Chapter 8. More Examples of Mistake-Proofing in Health Care
Introduction
This chapter features 34 additional examples of mistake-proofing in health
care. The examples in this chapter are more expensive and technology-based
than those described in Chapters 5-7, although some very simple examples are
also included. They are provided as both a catalog and a catalyst for reducing
human errors in health care.
Return to Contents
Example 8.1—Infant Abduction Prevention
Mistake-proofing often involves electronic sensors to ensure high-quality
industrial production. Electronic sensors are also used in health care applications.
In this example (Figure 8.1), an electronic device, or "tag," is
designed to be clamped to the infant's umbilical cord. The arrow in the photo
points to the cord clamp, which secures the tag to the infant. The tag ensures
that the infant is not removed from the nursery. If the infant is removed without
authorization, alarms sound, specified doors lock, and the elevators automatically
return to the secured maternity floor; the elevator doors remain open.
Return to Contents
Example 8.2—Bar Coding
Bar coding is one of the more common and effective information enhancement
and mistake-proofing devices. It is particularly useful in ensuring a match
between a patient and their treatment, medicines, and supplies (Figures
8.2 and 8.3).
One of the contributors to this example emphasized the importance of radiologists
matching the film they are reading to the right patient:
Bar codes are attached to every order so that the radiologist can electronically
identify the patient and be sure that the correct patient [information] has
been entered into the digital dictation system.
Another contributor stated:
Each specimen is labeled with a bar code that is specific to that patient
and the test that has been ordered. The instruments in the laboratory are programmed
to identify the bar code that ensures positive patient identification and to
verify that the correct test is performed.
Bar coding, however, is a setting function. Therefore, it is only as effective
as the regulatory function to which it is linked. Many of the control methods
used with bar coding are warnings or sensory alerts. The control methods of
shutdown and forced control are infrequently used.
AuBuchon discussed this shortcoming of bar coding systems for patient identification:
A disadvantage that we ran into when we began using the system on a trial
basis is that the system doesn't have to be used... ultimately, our anesthesiologist
said, 'You know, this is a really neat system, but I won't use it. He said
[that with] the Bloodloc™, I have got to use it, I have got to do something,
we have got to take it off, and that's the whole idea. It's a barrier. It
prevents the transfusionist from getting to a unit of blood that they are
not supposed to get to.' So we have continued using that older system rather
than the new, fancy system.1
The use of bar codes does not automatically prevent errors from occurring.
Staff should check that assigned bar codes match. In Figure
8.3, a line of
red laser light is hovering in the gap between two bar codes, increasing the
odds of reading the wrong bar code by mistake.
Given the prevalence of patient identification errors, bar coding is a very
promising direction in mistake-proofing.
Return to Contents
Example 8.3—Computer-Aided Nutrition and Mixing
Software is used to profile total parenteral nutrition (TPN) solutions (Figure
8.4). A patient's nutritional needs (protein, sugar, fat, vitamins, and electrolytes)
are entered into the software application. The software sends a message to
an automixer that compounds the ingredients to create the base solution. The
software issues a warning if certain concentrations of ingredients are exceeded
based on literature values.
Return to Contents
Example 8.4—Equipment Collisions
In hospital operating suites full of large, expensive equipment, there is
always the danger that units of equipment will collide with each other. Equipment
requires a wide range of motion while in operation. Collision detection systems
warn and, in some cases, can lock if they sense an impending collision. The
equipment in Figure 8.5 is situated in an angiographic suite and outfitted
with electronic and manual locks to prevent collisions.
Return to Contents
Example 8.5—Flawless Equipment Setup
When creating x-ray film, it is very important that the tube is centered to
the film and is situated the correct distance from the film. The position locks
(Figure 8.6) enable the tube to be centered quickly and correctly by only locking
at the correct positions.
Return to Contents
Example 8.6—Mistake-Proof Mistake-Proofing
Transport monitors, which employ flashing and audible alarms, warn all health
care workers of high/low heart or breathing rates. A misplaced blood pressure
cuff on the lower arm below the elbow, as in Figure 8.7, would result in inaccurate
blood pressure readings and trigger flashing and audible misplacement alarms.
Return to Contents
Example 8.7—Private Files
Often, mistake-proofing is accomplished by providing barriers that prevent
people from taking the wrong action. In Figure 8.8, a portion of the file cabinet
drawer can be locked. This mistake-proofing is neither mysterious nor subtle.
Return to Contents
Example 8.8—Computer Drug Interaction Checker
Software that checks for drug interactions (Figure 8.9) falls under Shingo's
concept of a successive-check.2 A successive-check
is a mistake-proofing device that facilitates checking work previously performed
by others and that, in a low-cost, relatively automatic way, notifies the user
that something is wrong. Shingo was of the opinion that defect detection and
rapid feedback following a mistake are nearly as effective as not making the
mistake at all. Even after an initial mistake, staff can recover before substantial
harm occurs. In this case, the pharmacist double-checks the prescriptions
submitted by doctors. It is clear that there is no resultant harm if an error
can be caught by the pharmacist before the patient receives the medicine, thereby
avoiding, at the very least, significant difficulties for the pharmacist, doctor,
and patient.
Return to Contents
Example 8.9—Computerized Physician Order Entry
According to Poon, Blumenthal, Jaggi, et al;3
Medication errors are the most common cause of preventable injuries in hospitals.
Computerized physician order entry (CPOE) systems can reduce the incidence
of serious medication errors by 55 percent, but only 10 percent to 15 percent
of hospitals use them.
CPOE is computer software that physicians and other health care providers
use to issue and record patient orders for diagnostic and treatment services
such as medications, laboratory tests, and diagnostic tests. Computers on wheels
(COWs) are available throughout hospitals so that staff can enter information
without having to go to a central location (Figure 8.10). CPOE provides several
mistake-proofing features:
- Informs providers of common dosages and overdose warnings via drop-down
menus.
- Eliminates the issue of legible handwriting.
- Conducts drug interaction and
allergy checking routines.
- Employs sophisticated systems that function as a clinical decision support system
(CDSS).a CDSSs are "active knowledge systems that
use two or more items of patient data to generate case-specific
advice."4
a. Go to
http://www.openclinical.org/dss.html#wyatt1991.
Return to Contents
Example 8.10—Sponge-Counter Bag
In aviation, significant effort is exerted to ensure that no foreign objects
are left inside fighter planes. This is done to prevent foreign object damage
(FOD). Changing G forces can make objects weightless. Subsequently, they could
fly through the cockpit and cause serious damage to people and equipment. FOD
is also a problem in surgery. Failing to remove foreign objects (tools or supplies)
from inside a patient can cause serious harm.
The sponge-counter bag (Figure 8.11) assists in keeping track of sponges removed
from a patient. Accounting for the sponges put into the patient is easier because
the sponges are not discarded immediately or put in a random pile.
Return to Contents
Example 8.11—Notebook Switches
Galsworth5 endorses the mantra that workers should
be able to "know by looking." The notebooks in Figures 8.12 and 8.13
enable users to do that. The dial on the notebook in Figure 8.12 and the switches
on the notebook in Figure 8.13 enable everyone to know the status of the paperwork
inside.
Colors indicate when medical staff have made entries that need to be processed
by administrative staff. A different color notifies the nurse when the work
is finished. No color is displayed when the work is completed, and no further
action is needed.
Return to Contents
Example 8.12—Plug Protection
In May 2004, a National Patient Safety Foundation (NPSF) LISTSERV® participant
inquired about the safest height for electrical wall outlets in pediatric rooms.
In his response, Matthew Rosenblum stated that he believes that other matters
are probably more important:
For example, how the cord is secured to the outlet and to the wall and how
the outlet is covered when no devices are plugged in. In this regard, there
are numerous products on the market for securing electrical cords to the
outlets and to walls. Also, many secure socket covers are
available.6
When an outlet is used properly, the plug fits without slowing the process.
The process is slowed only when an error occurs; then the mistake-proofing
device brings the process to a halt. Figures 8.14-8.18 illustrate various mistake-proofing
methods employed to make wall sockets safer.
Return to Contents
Example 8.13—Instructions Getting in the Way
The card shown in Figure 8.19 is not the strongest example of mistake-proofing.
It does, however, put knowledge in the world. Also, it is designed to stand
out against a noisy background. At a minimum, someone (a patient or family
member, perhaps) will have to move it out of the way in order to use the table
space.
A card on the overbed table (Figure 8.20) provides information to patients
about what patient safety behaviors to expect from staff and encourages them
to hold staff accountable for complying with those behaviors.
This example is similar to the time-out example (Chapter
7, Example 7.4).
It also has some common features with a proxy ballot that was mailed to a retirement
fund (Figure 8.21). The ballot was designed so that it would not fit in the
envelope until a small portion of the page containing the mailing instructions/checklist
was torn off.
Return to Contents
Example 8.14—Monitoring Glucose
In the past, glucose monitoring required that patients follow strict clinical
procedures to determine their blood glucose levels. Today, most of the precise
actions and calculations are designed into a portable glucose monitor that
is user-friendly and more mistake-proof.
Return to Contents
Example 8.15—Unit Dosing
Robotics, bar coding, and packaging medicines in plastic bags containing a
single dose, or "unit dose," form a powerful combination of mistake-proofing
devices. Individually, none of them would be very effective. The unit dose
package enables the machine to select a single dose to be delivered to a patient.
The unit dose package also provides a convenient way to associate bar codes
to a specific pill for use in the pharmacy and throughout the medication delivery
system. Bar codes make the packages containing the pills machine readable (Figure
8.22). The machine in Figure 8.23 provides the automation that makes converting
bottled medicines into unit doses less expensive, less labor intensive, and
more reliable.
Return to Contents
Example 8.16—Kits
The Massachusetts team from the Patient Safety Improvement Corps (PSIC) reported
their efforts in reducing central line infections.7 They
recommended a variety of changes to the central line insertion process. Included
in their recommendations is a customized kit (Figure 8.24) that standardizes
available supplies, including drapes and other site preparation materials.
Table 8.1. Cost comparison between two methods of reducing central line infections
Savings will equal the difference in total
episodic costs of the two methods:
([B]$6,525,000- [A]$3,240,000 minus the difference in equipment costs ([A]$147,840-[B]$55,552=$92,288) |
Method A: Previous Method |
Annual equipment cost
2,240 cases x $24.80/kit = $55,552.
Annual infection cost
$45,000/episode x 145 expected episodes = $6,525,000
Total Cost = $6,580,552 |
Method B: Using Custom Kit |
Annual equipment cost
2,240 cases x $66/kit =$147,840
Annual infection cost
$45,000/episode x 72 expected episodes = $3,240,000
Total Cost = $3,387,840
Net Savings = $3,192,712 |
According to the calculations in Table 8.1, the annual cost increase is substantial:
$92,288. Yet, if the number of infections can be reduced by only 3 episodes
out of 145 (a 2 percent decrease), the change will be cost-justified. The team
forecasted infection rates would be cut in half, a result that was supported
in their preliminary findings. The net savings appeared to be far more substantial
than the cost increase.
Source: Example and photos courtesy of an anonymous contributor. Used with permission.
Return to Contents
Example 8.17—Bacteria-Detecting Bandages
Benjamin Miller8 developed the technology to produce "smart
bandages" that indicate an infection by changing color (Figure 8.25).
The "smart bandage" is in the early stages of development, so actual
commercial products may still be years away. In its current form, the technology
is in a chip that reveals the existence of different bacteria by changing colors.
As a consumer product, a small chip would be embedded in a regular bandage.
Computer connectivity is another future possibility.
Return to Contents
Example 8.18—Urinalysis Test Strips
The old method of reading urinalysis test strips required health care workers
to make subjective decisions. Timing and color perception were critical to
error-free results. The machine in Figure 8.26 analyzes urine test strips and
prints out the results. In addition to the obvious mistake-proofing associated
with the automatic nature of the machine, the strip can be inserted in only
one direction, and the results can be printed out and placed in the patient's
medical chart. A transcription of the results is not necessary.
Improperly handled or inadequately maintained samples can result in inaccurate
diagnosis and treatment. The sample transport kit in Figure 8.27 maintains
urine specimen integrity without refrigeration for up to 72 hours at room temperature.
Return to Contents
Example 8.19—Controlled by Connections
In Figure 8.28, a benign failure protects patients. Only rubberized specula will
fit as attachments to this loop electrosurgical excision procedure
(LEEP)b machine. Standard metal specula cannot
be attached. If a metal speculum could be inadvertently attached to the machine
and used, it would result in burns or electrocution.
b LEEP is "a
way to test and treat abnormal cell growth on the surface tissue of the cervix.
LEEP is prescribed after abnormal changes in the cervix are confirmed by Pap
tests and colposcopy."
Go to: http://www.clevelandclinic.org/health/healthinfo/docs/0600/0642.asp?index=4711.
Return to Contents
Example 8.20—Child-proofing
Child-proofing is mistake-proofing. Since the bottle in the foreground of
Figure 8.29 is not child-proofed, it is kept inside a child-proofed medication
container when not in use to prevent accidents. In this example, an entire
demographic group is unable to open a container, the exact benign failure for
which it was designed.
Return to Contents
Example 8.21—Hemoglobin Testing
Precision in hemoglobin testing is important. Appropriate diagnosis and treatment
are based on the results. Automatic hemoglobin testing devices (Figure 8.30),
which perform the analyses in under 1 minute, have replaced analyses that relied
on visual judgment or time-consuming, complicated methods for their precision.
Return to Contents
Example 8.22—Auto Shut-Off Treadmills
The treadmill in Figures 8.31 is used in rehabilitative therapy. It is equipped
with an emergency stop button and automatically slows to a stop if the patient
trips or falls.
Return to Contents
Example 8.23—Visual Systems
Figures 8.32 and 8.33 are more examples of how to "know by
looking."5 Visual systems make a system's
status visible to all. Norman9 encourages visibility
to reduce errors: "make relevant parts visible." In Figure
8.32, the goal was to encourage employee donations in a workplace. The visibility
of the status of the blood supply made a dramatic difference. Employee donations
grew 300 percent. The sign served as a simple gauge to indicate inventory levels
and mitigated the human perception, or error, of believing that the blood supply
was more than adequate. The gas gauge depicted in Figure 8.33 is another visual
cue to the status of a machine.
Return to Contents
Example 8.24—Needleless Systems
Needleless systems are used throughout the hospital to prevent needle sticks.
The display panel in Figure 8.34 informs the nurse if there is air in the system.
Safety-engineered products for intravenous (IV) therapy have proven effective
in protecting health care workers from exposure to bloodborne pathogens (Figure
8.35). In a retrospective review, the Exposure Prevention Information Network
(EPINet) at the International Health Care Worker Safety Center at the University
of Virginia in Charlottesville showed that the rate of percutaneous injuries
among nurses declined from 19.5 per 100 occupied beds in 1993 to 9.6 per 100
occupied beds in 2001, a decrease of nearly 51 percent.10
Because these figures only include the first few months of legally mandated
safety device use, they don't fully reflect the effect of the Needlestick
Safety and Prevention Act,11 which mandated the use
of needleless IV systems in all health care settings.
Safety-engineered devices prevent accidental needle sticks in two ways: primary
prevention and secondary prevention. The most direct method of preventing needle
stick injuries is through primary prevention techniques that eliminate the
need to introduce sharps into the workplace, reducing the total number of sharps
used.
Return to Contents
Example 8.25—Dress Code Cued by Floor Tile
The patterned tile in the hallway (Figure 8.36) is a sensory alert that surgical
attire must be worn past this point. The tile adds a visual cue about what
to do, but it only works for those who have been taught what the tiles mean.
Patients, visitors, or new staff members will not be aware of this convention,
thereby limiting its effectiveness. Fortunately, patients are usually sedated
and recumbent in this hall, and visitors are prohibited.
Return to Contents
Example 8.26—Internet-Aware Refrigerator
Undergraduate engineering students at Virginia Military Institute (VMI)—advised
by a biomedical engineer, a computer engineer, and a physician—designed
a medical, Internet-aware, insulin refrigerator for patients living alone.
The small refrigerator (Figure 8.37) is monitored by a microcontroller that
is connected to a standard telephone outlet. If the refrigerator door is not
opened in a 16-hour period, the microcontroller sends an E-mail or a pager
alert to a designated caregiver. The system has battery backup in case of a
power outage. The system can be retrofitted to standard refrigerators.
Return to Contents
Example 8.27—Resources with Which to Err
Sometimes, mistake-proofing can be thought of as the removal of the materials
required to make errors. In the United Kingdom, the National Patient Safety
Agency, in its first patient safety alert, warned that potassium chloride solution
in its concentrated form should be removed from all general wards and replaced
by diluted products. Go to Chapter 7, Example 7.7.
Return to Contents
Example 8.28—Keeping Time
Mistake prevention in the work environment involves reducing ambiguity. As
far as time is concerned, variation is ambiguity. Clock systems (Figure 8.38)
eliminate variation. A receiver takes signals from global positioning system
(GPS) satellites and communicates the signals to other clocks in the system,
including those in computers.
The clocks in Figure 8.39, produced by different manufacturers, set themselves
accurately. When observed, the variation between them was approximately one-half
second.
Return to Contents
Example 8.29—Distinct Labeling
Businesses try to build an image for their product lines by using similar
packaging. Figure 8.40 illustrates a consistent image that leads to brand awareness
but may also lead to packaging that offers minimal distinctions between products.
Figure 8.41 shows that, while patterns and graphics can unify a company's
product line, individual product packaging can be visually distinct. Even
within the same product line, different dosages can be made
distinct.c
c Information design for
patient safety. A guide to the graphic design of medication packaging is available
from the UK's National Patient Safety Agency at
http://www.npsa.nhs.uk/site/media/documents/1539_Information_Design.pdf.
Return to Contents
Example 8.30—Free-Flow/No-Flow Protection
Infusing too much or too little fluid can lead to problems. The free-flow
protection on the IV pump in Figure 8.42 causes a benign failure. It is a simple
V-shaped piece of plastic (Figure 8.43) loaded on the machine. The flow of
medication to the patient stops if a tube is removed from the machine.
Some infusion pumps also offer downstream occlusion alarms that alert staff
that the tubes are blocked or that the clamp has not been opened, preventing
the fluid from infusing.
Return to Contents
References
1. AuBuchon J. Practical considerations in the implementation
of measures to reduce mistransfusion. Best practices for reducing transfusion errors—OBRR/CBER/FDA Workshop. Food and Drug Administration, Center for Biologics
Evaluation and Research and Office of Blood Research and Review. Bethesda,
MD; 2002 Feb 15. http://www.fda.gov/cber/minutes/0215bloo.htm. Accessed: Sept. 2005.
2. Shingo S. Zero quality control: source inspection
and the poka-yoke system. New York: Productivity Press; 1985.
3. Poon EG, Blumenthal D, Jaggi T, et al. Overcoming barriers
to adopting and implementing computerized physician order entry systems in U.S. hospitals.
Health Affairs 2004 July;23(4):184-90.
4. Wyatt J, Spiegelhalter D. Field trials of
medical decision-aids: Potential problems and solutions. In: Clayton P, ed.
Proceedings of the 15th symposium on computer applications in medical care, Washington,
1991. New York: McGraw Hill; 1991.
5. Galsworth GD. Visual workplace: visual thinking.
Presentation at 16th annual Shingo Prize Conference. Lexington, KY: May 2004.
6. Rosenblum M. Written correspondence. NPSF
LISTSERV®; 2004 27 May.
7. Alper E, Brush K, McHale E, et al. Prevention of
central line infections. Public-private collaboration,
http://www.patientsafety.gov/psic/StatePresentations/2004-2005/Massachusetts.ppt.
8. Smart bandages. Popular Mechanics 2002 May;
179(5):30.
9. Norman DA. The design of everyday things. New York:
Doubleday; 1989.
10. Jagger J, Perry J. Comparison of EPINet data for
1993 and 2001 shows marked decline in needlestick injury rates. Adv Exposure
Prev 2003;6(3):25-27.
11. Needlestick Safety and Prevention Act. Public
Law 106-430, 106th Congress; 2000 Jan 24.
Return to Contents
Proceed to Next Section