**Table 1.**

*A summary of several different microfluidic substrate types and their respective characteristics.*

*Perspective Chapter: Microfluidic Technologies for On-Site Detection and Quantification… DOI: http://dx.doi.org/10.5772/intechopen.105950*

includes whole-sample mediums, such as whole blood, saliva, and mucosal liquids, a potential biohazardous risk to the collector could exist [24, 25]. These potentially biohazardous samples can be collected by trained personnel swabbing the patient's sinuses or mouth, as well as through the drawing of blood [25]. These collection techniques are used to require trained operators to wear personal protective equipment (PPE) to minimize potential transmission and required processing off-site. In the early stages of the pandemic, it was common for diagnostic tests being administered by a healthcare professional wearing PPE to take more than a week to produce results. Many of these early tests still required millilitre-volume samples that lead to lowered detection-sensitivity issues [17]. Conversely, in the later stages of the pandemic, more and more diagnostic tests were being self-administered, and taking less time, as diagnostic technology continued to advance. However, these self-administered tests were often restricted to qualitative detection methods that could not yet provide a measure of an individual's "active infectiousness". In theory, microfluidic technology is expected to allow for a single portable microfluidic analyzer to both qualitatively and quantitatively diagnose large numbers of samples, with only one operator.

An equally important feature of microfluidic systems is its ability to facilitate on-site and on-device pre-treatment of samples. Some form of molecular separation or enhanced sample concentration was usually required to accurately process samples for diagnostic purposes [2]. In addition, this pre-treatment usually required some equipment that was both non-portable and difficult to operate. However, microfluidic technology has allowed for some of these pre-treatment techniques to be automated and contained within the test itself, for improved on-site capabilities [11]. The separation of unwanted molecules, from the sample, and the enhanced concentration of the target analyte could now be performed on-device and increase accuracy and sensitivity. One way in which microfluidics can facilitate this is by introducing a mixture of channel designs and reagents into the microfluidic device to initiate this pre-treatment of samples, directly inside the microfluidic device. The self-containing design principle intrinsic to most on-site microfluidic systems readily facilitates this feat and can be integrated into a myriad of diagnostic techniques such as qPCR [16]. Furthermore, by integrating these on-device pre-treatment steps into each individual test, the microfluidic analyzer would be able to capture and process data in real-time. Thereby, reducing some operational costs while improving the assay's sensitivity and limit of detection (LOD) significantly. By enhancing the modularity of diagnostic tests through microfluidic integration and a portable microfluidic analyzer, not only can the number of operators and trained personnel be reduced but can also provide quantitative analysis.

#### *4.2.1.3 Logistics*

Diagnostic platforms integrated with microfluidic systems may offer advantages over other non-microfluidic platforms such as portability, size, structural integrity, and reproducibility that can reduce logistical costs associated with the mass-production and distribution of tests. One of the most important measures of on-site diagnostic validity is the ability to efficiently reproduce, store and distribute tests without affecting quality or accuracy of the device [19].

Paper-based and polymer-based microfluidic designs offer a greater amount of reproducibility and modularity compared to non-microfluidic tests [2]. The moulds used to imprint upon the paper and polymer substrates in microfluidic

#### *Perspective Chapter: Microfluidic Technologies for On-Site Detection and Quantification… DOI: http://dx.doi.org/10.5772/intechopen.105950*

devices can be readily fabricated through methods such as lithography and etching [2]. These fabrication methods allow for microfluidic devices to remain relatively small and highly reproducible. Despite the moulds requiring expensive equipment to fabricate, only a few moulds need to be made to continuously produce a large number of microchannels for the intended devices. Thus, testing devices can be manufactured at high capacity. By separating the testing platform and the analyzer, one can significantly reduce costs associated with quality control and large-scale production.

Storage of the diagnostic tests depends on both the reagent lifespan used in the tests as well as the rigor of the test itself. Reagents used in diagnostic analysis often have expiration dates that are meant to limit quality control issues in the reagents themselves, which may lead to inaccurate testing results [22]. In general, most diagnostic reagents have a recommended shelf-life of 1 year, which would provide ample time for a microfluidic-based device to be stored and distributed with little worry. The additional rigor added to diagnostic devices through microfluidic integration refers to the thermostability and structural integrity of the devices themselves. If a test must be stored and transported in a temperature-controlled environment, the logistical costs for transporting those tests would significantly increase. Similarly, if the tests are not structurally sound, the need for more delicate transportation would also add to the costs of storing and transporting the tests. Thus, microfluidics offers a useful advantage in both scenarios as the self-containing principle behind ideal microfluidic diagnostic devices, can be applied. Therefore, through microfluidic integration, these important logistical metrics may be readily met, promoting continued massproduction of diagnostic tests.

### *4.2.2 Decreased turnaround time*

Prior to the advent of the technological revolution of the 2000s, microfluidic technologies were extremely limited in their automation capacity due to many of the automation-driven adaptations being hindered by the cost of computational processing, the size of the equipment required to process the experiments, and the time required for the experiment to complete [21]. The increased accessibility for computational processing and digital analysis over the last two decades has allowed microfluidics to achieve significant improvements in diagnostic technology, such as increased automation and high throughput analysis. Turnaround times in both microfluidic and non-microfluidic diagnostic tests have significantly improved over the last 2 years, from times greater than 1 week to less than 30 minutes [25]. Microfluidic automation and high throughput processing improved transmission prevention, and potentially improved treatment outcomes might be possible by more accurately diagnosing SARS-CoV-2 at an earlier stage.

#### *4.2.2.1 Automation*

Microfluidic integration of digital automation can allow for several simultaneous processes to occur, which can exponentially decrease the time it takes to analyze a certain number of samples. Greater computational processing power and more robust software and hardware have provided microfluidic technologies with the means to explore previously unfathomable feats in diagnostic testing [23, 24]. Inconsistencies such as human error in sample analysis or device fabrication can more easily be avoided, further reducing costs to the producer. This in turn

will create a greater incentive for microfluidic research, further driving down prices and effectively creating a feedback loop. Equipment can be coded to operate machinery, treat samples, and even provide qualitative and quantitative insight that would have been difficult for operators to observe. Through automation, the reliance on manual operation, experienced operators, and expensive equipment for diagnostic techniques is severely mitigated and can be adapted to process samples at speeds impossible for humans to do, otherwise known as high throughput processing.

#### *4.2.2.2 High throughput processing*

The benefits of high throughput processing achieved through microfluidic integration can be easily observed in microfluidic-based techniques such as droplet microfluidics. In droplet microfluidics, hundreds if not thousands of individual droplet microenvironments can be formed to initiate microscopic bioreactions in series, which can be later processed or analyzed in parallel [23, 27]. The processing speed in which high throughput analyses works, coupled with the micro-scale volumes used in microfluidics, presents a unique advantage that is not easily achieved in other detection methods [24]. Therefore, by integrating high throughput analysis through microfluidics, exponentially shorter turnaround times can be achieved.

### *4.2.3 Accuracy and LOD*

The effectiveness of conventional non-microfluidic diagnostic techniques can often be hindered by their limit of detection (LOD) and their accuracy [19]. Microfluidic devices can, in some cases, overcome these limitations. Accuracy is defined by the ability of the test to correctly discern between a state in which a target condition is met and when that target condition is not met [2, 19]. By establishing benchmarks for accuracy in non-microfluidic diagnostic techniques, we can compare it to the accuracy seen in similar microfluidic techniques. In many situations, since microfluidic integration does not necessarily change the detection method primarily used, the accuracy might not significantly improve. In most immunohistochemical assays, the accuracy of a diagnostic test varies significantly with differences in viral loads [17, 18]. As the concentration of virus particles increases, so does the accuracy of most immunohistochemical tests, not necessarily in a linear manner. Conversely, through microfluidic implementation, a pre-treatment step can be applied to concentrate and separate the target molecules from other unwanted material into standardized volumes [22]. In doing so, the variance in accuracy from samples with low viral loads can be mitigated. The modularity of microfluidic systems may offer a means to improve the LOD, as well as improved accuracy.

Target molecules at low concentrations found in macro-scale volumes of sample fluid are not always readily detected by non-microfluidic devices [23–25]. This limitation of non-microfluidic devices becomes more important when dealing with molecules that are not easily replicated by conventional means, such as protein biomarkers found on the surface of a SARS-CoV-2 cell. As a result, single-cell-resolution droplet microfluidic systems can be designed to generate thousands of microlitre droplet microenvironments to capture biomarkers, in extremely low concentrations, normally too difficult to detect through conventional means [27].

*Perspective Chapter: Microfluidic Technologies for On-Site Detection and Quantification… DOI: http://dx.doi.org/10.5772/intechopen.105950*

**Figure 7.** *The 7 pillars for assessing effective on-site diagnostic devices. Adapted from Ref. [2] with permission.*
