PATHPOD – A loop-mediated isothermal amplification (LAMP)-based point-of-care system for rapid clinical detection of SARS-CoV-2 in hospitals in Denmark

Sensitive and rapid detection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has been a vital goal in the ongoing COVID-19 pandemic. We present in this comprehensive work, for the first time, detailed fabrication and clinical validation of a point of care (PoC) device for rapid, onsite detection of SARS-CoV-2 using a real-time reverse-transcription loop-mediated isothermal amplification (RT-LAMP) reaction on a polymer cartridge. The PoC system, namely PATHPOD, consisting of a standalone device (weight less than 1.2 kg) and a cartridge, can perform the detection of 10 different samples and two controls in less than 50 min, which is much more rapid than the golden standard real-time reverse-transcription Polymerase Chain Reaction (RT-PCR), typically taking 16–48 h. The novel total internal reflection (TIR) scheme and the reactions inside the cartridge in the PoC device allow monitoring of the diagnostic results in real-time and onsite. The analytical sensitivity and specificity of the PoC test are comparable with the current RT-PCR, with a limit of detection (LOD) down to 30–50 viral genome copies. The robustness of the PATHPOD PoC system has been confirmed by analyzing 398 clinical samples initially examined in two hospitals in Denmark. The clinical sensitivity and specificity of these tests are discussed.


Introduction
The pandemic of coronavirus disease 2019 (COVID-19) has been ongoing since early January 2020, initiating from the outbreak in Wuhan, China [1]. We started our CORONADX project in April 2020 (https://coronadx-project.eu/), an EU-funded project to develop a new SARS CoV-2 detection point-of-care (PoC) system requiring minimum or no training of the operators and eliminating the need to send samples to central laboratories. More rapid and widespread tests would be crucial to mitigating the COVID-19 pandemic and limiting lockdowns [1][2][3][4]. Our goal has been to provide systems that will give a reliable response in less than 60 min. In the first year of the pandemic, the vaccines for COVID- 19 were not yet available [5]. Limitations in vaccine supply still remain an issue. New variants continue to emerge, such as B.1.617 (delta) [4], and currently Omicron [6] (BA. 5). Thus, rapid and highly sensitive detection still plays an essential role in quarantine and isolating the cluster to prevent the spreading of the highly contagious virus and its variants [7][8][9]. The current standard detection method for SARS-CoV-2 is real-time reverse-transcription polymerase chain reaction (RT-PCR) [10][11][12], which can take up to  h to deliver the results to the patients due to the cumbersomeness of the sample transportation and laboratory work [1], (Fig. 1-i). PoC testing, which can perform the test at sites rapidly and cost-efficiently, is a potential candidate to overcome the drawbacks of conventional RT-PCR [13].
Loop-mediated isothermal amplification (LAMP) reaction is a novel nucleic acid amplification technique that amplifies DNA with high specificity, efficiency, and rapidity under isothermal conditions. This method uses a set of four specially designed primers and a DNA polymerase with high strand displacement activity to synthesize up to 10 9 copies from target DNA in less than an hour at a constant temperature of about 65 • C [14][15][16][17][18]. The reaction releases a byproduct named pyrophosphate (P 2 O 4− 7 , see reaction 1, Fig. 1-iii), which can be used for monitoring the reaction as it can react with Magnesium ion (Mg 2+ ) to form the precipitation of Magnesium pyrophosphate (reaction 2, Fig. 1-iii) [14].
Here we report the study, development, use and clinical validation of the PoC testing device (PATHPOD) to detect the SARS-CoV-2 virus based on the real-time reverse-transcription loop-mediated isothermal amplification reaction (RT-LAMP) and lab-on-a-chip technology [19][20][21][22]. The electronic part of the device was developed using open-source hardwares and softwares, which are readily available even in low-resource regions [19]. The device weighs approximately 1.2 kg and can be used as a standalone or with a computer, preferably to be operated at remote places such as airports, nursing homes, mobile labs, etc. (Fig. 1-ii). The device can operate as a sample-in-result-out for less than 1 h compared to an overall turnaround time from swab sampling to reporting results to the clinicians of 2-3 days of the current standard RT-PCR carried out at central labs. The cartridges used with the device are made from polymer, using injection moulding, and have 12 wells for ten samples and two controls (positive and negative controls). The COVID 19 assays were first developed and demonstrated in tubes and then adapted to the cartridge and the PoC device. The PoC system was then validated using samples obtained from two hospitals in Denmark.
For instance, Chaouch et al. [23] provided a comprehensive review on the PoC detection techniques for SARS-CoV-2, discussing the advantages and limitations of various methods including LAMP, CRISPR, and antigen-based techniques. While the review highlights the rapidity and simplicity of LAMP-based PoC devices, it emphasizes the continuing need for advancements in terms of real-time detection and ease of operation.
Pang et al. [24] introduced a LAMP-based diagnostic platform that provides results within 45 min, but this platform requires technical expertise to operate and does not offer real-time monitoring. Similarly, Zhang et al. [25] reported a mobile laboratory that utilizes LAMP for SARS-CoV-2 detection, yet its system also lacks real-time monitoring and requires trained personnel for operation.
Fozouni et al., [26] developed a PoC platform named Combinatorial Arrayed Reactions for Multiplexed Evaluation of Nucleic acids (CAR-MEN) which used LAMP for detection but lacks ease of operation and rapid detection. Ganguli et al., [27] designed a device named ampli-LUTE which, despite its potential for scalability, requires pre-processing of samples and lacks real-time detection. Huang et al., [3] presented a microfluidic-based PoC device but lacks real-time data analysis capability and requires separate extraction of RNA.
Our work offers significant contributions to the existing literature by addressing these limitations, as we present a LAMP-based PoC device that provides real-time detection in an easy-to-operate manner, requiring no technical expertise.
Quick and inexpensive home tests for COVID-19 detection, specifically antigen tests, have become widely available and provide results within minutes [28]. While these tests offer convenience and rapidity, they generally have lower sensitivity and specificity compared to molecular tests [28,29]. Furthermore, antigen tests primarily detect the presence of viral proteins, which could be less effective in identifying new virus variants [30].
In contrast, PATHPOD, based on the RT-rLAMP technique, exhibits high sensitivity and specificity, making it a more reliable detection system [15]. Additionally, it offers real-time monitoring capabilities, an advantage not provided by most rapid antigen tests. Furthermore, its molecular-based detection method provides an advantage in terms of variant detection, as it can be rapidly adjusted to detect new genetic sequences associated with emerging SARS-CoV-2 variants.
There are other works reported on LAMP reaction for the detection of SARS-CoV-2. However, these works were not in PoC diagnostics perspectives and could not monitor the reaction in real-time [31][32][33][34]. Recently, Ganguli [27] reported a lab-on-a-chip system which used LAMP reaction and fluorescence for end-point reading detection of SARS-CoV-2. Ganguli's work has a limitation is that it was not only for an end-point detection but also the samples in their works were not clinical samples. In this line of work, i.e. using fluorescence for pathogen detection, we reported [35] recently another completely different prototype at lab scale (not a PoC device).
It is vital to emphasize that our current device and investigation in this work are the first comprehensive and detailed study on a LAMP PoC device for real-time, rapid and cost-efficient detection of the SARS-CoV-2 virus. The device described in this manuscript can perform real-time and onsite detection. To make the manuscript compact, the main manuscript has nine figures and is supported by 13 extended figures, 2 Videos (presented separately for easy access while reading the paper), plus the Supplementary materials with five figures and two tables.

Cartridge fabrication and working principle of the device
Fig. 2-i shows the 3D design (side, top and back views) of the polymer cartridge (having a diameter of 50 mm). Extended Fig. 1 shows the polymer injection molded step and cartridge assembling components, Video S1 in the Supplementary material shows the injection molding process. Fig. 2-ii shows the components of the cartridge after assembling before filling the reagents and samples (see extended Fig. 2 and Video S2 for more information on how to fill the reagents). Fig. 2-iii illustrates the working principle of the cartridge when it locates inside the PoC device. The refractive indices of air and COC polymer are 1 and 1.53, respectively. As a result, according to Snell's law [37,38], the critical angle between these two media is 40.8 • . The angle of the pyramid structure within the cartridge is designed to be larger than 40.8 • . In this case, it is 41.5 • . The incident light (from a LED light source, λ = 535 nm) is hence deflected via total internal reflection (TIR) at 90 • into the plane of the sample well where the light is absorbed due to precipitation of 4is a byproduct of the RT-rLAMP reaction (see  After passing through the sample well, the light is deflected upward, and the intensity of the light is measured using a phototransistor. The reduction of obtaining light intensity is calculated and plotted in real-time. The detection principle is thus based on light diminishing as the result of a precipitate forming (magnesium pyrophosphate) during the amplification of DNA. Green light (535 nm) is sent from a LED through the sample, and a phototransistor (BPW-77NA) is used to measure the intensity in real-time, sampling one time per second. Optical structures ( Fig. 2-iii), embedded directly in the transparent polymer of the injection moulded part of the PATHPOD cartridge, deflect the light coming from the top, in the plane of the cartridge, in 4 directions, see Fig. 2-iii and iv. By sequentially turning on and off the 6 LEDs and 5 phototransistors, it is possible to sample and detect all 12 wells individually.

PoC device
Fig. 3-i shows a schematic of an exploded view for the PATHPOD device. Fig. 3-ii is a photo of the PATHPOD prototype. Detail on the fabrication and electronic part of the PATHPOD can be seen in extended Figs. 3 to 7 (including designs, circuit drawing in Fritzing, printed circuit boards and final products after assembling). The users can follow the nucleic amplification reaction and results in real-time on a monitor (screen) when the PATHPOD device is connected to a computer (extended Fig. 7). Alternatively, the PATHPOD instrument itself can also be used as a standalone instrument (requires no computer/laptop), where the nucleic amplification results are displayed on the front panel LEDs of the device. A detailed list of the materials used in producing the system is shown in extended Fig. 8.
Extended Fig. 9 and 10 show the software flowchart for standalone operation, and pc connect operation, respectively. The operation starts with entering the sample ID, selecting the operation mode (1 for the COVID test), and then heating up to reach LAMP reaction temperature (65 C • ). After inserting the cartridge, the reaction will take 50 min at 65 • C and followed by a 5-minutes termination process, where the cartridge is heated to 85 • C to inactivate any remaining enzyme. Finally, the test results will be displayed on the bicolored LEDs showing positive samples as blinking red LEDs (which is helpful in case of color blindness people reading the results), while the negative samples are shown as green LEDs, with constant light. The results are also displayed on LCD, as shown in extended Fig. 7. All the measurement data, including sample ID, running mode, timing, detection sensor data, and heating temperature and reported results, were also recorded as Comma separated value file (.csv) in real-time on the system memory MicroSD card that can be used for further analysis, e.g. review on a PC for advanced data analysis using our developed software.
The software for advanced data analysis is written in Python. Extended figures 11-12 show the PC software interface of the PATHPOD system and the algorithm for data processing and results showing, respectively. The external PC software also provides a lot of extended functionalities for advanced data handling applications such as input user, sample information including sample ID and information of each test well, review measurement data as plotting graphs such as sensor data for LAMP, heating monitor, absorption and derivative. It is also possible to review and analyze data recorded on the system memory (MicroSD) card when the system is used in standalone mode and extract the report in the same format as in real-time mode. The software includes a custom mode for analyzing new test sample experiments (not only for testing COVID-19) with custom setting up different LAMP reaction times, a new threshold signal cut-off for negative and positive samples of both absorption and derivative. Finally, all process data will also be saved as an excel file on the computer or the memory card for future usage and report, review and analysis. Extended figure 13 shows the results of testing the stability of the systems and heater calibration.
The tests show that the PATHPOD system can be used continuously for at least 40 h without any issues. Moreover, in the Supplementary Information, we provide additional figures to further illustrate our findings and support the main manuscript. Fig. S1 presents the genome structure of the SARS-CoV-2 virus, highlighting different genes and their functions. This information is essential for understanding the molecular basis of the virus and its interaction with the host. Figs. S2 and S3 detail the optimization of the RT-LAMP reaction on the PATHPOD system, focusing on the MgSO 4 and Betaine concentrations, respectively. These optimizations were critical for ensuring the efficiency and specificity of the assay. Fig. S3 also demonstrates the real-time monitoring of the reaction by measuring the fluorescence signal of the DNA intercalating dye, Syto 9, using an Mx3005P (Stratagene, AH diagnostics, Denmark).

Assay development and initial performance evaluation
The sensitivity of the newly developed RT-LAMP assay was initially determined using a commercial thermocycler (Stratagene, AH Diagnostics, Aarhus, Denmark); the results are shown in Fig. 4a. The preliminary experiments were conducted in tubes using a fluorescence DNA intercalating dye (SYTO-9). A sensitivity of approximately 30-50 copies of pure plasmid DNA was achieved within 30 min. No falsepositive signals were observed even when prolonging the reaction with time up to 80 min.
The developed assay was then adapted to PATHPOD, and the sensitivity was analyzed (Fig. 4 b). We make a serial of 10-fold dilution of a commercial plasmid DNA that contains SARS-CoV-2 genome fragment with a known concentration (copies/µl according to manufacturer). In the PATHPOD system, the developed RT-LAMP assay for COVID 19 could detect 30-50 copies of pure plasmid DNA per reaction within 40 min. No false-positive signals were observed even when prolonging the reaction up to 50-60 min. The results for optimization of the RT-LAMP reaction, such as sample volume, MgSO 4 concentration, etc., are shown in the Supplementary material (Fig. S2).
In order to improve the accuracy and reliability of the PATHPOD device's interpretation of experimental data, we have incorporated the use of derivative curves in our analysis.
These derivative curves represent the change rate of the phototransistor sensor signal over time, which helps to account for dynamic changes during the amplification reaction, ultimately contributing to a more accurate diagnosis. In order to reduce the random noise, the derivative was calculated from an average of every 30 data points of the sensor signal. The baseline for the transmittance signal and the derivative calculation were obtained at the 10th minute after the reaction initiation, allowing the optical components and cartridge to attain reaction temperature and get stabilized.
The thresholds for signal cut-off for negative and positive samples were determined based on a combination of both threshold signal in terms of negative transmittance of the incident light and the first order derivative of the obtained sensor signal. This combination ensures a high degree of accuracy and confidence in the final results, with a confidence level > 95%.
This methodology allows us to account for well-to-well differences, cartridge to cartridge variations, and system-to-system performances. Furthermore, this approach aids in quality control for cartridges and helps to validate their performance.
The use of derivative curves in this context, although not commonly employed in traditional RT-PCR or LAMP diagnostics, provides additional data that enhances the reliability of the PATHPOD device. Our device is not only capable of providing rapid results but also ensures  high accuracy, which is crucial in managing the spread of COVID-19. Please refer to Fig. S4 and S5, and Table S2 in the Supplementary materials for more detailed information about the data analysis using derivative curves.

Limit of blank (LOB)
In order to determine the precision of the developed assay, the limit of blank (LOB) of the reference method (RT-PCR) was initially determined using 13 negative controls on a qPCR system using Luna@Univeral One-Step RT-qPCR kit (New England BioLab) and primers for SARS-CoV-2 detection approved by CDC (US). The C t value for 8 out of 13 negative controls varied between 36 and 42 whereas 5 samples did not have any C t up to 45 cycles. The estimated LOB was approximately Ct 35.

Limit of detection (LOD) for the LAMP reaction: SARS-CoV-2 plasmid controls
Moreover, in order to test the sensitivity and LOD of RT-LAMP assay in the conventional thermocycler, SARS-COV-2 plasmid control was analyzed in Log 2 dilution series between 4000 copies and 31.25 copies/ reaction (Fig. 5-i). The correlation between the copy numbers and Tt (Threshold time) values in the conventional thermocycler was satisfactory. It was possible to detect the standard plasmid control down to 31 copies per reaction, however, with a 50% confidence interval. The LOD was at 125 copies/reaction that has a 95% confidence interval (shown in Fig. 5-i).
In the PATHPOD, the sensitivity test was performed using plasmid control in dilutions. To determine the limit of detection (LOD) of the RT-LAMP assay in the PATHPOD system, a serial dilution of 2019-nCoV_N_-Positive plasmid control was used. The tested concentrations were between 24,000 copies/reaction and 15 copies/reaction. There were 339 test samples overall in that concentration range. Fig. 5-ii shows positive signals at 3000, 300, 150 and 75 copies/reaction in two repetitions within 50 min. The reaction was positive only 1 out of 2 repetitions with 30 copies. Thus, the LOD of RT-LAMP was 30-75 copies/reaction (or equivalent to 5.000-12.500 copies/mL of the sample) in the PATHPOD system, that is comparable with Mx3005P (Stratagene, AH diagnostics, Denmark) standard thermocycler system. The LOD of this study was better than recent reports, 100 copies/reaction [2], 120 copies/reaction [39], which used purified RNA template.
The PATHPOD response was satisfactory at higher concentrations; however, the response at lower concentrations varied considerably. The lower concentrations gave false negative (FN) results significantly between concentration 15-120 copies per reaction. The sensitivity was 100% for 150-300 copies (6 repetitions each) and 44% for 120 copies. Six repetitions of 75 copies also gave 100% true positive results (Fig. 5ii). This result infers that for low concentrations the LOD of PATHPOD is between 75 and 150 copies/reaction.

Specificity of the RT-LAMP for SARS-CoV-2
The specificity of RT-LAMP was determined using 10 RNA and DNA

Table 1
Detection of SARS-CoV-2 in PATHPOD and RT-PCR with dilutions of inactivated SARS-CoV-2 virus.

Analytical sensitivity with inactivated SARS-CoV-2 virus and purified RNA
In collaboration with Statens Serum Institut (SSI), the performance of PATHPOD was evaluated to determine the sensitivity of the PATHPOD system to detect SARS-CoV-2. For performance evaluation, three different panels of quantified inactivated SARS-CoV-2 virus prepared by SSI were used as a target. The inactivated SARS-CoV-2 virus provided by SSI was prepared at Log 10 dilutions in a background of negative oral swab samples that were confirmed as negative for SARS-CoV-2 by RT-PCR at SSI. These dilutions were subjected to heat lysis according to PATHPOD boiling method protocol and tested using the PATHPOD at both DTU and SSI. The RT-PCR tests were performed in parallel at SSI. The preliminary LOD of the PATHPOD system was determined initially. The results show that the use of the PATHPOD could detect inactivated SARS-CoV-2 virus at 1:100,000 dilutions in all three panels of inactivated SARS-CoV-2 virus in the preliminary LOD studies, with the LOD of 60 -120 copies. The results of the LOD for the PATHPOD system were comparable to RT-PCR results (Table 1). H1, H2 and H5 are short names of the three SARS-CoV-2 viruses isolated from Denmark and cultured at SSI. These virus cultures are used as references for quality control of the performance of different RT-PCR used in Hvidovre Hospital and Vejle Hospital in Denmark. In Table 1, the time intervals of 1 h (h), 24 h (h), and 5 days represent the duration between when the sample was subjected to thermolysis and when it was subsequently analysed. This experiment was designed to assess the stability of the RNA postthermolysis, an aspect that is critical in determining the flexibility and timeframe within which the analysis can be conducted.
The stability of RNA after lysis could have practical implications in various testing scenarios, particularly in situations where immediate analysis might not be feasible or desirable. By understanding the stability of the RNA over time, we can ensure that our test remains accurate and reliable, even when analysis is not conducted immediately after thermolysis. This information provides invaluable insights into the operational flexibility of our testing procedure and enhances its adaptability to a range of different scenarios.
In addition, the sensitivity of the PATHPOD system was also determined with one of the RT-PCR confirmed positive clinical sample wherein, MagPure purified RNA samples from patient swab was used to prepare the dilutions (Clinical sample with the reference number A-39 was provided by Hvidovre Hospital, Denmark).
The sample A-39 originally had a Ct value of 8.22 as determined by RT-PCR (Altona PCR kit). The relative copy number for these clinical samples was determined based on the standard graph generated using the commercial plasmid control. The sample A-39 with Ct 8.22 had a relative concentration of 29,176 copies/microliter. Serial dilutions were initially made in Log 5 scale (Fig. 7a) and further in Log 2 scale (Fig. 7b).
By using the PATHPOD system, it was possible to detect RNA purified from the A-39 sample at 1:8000 dilutions, and that corresponds to ~22 copies/reaction. Therefore, the LOD with purified RNA is in the range 20-40 copies/reaction.

The relative level of detection (RLOD) studies
RLOD of the PATHPOD was determined by testing one of the inactivated SARS-CoV-2 viral panel H5 (SSI reference name) at preliminary LOD concentration. Inactivated SARS-CoV-2 viral panel H5 was tested with 20 repetitions at 1:50,000 dilution. The PATHPOD could detect 19 samples as positive out of 20 repetitions. This dilution was considered as LOD of the PATHPOD (Table 2 and Fig. 8).

Clinical sensitivity
For the clinical evaluation of the PATHPOD, relevant SARS-Cov-2 positive and RT-PCR confirmed negative clinical samples were collected from two hospitals in Denmark. Two separate sets of the samples were examined at DTU: (i) One set consisted of purified RNA from SARS-CoV-2 clinical samples, (ii) Another set of the same samples was prepared by boiling method wherein clinical samples were subjected to heat lysis at 95 • C for 5 min and shipped frozen to DTU. Current diagnostic methods of SARS-CoV-2, such as RT-PCR and other studies using RT-LAMP normally use purified RNA samples as a template for amplification. The advantage of this method is that the sample template contains no or very few inhibitors. However, the method requires advanced laboratory facilities, and well-trained personnel, which is costly and time-consuming. The boiled sample method used in this study for sample preparation is straightforward, requiring only a heat block or water bath for the lysis of the cells and virus particles. It is thus applicable for POC field applications. Although this method does not eliminate inhibitors in the samples, it is compatible with LAMP or RT-LAMP, as LAMP is more resistant to inhibitors [40]. The efficiency of RT-LAMP using RNA samples prepared from the boiled sample method (LOD of 30-75 copies/reaction in this study) is comparable with RT-LAMP using purified RNA samples [2,25,41,42]. Moreover, a combination of the boiled sample method with RT-LAMP assay on PATHPOD takes less than 1 h from sample preparation to the final result for ten samples. This combination is much faster and more economical than current RT-PCR or other conventional nucleic acids (NA) based methods. Therefore, this method can be used on a routine basis, being fast, easy, inexpensive and efficient, such as for screening staff on the medical front-line, border control, as well as in rural districts with mobile clinical labs etc.
A total of 398 clinical samples were included in the study, of which 300 samples were purified RNA from SARS-CoV-2 clinical samples, and 98 samples were prepared by the boiling method as specified in the PATHPOD sample preparation protocol. Out of 398 samples initially examined and reported as SARS-CoV-2 (true) positives or negatives by Hvidovre Hospital and Vejle Hospital, 200 samples were confirmed positive by RT-PCR and 198 were negative samples. Of 200 positive samples, 151 were purified RNA, and 49 samples were prepared by the boiling method. In the preliminary assessment, the PATHPOD showed 73.4% sensitivity, 96.2% specificity and 89.2% accuracy with samples prepared via the boiling method. With the purified RNA, the PATHPOD showed 87% sensitivity, 98.3% specificity and 92.5% accuracy ( Table 3).
The influence of the reaction time in the PATHPOD was assessed based on the sensitivity and negative predictive value (NPV) at different intervals for the above 398 clinical samples (Fig. 9a: Assessment of the sensitivity and negative predictive value (NPV) of the PATHPOD against different intervals of reaction time).
The effect of SARS-CoV-2 viral concentration/viral load in the patient samples on the clinical sensitivity of the PATHPOD was also assessed. In Table 4, we present a comprehensive view of the sensitivity of the PATHPOD system against different Ct cut-off ranges. The samples are ordered based on the Ct value ranges to provide a systematic overview. This arrangement enables an accurate representation of the diversity of the samples tested and the varying response of the PATHPOD system to these samples.  (Table 4). RT-PCR analyses were done using Luna® Universal Probe One-Step RT-qPCR Kit 2x. Using the PATHPOD system, it is possible to detect the samples having Ct values below 30 with 100% sensitivity. However, the sensitivity of the PATH-POD system was lower than the samples with Ct values above 30 ( Fig. 9b and Table 4).

Conclusions
We show, for the first time, in this comprehensive study (2.5 years of studies involving hundreds of SARS-COV-2 plasmid controls and 398 clinical samples from 2 hospitals in Denmark) a PoC system for rapid detection of SAR-CoV-2 virus in less than an hour using reverse transcriptase loop-mediated isothermal amplification. The results confirmed that the method is faster in time and still maintains comparable detection sensitivity and specificity to the standard RT-PCR method. Additionally, this study opens up the possibility for cost-efficient detection and screening of patients with COVID-19, aiding the mitigation strategies to minimize the risk of transmission of SAR-CoV-2 and the spread of the COVID-19 pandemic, especially in low-resource regions.

Declaration of Competing Interest
The authors declare no conflicts of interest.

Table 3
Clinical sensitivity of the PATHPOD.