Chapter 9. Summary
Mistake-Proofing the Design of Health Care Processes -
Chapter 9. Summary
The examples in this book represent only a fraction of the current mistake-proofing methods and devices in the health care industry and only hint at the possibilities of how mistake-proofing could be applied. The implementation of mistake-proofing does not require starting from a standstill. Instead, existing solutions should be implemented wherever appropriate throughout each health care organization. Where ready-made solutions do not exist, designing, fabricating and installing new devices will be required.
Mistake-proofing is a change of focus, requiring more attention to the detailed design of processes, so that the easy way (or, ideally, the only way) to perform a task is the correct, efficient, and safe way. Mistake-proofing involves changing the physical attributes of a process. Consequently, mistake-proofing devices usually can be photographed.
Implementation of mistake-proofing in health care settings will be accomplished by putting knowledge in the world, designing benign failures, preventing failures in the work environment, detecting errors, preventing errors, and preventing the influence of errors. It will require the employment of devices that mistake-proof the actions of care providers, patients, and patients' family members.
Tables 9.1-9.5 recap the composition of the mistake-proofing examples presented in this book as they were categorized in Chapter 1. Although the selection of these examples was not intentionally biased, a distinct and restrictive definition of what does and does not constitute a mistake-proofing device affects these findings. Mistakeproofing is relatively narrowly defined here when compared with other authors' definitions.1,2 For example, Godfrey, Clapp, Nakajo, et al, include actions such as "train laboratory technicians to... empower all employees to... encourage patients to... clarify with physicians..."1 You cannot take a picture of these actions, so, while they may be worthwhile and effective actions, they would not be included here. Therefore, the proportions of examples reported in the tables do not provide a carefully constructed statistical sample that warrants populationwide conclusions. These tables suggest areas that lack medical mistake-proofing examples and call for new contributions to the body of knowledge.
Preliminary data from the example collection process suggest that many of the mistake-proofing examples included here have been broadly implemented in health care. Many device examples were submitted by people from differing organizations and geographical regions, and several were featured on commercial equipment or supplies. No locally developed devices were reported more than once. Further research is necessary to definitively determine if the implementation of certain commercially available mistake-proofing devices is widespread, as the preliminary data suggest. Findings of widespread implementation would be encouraging, suggesting that the health care industry is amenable to these devices.
Table 9.1 shows how the devices from this book are distributed among Tsuda's3 four approaches to mistake-proofing. One-half of the devices are designed to directly prevent mistakes by prohibiting them from taking place. Another 28 percent represent changes to the work environment intended to prevent mistakes in indirect ways, by removing ambiguity and making correct actions more obvious. Twenty percent of the devices rapidly detect errors, enabling staff to respond quickly and prevent more serious errors. Among those collected, only a few examples of preventing the influence of mistakes were identified.
Table 9.1 Mistake-proofing devices categorized by Tsuda's3 four approaches to mistake-proofing
|Approach||Count||Percent of total|
|Mistake prevention in the environment||42||28.0|
|Preventing the influence of mistakes||5||3.3|
Table 9.2 shows the distribution of devices that utilize the different setting functions identified by Shingo4 and Chase and Stewart.5 More than one-third of the devices, 35.3 percent, are physical setting functions. This percentage would not be unusual for any mistake-proofing application or, for that matter, any industry. The more interesting number is the 36.0 percent of information enhancement setting functions.
Chase and Stewart wrote about this type of device over a decade ago.5 They added information enhancement devices to those proposed by Shingo4 in the belief that this type of mistake-proofing would be needed in services. The fact that over one-third of the devices are in this category supports their belief.
|Setting Function||Count||Percent of Total|
|Grouping and counting||24||16.0|
Table 9.3 indicates the distribution of the collected mistake-proofing devices when categorized by control function. Shutdown and sensory alert devices are the most common control functions. The overall distribution of devices is somewhat evenly distributed among the control functions.
|Control Function||Count||Percent of Total|
Note: Numbers may not total 100 due to rounding.
Table 9.4 divides the mistake-proofing devices discussed in this book into the six categories defined by Chase and Stewart.5 These categories are divided into those concerning errors committed by customers (non-health care personnel) and errors committed by service providers (health care personnel). Of the collected examples, 24.66 percent address errors that would be committed by customers. Of these, almost 90 percent are mistake-proof aspects of the service encounter.
Few examples exist in the areas of preparation and resolution. The remaining 75.33 percent focus on the errors of health care personnel. Not surprisingly, the vast majority of provider devices, 62.50 percent of the total and 84.07 percent of the provider devices, address task performance errors, and 14.16 percent address errors associated with the tangibles delivered to patients. Only two (1.77 percent) devices collected ensure that patients were treated in a respectful and professional manner. This does not mean that patients were treated badly, only that few physical devices aided in providing proper treatment.
This analysis suggests the existence of a broad area of opportunity to identify or create additional mistake-proofing devices that address customer preparation, customer resolution, and provider treatment. The realization of these opportunities will result in a perception of more patient-centered care by everyone involved.
|Type of device||Device count||Percent of devices segregated by customer or provider||Percent of total devices|
One of the more surprising findings of this project has been the scarcity of locally developed or "do-it-yourself" examples (Table 9.5). Locally developed devices custom-made by process users are pervasive in industrial companies that have implemented mistake-proofing. The relatively few examples in health care may be partially explained by the fact that most industrial companies have a machine shop and tool and die makers readily available to fabricate any mistake-proofing device they need. To compensate, health care providers will need to develop external sources of expertise.
|Source of device||Count||Percent of total|
Although some mistake-proofing devices that will be needed in medicine will be created in-house or in an individual's garage or workshop, others will require more sophisticated design and production help. Competencies in inventive processes, design, fabrication, and assembly will be needed in some cases, and not all medical organizations will have these capabilities. These competencies usually will be found in engineering, maintenance, or biomedical engineering departments. In the absence of these departments, organizations must find other sources of supply.
One place to begin the search for help in developing a prototype for minimal cost is the engineering school at local colleges. Occasionally, engineering students may undertake projects as part of a class. Engineering programs will typically have two types of classes where devices could be designed and fabricated: "senior capstone design" courses and independent research courses. Organizations should expect to provide funding for required materials, but they may be able to avoid labor costs and profit margins. Squire6 suggests that:
... the school be physically close ... you want to be able to go there and explain the idea ... undergraduate engineers have a tendency to go off on their own, and without being available to see the development, you may end up with something very different than you envisioned.
Convincing an engineering school to adopt the project will also depend on the level of difficulty and whether the project requires a combination of competencies that would be beneficial to the students. This approach requires diplomatic treatment of intellectual property issues and commercial contingencies.
Karen Cox, a Patient Safety Improvement Corps (PSIC) 2004 participant, spoke of needing a farmer to weld a piece of equipment to solve a problem in the area of human factors and forcing functions.
The hooks that hold the containers in the infusion pump in Figure 9.1 are randomly arranged. One hook is occupied by a container that is connected to the smaller pump at left. The tubes are thoroughly tangled.
Karen Cox wanted a hook immediately above each of the infusion pumps so that it would be clear which medications were running through each of the four pumps (Figure 9.2).
If a device is not appropriate for an engineering class project, an organization should continue to explore its options. One possibility is to consider networking with local chambers of commerce or with members of civic organizations such as the Rotary Club or Optimist Club in order to develop contacts with local factory engineering managers. Engineering managers are likely to have experience obtaining custom tool design, fabrication, assembly, and installation in the local area. Local machine shops (sometimes listed under "Machinery-custom" in the phone book), metal fabricators, and systems integrators also can help.
Fabrication is an industrial term generally applied to the building of metal machines and structures. Fabrication shops and machine shops have overlapping capabilities, with fabrication shops concentrating on metal forming and welding. Go to: http://en.wikipedia.org/wiki/Fabrication_(metal).
Assembly is the stage of production in which components are put together into an end-product appropriate to the process concerned. Go to http://www.eyefortransport.com/glossary/ab.shtml.
A machine shop is a workshop where metal is cut and shaped by machine tools.
A systems integrator is an individual or company capable of making diverse components work together as a system. The word system usually implies the inclusion of a computer or microprocessor component to the project. Sources for more information include:
- A Directory of System Integrators in the Medical Industry for Factory Automation, Process Control, and Instrumentation is available at http://www.automation.com/sitepages/pid121.php.
- Medical DeviceLink—a Web site associated with the medical device industry provides a directory of North American Suppliers of Automation and Custom equipment and Software. Go to http://www.devicelink.com/company98/category/Manufacturing_Equipment_and_Software/AutomationCustom_equipment.html.
- Automation Resources Inc. offers "online resources for industrial automation, process control & instrumentation" at http://www.automationtechies.com.
- The Control and Information System Integrators Association (CSIA) provides a search feature that enables users to search for experienced CSIA member integrators according to industry, application, location, and service. Go to http://www.controlsys.org/about/member_directory.htm.
The CSIA also provides a free, two-volume guide to selecting and working with a systems integrator that covers most aspects of finding the right systems integrators, and highlights the nuances of navigating a project that otherwise might be initially overlooked. These are available at: http://www.controlsys.org/find/howto_guides.
C. Martin Hinckley's book, Make No Mistake! An Outcome Based Approach to Mistake-Proofing,7 contains extensive descriptions of, and supplier information about, sensors and other technologies that are useful in mistake-proofing.
The discussion in these nine chapter has introduced the concept of mistake-proofing and provided a rationale for using mistake-proofing to reduce errors in health care. It has also delineated a set of concepts, a vocabulary, and tools to assist organizations in taking action. This book contains 150 examples provided by the health care industry, as well as examples provided by manufacturing industries and people in everyday life. Anecdotal evidence indicates that, after they learn about mistake-proofing, readers are more likely to start noticing mistake-proofing examples around them and employ mistake-proofing to develop solutions. Gosbee and Anderson8 found that root cause analysis (RCA) teams who have been exposed to human factors engineering case studies often change their focus to "underlying design-related factors," such as mistake-proofing, as remedial actions. Initiating this change in focus is the goal of this publication.
As you complete Failure Modes and Effects Analyses (FMEAs) and RCAs or witness errors, you will envision new ways to solve problems and create novel mistake-proofing devices. As these ideas are implemented as locally developed mistake-proofing devices, please spread the news of their existence. Submit them as indicated below or publish them in some other venue so that others can benefit from the solution. Modesty, minimizing contributions, or assuming that others have thought of a locally developed solution does not serve the greater good. Some of the best mistake-proofing will be exceptionally simple and inexpensive. All solutions will be developed locally by someone before they become off-the-shelf solutions. Be that someone.
The examples presented here do not by any means represent an exhaustive listing of devices currently in use. Example contributions are welcome. Contribute mistake-proofing examples by visiting http://www.mistake-proofing.com and clicking on "Submit Example." Select the preferred submission method and add to the database of mistake-proofing examples. Comments on the devices featured in this book are also welcome.