Chapter 5. Examples of Alternative Approaches to Mistake-Proofing
Mistake-Proofing the Design of Health Care Processes -
Chapter 5. Examples of Alternative Approaches to Mistake-Proofing
As discussed in Chapter 1, mistake-proofing can use a variety of setting functions, control functions, or categories. This chapter provides 67 examples of mistake-proofing organized into 21 sets. Each set highlights alternative approaches to similar problems. Since each organization's processes are subtly but distinctly unique, different organizations can use a variety of related approaches to address specific situations.
The inclusion of the examples in this chapter follows the pattern of books on mistake-proofing in manufacturing.1,2,3,4 Some of these examples will be directly applicable to errors that confront organizations. Others might suggest how to approach a novel problem that is a source of concern. The following are issues for consideration as the mistake-proofing examples in this chapter are discussed:
- Consider the most appropriate circumstances for each mistake-proofing example.
- Consider whether processes have adequately addressed issues raised by the examples.
Pella Window engineers develop and test seven solutions before identifying and implementing the best one.
Example Set 5.1—One Exposure Only, Please
Unintentionally exposing x-ray film to light destroys images that are vital to proper patient care. These images can usually be recreated, but they cost time and money.
The entrance to the hospital darkroom door (Figure 5.1) has only one opening in a revolving drum. Passing through the vestibule ensures that the contents of the darkroom are not unintentionally exposed. The door creates a failure in that you cannot enter the darkroom under incorrect conditions.
Figure 5.2 shows a film bin equipped with a special locking mechanism. The mechanism includes a light sensor that will not allow the bin to be opened when light is present in the darkroom.
Example Set 5.2—Variations in Scald Prevention
Data from the National Safe Kids Campaign indicate that 4,000-5,000 children are scalded each year, receiving third-degree burns that cover at least 12 percent of the body.5 Most of these events do not take place in medical facilities, although fatalities have occurred in medical facilities. Listed below is a variety of devices designed to reduce the chance of scalding.
Figure 5.3 illustrates a device that uses color-changing plastics to warn when water reaches dangerous temperatures, and scalds can occur.
The color of the circular ring changes from purple (left) to pink (right) when water exceeds 40°C, 104°F. A triangular ring (not shown) changes at 37°C.
The anti-scald plug shown in Figure 5.4 works the same way as the circular ring, but with one crucial difference: the anti-scald plug, by requiring the user to place the device in order to fill the tub, enforces its own use. In other words, the user must use the device to fill the tub, and using the device prevents scalding.
The anti-scald valve in Figure 5.5 is attached to the end of a faucet. It contains a valve that closes when water reaches 117°F. This device is easy to install. It is simply threaded on to the end of the faucet, replacing the standard filter or aerator. It creates a "shutdown" that prevents the possibility of scalding. This device lacks an override valve that is available on some models. A valve would allow hotter water to be obtained, if necessary, without disassembling part of the faucet.
The device in Figure 5.6 is inserted into the end of the shower pipe and works much like the previous device. It also contains a valve that closes when water becomes too hot. This device installs in a few minutes using an Allen wrench and is completely hidden once installed.
The device in Figure 5.7 is a thermostatic mixing valve. It provides "forced control" by automatically mixing cold water with hot water to reduce the water temperature. The maximum water temperature can be set and adjusted. Installation of this type of valve is more troublesome than the other devices mentioned here. In some cases, the mixing valve is located behind the wall of the shower, and installation or adjustment requires carpentry and plumbing.
Example Set 5.3—Medical Gas Connections
Figure 5.8 illustrates the extensive mistake-proofing that medical gas tanks undergo. Here, the fittings that connect medical gases from the tanks in the back room to the tubes leading to the patient have undergone extensive mistake-proofing.
Tanks display the color-coding scheme. The color coding serves as a "sensory alert" to ensure that the correct connections are made between the tanks and the valves in the patients' rooms. The tanks are also fitted with pin-indexed connectors to prevent incorrect connections.
The generic regulator in Figure 5.9 has a dial that indicates how many liters of oxygen are delivered. The dial clicks as each liter of oxygen is delivered. Hinckley2 points out that converting adjustments to settings is a very powerful type of mistake-proofing because it requires far less attention to detail and can accelerate the process dramatically.
In this case, however, the design has a drawback which is indicated in a warning box in the instructions: "There is NO FLOW between settings. To obtain the desired oxygen flow, the indicating pointer must point to a specific number on the dial."6 Eliminating the possibility of one mistake can create an opportunity for another.
...a physician treating a patient with oxygen set the control knob to between one and two liters per minute, not aware that the numbers represented a discrete rather than a continuous setting. No oxygen was flowing through, yet the knob rotated smoothly, giving the suggestion that the intermediate setting of the machine was possible. The patient became hypoxic before the error was discovered. A design solution would have been a rotary control that snaps into a discrete setting along with some indication of flow. 7
If the probability of the second mistake is lower than the probability of the first, on average, patients will benefit. Designing the device so that flow continues at the rate of the last setting, regardless of whether or not the indicator sits between settings, might be safer.
Example Set 5.4—More Connections
Often, connections are pin-indexed and color-coded. In Figure 5.10, the holes for the pins for medical air are located at 12 o'clock and 5 o'clock. Using this system, all of the gases have a pin at 12 o'clock. The other pin is different for each gas.
In their new facility in West Bend, WI, St. Joseph's Hospital has gone further by standardizing the location of each gas outlet on the head wall. Each gas outlet is located in the same place on each head wall in the hospital.
How much color-coding is too much? In this case (Figure 5.11), the clear nozzle may be the most mistake-proof choice.
It is not possible to mount the nozzle on the wrong regulator because it is universal, not color-coded. If you have to stock yellow, green, and every other colored nozzle in each location where they may be used, you not only incur additional inventory costs, you may actually cause reportable violations of operating policies or procedures that would be impossible if only clear nozzles were stocked.
Errors occur where things are almost the same but are subtly different. What can be made identical should be. What cannot be made identical should be made obviously, even obtrusively, different (Figure 5.11).
The regulator in Figure 5.12 is attached to the wall with a vinyl-covered steel cable in order to avoid converting the regulator into a projectile that flies across the room when it is disconnected under pressure. This is an example of what Tsuda8 refers to as "preventing the influence of mistakes."
The tubing can attach to a regulator with (Figure 5.13A) or without (Figure 5.13B) the nozzle attached. If either option is acceptable, this design is mistake-proof. If one option is preferred, the design should be avoided.
The tubing can attach to a regulator with or without a nozzle attached.
Example Set 5.5—Variations in Tube Identification
Ensuring that labels on samples and test tubes are correct is a very important aspect of reducing medical errors. Often, printed labels with complete information are not available at the bedside and must be retrieved from the nurses' station. This delay can contribute to the introduction of errors into the system. The following three examples represent approaches to accurately identify tubes and samples.
When multiple tubes are drawn at the same time, writing a label for each can be time-consuming. The Bloodrac™ is produced by the makers of the Bloodloc™ (Chapter 7, Example 7.8). It enables several tubes to be stored together with one handwritten label. The identification is usually the patient's name and the Bloodloc™ code on the patient's wristband.
Another approach to reducing the effort required to accurately label multiple tubes is the use of a wristband with pre-printed, peel-off, self-adhesive labels (Figures 5.14 and 5.15). All the labels have the same unique identification number. As they are removed, the same number appears underneath each label.
These labels temporarily label the tubes until permanent labels containing complete information can be printed and affixed. Error rates are reduced because the preliminary labels are only available at the patient's bedside.
Although the tube in Figure 5.16 receives the temporary label with little chance of error, the possibility of errors occurring when matching permanent labels with temporary ones would also need to be addressed in the process.
Example Set 5.6—Variations in Esophageal Intubation Detection
Esophageal intubation is a common error that occurs when the intubation tube is inserted in the patient's esophagus instead of in the trachea. There are several approaches to detecting this error in hospital settings. Figures 5.17-5.21 illustrate a variety of low-tech approaches that could be employed where power requirements or space prohibit more sophisticated approaches.
After the patient is intubated, a staff member squeezes the bulb (Figure 5.17) and places it over the end of the tube. If the bulb fails to re-inflate to its original shape, the tube is in the esophagus. If it re-inflates fully, the tube is placed correctly.
The bag in Figure 5.18 has a detector that changes color in the presence of carbon dioxide. If the detector fails to change color, then the tube is in the esophagus. If it changes color, the tube is in the trachea. If the mistake-proofing device fails, the device will indicate a situation requiring corrective action.
The round cylinder at the end of the tube in Figure 5.19 is a whistle. As the patient breathes in and out the whistle makes an audible, wheezing sound. The staff can hear the patient's breath.
The device in Figure 5.20 is essentially a large caliber syringe, but instead of pushing fluids out, it pulls in anything available. If the plunger cannot be pulled out easily, or if stomach contents come out, the tube is in the esophagus. If the plunger pulls out easily and completely, the tube is in the trachea as it should be.
Example Set 5.7—Variations in "Take Your Medicine," Part I
The following examples illustrate a common problem in prescribed medications. The instructions from the bottle shown in Figure 5.21 are:
Take 5 tabs once daily x 3 days,
then 4 tabs once daily x 3 days,
then 3 tabs once daily x 3 days,
then 2 tabs daily x 3 days,
then 1 tab daily x 3 days.
Complying with these instructions requires the patient to pay careful attention to detail and have a good memory or to use some additional mechanism for tracking necessary changes in dosages.
A typical prescription bottle may not adequately convey detailed instructions. One approach to managing dosage changes is to print the dosage for each day on a calendar. This approach is cumbersome and increases the probability of making an error in following the prescribed instructions.
The packaging, however, puts lots of knowledge in the world (Figure 5.22) because the rows are labeled by day, and the instructions for each day are printed on the line below each row of pills. The act of taking a pill creates a record of where the patient is in the sequence. No calendar or other aid is needed.
Example Set 5.8—Variations in "Take Your Medicine," Part II
For some, taking medications four times daily is not a problem. For others, determining if they have taken their medications is more problematic. Several approaches have been created to help ensure that medications are taken as prescribed. Clearly, some devices require less attentiveness, and some approaches are more cost-effective than others. The effectiveness of each approach depends on the individual involved.
Figure 5.23 shows a daily pillbox. It is simple, straightforward, and easy to use. This mistake-proofing device is not particularly mysterious or clever. Yet, it is common enough that it must work for someone. A weakness of this device is that it provides only a visual cue to take medications. It provides no mechanism to know which pills to take at a particular time of day.
Birth control pills are sold packaged for use as a 1-month supply. The packaging indicates whether or not patient has been taking their medication consistently and are current, assuming they know what day it is (Figure 5.24).
If clever packaging is not enough, the wristwatch in Figure 5.25 can remind the user to take medications up to six times a day. Instead of an audible alarm, the watch vibrates discreetly.
The logical extension of the simple pillbox is shown in Figure 5.26. This system has seven boxes. Each box contains four compartments and a timer to remind the patient to take the medicine in the next compartment.
If a timer is not enough, the medication dispenser in Figure 5.27 dispenses the medications and detects when they are removed. If the pills are not removed on a timely basis, the dispenser places a phone call to a family member or caregiver who can follow up with the appropriate party.
Example Set 5.9—Variations in "Take Your Medicine," Part III
The next step in this progression towards more strict control of medication dosing would be a device to ensure that medications are actually administered. Even the most sophisticated dispenser ensures only that the medication is removed from the dispenser. The mistake-proofing devices that follow take a different approach. Instead of controlling the traditional process, new processes are developed. The time-release capsule (Figure 5.28) accomplishes the same goal as prescribing multiple doses of a medicine but in a more mistake-proof way. The patients need only to take a single pill.
The need to remember to take even a single pill has been further reduced. Patches are used to administer medications over much longer periods of time (Figure 5.29). The design of the medication delivery method reduces the need to remember to take medications.
Going one step further, the Norplant® system is embedded in a woman's body for 5 years. This contraceptive is a set of six small capsules that are placed under the skin of the upper arm (Figure 5.30). They eliminate the need to remember to take medication. One drawback is that the implants must be removed after 5 years.
Example Set 5.10—Examples from the Built Environment, Part I
New York-Presbyterian Hospital covers a city block with interconnected buildings featuring addition after addition (Figure 5.31). The interconnections inside these buildings can cause navigational problems. Those familiar with the layout of the hospital know that Building A connects to Building B on the first and third floors, but not on the second, or they may know that floor 15 of Pavilion X connects to floor 16 of Y Building via the skybridge. Patients who are newcomers to these buildings find it difficult to navigate the hospital.
One way to improve a patient's ability to find their way around is to mark well-traveled paths throughout the facility. Paths could be marked using solid tape or painted lines (Figure 5.32) or, like Hansel and Gretel's bread crumbs, they could be marked with strategically placed icons that lead to the desired destination. The colored shoes (Figure 5.33) are from New York Presbyterian Hospital.
Magnetic fire doors (Figure 5.34) are widely used in many types of commercial buildings. They are linked to fire detection systems so that the doors are released, then close when the fire detection system is activated. Closing the door does not prevent fires. It temporarily prevents the influence of mistakes, the fire, from spreading further.
During a fire or other emergency, many factors make it easy for those descending to descend too far down the stairwell and miss the street-level exit. The gate in Figure 5.35 provides a very strong cue that descending past that point is discouraged. This gate, like the door in Figure 5.34, closes automatically when the fire alarm is tripped.
Example Set 5.11—Examples From the Built Environment, Part II
St. Joseph's Hospital has designed every patient room so that the sink is clearly visible (Figure 5.36). Patients are encouraged to watch and ensure that staff members wash their hands before interacting with them. In most rooms, the doors open toward the sink (Figure 5.37), further encouraging handwashing.
Each patient room is also provided with a nurse's alcove that has all of the necessary supplies (Go to Chapter 7, Example 7.27) and a computer for electronic documentation of care accomplished before he or she moves on to another patient. The door to the alcove has a window so that the nurse can see the patient while completing the chart (Figures 5.38 and 5.39).
The concept of being able to see the patient is taken to the extreme in St. Joseph's intensive care unit (ICU). The ICU is located along the curved back of the building. It is engineered so that each patient's face can be seen from the work area. Being able to see patients is considered so important that the only exception to room standardization in the entire hospital occurs here. The last two rooms (most distant in Figure 5.40) had to be rotated 180 degrees so that the patients could be seen.
Example Set 5.12—Examples from the Built Environment, Part III
At St. Joseph's Hospital, the bathroom is placed near the head of the bed for fall prevention (Figure 5.41). In many hospitals, however, the room plans for every other room are rotated 180 degrees so that two rooms share a common plumbing wall. Often, this requires the patient to cross the middle of the room where no handrails are available to support the patient and prevent falls. At St. Joseph's hospital, the patient is provided with handrails from the bedside to every part of the bathroom (Figures 5.42-5.45).
If human beings are prone to perform automatically, as if on auto-pilot, perhaps making environments where that auto-pilot will be correct is worthwhile. That is the strategy for headwalls. Every room in the hospital has a standardized arrangement; the Emergency Room (ER), ICU, and medical/surgical rooms are all identical. Subsequently, ER activities can spill over into adjacent rooms in ICU during extremely busy times (Figures 5.46A and 5.46B).
Example Set 5.13—Getting X-Rays Right
The flasher plate (Figure 5.47) is a small window on the x-ray film cassette that prevents film from becoming exposed during the flashing of the patient's name onto the film. The cassette is inserted into the name flasher, then the flasher plate is automatically pulled back to expose the film to a small light that exposes the patient's name on the film (Figure 5.48). The flasher plate will only move when it is inserted into the name-flashing device.
The film is marked as 'left' or 'right' for reference to prevent the radiologist from misinterpreting the results when reading the film (Figure 5.49).
In addition to marking the left and right sides, a technologist's initials, film series, and other information can be recorded with a blunt writing instrument.
Figure 5.50 shows a dental x-ray film holder that is inserted in a patient's mouth. The hoop provides a target for positioning the x-ray machine so that the dental technician aligns it correctly each time.
Example Set 5.14—Exposure Control
Figure 5.51 illustrates two mistake-proofing techniques to prevent radiation exposure. The lighted sign over the x-ray room door alerts personnel that an exposure is in progress. When the x-ray machine is emitting x-rays, the exposure light alerts personnel that they should not enter the room. The door interlock prevents exposure if the x-ray room door is left open. A small switch in the door frame will not allow an exposure to occur if the door is left open.
In Figure 5.52, the mistake-proofing takes the form of personal protective equipment. The apron protects the technician from overexposure to radiation.
On older x-ray units that have a cord attached to the exposure button, the relatively short cord length does not allow the technologist to make an exposure while outside the protection of the control booth (Figure 5.53).
Example Set 5.15—Bed Alarms and Fall Reduction
A number of mistake-proofing devices have been developed to protect patients at risk for falls when getting out of bed. Bed alarms (Figures 5.54 and 5.55) are designed to notify caregivers when a patient gets out of bed, come in a variety of sizes and shapes, and perform a variety of functions.
In some cases, the alarm is built into the bed. In other cases, it is added-on as needed. In one model, a pad is placed under the bottom sheet and detects the patient's body weight. Go to http://www.abledata.com/abledata.cfm?pageid=113583&top=0&productid=84241&trail=0.
Some sensors can be placed under the mattress. The alarms come with many different sound alert options, including voice warnings that can be recorded by a loved one. One version will monitor how long a patient has been out of bed and telephone a caregiver if the bed remains unoccupied for too long. Go to http://www.bedmonitors.com/bedmonitors_how_works.htm.
The alarm in Figure 5.56 tethers the patient to the bed. The device is strapped to the patient. The cord is attached to the bed. The alarm will sound if the patient moves in a way that pulls the cord, detaching the tether from the device.
The bed alarm in Figure 5.55 is placed on the floor beside the bed so that an alarm sounds when a patient's feet touch the floor. Other approaches considered by manufacturers include the use of lasers and other sophisticated sensors to detect when a patient tries to get up. Go to http://www.norto.com.au/Norto-Emfit-Safebed.htm.
Example Set 5.16—Sharps
Protecting staff, patients, and visitors from the biohazard of "sharps" (exposed needles) takes many forms. The syringe in Figure 5.57 is equipped with a cover for the needle. The cover is inserted into the square base so that it stands vertically. The needle can be inserted into the cover using one hand. The other hand can be kept safely out of the way. When the cover is used as intended, a misjudgment in the insertion process will not result in a needle stick.
The syringe in Figure 5.58 has a hinged cover so that it can be easily closed with one hand and with a motion that provides minimal opportunity for a needle stick. The cover clicks into place and cannot be removed.
The needle in Figure 5.59 is nearly self-blunting. It is inserted normally, but when the needle is withdrawn, a sleeve containing a steel tip cover is held against the patient's skin. The tip of the needle catches the cover, pulling it from the sleeve. Devices that require less effort and involve a more natural motion will usually be more effective.
The scalpel in Figure 5.60 has a spring-loaded, retractable blade. Push a button near the back of the handle and the blade retracts.
A workbook on sharps safety by the Centers for Disease Control and Prevention (CDC)9 states: "A passive safety feature is one that requires no action by the user."
...Few devices with passive safety features are currently available. Many devices currently marketed as self-blunting, self-resheathing, or self-retracting imply that the safety feature is passive. However, devices that use these strategies generally require that the user engage the safety feature... Although devices with passive safety features are intuitively more desirable, this does not mean that a safety feature that requires activation is poorly designed or not desirable.9
Example Set 5.17—Controlling the Controls
The IV pump in Figure 5.61 features a small button on the back of the machine that can be used to lock the controls so that others who are unfamiliar with the equipment cannot tamper with the settings on the control panel on the front of the machine.
The switch on this IV pump has a clear plastic cover that prevents inadvertent bumping of the switch. Donald Norman10 recommends making it harder to do what cannot be reversed. The cover makes the equipment in Figure 5.62 more difficult to use, but apparently, the errors prevented more than compensate for the inconvenience.
Example Set 5.18—Software
Although most mistake-proofing devices can be photographed, software applications cannot. A number of logical checks can be performed, however, after information has been entered into an application via computer. The examples in this set show how software interfaces can be used to mistake-proof some aspects of medical processes.
The keypad on the radiology equipment in Figure 5.63 requires patient information to be entered prior to any exam.
Programmed protocols are pre-programmed instructions to the machine that control how the exam will be set up (Figure 5.64). This enables the operator to select the exam and be confident that all the correct settings are in place. It also ensures that exams are performed in a consistent manner, regardless of who is operating the machine.
The system shown in Figure 5.65 reviews patient histories and gives an alert if the blood type differs from the type previously entered or if the patient's blood contained an antibody. The system will also give a warning if a blood type different from the patient's is cross-matched for transfusion. An alert is given if the patient needs special blood products.
For stereotactic breast biopsies, this unit (Figure 5.66) contains protective software that, by calculating the needle's position, prevents the improper insertion of a biopsy needle that could result in patient injury.
Example Set 5.19—Refrigeration Feedback
Blood bank refrigerators are equipped with temperature monitors that sound an alarm if and when the temperature is out of the safety range (Figure 5.67). The alarms also produce continuous chart recordings (Figure 5.68) and visual digital readings (Figure 5.69).
The read-out "Status OK" indicates that the refrigerators in this blood bank are operating properly. When the temperature or other operating parameters are not correct, a message scrolls across the display, indicating which refrigerator is out of specifications and which specification is violated.
The alarm system in Figure 5.70 is functional but not as sophisticated as the other examples in this set. It features an audible alarm and lights to indicate which zone is down. This system benefits tremendously from the posted instructions, a simple job aid that puts knowledge in the world. It will be very useful when the alarm goes off. The instructions indicate which refrigerator corresponds to each zone and provides information about how to silence the alarm and troubleshoot the problem. Troubleshooting begins with ensuring that the door is sealed.
Example Set 5.20—Mistake-Proofing Patient Interactions
As Chase and Stewart11 indicated almost 15 years ago, the actions of the "customer" need to be mistake-proofed. These two examples show how software can reduce variation in processes for which patients' cooperation and precise responses are critical to successful outcomes. Both examples show that mistake-proofing the actions of health care staff can only partially lead to truly mistake-proofing processes. The actions and behaviors of patients, family, and loved ones must also be mistake-proofed.
It is very important that patients lie as motionless as possible during a CT exam. Breathing instructions are prerecorded and embedded in multimedia software. When the operator begins the exam, the correct breathing instructions are automatically played to the patient at the correct time without further operator intervention, helping ensure patient cooperation and optimal results (Figure 5.71).
The blood donation software application in Figure 5.72 is optimized by eliciting donor information through a Web-based survey. The system provides on-screen text with added privacy via earphone audio and other options, color pictures to emphasize important aspects of questions, and touch screens to eliminate "keyboard phobia" and mouse aversions. The system contains donor self-interview and staff-review modules. It prevents the production of a donor record until the survey has been completed and a staff member has judged the information to be acceptable.
Example Set 5.21—Wristbands
Wristbands (Figures 5.73-5.75) have been used extensively in medicine to provide sensory alerts for many different patient conditions. Wristbands provide a physical space for patient information to reside; efforts are underway to put as much information onto a wristband as possible. Color coding is widely employed to indicate allergies, fall risks, do not resuscitate (DNR) orders, etc. Improved printer functionality enables the placement of multiple symbols and photos on the wristband, along with a patient's name and date of birth. The wristband is also the locus for more sophisticated patient identification technologies such as bar coding. There is a magnetic data storage device (capable of storing medical records) mounted on the wristband in Figure 5.75.
Some hospitals place yellow wristbands on DNR patients. There is some concern that yellow LIVESTRONG™ bracelets that help support the Lance Armstrong Foundation's efforts to fund cancer research may be mistaken for a DNR wristband. While no one has ever died because of confusion between the two, some hospitals are reportedly taping over LIVESTRONG™ bracelets with white adhesive tape to be safe.12
This chapter discussed related sets of mistake-proofing devices. Chapter 6 is concerned with medical and nonmedical applications of mistake-proofing.
5. Bynum D. Domestic hot water scald burn lawsuits—The who, what, when, why, where, how. In: Petri VJ, et al. Annual ASPE Meeting. Indianapolis, October 25-28, 1998. See also http://www.tap-water-burn.com/#4. Accessed January 2007.
9. Centers for Disease Control and Prevention. Workbook for designing, implementing, and evaluating a sharps injury prevention program. http://www.cdc.gov/sharpssafety/appendixB.html. February 12, 2004. Accessed Sep 2005.