A prospective, randomised clinical evaluation of ... - Wiley Online Library

Anaesthesia, 2004, 59, pages 80–87 .....................................................................................................................................................................................................................

APPARATUS

A prospective, randomised clinical evaluation of a new safety-orientated injectable drug administration system in comparison with conventional methods* C. S. Webster1,2, A. F. Merry1,2, P. H. Gander3 and N. K. Mann1 1 Department of Anaesthesia, Green Lane Hospital, Private Bag 92-189, Auckland, New Zealand 2 Department of Anaesthesiology, School of Medicine, University of Auckland, Auckland, New Zealand 3 The Sleep ⁄ Wake Research Centre, Massey University, Wellington, New Zealand Summary

Fifteen anaesthetists were observed while providing anaesthesia for 15 pairs of adult cardiac surgical operations, using conventional methods for one of each pair and a new drug administration system designed to reduce error for the other. Aspects of each method were rated by users on 10-cm visual analogue scales (10 being best). The new system was rated more favourably than conventional methods in terms of safety (median [range] ¼ 8.1 [6.8–9.7] vs. 7.1 [2.6–9.3] cm; p ¼ 0.001) and usability (8.5 [5.9–9.4] vs. 7.5 [3.2–9.8] cm; p ¼ 0.027). The new system saved preparation time both before anaesthesia (median [range] ¼ 180 [32–480] vs. 360 [120–600] s; p ¼ 0.013) and during anaesthesia (10 [2–38] vs. 12 [10–60] s; p ¼ 0.009). Prefilled syringes for the new system increased costs by e23.00 per anaesthetic (p ¼ 0.041), but this increase is likely to be offset by the potential of the new system to decrease costly iatrogenic harm by preventing drug error. Keywords

Safety. Anaesthesia; intravenous. Drug labelling. Risk reduction behaviour.

. ......................................................................................................

Correspondence to: Mr C. S. Webster E-mail: [email protected] *Some preliminary results of this study were presented at the Annual Scientific Meeting of the Australian and New Zealand College of Anaesthetists, Melbourne, Australia, May 2000. Accepted: 5 July 2003

Iatrogenic harm is very costly in terms of increased hospital stay, resources consumed, injury to patients, lives lost and careers ruined [1–4]. The Institute of Medicine Report has estimated that between 44 000 and 98 000 deaths each year in the United States of America are due to preventable adverse events in health care [3]. Drug administration error has been identified as a leading cause of such harm [3, 5], and is of particular concern in anaesthesia, because of the sheer number of drugs that anaesthetists are required to administer [2, 5, 6]. In a prospective study of 10 806 conventional anaesthetics, our group recently estimated a rate of 1 error per 200 anaesthetics in the administration of drugs by injected bolus [7]. In addition, drug administration error has been identified as the most common cause of awareness under 80

anaesthesia and is a persistent cause of incidents in the recovery room [8, 9]. Much is known about the nature of safety in complex, high-technology organisations [10–13] and recently there have been widespread calls for this expertise to be applied in medicine [14–17]. Therefore, members of our group have developed a new administration system for injectable drugs with the express aim of reducing error and facilitating safe practice [18]. The system’s design is based on lessons from empirical incident reporting [7], the psychological mechanisms underlying human error [10] and the principles of safe-system design developed in the high-risk chemical, nuclear and aviation industries [11, 12, 19, 20]. The new system includes specialised trays (which support aseptic technique and promote a well-organised anaesthetic 2004 Blackwell Publishing Ltd

Æ

Anaesthesia, 2004, 59, pages 80–87 C. S. Webster et al. Evaluation of a drug administration system . ....................................................................................................................................................................................................................

workspace), colour- and bar-coded labelling of syringes (which facilitate the selection and tracking of drugs), prefilled syringes for the most commonly used anaesthetic drugs (to remove a key error-prone step in the anaesthetic and to save time), and automatic visual and auditory verification of syringes using a computer and bar-codes just before each drug administration. The new system also compiles a timed record of drugs administered throughout the anaesthetic. A more detailed discussion of the design and operation of the new system and its components is available elsewhere [18, 21–23]. In a pilot study, a prototype of the new system was evaluated in a highfidelity human-patient anaesthesia simulator and was judged as offering a significant safety advantage over conventional methods [21]. In this study, we aimed to evaluate the new system in a clinical setting during a larger number of anaesthetics before introducing the system into wider use. Methods

The system used in this study (the ‘Mark II’), is a refined version of the prototype (the ‘Mark I’) drug administration system [21]. The refinements, based on the simulator evaluation, include: 1 simplification of the drug tray such that one, two or three trays can be used in a modular fashion depending on the complexity of the anaesthetic; 2 rolls of labels coloured by class of drug according to international standards and presented in dispensers on a small shelf above the anaesthetic drug-trolley work surface [24–26]; 3 pharmacologically validated prefilled syringes for the most commonly used anaesthetic drugs; 4 barcodes (containing expiry dates) on prefilled syringes, allowing the announcement of an expired-drug warning upon scanning; 5 the provision of a portable hand-held barcode scanner on a long flexible cord; 6 voice announcements restricted to drug name, without concentration, to avoid confusion when the contents of a syringe are diluted; and 7 the addition of a mute feature that enables the anaesthetist, with a single key-stroke, to switch off the voice announcement of the scanned drug for the next administration only. At the time of the study, six of the most commonly used drugs in cardiac anaesthesia were available in the Mark II prefilled format: fentanyl, pancuronium, calcium, ephedrine, midazolam and magnesium. All prefilled syringes, except fentanyl, were readily available from the Mark II drug trolley. Other drugs used during Mark II anaesthetics were drawn up conventionally. 2004 Blackwell Publishing Ltd

Use of the system’s hardware is governed by the following procedural rules: 1 that each anaesthetic should start with an empty workspace and trays; 2 that all empty ampoules or prefilled syringes should be retained on trays and separated from drugs in use as a supplementary record of the anaesthetic; 3 that the Mark II labels should be used on all conventionally filled syringes; and 4 that the labels should be read and each syringe scanned with the barcode reader before each drug administration. This prospective randomised study replicated the methods of a previous simulator-based evaluation [21] and was carried out in the cardiothoracic surgical unit at Green Lane Hospital, Auckland, New Zealand with local research ethics committee approval and written informed consent from participating patients. Volunteer anaesthetists were recruited to take part in the study on the basis of availability and with informed consent obtained verbally. Each participating anaesthetist was briefed on the features and operation of the Mark II system before the study, and each participating senior anaesthetist used it to perform at least one adult cardiothoracic anaesthetic with the assistance of a member of the study group. Participants were then observed while providing anaesthesia for coronary artery or heart valve surgery. A conflict of interest (as detailed below) precluded AFM’s participation in any study anaesthetic and CSW carried out all observations. To control for order effects (including time of day and fatigue) each anaesthetist was asked to use the Mark II system in the morning and conventional methods in the afternoon or vice versa, decided at random. When two or more anaesthetists worked together, randomisation was based on the primary anaesthetist (designated as the one with the greatest responsibility for patient care). The study observer was present in the anaesthetist’s work area throughout the anaesthetic and recorded the preparation times for intravenous drugs used. Preparation times for infusion syringes were recorded separately. Techniques used by the anaesthetist to prevent inadvertent administration of drugs perceived as potentially dangerous, such as protamine and metaraminol, were also recorded, as were the use of labels and any problems that occurred in the work environment that could affect the conduct of the anaesthetic. Comments made by anaesthetists that were related to intravenous drug administration were also recorded. When the Mark II system was used, failures to comply with operational rules were recorded: these included not keeping used ampoules or empty syringes, not scanning labelled syringes with the barcode reader before administration, not using labels, and not using drug trays. Difficulties experienced with the use of any aspect of the Mark II system were also recorded. 81

Æ

C. S. Webster et al. Evaluation of a drug administration system Anaesthesia, 2004, 59, pages 80–87 . ....................................................................................................................................................................................................................

Visual analogue scales completed after each conventional and Mark II anaesthetic Title Range 1. Acceptability Not at all – entirely 2. Practicality Not at all – entirely 3. Ease of use Not at all easy – very easy 4. Likelihood of forgetting to give agents Very likely – very unlikely 5. Likelihood of giving the wrong agent Very likely – very unlikely 6. General propensity for error Very likely – very unlikely

Table 1 Questions for visual analogue

scales and free-form comments used in the evaluation of Mark II and conventional systems.

Visual analogue scales completed after each Mark II ⁄ conventional pair of anaesthetics Title Range 7. How safe is the Mark II system compared with existing Unsafe – safe methods? 8. How clinically acceptable is the Mark II system compared with Not at all – entirely conventional methods? 9. Rate the organisation and layout of the Mark II system Very poor – excellent compared with conventional methods Free-form comments Please record any aspects of the Mark II system you consider unacceptable Any comments about any aspects of the Mark II labels? Any further comments about the Mark II system?

The total cost of intravenous drugs, syringes and drawing-up needles was calculated for each anaesthetic. At the end of each anaesthetic (whether carried out with the Mark II system or conventional methods), the primary anaesthetist was asked to rate, on 10-cm visual analogue scales (VAS), aspects of the method just used (Table 1). Each primary anaesthetist was also invited to record on their VAS form any comments about good or bad aspects of the system just used, as well as suggestions for improvement. In addition, at the end of the study day, after a pair of observed anaesthetics, the primary anaesthetist was asked to complete a further three VAS scores that directly compared the Mark II system with conventional methods, and was given a further opportunity to make comments. On all VAS scales, a score of 10 was best and a score of 0 worst. Our objective in this study was to compare the Mark II system with conventional methods of drug administration in terms of safety and usability, using the same endpoints as the previous simulator-based evaluation [21]. Therefore, our primary outcome variables were: a single ‘safety score’ (calculated for each primary anaesthetist by taking an average of VAS scores 4–6 in Table 1) and a single ‘usability score’ (calculated by taking an average of VAS scores 1–3 in Table 1). All group comparisons were performed using the Wilcoxon signed-rank test. One-sample Wilcoxon signed-rank tests were used to identify differences from equality in the direct VAS comparison of the Mark II with conventional methods. Analyses were carried out using SYSTAT 10 (SPSS Inc., Chicago, IL, USA). A probability value of < 0.05 was designated as significant. Based on the effect size for safety ratings in the previously published simulator-based study [21], a group size of 15 was 82

calculated to provide a 90% power to detect a difference at the p ¼ 0.05 level. Qualitative free-form comments and observed events were grouped on the basis of related themes and used to supplement quantitative data in the validation of the Mark II system [27]. Results

One patient declined to give consent and two were not studied after giving consent because their procedures were postponed to an indefinite later date. Eleven senior anaesthetists (nine consultants and two senior registrars) and four registrars participated in the study. After randomisation, four patients were not included in the analysis because an unexpected change of anaesthetist between the first and second anaesthetic precluded matching pairs of observations and so no group comparison could be performed. Analysis was therefore performed on observations involving one pair of patients for each of 15 anaesthetists. Eight Mark II observations were performed in mornings and seven were performed in afternoons. On one day of observation, after being randomised to use the Mark II system in the morning, one anaesthetist insisted on continuing to use the Mark II system in the afternoon. The second observation of the study pair was conducted with the same anaesthetist and using conventional anaesthetic methods on an afternoon of a subsequent day. The postponement of the procedures of two patients (mentioned above) necessitated the completion of a further two pairs of observations on an afternoon of a subsequent day. Mark II and conventional anaesthetics were similar in terms of the operations performed, duration of anaesthesia, the number of syringes and drugs used, and in terms of 2004 Blackwell Publishing Ltd

Æ


Table 2 Characteristics of study anaesthetics. Values are number

Mark II system (n = 15) Surgery type Coronary artery bypass and graft Heart valve replacement or repair Details of anaesthetics Syringes setup before anaesthetic Additional syringes used during anaesthetic Infusions administered Boluses administered Number of different drugs used Duration of anaesthetic; min

13 2

11 4

8 [5–12] 9 [5–13] 2 22 16 229

Conventional system (n = 15)

[1–5] [12–31] [10–20] [180–454]

7 [4–11]* 9 [6–12] 2 21 15 235

[1–4] [16–29] [10–17] [185–300]

10

Average visual analogue score (cm)

or median [range].

9 8 7 6 5 4 3 2 1 0 M

*One data point not recorded.

C Usability

M

C Safety

Figure 1 Combined box plot and scattergram of visual analogue

the number of individual drug boluses and infusions administered throughout (Table 2). Setting up drugs at the beginning of an anaesthetic was significantly faster with the Mark II system than with conventional methods (median [range] ¼ 180 [32–480] s vs. 360 [120–600] s; p ¼ 0.013). The use of prefilled syringes in the Mark II group also significantly decreased the median time required to prepare supplementary syringes of drug during an anaesthetic when compared with conventional methods (median [range] ¼ 10 [2– 38] s vs. 12 [10–60] s; p ¼ 0.009). The preparation of drug infusions took a median of 120 s in both groups. The cost of drugs, syringes and drawing-up needles used during an anaesthetic varied widely between anaesthetists. However, in Mark II anaesthetics, the use of prefilled syringes increased costs by a median of e23.00 compared with conventional methods (median [range] ¼ e178.00 [102–428] vs. e155.00 [111–390]; p ¼ 0.041). Median [range] usability VAS scores (Fig. 1) were 8.5 [5.9–9.4] cm for the Mark II system and 7.5 [3.2–9.8] cm for conventional methods (p ¼ 0.027). Median [range] safety scores were 8.1 [6.8–9.7] cm for the Mark II system and 7.1 [2.6–9.3] cm for conventional methods (p ¼ 0.001). One form was not returned in the direct comparison of the Mark II with conventional methods conducted at the end of each study day. This analysis has therefore been performed with 14 responses. The Mark II system compared favourably with conventional methods (Fig. 2) in respect of safety (median [range] 8.6 [6.7–10] cm, p ¼ 0.001), clinical acceptability (8.3 [6.0–9.9] cm, p ¼ 0.001), and organisation and layout (8.4 [5.3–10] cm, p ¼ 0.001). The Mark II pre-administration computerised check led to the prevention of a potential error on four separate 2004 Blackwell Publishing Ltd

scores for usability (p ¼ 0.027) and safety (p ¼ 0.001) for the Mark II (M) and conventional (C) anaesthetic systems (10 cm being best, n ¼ 15). Box length indicates the interquartile range. The line inside the box indicates the median. Whiskers encompass those values falling within limits located between the ends of the box and 1.5 times the interquartile range beyond the ends of the box. Outliers are shown beyond these limits. See Table 1 for the questions used in the VAS scores.

O

C

S

0

1

2

3

4

5

6

7

8

9

10

Visual analogue score (cm) Figure 2 Combined box plot and scattergram of visual analogue

scores for anaesthetic procedures, with the Mark II system in direct comparison with conventional methods in respect of (S) safety (p ¼ 0.001), (C) clinical acceptability (p ¼ 0.001), and (O) organisation and layout (p ¼ 0.001; 10 cm being best, n ¼ 14). Box plot description as for Fig. 1. A score of 5 cm would imply an equal rating; scores > 5 cm favour the Mark II system; scores < 5 cm favour conventional methods. See Table 1 for questions used in the VAS scores. 83

Æ


occasions when expired prefilled syringes (two of pancuronium and one each of midazolam and calcium), were detected on scanning because of the announcement of the expired drug warning. Despite this, the voice announcement was silenced with the mute key on five occasions: twice during the administration of metaraminol, and once each during the administration of calcium, magnesium and heparin. In addition, the drug administration record, automatically compiled by the Mark II system, was referred to on four separate occasions to complete details of the hand-written anaesthetic record. A number of observations related to colour-coding problems and the drawing up of drugs and labelling of syringes were made during anaesthetics conducted with conventional methods. In three anaesthetics, fluorescent ‘medication added’ stickers, overwritten with ball-point pen, were used to label the majority of syringes, even though conventional colour-coded labels were available. On one occasion, two differently coloured user-applied labels were available for the same drug (midazolam). On

another occasion, five syringes were identified by abbreviated drug names only, written on syringes in black marker pen. In a further three anaesthetics, the anaesthetist used a combination of syringe size and drug ampoules left on the end of the drawing up needle to identify syringes. This system appeared to break down as the anaesthetic progressed, with the appearance of some completely unlabelled syringes containing drug but without ampoules on their needles. Observations and comments related to procedural and environmental aspects of study anaesthetics are summarised in Table 3. Six anaesthetists made positive comments about time savings and improved safety associated with the use of the new system. Two anaesthetists commented that all drugs should be prefilled for ready administration. However, another two anaesthetists preferred to draw up drugs only as and when needed as a safety strategy. With the use of both the Mark II and conventional methods, drug syringes were often stored in locations other than the drug tray, but the mixing of drugs

Table 3 Comments and observations for procedural aspects of Mark II and conventional systems. The frequency of individual events

and comments is one unless otherwise indicated in brackets. Mark II system Procedures Comments made by anaesthetist Mark II prefilled syringes save time – could potentially make 15 min of setup into 3 min; helps greatly in setting up quickly and accurately (2) Would like to see all drugs in prefilled syringes, including metaraminol, protamine and potassium (2) There is a tendency to give the drug then scan the syringe

Observations Kept some or all empty ampoules and syringes (7) Drug syringes at head of patient (2); on anaesthetic machine (3); attached to patient’s central venous line (2) Drugs mixed in the same syringe: cephazolin and tranexamic acid (3); vecuronium and pancuronium; propofol and fentantyl for infusion Safety Comments made by anaesthetist New system will reduce error rate; new system great improvement in terms of safety (2) Mark II trays should be used to separate syringes into those needed pre ⁄ post bypass Observations Prepared diluted metaraminol in 1 ml syringe (4) Registrar cut finger on glass drug ampoule Environmental factors Observations Transoesophageal echocardiography machine present – up to five anaesthetists crowded into cramped anaesthetic workspace (2) Confusion over how much protamine had been given when second case started before first finished and senior anaesthetist out of room

84

Conventional system

Observations Drugs mixed in the same syringe: cephazolin and tranexamic acid (8); cephazolin, tranexamic acid and dexamethasone (2); vecuronium and pancuronium Drug syringes on anaesthetic machine (7); attached to patient’s central venous line (3) Kept some or all ampoules (8) Violation of sterile technique during draw up of drug when finger touched needle Refilled fentanyl syringe with potassium without relabelling Open etomidate ampoule carried over from previous case Registrar uncertain how much heparin had been drawn up while senior out of room Observations Prepared diluted metaraminol in 1 ml syringe (4) Located protamine, metaraminol and heparin away from rest of drugs (3) Drew up drugs as needed – particularly protamine (2) Senior anaesthetist cut finger on glass drug ampoule Observations Case delayed for 1 h because theatre in use with long morning case Burette overflow (2); malfunctioning heart rate monitor No anaesthetic technician available to do blood gas analysis Pressure problem with perfusion machine Patient draped in the wrong position for vein harvesting because of confusion over notes

2004 Blackwell Publishing Ltd

Æ


in a single syringe occurred less often with the Mark II system. On one occasion, violation of sterile technique occurred during the drawing up of a drug, and two anaesthetists cut themselves while opening glass drug ampoules. In addition to its benefits, use of the Mark II system did not appear to create any new problems or difficulties in the wider operating theatre environment. However, a number of problems with established operating theatre systems were observed, including overcrowding of the anaesthetic workspace, case scheduling problems, communication breakdown and equipment malfunction. Discussion

To our knowledge, this study is the first prospective, randomised, clinical trial of an intervention designed to reduce drug administration error in anaesthesia. It represents a further step in a programme of work intended to reduce error in anaesthesia through better system design [7, 18, 21]. The study’s primary finding was that in clinical use, the new system was judged as significantly safer and more usable than conventional methods of drug administration. In addition, the new system saved time. Ratings of the new system (Mark II) in this study were more favourable than those for a prototype (Mark I) used in a high-fidelity simulator evaluation [21], suggesting that refinements made before the clinical trial may have improved the system. In particular, it is interesting to note that the Mark II system was judged to be more usable than conventional methods in our study – a result in contrast with the prototype evaluation, in which usability was judged as comparable with conventional methods. This study has a number of limitations. As blinding was not possible, observer bias cannot be ruled out. The data from four observed anaesthetics could not be used because an unexpected change of anaesthetist made it impossible to perform a paired comparison in these cases. One anaesthetist violated the randomisation schedule by insisting on using the Mark II system all day on the planned day of observation, and the schedule was also violated twice because of case postponement. However, we do not believe that these difficulties biased our results. On four occasions, anaesthetists were observed to use the drug administration record, automatically compiled by the Mark II system, to complete their hand-written anaesthetic record. When an anaesthetist is busy giving a number of different drugs, particularly during a crisis, he or she is caught in a double-bind. It is not possible to administer drugs and manually record their details simultaneously. Discrepancies between what gets recorded retrospectively and what actually happened are well known in anaesthesia [28]. The drug administration 2004 Blackwell Publishing Ltd

record aspect of the Mark II system has now been expanded considerably – a complete automated anaesthetic record, including physiological data, patient and clinician information, procedural events and clinical notes is now compiled [18]. The use of prefilled syringes increased the drug cost of anaesthetics in this study. However, the median increase (e23.00) was only 8% of the variation in drug cost of conventional anaesthetics (e111.00–390.00), and this increase is considerably smaller when compared with the total cost of an anaesthetic, including all equipment and salaries. Studies have shown that an awareness of the cost of drugs can lead to considerable savings, simply because anaesthetists can then make more use of cheaper alternatives; this can be done without compromising patient care [29]. Therefore, it seems likely that in the medium term, some rationalisation of drug usage could offset the cost of prefilled syringes and the new administration system. To promote awareness of the cost of drugs, the new system displays the cost of individual drugs and a running total of the cost of the anaesthetic so far. In addition, it is well known that harming patients is very costly and inefficient [3], and drug administration error currently occurs at a rate that is unacceptably high [7]. In the long-term, the prevention of errors is likely to translate into considerable savings from fewer cases of iatrogenic harm. Furthermore, large-scale studies indicate that even minor intra-operative incidents (including drug errors) are associated with longer hospital stay and greater resource utilisation [30], both of which are expensive. Taken together, this evidence suggests that a safe anaesthetic is in fact a cost-effective anaesthetic. Therefore, we believe that the new administration system is likely to more than pay for itself in the long term. Recently, an analysis of the task of administering a bolus of intravenous drug identified the involvement of 41 distinct steps [31]. The use of the Mark II system and prefilled syringes would remove the need for 23 (56%) of these 41 steps. This reduced task complexity is consistent with principles of safe system design, and is likely to be particularly advantageous during an emergency [22, 32]. Two well-known occupational hazards in anaesthesia that are avoided with prefilled syringes are cuts from glass ampoules (seen twice in this study) [33–35] and needlestick injury, which can occur even with the use of ‘safety’ needles [36, 37]. Prefilled syringes also eliminate three known hazards for the patient: violation of sterile technique during the drawing up of a drug (seen once in our study); drug error because of the mislabelling of a syringe under time pressure in the operating theatre [5]; and the persistent problem of the injection of minute glass shards which can contaminate the contents of glass drug ampoules upon opening [38] (a filter is used routinely during the 85

Æ


manufacture of prefilled syringes [22]). However, we consider it unwise to present potentially very dangerous drugs, such as protamine, potassium and metaraminol in a prefilled format, because of the possible increased risk of inadvertent administration. Most anaesthetists have their own special strategies for avoiding error with dangerous drugs, and from a systems-safety perspective it would seem sensible to allow them to continue to handle these drugs in a separate and distinct manner. However, the advantages of having the bulk of remaining drugs in a prefilled format clearly go beyond convenience and time saving. High risk, complex work environments can be understood as extended cognitive systems [39] in the sense that each individual must act in a precise and timely manner in coordination with many other members of a team. Details of the status and nature of procedures in the operating theatre are acquired from many sources, including records such as patient notes and the anaesthetic chart, the position and state of equipment, and the actions and conversations of clinicians. Cognitive task analysis identifies one of the most dangerous states of a complex system as one in which there are no external indicators of the status of current tasks, i.e. when this information resides only in the head of an individual [20, 39]. Our study saw two instances of this situation when a consultant left the room with a registrar caring for the patient, without communicating or recording the details of heparin and protamine use. In addition, a patient was draped in an incorrect position for vein harvesting because staff did not know the particular surgical approach to be used. While the new drug administration system attempts to ‘externalise’ as much information about the anaesthetic as possible [18, 21], it seems clear that other organisational aspects of the operating theatre could benefit from similar techniques [40]. It is also likely that efforts to increase external information sources in the operating theatre will be tempered to some extent by the culture prevailing among individual groups of anaesthetists. The mute feature of the Mark II system, which silences the drug voice announcement for the next administration, was put in place at the request of anaesthetists in the simulator trial [21], who felt the need to occasionally administer a drug without sharing this information with others, particularly the surgeon. The mute feature was used on five occasions in this study (although no complaints were made about the voice when not muted). Such a ‘need to know’ approach to communication runs counter to the principles of crew resource management that are now central to operational safety in the aviation industry [27]. Despite the fact that anaesthesia often looks to the aviation industry for ways to improve safety, the anaesthetic workspace in a modern operating theatre seems far less technologically and environmentally integrated and more prone to failure than the cockpit of an aircraft [6, 27]. 86

Despite the fascinating and unique nature of the environment, the anthropology of the operating theatre remains almost entirely unexplored. Planned on-going development of the new drug administration system includes the addition of pre-operative assessment and postoperative medication prescription modules that can be seamlessly integrated with theatre systems through the hospital’s computer network. In addition, such planned integration of the system will allow on-line access to patient information (including allergies and laboratory test results), automatic alerts for allergies and contraindicated drugs, and integration with hospital stock control and billing systems. Future research will be needed to validate these developments in keeping with the existing standards of continuous quality improvement [41]. In conclusion, this study provides evaluative evidence in the clinical setting, which supports that from a previous simulator-based study, that the new drug administration system has the potential to reduce drug error in anaesthesia, thus addressing one aspect of the important and costly problem of iatrogenic harm in health care. Acknowledgements

Grants from the Australian and New Zealand College of Anaesthetists and the New Zealand Lottery Grants Board (AP68583) supported this project. We thank all the anaesthetists who took part in the study, Louise Larsson for technical support, Diana Emmens for critical review of the manuscript and Dr Cornelis Kruger for assistance with the calculation of costs. Mr Webster contributed to this work during the tenure of a three-year Training Fellowship Award from the Health Research Council of New Zealand. Patents and design registrations have been applied for in respect of certain aspects of the new drug administration system. Professor Merry has financial interests in the new drug administration system through share ownership in Safer Sleep Ltd, Auckland, New Zealand. References 1 Ferner RE. Medication errors that have led to manslaughter charges. British Medical Journal 2000; 321: 1212–16. 2 Merry AF, McCall Smith A. Errors, Medicine and the Law. Cambridge: Cambridge University Press, 2001. 3 Institute of Medicine. To Err Is Human – Building a Safer Health System. Washington, DC: National Academy Press, 2000. 4 Department of Health. An Organisation with a Memory – Report of an Expert Group on Learning From Adverse Events in the NHS. London: The Stationary Office, 2000. 5 Currie M, Mackay P, Morgan C et al. The ‘wrong drug’ problem in anaesthesia – an analysis of 2000 incident reports. Anaesthesia and Intensive Care 1993; 21: 596–601.


Æ


6 Webster CS. Why anaesthetising a patient is more prone to failure than flying a plane. Anaesthesia 2002; 57: 819–20. 7 Webster CS, Merry AF, Larsson L, McGrath KA, Weller J. The frequency and nature of drug administration error during anaesthesia. Anaesthesia and Intensive Care 2001; 29: 494–500. 8 Bergman IJ, Kluger MT, Short TG. Awareness during general anaesthesia – a review of 81 cases from the Anaesthetic Incident Monitoring Study. Anaesthesia 2002; 57: 549–56. 9 Kluger MT, Bullock MFM. Recovery room incidents – a review of 419 reports from the Anaesthetic Incident Monitoring Study. Anaesthesia 2002; 57: 1060–6. 10 Reason J. Human error – models and management. British Medical Journal 2000; 320: 768–70. 11 Perrow C. Normal Accidents – Living with High Risk Technologies. New York: Basic Books, 1984. 12 Sagan SD. The Limits of Safety – Organizations, Accidents, and Nuclear Weapons. Princeton, NJ: Princeton University Press, 1993. 13 Reason J. Managing the Risks of Organisational Accidents. Aldershot: Ashgate, 1997. 14 Berwick DM. Not again! Preventing errors lies in redesign – not exhortation. British Medical Journal 2001; 322: 247–8. 15 Birks RJS. Safety matters [editorial]. Anaesthesia 2001; 56: 823–4. 16 Carthey J, de Leval MR, Reason JT. Institutional resilience in healthcare systems. Quality in Health Care 2001; 10: 29–32. 17 Reason JT, Carthey J, de Leval MR. Diagnosing ‘vulnerable system syndrome’ – an essential prerequisite to effective risk management. Quality in Health Care 2001; 10: ii21–5. 18 Merry AF, Webster CS, Mathew DJ. A new, safetyoriented, integrated drug administration and automated anesthesia record system. Anesthesia and Analgesia 2001; 93: 385–90. 19 Takano K, Reason J. Psychological biases affecting human cognitive performance in dynamic operational environments. Journal of Nuclear Science and Technology 1999; 36: 1041–51. 20 Chiles JR. Inviting Disaster – Lessons from the Edge of Technology. New York: Harper Collins, 2001. 21 Merry AF, Webster CS, Weller J, Henderson S, Robinson B. Evaluation in an anaesthetic simulator of a prototype of a new drug administration system designed to reduce error. Anaesthesia 2002; 57: 256–63. 22 Webster CS, Merry AF, Ducat CM. Safety, cost and predrawn emergency drugs. Anaesthesia 2001; 56: 818–20. 23 Webster CS, Mathew DJ, Merry AF. Effective labelling is difficult, but safety really does matter. Anaesthesia 2002; 57: 201–2.


24 Anonymous. Standard Specification for User Applied Drug Labels in Anesthesiology (D4774–94). Philadelphia: American Society for Testing and Materials, 1995. 25 Anonymous. Standard for User-Applied Drug Labels in Anaesthesia and Critical Care (Z264.3–98). Etobicoke: Canadian Standards Association, 1998. 26 Anonymous. User-Applied Labels For Use on Syringes Containing Drugs Used During Anaesthesia (AS ⁄ NZS 4375). Wellington: Standards New Zealand, 1996. 27 Helmreich RL, Merritt AC. Culture at Work in Aviation and Medicine. Aldershot: Ashgate, 2001. 28 Galletly DC, Rowe WL, Henderson RS. The anaesthetic record – a confidential survey on data omission or modification. Anaesthesia and Intensive Care 1991; 19: 74–8. 29 Lin YC, Miller SR. The impact of price labeling of muscle relaxants on cost consciousness among anesthesiologists. Journal of Clinical Anesthesia 1998; 10: 401–3. 30 Niskanen MM, Ruokonen ET. Association between intraoperative incidents and postoperative outcome and resource utilisation. Anaesthesia 2002; 57: 1052–9. 31 Fraind DB, Slagle JM, Tubbesing VA, Hughes SA, Weinger MB. Reengineering intravenous drug and fluid administration processes in the operating room – step one: task analysis of existing processes. Anesthesiology 2002; 97: 139–47. 32 Ducat CM, Merry AF, Webster CS. Attitudes and practices of New Zealand anaesthetists with regard to emergency drugs. Anaesthesia and Intensive Care 2000; 28: 692–7. 33 Stranix ND, Patel A. More painful spikes. Anaesthesia 2000; 55: 412. 34 Palmer J, Ball D. Opening propofol ampoules. Anaesthesia 1998; 53: 416–7. 35 Ali PB. Persistent problem with propofol ampoules. Anaesthesia 1997; 52: 1020. 36 Josefson D. Medical students at risk from needlestick injury. British Medical Journal 1999; 318: 144. 37 Singleton RJ, Ludbrook GL, Webb RK, Fox MAL. Physical injuries and environmental safety in anaesthesia – an analysis of 2000 incident reports. Anaesthesia and Intensive Care 1993; 21: 659–63. 38 Lye ST, Hwang NC. Glass particle contamination – is it here to stay? Anaesthesia 2003; 58: 93–4. 39 Hutchins E. Cognition in the Wild. Cambridge, MA: Institute of Technology Press, 1999. 40 Naikar N, Sanderson PM. Evaluating design proposals for complex systems with work domain analysis. Human Factors 2001; 43: 529–42. 41 Anonymous. Risk Management (AS ⁄ NZS 4360). Wellington: Standards New Zealand, 1999.

87