The development of point-of-care medical diagnostic devices
Much has been written about the impact of COVID-19. It has significantly altered how we live our lives and made us question the balance of prevention vs cure in our healthcare systems. With the need to track the virus, the importance of rapid diagnostic testing and rapid point-of-care (PoC) testing is becoming increasingly apparent.
Outside of the immediate pandemic, PoC testing has the potential to bring benefit to patient, clinician and payer in the right application. In current times, there is, unsurprisingly, an even greater interest in the development of new diagnostic platforms.
These new platforms are disruptive products, allowing tests to be run outside of the traditional lab setting, and, while this decentralised approach can be more convenient for the patient, it raises questions around who the user is, the quality control of results and complications for the end payer. To develop a successful new product not only requires an enabling technology, but also foresight into how these products will be adopted and used and paid for. As the focus on diagnostic testing continues to evolve, KD has been considering what it takes to develop a successful PoC diagnostic device.
New assays or detection technologies are often the driver for a diagnostic platform, most usually targeting an improvement in accuracy, cost, speed, throughput or multiplex capability. However, in the US, where the patient is used to paying directly for healthcare, they are increasingly predisposed to the notion of choice and as companies begin to target the patient directly, success will depend on the marketing and ease of use of the tests themselves. Because of this, companies need to place more focus on the user experience, keeping costs in check through targeting features around key functions.
In many situations, this may mean a close look at the sample collection kit itself, which is often seen as secondary to the test, but is the user’s main ‘touch point’ with the product. This is particularly evident in the current pandemic where the sample collection is not an easy thing to do and the patient often requires several attempts before getting a sample successfully. This may drive an approach to kitting which could involve the inclusion of multiple swabs for one patient.
The need to understand the application is also wrapped up in the business model. In the current pandemic, for example, it is unlikely that a (relatively expensive) PoC test is going to be the answer to the majority of testing as it is almost certainly more cost effective to have people self-isolate while waiting a day for a result from a lower cost, high throughput lab facility. However, there may still be significant markets for PoC testing where there is value in an immediate result such as for airport screening, or in private testing markets for companies or sporting events. Again, understanding the use of the device will drive the features and presentation of the product.
A particular challenge in a diagnostic product development is the degree to which the instrument, disposable and assay interact with one another. The disposable complexity in particular will be dictated by the assay process steps and, for assays with a greater complexity, a good starting point is to map out the assay process at an early stage and use this as a tool for driving the development – from the sequence and timings that the instrument has to implement through to the number of reaction chambers and reagents on the disposable.
This document then becomes the main interfacing tool between the instrument, disposable and assay teams, and often plays a key part in finding the right compromise between assay complexity and the disposable cost.
Slight changes in the assay can have a significant impact on the disposable complexity and cost, and yet the assay cannot be fully realised without a platform to test it on. This can create something of a circular dependency. With this in mind, the transition of the assay from manual ‘benchtop’ process to something more integrated will often benefit from the development of simple custom ‘labware’ that, while operated manually, can represent the reaction chambers and format of the final intended disposable. These bits of labware can be used to test the viability of the assay in the ‘disposable format’, explore simplifications to the assay process, test for material compatibility, optimisation of the assay and be used to conduct early reagent stability tests before a full disposable is available. Effective use of this kind of labware can mean that the integration of the assay onto the final disposable is a straightforward affair in the later stages of the development.
The disposable is the primary interface to the assay but, perhaps more importantly, can be a significant part of the consumable costs and often the reason that a PoC test is inherently more expensive than a test run on a higher throughput instrument. As such, it is central to all the trade-off decisions in the system and will influence all aspects of the assay performance from speed and sensitivity through to the repeatability and shelf life. Ultimately the disposable design will quickly ‘bake in’ the cost of the disposable early in the development and an overly complicated disposable will be difficult to cost-reduce later in the programme, no matter what volumes it is manufactured in.
To prevent the costs of the disposable rising, early focus needs to be on how it implements the assay process, which will influence the format and complexity of the disposable. A lot of work in the early stages needs to be a collaborative process of translating the assay process into steps that can be implemented on the disposable, and testing these experimentally.
The detection system
The detection system is core to the generation of results data. Establishing a mature detection system early on in the development will yield many benefits by setting a ‘standard’ by which all assay optimisation can be measured against. This means that the development team can make ‘apples to apples’ comparisons between different reagent formulations or processing parameters throughout the development. It also allows the team to start to build a library of raw results data that can be used as a ‘training’ data set for later stage optimisation of any data processing algorithms or threshold evaluations.
The successful mix
A successful development is often underpinned by the degree to which the teams are happy to engage in a collaborative process in the early stages of the development. This requires a combination of the right skills, from the assay development, fluidics and disposable design, through to the software and electronics and mechanical design of the instrument, but also the degree of overlap that allows the different groups to understand and discuss trade-offs in the system. Ultimately, much of this comes down to the mindset of the team and a collaborative approach to the process rather than giving into the temptation to design the separate subsystems in isolation.