Update on Rapid Diagnostics for COVID-19: A Systematic Review

: An accurate and rapid diagnosis of COVID-19 is an effective strategy for pandemic control, allowing disease screening and timely therapeutic intervention. We analyzed scientiﬁc reports about rapid tests for the diagnosis of COVID-19 to assess their reliability parameters. Medical Subject Headings terms or keywords related to point-of-care and rapid diagnostic testing for SARS-CoV-2 and COVID-19 were searched in data published from November 2020 to November 2021 in PubMed and Google Scholar databases. Notable differences were observed in sensitivity among direct tests that used different samples, and good accuracy was reported in a signiﬁcant number of studies (>80%). Pediatric samples and samples with high Ct values (RT-PCR) had suboptimal sensitivity (range 45.4% to 66%). Further, a lack of sensitivity (<46.2%) was observed in point-of-care tests and in rapid diagnostic tests for antibody detection in the ﬁrst days after infection, with increasing values in postinfection analysis (>60%). For serological detection of IgM or Antigen rapid diagnostic tests, no cross-reactivity was found with other coronaviruses. Therefore, although these tests are very important in facing the pandemic, they still need to be improved to test cross-reactivity against other pathogens, especially against other coronaviruses.


Introduction
Severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2) was discovered in November 2019 in Wuhan, China, and is the causative agent of coronavirus disease (COVID-19) [1]. A few months after its discovery, the World Health Organization (WHO) declared the COVID-19 outbreak a global pandemic [2]. However, more than a year and a half into the pandemic, the number of reported cases and deaths gradually increased, reaching the mark of more than 259 million cases and more than 5 million deaths [3].
Rapid and accurate diagnosis of COVID-19 is crucial for disease screening and managing public health preparedness [4]. Real-time reverse transcription-polymerase chain reaction (rRT-PCR), a molecular test for virus nucleic-acid detection, is the most suitable method to detect SARS-CoV-2 [5]. However, rRT-PCR may yield false-negative results if there is a low amount of viral genome in the sample or a lack of an adequate window period of viral replication [6]. In addition, performing rRT-PCR requires laboratory infrastructure and can take several days to deliver results [7].
Rapid diagnostic tests (RDTs) are an alternative for fast and inexpensive diagnosis of COVID-19 because they are easy to use and results are available in a short time. Furthermore, they are easy to interpret, allowing testing to be possible in near-person, decentralized healthcare settings. In addition, antigen RDTs (Ag-RDT) have good analytical performance. Therefore, they can be used at the point of care (POC) to help increase testing and reduce the spread of infection through early self-isolation [7][8][9][10][11][12][13]. On the other hand, detecting antibodies produced in response to infection with SARS-CoV-2 helps identify asymptomatic patients [14,15]. Moreover, they are employed in serum-surveillance studies and investigations of the ongoing outbreak [5].
The importance of mass testing has triggered the development of numerous RDTs, which have been readily implemented during the COVID-19 pandemic [8]. However, questions about the importance of constant test selection based on its accuracy persist.
In our first review, we present an overview of studies reporting the use of POC and RDT tests to diagnose COVID-19 in the first year of the pandemic [16]. However, we observed insufficient data regarding the performance of the analyzed tests and a lack of validation of cross-reactivity with other pathogens in the tests launched in the market. Here, we carried out a new analysis of scientific data published on rapid tests available for the diagnosis of COVID-19 to verify whether the analyses of the available tests have been improved considering the discoveries about SARS-CoV-2, one-year after our first analysis.

Literature Search
In this systematic review, all studies published from 01 November 2020 until 30 November 2021 were retrieved from PubMed and Google Scholar databases. We considered subject titles and headings/subtitles (when applicable) in addition to keywords and abstracts to identify the searched terms: "point of care sarscov2 diagnostic or diagnosis", "point of care COVID-19 diagnostic or diagnosis", "rapid test COVID-19 diagnostic or diagnosis" and "rapid test SARS-COV-2 diagnostic or diagnosis".

Selection Criteria
Case-control, cohort studies, and randomized clinical trials with experimental data in POC or RDT tests for SARS-CoV-2 efficiently performed analyses and provided robust data on their results were eligible for inclusion. WHO and the United Nations Children's Fund (UNICEF) point out that RDT should preferably provide results between 15 and 30 min to better deal with the COVID-19 pandemic [7,9]. Therefore, only rapid POC and RDT tests that deliver results within 30 min were included. Duplicate articles, preprints, and other articles with uncorrelated themes such as editorials, case reports, modeling studies, and studies that did not present POC or RDT data for COVID- 19 were not included in the present systematic review. There were no language restrictions.

Data Extraction and Quality Assessment
One independent investigator performed a systematic search in the databases. Seven investigators independently analyzed the title, abstract, and full-text articles to apply the selection criteria following a script with the same parameters. We extracted authors, year of publication, test type, assay method, diagnostic criteria (sensitivity, specificity/accuracy), time detection, sample size, and cross-reaction analysis for each study. Two reviewers independently analyzed and resolved the methodological quality of the eligible studies.

Results
A total of 2135 million publications were retrieved from databases, a sample robust enough to identify all relevant articles on the addressed topic, such as our previous study of the first year of the pandemic [16]. After screening the titles, abstracts, and/or keywords, 520 studies were selected, and 52 were chosen for full-text review. Articles selected in this review were published from 01 November 2020 until 30 November 2021. Details of the total number of articles screened, assessed for eligibility, extracted, and included in the final analyses are shown in Figure 1.

Results
A total of 2135 million publications were retrieved from databases, a sample robust enough to identify all relevant articles on the addressed topic, such as our previous study of the first year of the pandemic [16]. After screening the titles, abstracts, and/or keywords, 520 studies were selected, and 52 were chosen for full-text review. Articles selected in this review were published from 01 November 2020 until 30 November 2021. Details of the total number of articles screened, assessed for eligibility, extracted, and included in the final analyses are shown in Figure 1.  Tables 1 and 2 summarize the analyzes used in this study.   Cross-reaction was tested on serum samples from a patient with Dengue or typhoid fever. No cross-reactivity was observed.
Diagnostic accuracy was reported in only 11 studies [22,23,26,27,31,32,34,[36][37][38]55], with values greater than 80% except for one study that reported 75.9% [36], demonstrating that the developed tests had a good performance in detecting SARS-CoV-2. Otherwise, cross-reaction analyses were verified in only nine studies, showing the need to improve the diagnostic performance and the information provided about them.
The sensitivity and specificity of these tests are shown in Figure 2.
The sensitivity and specificity of these tests are shown in Figure 2.
The sensitivity and specificity of these tests are shown in Figure 3.
The sensitivity and specificity of these tests are shown in Figure 3.

Discussion
The COVID-19 pandemic has been a global public health challenge for over a year. The need for rapid screening of SARS-CoV-2 infection has led to a constant search for the development of diagnostic tests [7,8]. Therefore, we conducted a systematic review of studies of POC and RDT tests for COVID-19 one year after our first analysis to verify whether there was progress in the tests' research considering the discoveries about SARS-CoV-2.
In this paper, we selected 52 studies that reported using POC and RDT tests to diagnose COVID-19. The data were divided into POC and RDT tests for direct detection

Discussion
The COVID-19 pandemic has been a global public health challenge for over a year. The need for rapid screening of SARS-CoV-2 infection has led to a constant search for the development of diagnostic tests [7,8]. Therefore, we conducted a systematic review of studies of POC and RDT tests for COVID-19 one year after our first analysis to verify whether there was progress in the tests' research considering the discoveries about SARS-CoV-2.
In this paper, we selected 52 studies that reported using POC and RDT tests to diagnose COVID-19. The data were divided into POC and RDT tests for direct detection of SARS-CoV-2 (Table 1) and POC and RDT serological tests to detect antibodies against SARS-CoV-2 ( Table 2).
Most of the selected POC and RDT studies showed good sensitivity and specificity. Our results indicate good performance for saliva samples, NPS/OP swabs, and throat samples, suggesting that these samples can be used as an alternative to NPS or nasal swabs commonly used for the diagnostic testing of SARS-CoV-2. [17].
Regarding the pediatric samples, we found lower sensitivity values when compared with samples from adult patients [22,24,25,28]. Furthermore, age was associated with antigen-test performance irrespective of Ct values and symptom duration [17]. Although children are more likely to be asymptomatic or have mild symptoms, they can transmit SARS-CoV-2, making screening essential to help contain the spread of the virus [69].
Regarding serological tests POC and RDT for detecting antibodies, most showed high sensitivity and specificity values, indicating good clinical performance. However, we observed lower sensitivity values in tests that performed the analysis in the first days after SARS-CoV-2 infection [57,61,62,64] than in postinfection studies.
The lower sensitivity verified may be related to the time required to develop antibodies. Although IgM and IgG seroconversion can occur simultaneously or sequentially and antibody titers can plateau after six days [75], most patients do not produce an antibody response until the second week after the initiation of symptoms [76].
Thus, for a more accurate diagnosis of COVID-19, tests based on direct detection are preferable in the first days of symptoms, about 3 to 7 days of infection, and antibody tests should be used after this period due to their greater sensitivity [8,59]. Moreover, combining techniques can increase the sensitivity in diagnosing SARS-CoV-2 infection [6,59,77].
On the other hand, it is of concern that only a few studies have reported cross-reaction assays. Although many manufacturers have tested cross-reacting for different pathogens and presented data of low homology between the proteins of SARS-CoV-2 and other human coronaviruses (used for manufacturing), cross-reaction is not commonly addressed [78][79][80][81][82][83][84]. Cross-reactivity has been reported in COVID-19 and Dengue cases [85][86][87][88], both in serology for COVID-19 in patients with Dengue and serology for Dengue among patients with COVID-19 [89].
In addition, coinfection between SARS-CoV-2 and other pathogens such as influenza and the Dengue virus has been reported [90][91][92][93][94][95][96][97]. The similar clinical symptoms of SARS-CoV-2 and other febrile illnesses favor misdiagnosis. Incorrect or later diagnosis can affect clinical management, exacerbate complications, and increase mortality. Therefore, it is essential to identify coinfections rapidly. Testing for SARS-CoV-2 in areas with overlapping outbreaks is not enough to rule out the possibility of coinfection, and PCR tests are required to confirm the infection.
Concerning the findings of our first review [16], 117 more studies were published, and 30 more trials were tested in one year. Likewise, cross-reaction was evaluated in 13.05% of studies in the first year analyzed and 28.85% in the second year. Although the percentage of articles that reported accuracy remained similar (26.09% in 2020 and 26.93% in 2021), the accuracy values obtained increased. Likewise, sensitivity and specificity parameters were improved. However, there was no statistical difference observed (Figure 4).
Several factors might be related to the heterogeneous performance of the tests, such as the quality of the individual smear, whether the sample is fresh or not, the storage conditions of the sample, anatomical collection site, viral load of the sample, day of onset of symptoms, among others [25,28,98]. These issues make it difficult to compare the tests directly.
conditions of the sample, anatomical collection site, viral load of the sample, day of onset of symptoms, among others [25,28,98]. These issues make it difficult to compare the tests directly. It also explains why we observe so many performance divergences reported by studies or test manufacturers. We still found few trials that revealed the accuracy of the tests, raising uncertainties about the accuracy of the COVID-19 POC and RDT tests.
Considering the importance of an accurate diagnosis, we expected to find more studies evaluating cross-reactions with other pathogens, especially with endemic coronaviruses, in this second year of the pandemic.
We believe next-generation tests should incorporate phylogenetic and structural analyses to select antigens used in their manufacturing process. For example, in the first semester of the pandemic, 45 epitopes with mutations in the main antigen of SARS-CoV-2 [99] were reported, denoting the need for the preemptive use of antigens with variants for the development of serological tests.
With the infection of hundreds of millions of people, mutations have become much more frequent, reaching more than 50 occurrences only in the spike protein of the Omicron variant, which is responsible for a robust escape from neutralization steps [100]. Therefore, the incorporation of mutations of importance in epitopes should be the main focus in developing tests that use antibodies/antigens.
Although the POC and RDT tests were critical to fighting COVID-19, our findings corroborate that the POC and RDT tests still need further validation regarding test accuracy and cross-reactivity with other pathogens [101].
Overall, we have identified some limitations in the analyzed studies: discrepant sample numbers for validations, differences in the samples used, and a low number of studies assessing accuracy and cross-reactivity. In addition, this systematic review also has limitations: we could not perform a meta-analysis due to the variability of the methods and results found in the studies.

Conclusions
Although more than a year of the COVID-19 pandemic has passed, POCs and RDTs tests are still needed since they contribute to the management of COVID-19. This study It also explains why we observe so many performance divergences reported by studies or test manufacturers. We still found few trials that revealed the accuracy of the tests, raising uncertainties about the accuracy of the COVID-19 POC and RDT tests.
Considering the importance of an accurate diagnosis, we expected to find more studies evaluating cross-reactions with other pathogens, especially with endemic coronaviruses, in this second year of the pandemic.
We believe next-generation tests should incorporate phylogenetic and structural analyses to select antigens used in their manufacturing process. For example, in the first semester of the pandemic, 45 epitopes with mutations in the main antigen of SARS-CoV-2 [99] were reported, denoting the need for the preemptive use of antigens with variants for the development of serological tests.
With the infection of hundreds of millions of people, mutations have become much more frequent, reaching more than 50 occurrences only in the spike protein of the Omicron variant, which is responsible for a robust escape from neutralization steps [100]. Therefore, the incorporation of mutations of importance in epitopes should be the main focus in developing tests that use antibodies/antigens.
Although the POC and RDT tests were critical to fighting COVID-19, our findings corroborate that the POC and RDT tests still need further validation regarding test accuracy and cross-reactivity with other pathogens [101].
Overall, we have identified some limitations in the analyzed studies: discrepant sample numbers for validations, differences in the samples used, and a low number of studies assessing accuracy and cross-reactivity. In addition, this systematic review also has limitations: we could not perform a meta-analysis due to the variability of the methods and results found in the studies.

Conclusions
Although more than a year of the COVID-19 pandemic has passed, POCs and RDTs tests are still needed since they contribute to the management of COVID-19. This study provides an overview of studies reporting the use of POC and RDT tests to diagnose COVID-19 published from November 2020 to November 2021. We identified heterogeneous performance across assays that may be related to various factors such as quality of the sample, viral load, and anatomical collection site.
Combining viral antigen or genome-detection methodologies with antibody-detection tests is recommended to increase diagnostic accuracy. However, information about test accuracy has been limited, and some cross-reactions with different pathogens have been found. Given these findings, further validations on cross-reactivity with other pathogens and the accuracy of the tests are needed.