Medical Device Link.

DESIGN

Image: iStockphoto

Finding the best design

Identifying usage errors is so crucial to risk-managed medical device development that several regulatory bodies have recommended incorporating Human Factors Engineering (HFE) into the risk management process. Because the purpose of HFE is to understand and optimise how people use and interact with tools, processes, products and environments, it is a natural and logical part of any risk management process. To identify known and foreseeable hazards, it is important to have a clear understanding of who will be using the device, for what purpose and where. HFE methods are best used throughout the product development process, from the early stages of concept development through to product testing and evaluation. This article describes three stages of the HFE process:

• Stage 1: Understand and define
• Stage 2: Develop, test and iterate
• Stage 3: Validate.

Stage 1: Understand and define

Before designing a product, it is imperative to understand

• the people who will be interacting with it
• the tasks for which it will be used
• the environments in which it will be used
• business, marketing and technical considerations

The information and knowledge gained in this stage can be used in many ways in the subsequent stages: from inspiring insights to establishing usability metrics with which to evaluate concepts. There are a number of techniques that can be employed to obtain this information, as described below.

Desk research

This is the review and analysis of secondary sources of information such as published reports, articles, books, blogs and databases. Depending on the product being developed, there may be a great deal of pre-existing data about the product. Rather than “re-inventing the wheel,” HFE specialists can gain an understanding of the historical context and current state of affairs by analysing what has already been done. For example, the United States Food and Drug Administration (www.fda.gov) and the United Kingdom Medicines and Healthcare products Regulatory Agency (www.mhra.gov.uk) maintain freely accessible databases that contain adverse incident reports for medical devices. This knowledge helps product designers and engineers avoid common problems with particular classes of devices and indicates an opportunity to innovate with new solutions.

In a recent design project, the company’s Human Factors (HF) team, while researching infusion pumps, discovered that there were many dosage errors resulting from nurses not entering decimal points correctly. Sometimes they entered 150 mg when they should have entered 1.50 mg, or vice versa. Depending on the therapy being administered, underdosing or overdosing could obviously result in injury or even death. Awareness of this problem highlighted an important risk factor that needed to be addressed during the design phase and also the requirement for possible user interface solutions.

Sometimes the client company has a great amount of information that can be reanalysed from a HF viewpoint, as was the case with the Human Growth Hormone (hGH) auto-inject device, easypod (Merck Serono www.serono.com). In a review of the client’s market research on earlier injection devices, a number of issues were identified that were extremely relevant to the new device. These were the users’ desire for ease of use, safe storage and the anxiety that parents and children experience about the injection process. These insights inspired the design team to include certain features and interactions that increased ease of use and decreased anxiety. For example, they created an interface that allowed the child and parent to manipulate aspects of the injection. These included “comfort settings” such as needle depth and injection speed, which gave them a sense of control and, in turn, reduced anxiety.

Field research

The field research technique involves observing and interviewing subject domain experts and users in the environments of interest. To understand what a medical device should do, how it should work, where it will “live,” who will touch it during its lifecycle, and what other products or devices it may interact with it, the HF team talks with the people who will interact with the device, directly observes their daily activities, and maps the communication networks and procedures that inform its use. This may include patients and clinicians, as well as those who will be involved with maintaining, monitoring, handling, installing, prescribing, evaluating or purchasing the device. For example, in developing the auto inject device, the team talked with endocrinologists and paediatricians to understand how the therapy works, and with patients and their parents to understand how it is administered. Researching the injection process, from the child’s and parents’ point of view helped the design team to empathise with a child’s fear and anxiety.

Figure 1: Visual block models of autoinject device were produced to test the versatility of handling and the “traffic light actuation button.”

They discovered that parents and children were unsure of when the drug injection was finished. Parents did not want to leave the device on any longer than necessary, because that would extend the time of
perceived pain and inconvenience. But, they also did not want to remove it before all the drug had been administered. This was for economic reasons: the drug is expensive and they do not want to “waste” it; and for therapeutic reasons: they want their child to benefit from the correctly administered dose. This insight inspired the design team to create a “traffic light” system that indicates three stages: when it is time to press the injection button (feedback that the device is pressed against the skin); when the device should be held stable to allow the drug to inject; and when it is time to remove the device (Figure 1). The feedback gives the parent and child visibility into the process and increases their sense of control. Children know when to “prepare” for the injection, which allows them to better manage their perception of pain. Neurologically, this phenomenon can be traced to brainstem and endorphin pathways that release the chemicals necessary to ensure the pain is physiologically masked.

Figure 2: Operational modes of hospital pump (Quantex).

In another example, during the development of an infusion pump, the HF team observed that the lighting levels in intensive care units varied between daylight (bright) and night (dark), and that at night, patients do not want to be disturbed by bright displays near their bed. Thus, the device needed to have a night-mode, in which only the most important information was illuminated and clearly visible from standard distances. Talking with the night-shift nurses, the team determined that the only thing the nurses needed to know from a distance was that the device was on. Once they were closer to the device, they could see more information by pressing the “light” button. These insights inspired the idea of different operational modes (day, night-standby and night-operational), which were introduced into a prototype and tested with representative users (Figure 2).

Doing research in the context-of-use evokes much richer information. It prompts people to mention things they may not have remembered outside the environment. It allows the HF team to analyse discrepancies between what people do and what they say they do, and to observe aspects of the environment that can feed into the device requirements such as lighting levels, dust and mobility.

Storyboarding

This technique involves illustrating in a sequential, narrative way how users interact with a product in certain environments or situations. Storyboards can be used as explanatory tools to show how things are currently done and how things could be done. Essentially, the sequential, narrative-style and illustrations (graphic or video) help people understand potentially complex interactions by telling a story. They are also useful tools for establishing information gaps.

Figure 3: A storyboard of the life of an oxygen cylinder (BOC Medical).

As part of the development of the CombiGuard oxygen cylinder (BOC Medical, www.bocmedical.com), the life of an oxygen cylinder was mapped to capture the different issues and interactions, from initial filling at the BOC plant, through transportation, storage and use in hospital, ambulance and home environments to return to BOC for refilling (Figure 3). These observations showed the variety of postural challenges faced by different types of users throughout the lifecycle of the product. This helped the design team identify the stresses and strains that users experienced whilst interacting with product, which in turn identified potential risk factors.

Stage 2: Develop, test and iterate

In this stage, the HF team translates its understanding of users, tasks and environments into requirements and then collaborates with designers and engineers to create concepts that meet the needs of the users and stakeholders. As concepts are created through a variety of means, including sketching, block models, wire frames of the user interface and prototyping, they can be evaluated and tested against the requirements, and the requirements can be updated to reflect what was learned from the evaluation. Techniques to be employed during Stage 2 are described below.

Rapid prototyping and testing

The approach here involves creating increasingly more realistic approximations of the final product so that concepts can be evaluated, improved and tested again. Concept ideas can be made more tangible in a variety of ways. These include sketches, computer aided design renderings, wire frames (similar to blueprints, but used to represent individual screens in a digital user interface), block models, Wizard-of-oz models,¹ and looks-like-works-like prototypes. The team members can then test the concepts by applying their expertise in interaction design to identify potential problems with them; in a sense they act as surrogate users and find the problems that users are likely to have with a product. Representative users can also be used to test concepts.

Because the goal at this point is to find problems with a product, it is not necessary to involve large numbers of users in the evaluation; once a problem is found, it does not need to be found again. Consequently, for most products, only six to ten users are needed. For more information on this topic, Jakob Nielsen’s research into usability testing methods and the “optimal” number of users to test is recommended reading.²

Figure 4: Early testing of foam block models and sketched concepts with parents and children.

During the design of the easypod autoinject device, several “block” models were created that showed the different shapes that the injection device could be configured into. This process helped clarify the technical and the mechanical possibilities, but it also helped to identify usability issues. The models were given to children and parents and the HF team observed how they held them, oriented them, and placed them against the skin (Figure 4). It was discovered that if the injection button was placed on the side of the device, most people tilted the device slightly to get a better view of the button; this behaviour resulted in more “wiggle” of device. This instability and movement during the injection process would result in a more painful injection by stimulating more nerve fibres. If, however, the injection button was placed on top of the device, people did not feel the need to tilt it, and they also tended to hold it in a position that was more comfortable and would result in less chance of a painful injection. It was also observed that different users held the device in different ways depending on if they were self injecting or injecting their child. Some preferred to hold the device in one hand and actuate with the other, some preferred to do everything with one hand. Thus, the shape and configuration of the device needed to allow both one-handed and two-handed operation. It also needed to be used by people with small hands (children) and large hands (adults).

Stage 3: Validate

Once the HF team is confident that most of the major problems with a prototype have been identified, a more robust, functional prototype can be taken into the desired environments and validated by users over a longer course of time (weeks or even months). These evaluations allow the designers and engineers to discover additional in situ risks that would not normally be seen in the types of quick usability tests conducted in Stage 2. Obviously, the device classification and the associated regulatory requirements determine the scope of these field trials.

Validation testing

Figure 5: Product being tested in an isolation chamber with laminar flow hood.

Validation testing involves observing and measuring how well or how poorly people can use a product for its intended purpose under the specified environmental conditions, over a period of time. For example, preparation of oncology drugs in hospital pharmacies is a specialised activity undertaken in environments that are unfamiliar to most designers and engineers, and may involve using isolation chambers and laminar flow hoods (Figure 5). It is important to test new devices intended for these environments in situ with people who have to perform these tasks daily to understand the pressures and potential safety issues.

Additional HFE benefits

Medical device development is highly regulated and necessarily process driven. New product development is an iterative process of design, testing and refinement; products are rarely right first time. Applying HFE practices throughout the process helps to reduce the number of iterations and plays a crucial role in the development of easy-to-use and risk-managed products. In addition, the insights uncovered during the HFE process can lead to product innovations that differentiate a product from its competitors.

---

Matt Pattison is Senior Human Factors Researcher, Heather McQuaid is Senior Manager and Interaction Designer, and Alun Wilcox is Director of Medical at PDD, a product innovation consultancy for medical/pharmaceutical and other markets, 85–87 Richford Street, London W6 7HJ, UK, tel. +44 20 8735 1111, e-mail: alunwilcox@pdd.co.uk, www.pdd.co.uk