and reduce work intensity for junior doctors. Such incidental benefits should provide sufficient financial incentive to tackle such an important factor in patient safety.
The anaesthetist and the anaesthetic assistant are often the last to scrutinise the preparations for each case and they play a crucial role in ensuring the safety of the patient. Future studies should take a more detailed approach, examining every stage of care of the emergency surgical patient. This would require considerable resources and the co-operation of all specialities. It is likely that a ‘whole system’ pathway analysis is required to effect significant improvements for the majority of patients.
One potential solution is to concentrate all emergency surgical patients in one ward before surgery. Nursing staff could then familiarise themselves with the demands of rapidly preparing such cases and the ward would become a focus for communication. Alternatively, clinical nurse specialists could be used to co-ordinate urgent and emergency cases. In the recent government report ‘An organisation with a memory’, targets were set to eliminate errors that occur repeatedly year after year [3]. Without improvements in organisation, the emergency surgical patient remains at risk from a disaster waiting to happen.
Acknowledgments
We thank all the members of the Department of Anaesthesia and Pain Management at University Hospital Lewisham for their help with this study. We are particularly grateful to Dr Sally Edwards for her advice and support.
