Time for a Paradigm Shift
In 1999 the Institute of Medicine called for a halving of error in health care within 5 years. Numerous other authoritative calls for improved safety have been made since, including legislative moves by Congress and the Food and Drug Administration. Despite this, the vast majority of drugs used in health care continue to be administered by traditional error-prone means, and drug error remains a hazard to patients everywhere. The problem is of particular concern in anesthesia, where large numbers of potent drugs are given, often in rapid sequence. Historically, system redesign in anesthesia has been successful in eliminating error, for example in the elimination of problems with the delivery of oxygen to patients. However, we believe that much of the low hanging fruit of the benefit of simple engineering solutions has now been plucked. Thus, rather than an entirely blameless culture of safety focused solely on systems, we propose a “just culture,” where accidents can be identified as blameless errors, or culpable violations. We all make errors, even when doing our best to avoid them—they are unintentional, blame is usually unhelpful, and they are the appropriate target of system redesign. We believe what is now required to further reduce error in drug administration is a more sophisticated approach, involving a better understanding of the nature of human error itself, and better compliance in the adoption of safety procedures and systems.
Congress has recently passed legislation that includes steps to prevent hospital medication errors. In part this was in response to the tragic deaths of 3 premature infants in Indianapolis in September 2006, after they were accidentally administered adult doses of heparin. The Institute of Medicine (IOM) has estimated that each year medication errors injure at least 1.5 million Americans and cost the health system more than $3.5 billion.1 Drug errors feature prominently in every large-scale study of iatrogenic injury conducted.2 In 1999 the IOM called for a halving of errors (including medication errors) in health care over the next 5 years. In 2004 the Food and Drug Administration (FDA) mandated the use of barcodes at unit-dose level on all medications;3 these are being phased in over 5 years. However, little else seems to have changed, and the IOM’s goal of a 50% error reduction certainly has not been achieved in relation to medication administration. In 2007, the vast majority of drugs used in health care are still administered by traditional means, and drug error remains a significant hazard to the health of patients everywhere.
The Extent of the Problem
There is no aspect of anesthesia that occupies a more important place in the safe management of our patients than the accurate administration of medications. It is therefore surprising how little has been published dealing with reducing medication error in anesthesia. A recent systematic review of the literature from 1978 to 2002 identified only 98 references on this subject, and only 1 involved a randomized trial (conducted with a human-patient simulator), only 2 could be considered experimental or quasi-experimental, and only 11 contained observational data.4 The landmark 1978 paper by Cooper et al., the starting point for the analysis in the above study, identified 359 incidents.5 The first, second, and fourth most frequent incident categories were breathing circuit disconnection, inadvertent gas flow change, and gas supply problems. The third most frequent was syringe swap. In 1984 a further critical-incident analysis published by Cooper’s team, showed a similar pattern of problems.6 The most frequently cited critical incident category was breathing circuit disconnection. The next 8 categories included both syringe swap and ampule swap. Drug overdose (via syringe and vaporizer) was also listed. Cooper’s group concluded that human error was the dominant issue in anesthesia safety and encouraged the specialty to direct patient safety efforts toward monitoring instrumentation and improvement in equipment using human-factors techniques. Today history has vindicated this vision. Engineering innovations have virtually eliminated problems with the delivery of oxygen to patients. A recent review of 4,000 incidents and over 1,200 medico-legal notifications reported by anesthetists in Australia revealed no cases of hypoxic brain damage or death from inadequate ventilation or misplaced tubes since the introduction of oximetry and capnography.7 However, no such systematic innovations have yet been widely adopted to reduce medication error.
We don’t know what the rate of medication error was in 1978, but recent data have shown that the magnitude of the problem today is more serious than previously thought. Using facilitated incident monitoring (which provides a denominator) and prospectively collecting data from over 10,000 anesthetics in New Zealand, approximately 1 error was shown to occur for every 130 anesthetics.8 A very similar rate was found in Seattle, using the same study method.9 Other studies from various countries and types of institution suggest that these estimates are of the correct order of magnitude10-13 (Table 1, page 47) and reflect the situation in anesthesia as it is widely practiced today, rather than any local aberrations in standard of care.
Orser’s group took a different approach. They sent an anonymous survey to all 2,266 members of the Canadian Anesthesiologists’ Society in 1995.14 Thirty percent of the members responded to the survey and 1,038 drug-related events were examined in detail. Most anesthesiologists had experienced >1 drug error. Syringe swap was the most common category of error. Fifteen of the errors (1.4%) resulted in major morbidity (including 4 deaths). In a similar survey in New Zealand, 89% of respondents admitted having made at least 1 drug error.15 The Canadian study provides valuable insights into the root causes of drug error. For example, although 86% of respondents were aware of the Canadian Standards Association labeling standards, and 86.9% agreed or strongly agreed that these labels reduced the incidence of drug errors, only 72% actually used them. Furthermore, fewer than half the respondents “always” read the label. These findings are not edifying for a specialty group with a claim to being leaders in safety, and there is no reason to believe that the practices described, and the attitudes that drive them, are confined to Canada. One of us administered a questionnaire to 210 delegates at a New Zealand anesthesiology conference, asking 12 questions concerning perceptions about the drug error problem.16 Respondents answered questions in relation to their own practice and anesthesia practice in general (hence their colleagues). The majority of anesthesiologists felt that drug error in anesthesia was a significant problem, and one the public was becoming increasingly intolerant of; however, few were concerned over the chance of harming an individual patient in this way themselves, and most felt that error was more of a problem with other anesthesiologists’ practices than with their own. Similarly, in Australia, anesthesiologists estimated the risk of awareness in their personal practices as half as likely than in that of their colleagues.17 These are classic examples of optimist bias, a common psychological phenomenon in which individuals, on average, view their own abilities as better than average (a statistical impossibility).
Lessons from Intensive Care
Table 1. Prospective estimates of rates of drug administration error in anesthesia (1978-present)
|Study||Sample||No. of Anesthetics||No. of Drug Errors||Drug Error Rate|
|Craig & Wilson10||6 Months||8312||12||0.14%|
|Kumar et al.11||April 1984-January 1985 and April 1985-January 1986||28965||31||0.11%|
|Short et al.12||1990||16739||26||0.16%|
|Fasting & Gisvold13||September 1996-October 1999||55426||63||0.11%|
|*Webster et al.8||February 1998-October 1999||10806||81||0.75%|
|*Bowdle et al.9||21 weeks||6709||41||0.61%|
Disguised-observer studies in the ICU literature offer lessons for the anesthesia professional in the operating room. The disguised-observer technique is known to accurately identify rates of error in hospital environments, and there are many similarities between the ways drugs are given in the ICU and in the operating room during anesthesia.
At 2 Dutch hospitals van dem Bemt et al. used the disguised-observer technique.18 The researchers observed 233 drug administrations to 24 patients over the 5-day study period. The error rate was 44.6% (104/233) when wrong-time errors were included and 33% (77/233) when wrong-time errors were excluded. A wrong-time error was defined as the administration of a drug >60 minutes earlier or later than prescribed.
If these data are even partially indicative of the problem in anesthesia, then it is considerably worse than that suggested by the studies summarized in Table 1. This possibility is reinforced by recent (as yet unpublished) work in New Zealand using direct observation in human-patient simulation involving complex anesthetic cases.
Why Medication Error in Anesthesia Continues to Occur
It is not difficult to inject 1 drug safely, but the challenge the anesthesia professional faces is to participate in the administration of perhaps half a million drugs during a professional lifetime. Doing this 100% accurately is very difficult. Many of our patients have diminished physiologic reserve to tolerate drug error. As they are sedated or anesthetized they cannot correct or detect drug errors themselves. They depend on us to do this, and this is a responsibility we should not take lightly.
Errors, Outcome, and Blame
The outcome of an error is largely determined by chance. You back out of your driveway and run over an unseen squirrel that dashes under your car. Or, you back out of your driveway and run over an unseen child that does the same thing. The error mechanism is identical in each case, but both you and society will judge yourself differently. The same can be said of drug errors. There is no moral difference between a drug error that causes no harm and one that results in death.
Recently one of us was asked to debrief a trainee who inadvertently administered 200 mg of dopamine as a bolus, using an unlabeled syringe (this was a look-alike problem, set up by a recent change in the formulations of 2 drugs). With help from his supervisor, he was able to respond to the sudden catastrophic rise in blood pressure, the patient’s life was saved, and in the end no harm ensued. In 1990, another anesthesiologist gave the identical drug in error (having also been set up, this time by having the ampule of dopamine placed in the compartment in the drug drawer labeled “Dopram”).19 The patient lost her life and the anesthesiologist was convicted of manslaughter.
The importance of an adverse event should be judged by its potential outcome rather than its actual outcome. The enormity of the potential outcome from a drug error does not justify recourse to the criminal law, but surely it does justify taking the problem seriously, reporting the incident, and (as a minimum) labeling one’s syringes and reading one’s ampules. This concept is encapsulated in a World Health Organization motto, which states, “To err is human; to cover up is unforgivable; to fail to learn is inexcusable.”
It is impossible to address drug error effectively without addressing the organizational culture of anesthesia. In Human Error James Reason advocated a blame free culture as necessary for effectively reducing error. In the end, few people are really comfortable with the notion that blame should be set aside completely. Today most authorities (including Reason) would probably advocate a “Just Culture.”20 This implies early triage of incidents into those in which blame may be appropriate, and those in which it is not. By definition, errors fall into the latter category. In fact, if the aim is to promote patient safety, the former category should be reserved for clearly egregious behaviors, such as leaving an anesthetized patient unattended, or working under the influence of alcohol or drugs.
Aviation, for most anesthesiologists and nurse anesthetists, is the obvious model for safety. There are lessons to be learned from aviation, as there are from high-reliability organizations in other fields such as the nuclear power industry.21 However, the metaphor of the anesthesia professional as a pilot, and the notion that “take-offs” and “landings” are like induction and emergence is limited. For a start, the system formed by the patient, the anesthesiologist and/or the CRNA, and the surrounding environment of the operating room (including personnel and equipment) is more complex than that which characterizes commercial aviation.2 It does seem that aviation has embraced a safety culture for decades, whereas some anesthesia providers seem to harbor an attitudinal barrier to safety.16,22 In a safety culture, accidents are interpreted as evidence of faulty system design. Both accidents and incidents are viewed as opportunities to redesign the work environment and improve safety. Such cultures, therefore, embrace a healthy incident reporting system. Individual errors may not be foreseeable, but the contributing factors can be anticipated and addressed.23 By contrast, a person-centered approach to error involves blaming individuals for their carelessness, forgetfulness, or other character weakness when things go wrong. Such an approach has been called the culture of denial and effort: denial, because it denies the psychological reality that error is a statistically inevitable consequence of human action; and effort, because it implies that with sheer effort alone all error can be avoided.24 It directs attention away from faulty work systems, leaving them untouched and able to predispose to further errors and failures in the future. The culture of denial and effort is the antithesis of the culture of safety and is clearly unhelpful and unsound. Despite this, the person-centered approach persists in health care (including anesthesia), and often hinders the adoption of safety systems and procedures.
In the end, perhaps the biggest single difference between anesthesia and aviation relates to the perception that expenditure on safety is justified. The numbers involved in a single airline accident grab public attention and demand a response. Individual anesthesiologists or CRNAs harm harm patients 1 at a time. Collectively and over time the harm mounts up, but because it is sporadic it is largely invisible. Imagine the public’s response to 5,000 plus cases of intraoperative awareness if they all occurred in 1 hospital in the first 2 weeks of January, instead of being spread out over the calendar year and the entire country.25
In anesthesia, and health care generally, the predominant cultural focus is on productivity. The current common demand on the part of hospital administrators for a “business case” or a “return on investment” (ROI) to justify expenditure on safety is misguided if it doesn’t factor in the wider picture which includes the very real cost of iatrogenic harm.24 It is reminiscent of the saga of the Ford Pinto.26 This car was designed in such a way that the fuel tank would rupture and explode in certain rear-end collisions, burning or killing its occupants. Ford knew about these risks. However, the business case was taken that it would be cheaper for Ford to continue to sell Pintos, let its customers burn, and to pay out the lawsuits on these somewhat infrequent cases, than to recall all Pintos and fix the problem. In the end it turned out cheaper than expected to fix the problem, so even the business case seems to have been flawed. In addition, the public outcry that followed the exposure of Ford’s commercial cynicism did enormous damage to the company’s reputation and sales. Ford earned the dubious distinction of being the first corporation to be charged with the criminal offence of reckless homicide. A similar situation often occurs in health care. Safety should be funded because it is the right thing to do, not because of any ROI directives. However, in health care, doing the right thing, first time to the right patient, usually turns out to be the best from a business perspective as well. Harming patients during their treatment, and then having to treat them for such harm, is extraordinarily inefficient and expensive. The savings from even one avoided case of significant iatrogenic harm would pay for a great deal of safety. Furthermore, the cost to the health care organization in terms of lost reputation can be many times larger than the cost of treating the harmed patient. Iatrogenic harm simply doesn’t pay.
How Should We Administer Drugs?
Table 2. Strong safety recommendations based on a systematic review of the entire literature on drug administration error in anesthesia and as validated against actual incident reports
|1.||The label on any drug ampule or syringe should be carefully read before a drug is drawn up or injected.|
|2.||Legibility and contents of labels on ampules and syringes should be optimized according to agreed standards in respect to some or all attributes of font, size, color, and the information included.|
|3.||Syringes should be labeled (always or almost always).|
|4.||Formal organization of the drug drawers and workspace should be used with attention to tidiness, position of ampules and syringes, separation of similar or dangerous drugs, removal of dangerous drugs from the operating rooms.|
|5.||Labels should be checked specifically with a second person or a device (such as a barcode reader linked to a computer) before a drug is drawn up or administered.|
Administering drugs is fundamental to anesthesia, and its importance should be elevated in programs of continuing professional development, in priorities for research, and in self-directed reading. For example, in a recent survey of anesthesiologists, only 19% reported having received specific training on how to administer drugs safely.16
A systematic literature review has brought together received wisdom on how best to reduce the risk of drug administration error during anesthesia.4 Five strong recommendations survived testing against actual incident reports (Table 2). These would seem to be a good starting point for action.
Furthermore, involvement of pharmacists in the operating room has been recognized as a core principle for improving drug safety in anesthesia.27 The preparation and labeling of drugs in a central pharmacy should decrease the incidence of error. In the pharmacy 2 people check each other’s work, multiple syringes are prepared for 1 drug at a time, and the environment is one in which distractions are few, order reigns, and time is available to check and recheck (and record the checking on a form). In addition, dispensing accuracy generally improves administration accuracy.
Table 3. Possible additional measures to promote safer drug administration in anesthesia
- The provision of all labels in a standardized format emphasizing the class and generic names of each drug, incorporating a bar-code and class-specific color-code consistent with international drug labeling standards.27,30
- The presentation of selected, commonly used drugs in pre-filled syringes prepared under quality assured conditions and pre-labeled as above.27
- The use of a bar-code reader to scan the drug at the point of administration immediately before it is given, linked to an auditory prompt (i.e., the computer speaks the name of the scanned drug) and a visual prompt (i.e., the computer displays the name of the drug, in prominent color-coded format) to facilitate checking of the drug’s identity.27
- Integration of scanned information into an automated anesthetic record, facilitating accuracy of the drug information in the record and reducing the cognitive load on the anesthesia professional.27
- The use of devices at the point of care to automatically measure the dose of drug administered.29
- The use of purpose designed drug trays to facilitate the layout of syringes and ampules and organization of the anesthesiologists’ or nurse anesthetists’ workspace.27
- Infusion syringe labels consistent with the standardized labels described above, which incorporate a dosing nomograph into the label itself, thus removing the need for look-up tables or dose calculations and reducing the cognitive load on the anesthesia provider.31
- The use of automated medication dispensing systems with features such as single-issue drawers and barcode scanners to facilitate safer dispensing of drugs in the operating room.
One thing is clear—we will not make progress while we continue to embrace idiosyncratic approaches to this problem. Health care organizations must establish sound techniques for drug administration, teach them to their residents, and provide role models of their use. There is an increasing range of solutions available for the problem of drug administration error in anesthesia (Table 3).27-31 Incident data, prospectively collected over a period of 5 years, have shown that the use of a system incorporating a number of these safety principles has been associated with a significantly lower rate of drug error per bolus administered.32
The Need for a New Paradigm
A few years ago a senior colleague made a drug error while anesthetizing a human-patient simulator. He admitted it and said, “I must try harder in the future.” Here is the heart of the problem. This is the person-centered view that impedes progress and is doomed to fail. The truth is that he was trying as hard as he could—he was under the direct observation of several of his peers, and was very motivated to perform as well as possible. That is the defining point about errors; we make them, unintentionally, even when we are trying not to. Trying harder will not substantially reduce error, but re-designing systems to make them inherently less error-prone will.
Berwick has popularized the quote, attributed to Einstein, that, “Madness is doing the same thing over and over again and expecting a different result.” We will not reduce drug error until we change the way we give drugs. This will include embracing technological solutions of one sort or another. However, it will also mean complying with these solutions. It is unlikely that forcing functions will ever make drug error in anesthesia impossible. It is certain, however, that redesigning the system can make errors much less probable—provided anesthesiologists and nurse anesthetists actually make the effort to take proper advantage of the innovations.
Dr. Stabile is an Adjunct Clinical Professor at Vanderbilt University School of Medicine, Nashville, Tennessee, and St. Louis University School of Medicine, St. Louis, Missouri.
Dr. Webster is a Research Fellow in the Department of Anaesthesiology at the School of Medicine at the University of Auckland, Auckland, New Zealand.
Dr. Merry is Professor and Head of the Department of Anaesthesiology at the School of Medicine at the University of Auckland, Auckland, New Zealand.
DISCLOSURE: All three authors own shares in Safer Sleep, LLC., a manufacturer of an automated anesthesia record system that includes a barcode-based drug administration system. This company is also a contributor to the APSF. Dr. Webster has received research grant support from this company, Dr. Merry is a Director of this company, and Dr. Stabile is Chief Medical Officer and Chairman of the Medical Advisory Board of Safer Sleep, LLC.
- Institute of Medicine. To err is human—building a safer health system. Washington DC: National Academy Press; 2000.
- Webster CS. Implementation and assessment of a new integrated drug administration system (IDAS) as an example of a safety intervention in a complex socio-technological workplace. University of Auckland: Auckland; 2004.
- Bar code label requirement for human drug products and biological products. Fed Regist 2004; 69: 9119-9171.
- Jensen LS, Merry AF, Webster CS, Weller J, Larsson L. Evidence-based strategies for preventing drug administration error during anaesthesia. Anaesthesia 2004; 59: 493-504.
- Cooper JB, Newbower RS, Long CD, McPeek B. Preventable anesthesia mishaps—a study of human factors. Anesthesiology 1978; 49: 399-406.
- Cooper JB, Newbower RS, Kitz RJ. An analysis of major errors and equipment failures in anesthesia management—considerations for prevention and detection. Anesthesiology 1984; 60: 34-42.
- Runciman WB. Iatrogenic harm and anaesthesia in Australia. Anaesth Intensive Care 2005; 33: 297-300.
- Webster CS, Merry AF, Larsson L, McGrath KA, Weller J. The frequency and nature of drug administration error during anaesthesia. Anaesth Intensive Care 2001; 29: 494-500.
- Bowdle A, Kruger C, Grieve R, Emmens D, Merry A. Anesthesia drug administration error in a university hospital. Anesthesiology 2003; 99: A1358.
- Craig J, Wilson ME. A survey of anaesthetic misadventures. Anaesthesia 1981; 36: 933-936.
- Kumar V, Barcellos WA, Mehta MP, Carter JG. An analysis of critical incidents in a teaching department for quality assurance—a survey of mishaps during anaesthesia. Anaesthesia 1988; 43: 879-883.
- Short TG, O’Regan A, Lew J, Oh TE. Critical incident reporting in an anaesthetic department quality assurance programme. Anaesthesia 1993; 48: 3-7.
- Fasting S, Gisvold SE. Adverse drug errors in anesthesia, and the impact of coloured syringe labels. Can J Anesth 2000; 47: 1060-1067.
- Orser BA, Chen RJB, Yee DA. Medication errors in anesthetic practice: a survey of 687 practitioners. Can J Anaesth 2001; 48: 139-146.
- Merry AF, Peck DJ. Anaesthetists, errors in drug administration and the law. N Z Med J 1995; 108: 185-187.
- Webster CS, Grieve DJ. Attitudes to error and patient safety. Prometheus 2005; 23: 253-263.
- Myles PS, Symons JA, Leslie K. Anaesthetists’ attitudes towards awareness and depth-of-anaesthesia monitoring. Anaesthesia 2003; 58: 11-16.
- van den Bemt PM, Fijn R, van der Voort PH, Gossen AA, Egberts TC, Brouwers JR. Frequency and determinants of drug administration errors in the intensive care unit. Crit Care Med 2002; 30: 846-850.
- Skegg PDG. Criminal prosecutions of negligent health professionals—the New Zealand experience. Med Law Rev 1998; 6: 220-246.
- Merry AF, McCall Smith A. Errors, medicine and the law. Cambridge: Cambridge University Press: 2001.
- Webster CS. The nuclear power industry as an alternative analogy for safety in anaesthesia and a novel approach for the conceptualisation of safety goals. Anaesthesia 2005; 60: 1115-1122.
- Gaba DM, Singer SJ, Sinaiko AD, Bowen JD, Ciavarelli AP. Differences in safety climate between hospital personnel and naval aviators. Hum Factors 2003; 45: 173-185.
- Reason J. Managing the risks of organisational accidents. Aldershot: Ashgate; 1997.
- Webster CS. The iatrogenic-harm cost equation and new technology. Anaesthesia 2005; 60: 843-846.
- Preventing and managing the impact of anesthesia awareness. Issue 2, October 6, 2004, Sentinel Event Alert, The Joint Commission. Available at http://www.jointcommission.org. Accessed August 31,2007.
- Flammang JM. Ford Pinto rear-impact defect. In Schlager N, editor. When technology fails—significant technological disasters, accidents and failures of the twentieth century. Gale Research Inc.: Detroit; 1994. p. 156-162.
- Merry AF, Webster CS, Mathew DJ. A new, safety-oriented, integrated drug administration and automated anesthesia record system. Anesth Analg 2001; 93: 385-390.
- Safer Sleep LLC. http://www.safersleep.com. Accessed August 31, 2007.
- DocuSys: Digital Medical Solutions. http://www.docusys.net. Accessed August 31, 2007.
- Standard specification for user applied drug labels in anesthesiology (D4774-94). American Society for Testing and Materials: Philadelphia; 1995.
- Merry AF, Webster CS, Connell H. A new infusion syringe label system designed to reduce task complexity during drug preparation. Anaesthesia 2006; 62: 486-491.
- Merry AF, Webster CS, Larsson L, Weller J, Frampton CM. Prospective assessment of a new anesthestic drug administration system designed to improve safety. Anesthesiology 2006; 105: A138.