How to KISS (Keep it Simple, Stupid)
Consider designing the readout first when developing point-of-need assays.
When I describe proof-of-concept results for a new point-of-need assay platform, people rarely ask about operational simplicity. Instead, they question me about cost, stability, mass production, and other issues that, which at least at the outset, are irrelevant to whether the assay is easy to perform and evaluate. In my view, these latter two metrics ought to be the highest priority at the proof-of-concept stage.
Ideally, a point-of-need assay will require only a single step by the user (adding the sample), function without specialized electronics or instruments, cost very little, provide sensitive and selective results in minutes, and give completely unambiguous readouts to anyone, anywhere, at any time. Rather than being a product of remodeling and simplifying existing assay platforms, perhaps this ideal type of point-of-need assay is achievable by first inventing new readouts for assays. After all, an assay that only a few trained individuals can interpret will have a much smaller global impact than an equally good assay with an output that is exceedingly simple to read and understand.
The need to improve the readout is especially true for quantitative point-of-need assays. The standard outputs of color, fluorescence, and electrochemistry are susceptible to contaminants in the sample that affect their sensitivity and reproducibility, and all require specialized electronic readers to enable quantification (specialized readers can be a deal breaker for end users in some point-of-need environments). New outputs, on the other hand, may offer new opportunities (1).
Distance-based measurements (for example, the distance that a colored sample travels on an assay platform) (2), bar-based measurements (for example, the number of bars that change color during an assay) (3), and time-based readouts are all emerging as alternatives to more traditional analytical signals for use in point-of-need assays (4, 5). Such readouts are simple and clear – the user need only compare distances, count colored bars, or measure time. They also take advantage of straightforward and familiar tasks, do not require specialized instruments to obtain a quantitative result, and are in the process of being generalized for a variety of classes of analytes.
Time-based readouts are exciting due to the high-resolution and accuracy of the quantitative measurement. Indeed, our research group has developed time-based readouts for paper-based assays, where the time for color to appear in one region of paper relative to another region correlates with the concentration of a target analyte. These time-based assays include sample pre-processing steps, signal amplification for trace level detection, and normalization of assays for effects of humidity and temperature on sample distribution rates in paper devices. The user does not need to know that these features exist, since he or she must only add a sample (of nearly any volume) to the paper, and then use a stopwatch to measure the time for two regions to become colored relative to one another. The assay is very easy to perform and the readout is straightforward. Moreover, the approach is compatible with detecting and quantifying small molecules, inorganic ions, enzymes, and proteins, and is capable of femtomolar detection limits. In addition, there are further opportunities for continually improving sensitivity through additional research.
Our time-based assay platform still is at the proof-of-concept stage, but it illustrates the new capabilities that can emerge when one designs the readout before developing the entire assay platform. For standard laboratory settings, it is logical to continue improving upon existing readouts and assay strategies. But the unique challenges associated with simple, inexpensive, point-of-need tests requires new – perhaps even backwards – approaches to designing assays. The concept “invent the readout first” is unconventional, but just might be backwards enough to circumvent intellectual or technological traps that often inhibit the development of effective point-of-need assays.
- E. Fu, “Enabling Robust Quantitative Readout in an Equipment-Free Model of Device Development”, Analyst 139, 4750−4757 (2014).
- D. M. Cate et al., “Simple, Distance-Based Measurement for Paper Analytical Devices”, Lab Chip 13, 2397−2404 (2013).
- Y. Song et al., “Multiplexed Volumetric Bar-Chart Chip for Point-of-Care Diagnostics”, Nature Communications 3, 1283 (2012).
- G. S. Lewis et al., “A Prototype Point-of-Use Assay That Measures Heavy Metal Contamination in Water Using Time as a Quantitative Readout”, Chem. Commun. 50, 5352–5354 (2014).
- G. S. Lewis et al., “A Rapid Point-of-Care Assay Platform for Measuring Femtomolar Levels of Active Enzyme Analytes Using Measurements of Time as the Readout”, Anal. Chem. 85, 10432–10439 (2013).
Scott Phillips started his independent career at Pennsylvania State University, Pennsylvania, USA, in 2008. His research interests include: (i) developing thermally stable detection and signal amplification reagents for use in point-of-care diagnostics; (ii) developing exceedingly inexpensive but high performance diagnostic devices for use in resource poor environments; and (iii) designing new classes of stimuli-responsive plastics that display amplified and autonomous responses for biomedical and environmental applications. He is a co-author of more than 60 papers, three book chapters and holds six patents.
