Feature Articles

Published: January 1, 2010
Find more content on:
Finding success through the IVD product definition process

The product definition process can overlook optimal solutions if existing practices or favored technologies are accepted without adequate questioning.

By: Rob Danby

The successful development of new IVD products depends on first establishing the right product definition. The product definition serves as a guide for the development project. It should contain an optimal product concept and a comprehensive list of requirements that help designers make choices throughout the development project and that allow marketers to understand what benefits the product will deliver and how it will fit into the market.
 
The product definition process precedes the start of the IVD development project and is the result of numerous tasks including broad stakeholder inquiries, observations, assessments, analyses, measurements, and selecting and applying expertise, creativity, and innovation. Various tools are available to support such activities. This article examines some of these tools and presents an example of how they can produce a breakthrough result.
 
A clear product definition sets expectations for the development project team and the marketing group, and facilitates communication with the product development investors. However, many factors can lead to a product definition that overlooks the optimal solution. For example, if an IVD company accepts certain existing practices or favored technologies without conducting sufficient questioning, alternative directions in innovation can remain undiscovered. Many companies do not frequently need to produce a new product definition, so they lack the skills and training to carry out such a systematic process. In some cases, a new product will take an IVD company beyond its core technical competence, while the product definition process itself requires a wide range of skills that a company may not have in house. As a new product's success hinges on the product definition process, engaging and working with a skilled outsource supplier is worth considering.
 
This article discusses the product definition process by breaking it down into three steps: understanding the real needs, exploring candidate solutions, and selecting an optimum solution. For each step, the article looks at the specific tools that produce systematic results and address the key questions that can change the course of an IVD development project. Using examples from an actual development project, the article shows how careful, unbiased application of the product definition process leads to an innovative solution and market success.
 
Product Definition Process

 

Figure 1. (click to enlarge) Three-step product definition process.
The product definition process comprises many tasks and can be divided up in many different ways depending on the innovation philosophy being followed and the market being addressed. This article follows a simple three-step model that has been used successfully in a number of IVD development projects (see Figure 1). Each step involves a list of tasks that can facilitate understanding needs, exploring solutions, and selecting a solution. Although the diagram in Figure 1 is linear, the product definition process can be iterative: testing a selected optimum solution with real users may reveal further improvements that could significantly affect the product definition.
 
For every IVD development project, the various tools associated with each step are selected based on their ability to collect data easily and their appropriate cost of use. In the example development project discussed in this article, the following tools are examined:
 
Step 1: Contextual inquiry, market gap analysis, needs-ranking matrix. 
Step 2: Functional analysis. 
Step 3: Concept-selection matrix.
 
Figure 1 includes a fourth step labelled “Develop,” which indicates that product development follows the product definition process.
 
Understanding the Real Needs

 

Figure 2. (click to enlarge) Market gap analysis is a graphical method of examining the market for gaps in benefits to users by plotting the performance of existing products on a series of two-axis graphs. This example shows throughput vs. cost per test to identify areas of importance that are not currently served.
This first step of the product definition process is aimed at clarifying the various needs and identifying the gaps in the IVD market that create an opportunity for a new product. Informal market research can often ascertain the possibility of a gap in the market and provide a spark of interest in developing a new product in the first place. A market gap analysis can be useful in more formally defining the opportunity space. For example, the market gap analysis for immunochemistry analyzers in Figure 2 shows throughput versus cost per test for 10 current systems on the market and identifies a couple of potential opportunities for further investigation.
 
After identifying the IVD market gaps, various methods such as quality function deployment (QFD), outcome-driven innovation, and others are used to capture and define user needs. Accurately capturing such users' needs is critical to product success.1
 
There are many different ways to gather information on IVD user needs, including the following: focus groups, individual interviews, and contextual inquiry. Such methods are based on structured interviews that focus on the customers' experiences with current products or precursor processes. Structured observation is also a powerful tool in contextual inquiry since users cannot always convert their actions into words. Needs statements are then extracted from the collected data and ranked in a weighted hierarchy.
 
A focus group is an interacting group of users of similar IVD products responding as a group to set questions designed to extract information about their current needs. The quality of the output from focus groups depends on the skills of the facilitator and the ability of reviewers to interpret responses.
 
Contextual inquiry is a method focused on user-centered IVD design principles, in which the developers engage with users at multiple levels to define and specify the requirements for a particular product. Such engagement aims at discovering user expectations, documenting met and unmet needs, and revealing data surrounding the context of use for the product. During such activities, developers use a structured framework to observe and interview users working in their places of employment. The results paint a picture of the context of use for the product, and identify where gaps are present or where additional services or features would be beneficial.2
 
In addition to user needs, certain real needs by IVD manufacturers emerge due to the involvement of other groups, which affect the development of a new product. For example, since complex systems will require maintenance and calibration, field service technicians have specific needs related to a new product. Regulatory authorities have defined minimum standards and specified design practices for IVDs, which must be complied with. Many new IVD products are being used at the forefront of research areas, so researchers may have needs that must be accommodated to secure the product's continued presence in the future. Finance and marketing departments may foresee a window of opportunity closing, so having a product launched in the market within a certain time period can often be a very real need, which influences the product definition and the product's ultimate success.
 
Figure 3. (click to enlarge) A needs trade-off matrix. When reading down the columns, a value of 2 means the column is more important than the row; a value of 1 means the column and row are of equal importance; and a value of 0 means the column is less important than the row.
Figure 3 shows an example of a needs trade-off matrix. Once a need has been defined, its importance relative to all other needs is determined by a one-to-one comparison with the entire set of needs. In this example, even though future proofing and development time are not user needs, they are real manufacturer needs that influence the product definition. While the scoring process can be arbitrary, it clarifies the consensus opinions and communicates to all parties the priority of each need. The weighted set of needs can then be prioritized and used in subsequent steps to guide and assess product concept development.3
 
The validity of the weighted needs in Figure 3 improves as the accuracy of the data on which they are based increases. However, collecting accurate data can be both time consuming and expensive. Balancing the need for accurate data against the cost of completing the activity is an important decision since the validity of the weighted needs affects the designer's approaches to product concept development.
 
By listing the needs criteria in order of importance, they can be used as a metric to assess different product concepts. In the needs comparison example in Figure 3, footprint has a lower weighting than cost of materials. Solutions that reduce the cost of goods at the expense of a larger footprint will receive a higher and therefore better assessment score.
 
Example IVD Development Project. The new product developer Applied Imaging Corp. (San Jose, CA) was working on perfecting software capable of reliably inspecting slides of tissue biopsies for various biomarkers and tagging the location of abnormalities for later assessment by pathologists. This system represented a significant advancement in cytology and cancer diagnosis. A pathologist would only need to inspect those locations on slides the software tagged as abnormal. The system could also greatly increase the productivity and effectiveness of biopsy assessment for cancer detection.
 
The system that Applied Imaging envisioned comprised a standard microscope platform, a charge-coupled device (CCD) camera, and computer hardware. The essence of this product's advantage was almost pure knowledge running on standard equipment.
 
However, one important need remained unresolved. If the desired efficiency gains from this system were to be delivered, there needed to be a reliable way to retrieve the slides from storage, place them on a microscope platform, and return them without human assistance. In other words, an automated method for managing and handling slides was needed.
 
Applied Imaging's core skills are biomarker detection and image analysis. No off-the-shelf slide handling system was available that could manage a work shift's worth of slides, position them on a microscope platform for imaging, and retrieve them on command at a later time. The lack of such a system was in effect a market gap. A key question was, “Does this project go beyond the core competence of Applied Imaging's in-house development group?”
 
After checking with various slide-handling system manufacturers, Applied Imaging decided to work with Invetech (Melbourne, Australia), an IVD product development company with a track record in slide-handling projects, such as automated slide stainers, cover slippers, and slide-based IVD projects. Invetech's track record demonstrated a level of competence in automated slide-manipulation projects. Invetech's assignment was to design a product that automates slide handling for one entire work shift and provides the necessary precision handling and high reliability for unattended operation (see Figure 4).
 
Figure 4. (click to enlarge) The addition of slide-warehouse and -handling capability to the inspection microscope to achieve a high-throughput system.
Invetech engaged in contextual inquiry for Applied Imaging to identify the various needs. While more complex products could require more rigorous analytical tools, the possible alternatives for this project were the following: QFD that begins with conducting in-depth structured interviews at the actual workplace, Kano analysis that follows a specific questionnaire on customer requirements, and outcome-driven innovation that emphasizes careful needs analysis and structured interviewing.1,4-7 In the product-definition process, the time and costs required for such different approaches determine the eventual tool choice. The investment in this phase of the product development project must be in line with the overall budget and schedule.
 
The contextual inquiry carried out for Applied Imaging produced many useful observations, including the following:
 
• Pathologists lay slides flat on a bench in a horizontal plane format from which they select slides for inspection.

• Bench space is often minimal.

• In those cases for which slide handling has been automated elsewhere in the work environment, the slides are almost always warehoused in racking systems with the slides placed in stacks.

• High throughput means at least one work shift of unattended operation.

• The ability to relocate the abnormal image on the slide is essential for product acceptance.

A key contextual inquiry observation noted that in numerous pathology labs, slides were stored in a planar format and were almost always selected from this flat format prior to microscope inspection. Pathologists preferred to manage their slides by having the slide face visible rather than on its edge. However, in pursuing a smaller footprint, conventional slide warehousing is designed around stacking the slides on their edges, which hide the slide faces. This observation encouraged the designers to explore a product concept that breaks with convention by sacrificing footprint to accommodate insights gained from the pathology labs.

 

Exploring Candidate Solutions

 

For this second step of the product definition process, a large number of inventive tools are available. One such tool is TRIZ, a theory of problem solving that originated in the former Soviet Union in the late 1940s and is based on research in innovation and invention.8,9
 
TRIZ is a systematic process and an extensive and evolving series of innovation tools that are currently gaining global acceptance. Based on extensive research into various inventions and patent registers, the TRIZ philosophy is focussed on the following key findings: 
 
• Innovations emerge from a small number of inventive principles.

• Technology trends are highly predictable.

• The best solutions transform negative elements into useful resources and reduce the need to trade-off positive elements by eliminating underlying conflicts.

The TRIZ toolbox contains eleven tools for generating solutions. The choice of tool depends on the type of solution being sought. The TRIZ philosophy is based on patterning and seeking solutions by examining similar problems.10

What other methods like TRIZ already exist? Patent searches, trade literature searches, reviews and specifications of other similar equipment, and similarities in previous projects provide creative inspiration and information on accepted solutions
in comparable situations. Seeking out expertise on specific development points is also a route to new solutions.

For complex systems, functional analysis is a powerful tool. This analysis looks at the product or system from the outside in and establishes the inputs, outputs, interfaces, and functional blocks from the outside to the heart of the system. The complex function required is made tractable by reducing it to an interconnection of very simple functions and following the input/output paths. The aim should be to draw a functional block diagram with as many of the interfaces as possible in a standard form. Each functional block should do a single job and should be described independent of a solution to achieve that function. Those interfaces in which non-standard form is essential have potentially far-reaching effects on the form of a viable concept.

 

Figure 5. Simplified functional analysis for a slide imaging system.
Example IVD Development Project. The functional block diagram for the automated slide imaging system is shown in a simplified form in Figure 5. In this system, all the interfaces between the blocks are either optics, power, or data, except for the functions involved in transporting the slides. While power, data, and optics interfaces can be designed based on industry-standard formats, the slide transport path needed additional effort. Such unusual inputs, outputs, or interfaces are areas that require investments in development.
 
By investigating existing slide-movement technologies available on the market, the product developers discovered a piece of unspoken designer's knowledge that slides should be moved only by the edges whenever possible. Slide movement systems commonly use gripper fingers, pusher arms, and racking systems in which slides lay only on their edges.
 
The glass for the slides is first formed into thin sheets, the sheets are then cut to size, and the edges are ground. But handling slides by the edges produces small amounts of glass dust that accumulates due to the slides being moved around hundreds of times and, in combination with complex mechanical systems, results in unreliability, maintenance issues, and service calls. A key question was what happens to reliability if glass dust accumulation is reduced or eliminated.
 
Figure 6. (click to enlarge) Edge grippers vs. face suction for moving slides.
Since thin sheet glass was introduced to replace mica, the purpose for handling slides by their edges has been to avoid having fingerprints in the viewing area. However, is this edge-manipulation practice essential for an automatic slide handler in which no fingerprints will be left and direct contact with the viewing area of the slide can be avoided without the constraint of edge-only handling? Another key question was, “What if this edge-only constraint is lessened?”
 
The answers to these questions led the system designers to a radical new concept for the slide-management system. The proposed slide movement solution involved a suction cup which contacted the cast surface of the slides, thereby avoiding the edges and allowing the slides to be stored in trays instead of racks (see Figures 6 and 7). This solution is the same kind of planar layout that lab pathologists use to select slides for viewing. This alternative concept in slide actuation and storage led to two types of concepts for the slide-handling robotics: one based on vertical stacks and another based on horizontal trays.
 
Selecting an Optimum Solution

 

At the end of step 1, a list of weighted needs was produced. At the end of step 2, a set of possible design concepts was considered. This final step of the product-definition process evaluates the concepts and decides which solution best satisfies the established needs.
 
Figure 7. (click to enlarge) Two families of solution possible based around the storage format and slide actuation principle used.
This objective can be achieved by assessing each concept individually for how well it meets each need. In Figure 8, each need is rated on a scale of 1 to 5. Each concept's overall score can be determined by multiplying the weighted needs by the assessment scores and then adding up the results for each need.
 
Example IVD Development Project. In the case of the automated slide-handling system, two groups of solutions were possible depending on the storage format for the slides. In assessing variations of these groups, the solutions based on a horizontal tray consistently outscored the ones based on a slide stack.
 
Development Outcome

 

Figure 8. (click to enlarge) Individual concept assessment format.
The choice of a planar horizontal storage and suction cup system instead of finger grippers allowed a simple two-axis robot system to actuate all the movements necessary to transfer the slides from the warehouse to the precise location on the microscope platform. With less complexity, the project finished in 13 months, and units were ready to ship in advance of regulatory approvals for the imaging software.
 
Users embraced the horizontal format, and the target of zero slide breakages was met. This product continues to be a market leader eight years after its launch and has been adapted to several other microscope configurations. As the design intended, handling slides using a suction cup generated no glass dust, and consequently the instrument has developed a reliable reputation. In Figure 9, a pathology lab technician loads a tray of slides while the system continues to run. This “load while active” feature supports the original aim of freeing up more operator time.
 
Figure 9. The final product.
During the product definition process, a key question was, “What happens if an IVD company breaks away from market-established practices?” Accepting a larger footprint and the unconventional handling and storage of slides in order to make gains in reliability, usability, design simplicity, and reduced time to market resulted in a successful product. The outsource provider was able to introduce unconventional ideas by applying a systematic product definition process and break with the conventional wisdom that in the lab, smaller is always better.
 
Conclusion

 

High-quality answers to the right key questions are important to achieving success when developing new IVD products. Using structured tools and a systematic process and working with an outsource development partner can provide a clearer path to the right questions and answers, leading to better initial product definition, shorter time to market, and products that lead rather than follow the field.
 
 
References

1. A Ulwick, (New York City: McGraw-Hill, 2005).

2. H Beher and K Holzblatt, (London: Academic Press Harcourt Place, 1998).

3. S Pugh, D Clausing, and R Andrade, (Reading, MA: Addison-Wesley Longman, 1996).

4. JR Hauser and D Clausing, “The House of Quality,” , May-June 1988: 63 -73.

5. Y Kai, (Six Sigma Operational Methods, 2007).

6. K Matzler and H Hinterhuber, “How to Make Product Development Projects More Successful by Integrating Kano's Model of Customer Satisfaction into Quality Function Deployment,” 18, no. 1 (1998): 25-38.

7. E Sauerwein, et al., “The Kano Model: How To Delight Your Customers,” International Working Seminar on Production Economics, Innsbruck, Austria, February 19-23, 1996.

8. DL Mann, (Devon, UK: Lazarus Press, 2007).

9. G Altshuller, (Worcester, MA: Technical Innovation Center, 1994).

10. J Terninko, A Zusman, and B Zlotin, (Boca Raton, FL: CRC Press, 1998). 5

 

 

 Rob Danby is manager of diagnostics at Invetech (Melbourne, Australia). He can be reached at rob.danby@invetech.com.au.
 
 
 
 
 
 
 
 
 
 
 

 


Your rating: None Average: 3.5 (2 votes)

Login to post comments