Full text of DOI:10.1111/j.1553-2712.2000.tb00462.x

ACADEMIC EMERGENCY MEDICINE • November 2000, Volume 7, Number 11
1183
Defining, Identifying, and Measuring Error in
Emergency Medicine
J
ONATHAN A. HANDLER, MD, MICHAEL
G
ILLAM, MD,
ARTHUR B. SANDERS, MD, RICHARD
KLASCO, MD
Abstract. The findings of a consensus committee reporting, and surveillance systems are addressed, as
created to address the definition, measurement, and is error reporting at the clinician, hospital, and over-
identification of error in emergency medicine (EM) sight group levels. Committee recommendations are
are presented. The literature of error measurement made regarding the initial steps EM should take to
in medicine is also reviewed and analyzed. The con- address error. The establishment of patient safety
sensus committee recommended adopting a standard boards at each institution is also recommended. Key
set of terms found in the medical error literature. Is- words: error; public health; emergency medicine;
sues surrounding error identification are discussed. safety. ACADEMIC EMERGENCY MEDICINE 2000;
The pros and cons of mandatory reporting, voluntary 7:1183–1188
1
7
,000 deaths per year. In a meta-analysis of pro-
D
EFINING AND MEASURING ERROR
spective studies of adverse drug reactions in hos-
pitalized patients, it was found that the overall in-
cidence of serious adverse drug events (ADEs) was
The term ‘‘medical error’’ has been the focus of
much discussion recently, stimulated by the Insti-
tute of Medicine (IOM) report ‘‘To Err Is Human.’’
1
6
.7% and the incidence of fatal ADEs was 0.32% of
5
hospitalized patients. Extrapolation of these data
led to estimates that in 1994 more than 2.2 million
hospitalized patients had serious ADEs and 6,000
had fatal ADEs, making ADEs between the fourth
and sixth leading causes of death in the United
The report brought to the attention of the public
that adverse events in medicine are common and
are one of the leading causes of morbidity and mor-
tality within the United Sates. The report esti-
mates that 44,000–98,000 patients hospitalized in
the United States each year die as a result of med-
ical errors, making this one of the leading causes
of deaths. Although the methodology of the IOM
5
States. Serious ADEs were those that necessitated
hospitalization, were permanently disabling, or re-
sulted in death. The U.S. Pharmacopeial Conven-
tion (USP) has developed a systematic taxonomy
2
,3
report has stimulated controversy, a recent re-
port from the United Kingdom’s National Health
Service confirms the magnitude and global impact
6
of medication errors. Such standardized definition
sets are needed in order to conduct meaningful re-
search and to quantify progress.
4
of medical errors.
It is clear from these data that medical errors
are a major cause of morbidity and mortality in the
United States. There are few data in the literature
about medical errors in emergency medicine (EM).
However, studies in the literature imply that EM
Between 3% and 4% of patients admitted to the
hospital have adverse events resulting in injury or
disability. About 30% of these adverse events are
thought to be preventable and represent subopti-
mal care. The total national cost for medical errors
are estimated to be $37–50 billion, with prevent-
able adverse events accounting for $17–29 billion.
Medication errors have been more extensively
studied and are estimated to account for about
7
is a key area for preventable medical errors.
1
In a recent major study on adverse events in
hospitalized patients, researchers selected a sam-
ple of hospitals in Utah and Colorado and ran-
8,9
domly sampled 15,000 discharges in 1992.
Trained nurse reviewers retrospectively reviewed
hospital records for any one of 18 indicators of pos-
sible adverse events. Criteria included subsequent
or previous hospital admission, hospital-incurred
From the Division of Emergency Medicine, Northwestern Uni-
versity Medical School, Chicago, IL (JAH); Division of Emer-
gency Medicine, Evanston Northwestern Healthcare, Evans-
ton, IL (MG); Department of Emergency Medicine, University
of Arizona, Tucson, AZ (ABS); and Micromedex, Inc., Engle- trauma, ADEs, transfer to a special care unit or
wood, CO (RK).
Received July 6, 2000; accepted July 19, 2000.
Address for correspondence and reprints: Jonathan A. Handler,
MD, Division of Emergency Medicine, Northwestern Univer-
another hospital, return to the operating room,
complication, myocardial infarction, stroke, pul-
monary embolus, neurologic deficit, fever, cardiac
arrest, death, unplanned procedure, and litigation.
Charts meeting one or more of these criteria were
sity Medical School, 216 East Superior, Suite 100, Chicago, IL
0611. Fax: 312-908-6004; e-mail: jah505@nwu.edu
6

1184
PROCEEDINGS
Handler et al. • IDENTIFYING ERROR IN EM
then reviewed by a trained internal medicine or questions that cry out for further research of errors
family practice physician who made a decision in EM.
about whether an adverse event occurred and
whether this event was due to negligence. Negli- Emergency medicine should adopt the defini-
gence was defined as medical care that fell below tions that are consistent with the Institute of
the standard expected of physicians in the com- Medicine report, the USP taxonomy, and the
munity. Judgments about adverse events were major studies in the medical literature. The def-
made on a six-point confidence scale where 3 rep- initions are listed in Table 1. While these terms
resented just under a 50% likelihood of error and are useful for categorization, many of the terms
4
represented just over a 50% likelihood of error. are charged with meanings that may interfere
Confidence scores of 4 or higher were then col- with our goal of improving emergency care for pa-
lapsed to make a decision about negligence. The tients. The panel proposes that EM focus on pre-
authors found that adverse events occurred in ventable adverse events and potential adverse
2
.9% of hospitalizations. Negligence was the cause events. A preventable adverse event is when an in-
of the adverse event in 33% in Utah and 27% in jury occurs as a result of substandard medical
Colorado. Death occurred in 6.6% of patients with care; the judgment is made that standard care
adverse events and 8.8% of patients with negligent would have reasonably prevented the injury. A po-
adverse events. Operative adverse events com- tential adverse event involves substandard care
prised 45% of all events, with 17% being due to that could have resulted in injury. Potential ad-
negligence. Adverse drug events comprised 19% of verse events are ‘‘near-misses.’’ By accumulating a
all adverse events, with 35% being due to negli- database of these near-misses, we can better un-
8
,9
gence. Adverse events in the emergency depart- derstand how preventable adverse events occur
ment (ED) were analyzed as a subgroup. The au- and build systems to prevent injury.
thors note ‘‘Although only 1.7% of adverse events
The identification of potential and preventable
were attributed to nursing staff and emergency adverse events can occur only in an environment
physicians, a notable 74.2% and 94.8%, respec- in which patient safety is a high organizational
tively, were negligent, mostly caused by failed di- priority, individuals are not blamed for mistakes,
agnoses. . . . Among locations, the emergency de- and there is a nonpunitive system for reporting
partment had the highest percentage of negligent problems in medical care to peer-review protected
adverse events (52.6%), and a remarkable 94.8% of committees that are empowered to institute
adverse events attributed to emergency physicians changes in the system.
9
were judged negligent.’’
The second major study used a similar meth- Emergency medicine should focus on identify-
odology. The Harvard Medical Practice Study ing preventable and potential adverse events.
investigated adverse events and negligence in
hospitalized patients in New York State. The re-
searchers reviewed more than 30,000 randomly se-
lected records from 51 acute care hospitals.
Nurse reviewers screened the charts for 18 defined many industries,
ERROR IDENTIFICATION
1
0–12
Error identification, a well-defined process in
13–16 13
is still in its infancy in EM.
criteria. Charts meeting the criteria were then re- The state of the art in error identification systems
viewed by two physician reviewers to determine is derived from the relatively small body of knowl-
whether an adverse event occurred and was due to edge in the medical field (especially the advance-
17
negligence. The researchers found that adverse ments made in anesthesiology ), as well as the
events occurred in 3.7% of hospitalized patients, body of knowledge and experience in systems with
with 27.6% due to negligence, and 13.6% of adverse effective error identification processes in place
13,15,16
events resulted in death. In this study 48% were (e.g., nuclear reactor systems
). Without error
associated with surgery and 19% with medications. identification an error may remain occult, so that
The ED accounted for only 2.9% of the total ad- rescue measures after the occurrence of an error
verse events. However, 70.4% of these ED adverse may not be undertaken and corrective system
1
0–12
events were judged to be negligent.
changes for greater patient safety are unlikely.
Both of these studies focused on the issue of This is especially true in the ED, the location of
malpractice litigation and its relationship to injury the highest proportion of negligent adverse events
caused by negligence. These studies, nevertheless, in the hospital.
have become very important in characterizing
9
–12
medical errors and iatrogenic injuries. Although Solutions for errors in medicine should focus on
the studies were flawed because emergency phy- changes to the system and processes rather
sicians were not the peer reviewers in determining than punitive targeting of individuals. A non-
the standard of care, the studies raise intriguing punitive system for identifying and reporting

ACADEMIC EMERGENCY MEDICINE • November 2000, Volume 7, Number 11
1185
TABLE 1. Recommended Definitions*
preventable and potential adverse events in EM
must be encouraged. The panel is firm in its be-
lief that lasting error reduction will be made (error of execution) or use of a wrong plan to achieve an aim
through system changes and the processes by (error of planning); the accumulation or errors results in acci-
Error—Failure of a planned action to be completed as intended
1
dents.
which medical decisions and care are made. Errors
made by an individual often reflect system-wide
Active error—An error that occurs at the level of the frontline
operator and whose effects are felt almost immediately.
1
,16,18–20
problems.
For example, the urine culture re-
1
sult on an ED patient is reported two days after
the patient’s discharge from the ED. No one in- Latent error—Errors in the design organization, training, or
forms the patient that she needs antibiotics, and
the patient returns a week later with urosepsis.
maintenance that lead to operator errors and whose effects
1
typically lie dormant in the system for lengthy periods of time.
This case would meet the definition of a prevent-
able adverse event as well as medical error. The was not what was intended; the wrong action is observable.
typical reaction is to blame the physician who or-
dered the culture for not following up on the re-
sults and to educate the physician to be more dil-
igent the next time. This approach is certain to fail
Slip errors—An error of execution when the action conducted
1
Lapse errors—An error of execution when the action conducted
was not what was intended; the wrong action is not observ-
1
able.
and result in similar errors in the future. Individ- Mistake—An error in which the action proceeds as planned but
ual punitive measures in this case will not solve fails to achieve its intended outcome because the planned ac-
1
tion was wrong; error of planning.
the system-wide problem of an inadequate follow-
up system for culture results in the ED.
Cognitive errors may contribute to a large pro-
portion of preventable and potential adverse
events. It is crucial, however, that these events be Patient safety—Freedom from accidental injury; ensuring pa-
recognized. A punitive approach will discourage re-
porting of potential adverse events. Further, it is
well known from other industries that error reduc-
Accident—An event that involves damage to a defined system
that disrupts the ongoing or future output of the system.
1
tient safety involves the establishment of operational systems
and processes that minimize the likelihood of errors and max-
1
imize the likelihood of intercepting them when they occur.
tion will not be achieved through punitive mea- Quality of care—Degree to which health services for individ-
sures but rather through changes in the medical uals and populations increase the likelihood of desired health
outcomes and are consistent with current professional knowl-
edge.
system and the medical decision-making process.
1
The panel recognizes the inherent complexity of
Adverse event—An injury resulting from a medical interven-
tion.
systems, and their recommendations should be
reevaluated as new information arises. Error
identification is not an easy process. If error iden-
tification were straightforward, it would not be the
subject of major investigation today. The difficulty
in error identification stems from the subjective
1,8–12
Preventable adverse event—An injury that occurs as a result of
medical error; with standard medical care the injury would not
1
have occurred.
Potential preventable adverse event (‘‘near-miss’’)—A medical
21
1,24
nature of medical care, combined with the com- error that could have resulted in injury.
1
6
plexity of the system. A defining characteristic of
a complex system is that change can have unpre-
System—A set of interdependent elements interacting to
4
1
achieve a common aim.
dictable effects. It is impossible to say with cer-
tainty that any action will have a positive impact, *For complete citations, see the reference list.
a negative impact, or no impact at all on the sys-
1
,13,18
tem. The panel recognizes that EM operates in a
complex systems environment, and consensus in
this document should be reevaluated as new data
arises.
cant injury.
These errors are discoverable by
audit (allowing enforcement), and they are the
types of errors for which punitive action can be
taken. Typically, there is at most limited confiden-
tiality to the reports, and limited or no immunity
from penalty. This type of reporting satisfies the
‘‘public’s right to know,’’ and it typically identifies
T
HE ‘‘WHO, WHEN, WHERE
, AND HOW’’
OF RROR DENTIFICATION
E
I
a series of many events that coincided to threaten
1
There are generally three methods of error iden- patient safety. The lack of confidentiality often
tification: mandatory reporting, voluntary report- brings attention to the individual rather than to
ing, and active surveillance systems.
the system deficiencies that led to the errors. The
In most industries where mandatory reporting resulting cycle of punishment may fail to root out
exists, a regulatory agency or governmental body system-wide problems that cause or contribute to
13,16,18
mandates reporting of errors resulting in signifi- error.

1186
PROCEEDINGS
Handler et al. • IDENTIFYING ERROR IN EM
Voluntary error reporting systems are usually group (often a governmental agency) can occur
used to identify ‘‘latent’’ errors (errors in system through mandatory or voluntary reporting sys-
1
,18
design that pose a risk to the patient).
An ex- tems, and active surveillance by audit and direct
ample is to name one medication ‘‘Norflex’’ and an- observation. Oversight groups can best recognize
other ‘‘Norflox,’’ which will predictably lead to con- health care system-wide problems, especially
fusion. Such latent errors are identified either by among widely disparate patients. They may have
a near-miss potential adverse event (error not the resources and power to implement and enforce1
leading to injury) or by recognition of the potential change throughout the entire health care system.
for danger before a mistake has actually occurred. Error identification by an oversight group is a
Identifying these errors may allow corrective sys- standard element of high-risk, complex, yet highly
1
,18
1,13
tems changes before an injury occurs. Hospitals, safe systems such as airline travel.
However,
safety ‘‘societies,’’ governmental bodies, and regu- medicine differs in that its errors are smaller-scale
latory agencies are all candidates to manage vol- (one person injured at a time), yet far more fre-
untary reporting systems. In most industries these quent. With the potential for nearly 100,000 inves-
systems are confidential (and sometimes anony- tigations per year, it is unclear that an oversight
mous), and they often confer immunity from pen- group will be able to handle health care’s volume
1
,13,18
alty.
Nonpunitive and confidential systems of errors. Confidentiality or anonymity, and im-
13,18
tend to have high compliance rates.
The avia- munity from penalty are significant concerns when
tion safety reporting system has received and reporting to a governmental organization. Over-
saved more than 500,000 voluntary reports of sight groups might find it easier to penalize indi-
13
near-misses.
Without individuals as targets, viduals than to enforce corrective system changes.
safety committees focus on systems changes that Even the fear of penalty may hinder compliance by
13,18
can prevent future error.
hospitals and clinicians.
The large bureaucracy
Active surveillance is the identification of error of an oversight agency may make error manage-
through observation. This can be direct observa- ment sluggish.
tion while providing care (such as an observer
Human and computerized active surveillance
watching clinicians ‘‘in action’’) or through chart systems hold great promise for error identification.
review, or observation of error ‘‘markers’’ (e.g., Errors can be flagged using ‘‘markers’’ such as the
tracking abnormal drug levels as a medication dos- use of reversal agents (e.g., naloxone) or other
9
,11,21–25
ing error marker).
Active surveillance may therapeutics (e.g., diphenhydramine, epinephrine),
identify error before injury actually occurs (e.g., abnormal laboratory data (e.g., hypoglycemia), a
software that will not allow a drug to be given to change in patient location or level of care (e.g.,
a patient who has a known allergy to it). Most ICU admission), an unplanned emergent proce-
active surveillance systems are highly objective, dure, an overlong length of stay for diagnosis, and
simply flagging all occurrences of an error diagnostic codes (e.g., allergic reaction diagnostic
2
2,24,25
1,22,25
marker.
code).
Making errors more visible when they
15
Error identification by clinicians allows those occur facilitates error identification. An example
with the greatest medical and situational knowl- is the color-coding of ophthalmologic medication
edge to provide insightful reports. Clinician re- caps. Pupillary dilation drops come in red-capped
porting may be biased by concerns of penalty or bottles, and pupillary constriction drops come in
1
3,18
embarrassment.
Clinicians may not have the green-capped bottles. If the pharmacy sends a
system-wide view necessary to recognize danger- green-capped bottle when cyclopentolate (a dilator)
ous trends among disparate patients. Clinicians is ordered, an error has probably occurred.
are unlikely to be able to implement change with-
out the assistance of a larger institution.
Rapid error identification facilitates corrective
action to prevent or limit injury. Therefore, it is
1
15
Hospitals can identify error through random imperative to create a system that can identify er-
chart sampling, quality assurance systems, report- ror at the earliest possible moment.
ing, or active surveillance systems. Hospitals may
In summary, error identification systems solely
best be able to implement active surveillance sys- involving clinicians may be easy to implement, but
tems. Hospitals may have a broad system view are unlikely to effect system change without the
that can identify some errors among disparate pa- involvement of the hospital or an oversight group.
tients. A hospital will not have the health care sys- Adding hospital resources and involvement to the
tem-wide view of a state or federal regulatory system adds some complexity, but the net effect
agency, nor the power to implement change outside will likely be positive. Members of the health care
1
the institution. Hospitals may be biased in re- system will probably resist the involvement of an
porting to protect themselves from penalty or neg- oversight agency due to concerns over bureaucratic
2
3
ative press.
sluggishness, immunity, and confidentiality. Vol-
Error identification involving an oversight untary reporting systems are typically confidential

ACADEMIC EMERGENCY MEDICINE • November 2000, Volume 7, Number 11
1187
or anonymous, confer immunity, and do not require anecdotal evidence suggests that review commit-
compliance audits. They will probably be easier to tees are more likely to decide that error has oc-
implement than mandatory systems. Voluntary re- curred if an outcome was poor than if the same
porting may identify error-prone systems before in- steps were taken but the outcome was unremark-
jury occurs. Active surveillance systems are likely able. The panel encourages objectifying the re-
to be highly effective, but they require significant views of cases by initially blinding reviewers to
resources and infrastructure to implement (such as outcome. Blinded to outcome, reviewers may focus
computer hardware and software).
on system correction rather than assignment of
blame.
Patient safety must be made a high organiza-
tional priority in each hospital. While the exact Systems to begin to reduce error should be im-
system for identifying and evaluating medical plemented today. Without action and only fur-
errors will vary with the needs of the institu- ther study, a year from now, no patient will have
tion, each institution should have an oversight benefited from the effort to improve patient safety
body that is peer-review protected and capable in EM. The panel recommends that each hospital
of carrying out the functions of a patient safety begin implementing systems to identify and cor-
board. The panel recommends that EDs work rect error within its ED. Each institution will need
within their hospitals to promote the IOM recom- to determine its own resources and focus until
mendation for the creation of a hospital safety pro- more is known about the scope of errors in EDs.
1
grams with defined executive responsibility. Each
hospital should create a committee or collection of Systems should be designed to catch error as
committees that can serve as a patient safety early as possible. The modern advance of tech-
board to champion patient safety for all patients, nology has made real-time data processing possi-
including those in the ED. This board would serve ble. Computerized systems for order entry, drug
the following roles:
dispensing, and electrocardiogram analysis are all
•
implement systems to monitor preventable and currently available, giving us the potential to rec-
potential adverse events
ognize and prevent error as it occurs. Early error
•
review reports of error in a peer-review-pro- recognition may allow corrective or rescue steps to
tected fashion
be taken before injury occurs.
•
implement system-wide changes to promote pa-
tient safety
The identification of potential and preventable
adverse events by clinicians and hospitals
The safety committee(s) should serve the role of consisting of active surveillance systems, vol-
monitoring the institution for preventable and untary, nonpunitive reporting, quality improve-
potential adverse events, reviewing reports of ment systems, and chart sampling is a reason-
error in a peer-review-protected venue and be able starting point for the institutional patient
empowered to make system-wide improvements safety board. Each hospital’s patient safety
to prevent future error. The panel emphasizes board will be responsible for creating error iden-
that the committee must be empowered to make the tification systems. Among many options, these may
necessary system changes. Due to the interdepart- include a secure internal web page that employees
mental nature of system-wide problems, it is ex- can use to anonymously report errors, an error hot-
pected that representatives from multiple depart- line for patients, a specialty review committee that
ments will serve on the committee(s).
monthly pulls charts for review, computerized sur-
The patient safety board may be wrapped into veillance systems, and the existing quality im-
existing hospital committees and structures. provement process. Importantly, a regular system
Smaller hospitals might join with larger hospitals of review is recommended to determine a baseline
to pool resources and perspectives.
Political factors may play a role in the ability of can help evaluate error-reducing interventions.
a committee to achieve results. The official recog- The panel recommends broad measures for er-
level of error. Following changes in this baseline
nition of a patient safety director who is respon- ror identification, but at this point it does not rec-
sible to the institution’s chief executive officer may ommend mandatory reporting systems. The ten-
facilitate cross-departmental cooperation for error dency of mandatory reporting to lead to punitive
reduction.
measures, the tendency to lack confidentiality or
anonymity, and the need for auditing to ensure
The panel encourages blinding reviewers to out- compliance make further work necessary before
come in error investigations to maintain a focus such systems can be recommended. The IOM cur-
on system improvement. It is human nature to rently suggests that the Forum for Healthcare
assign blame when suffering has occurred. Strong Quality Measurement and Reporting manage

1188
PROCEEDINGS
Handler et al. • IDENTIFYING ERROR IN EM
1
3. McDonald CJ, Weiner M, Hui SL. Deaths due to medical
mandatory reporting efforts. The panel currently
defers further recommendations on mandatory re-
errors are exaggerated in Institute of Medicine report. JAMA.
2
000; 284:93–5.
porting until more progress is made in that pro- 4. Department of Health. An organisation with a memory. Re-
cess. The panel sees benefit in creating a central-
ized agency for tackling the broader issue of error
in EM (e.g., a National Board for Quality Care in 2000.
Emergency Medicine), but is deferring recommen-
dations until future research and development can
guide that discussion.
port of an expert group on learning from adverse events in the
NHS, chaired by the Chief Medical Officer. London, United
Kingdom: Copyright Unit, Her Majesty’s Stationery Office,
5
. Lazarou J, Pomeranz BH, Corey PN. Incidence of adverse
drug reactions in hospitalized patients: a meta-analysis of pro-
spective studies [see comments]. JAMA. 1998; 279:1200–5.
6. National Coordinating Council for Medication Error Re-
porting and Prevention (NCC MERP). NCC MERP Taxonomy
of Medication Errors. Rockville, MD, 1998.
C
ONCLUSIONS
7. Chin MH, Wang LC, Jin L, et al. Appropriateness of medi-
cation selection for older persons in an urban academic emer-
gency department. Acad Emerg Med. 1999; 6:1232–42.
To see the future of error reduction in medicine,
visit a department store. Salespeople, tags, and
signs help you choose the right items. Store design in Utah and Colorado. Med Care. 2000; 38:250–60.
8
. Studdert DM, Thomas EJ, Burstin HR, Zbar BI, Orav EJ,
Brennan TA. Negligent care and malpractice claiming behavior
9
. Thomas EJ, Studdert DM, Burstin HR, et al. Incidence and
lowers the chances you will head away from some-
thing you need. Bar codes ensure rapid and
types of adverse events and negligent care in Utah and Colo-
rado. Med Care. 2000; 38:261–71.
accurate item identification. Laser scanning de- 10. Brennan TA, Leape LL, Laird NM, et al. Incidence of ad-
verse events and negligence in hospitalized patients. Results
vices reduce the errors that were common in hand-
of the Harvard Medical Practice Study I. N Engl J Med. 1991;
entering prices at the counter. Credit card units
3
24:370–6.
allow rapid identification of credit limit, prevent- 11. Leape LL, Brennan TA, Laird N, et al. The nature of ad-
verse events in hospitalized patients. Results of the Harvard
ing you from spending more than allowed by the
Medical Practice Study II. N Engl J Med. 1991; 324:377–84.
card. A combination of security cameras, security
12. Localio AR, Lawthers AG, Brennan TA, et al. Relation be-
personnel, and sensors at the door prevent you tween malpractice claims and adverse events due to negli-
gence. Results of the Harvard Medical Practice Study III [see
from ‘‘accidentally’’ walking out of the store with-
comments]. N Engl J Med. 1991; 325:245–51.
out paying for an item. A sign at the exit encour-
aging submission of comments and suggestions al- mishaps: lessons from non-medical near miss reporting sys-
lows management to identify additional problems
and errors. The benefits of these systems must
more than pay for their costs or they would not be 15. Nolan TW. System changes to improve patient safety.
instituted. Customers receive better service and
are more satisfied while the companies reap
greater profits.
Error identification is possible in medicine as
well, in many cases using the same systems and
technologies. Identification of systems problems 19. Reinertsen JL. Let’s talk about error [editorial]. BMJ.
that cause error is the first step in saving the lives
of nearly 100,000 people per year. The stakes may
be high, but so is the potential payoff. By setting association of litigation and risk management protocol. BMJ.
the example in an environment as chaotic, fast-
paced, and difficult as the typical ED, EM will set
13. Barach P, Small SD. Reporting and preventing medical
tems. BMJ. 2000; 320:759–63.
1
4. Helmreich RL. On error management: lessons from avia-
tion. BMJ. 2000; 320:781–5.
BMJ. 2000; 320:771–3.
1
6. Reason J. Human error: models and management. BMJ.
2
000; 320:768–70.
17. Gaba DM. Anaesthesiology as a model for patient safety
in health care. BMJ. 2000; 320:785–8.
1
8. Cohen MR. Why error reporting systems should be vol-
untary [editorial]. BMJ. 2000; 320:728–9.
2
000; 320:730.
2
0. Vincent C, Taylor-Adams S, Chapman EJ, et al. How to
investigate and analyse clinical incidents: clinical risk unit and
2
000; 320:777–81.
2
1. Thomas EJ, Brennan TA. Incidence and types of prevent-
able adverse events in elderly patients: population based re-
the standard for error identification for the entire view of medical records. BMJ. 2000; 320:741–4.
2
2. Nightingale PG, Adu D, Richards NT, Peters M. Imple-
medical profession.
mentation of rules based computerised bedside prescribing and
administration: intervention study. BMJ. 2000; 320:750–3.
References
2
3. Pietro DA, Shyavitz LJ, Smith RA, Auerbach BS. Detect-
ing and reporting medical errors: why the dilemma? BMJ.
. Kohn LT, Corrigan JM, Donaldson MS (eds). To Err Is Hu- 2000; 320:794–6.
man: Building a Safer Health System. Institute of Medicine 24. Weingart SN, Wilson RM, Gibberd RW, Harrison B. Epi-
demiology of medical error. BMJ. 2000; 320:774–7.
1
Report. Washington, DC: National Academy Press, 1999.
2
. Leape LL. Institute of Medicine medical error figures are 25. Bates DW. Using information technology to reduce rates
not exaggerated. JAMA. 2000; 284:95–7.
of medication errors in hospitals. BMJ. 2000; 320:788–91.