Europe and the entire world have faced the second wave of Coronavirus Disease 2019 (COVID-19) pandemic which was characterized by an increased number of infections and related deaths worldwide, thus still highlighting critical issues in the management of this health emergency (1,2). At the time of the writing of the present review article, 113,523,131 laboratory-confirmed COVID-19 cases and 2,519,454 COVID-19-related deaths have been recorded worldwide, highlighting the impressive impact of this pandemic globally (3).

Despite the optimism deriving from the approval of two new mRNA vaccines and of one recombinant vaccine against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections by the Food and Drug Administration (FDA) and the European Medicines Agency (EMA) (5,6), a third wave of infections, already observed in the United States, the United Kingdom and other countries with greater proportions than the one just concluded, is expected in the coming months (3,6,7). In this regard, several governments worldwide have already begun to adopt social distancing measures and the lockdown of collective activities in order to avoid a drastic increase in the number of infections (8,9). In addition, a great concern is also represented by the need to differentiate COVID-19 cases from seasonal flu that could clog hospital emergency services, slowing down the diagnostic and therapeutic procedures for patients with COVID-19 (10). These reasons have led the scientific community to question which diagnostic strategies are optimal in order to efficiently combat the imminent increase in COVID-19 infections, as well as to perform differential diagnoses between COVID-19 infections and seasonal flu.

In this context, population screening strategies have been proposed and are currently being implemented for the effective monitoring of the COVID-19 epidemiological curve and to screen the immunized population; however, it is not yet clear which is the most effective method for these surveillance programs (11-14).

Since the beginning of the pandemic, immense efforts have been made for the development of effective diagnostic strategies which may be used to accurately identify SARS-CoV-2-infected patients, thus limiting the risk of contagion and promptly treating any respiratory symptoms, avoiding serious consequences for individuals (15).

At present, several diagnostic methods have been approved by regulatory agencies worldwide; however, there is still confusion regarding the correct tests to be used based on the patient's medical history or the investigation purpose (16). As regards Europe, there are 365 different commercialized devices CE-IVD- and FDA-approved or used for research purposes. Of these devices, 168 are immunoassays, 192 are PCR-based methods, three are NGS-based methods and two commercialized tools are based on different technologies (17). Therefore, it is evident that the selection of the optimal diagnostic test can be difficult.

On these bases, the present review aimed to provide the rationale for the correct use of SARS-CoV-2 diagnostic tools currently available by setting out the decision-making principles useful for the correct choice of the most appropriate test.

As will be discussed in the following paragraphs, several diagnostic tests are currently available for the early identification of SARS-CoV-2 infection, for the monitoring of the presence of any infections among healthcare workers, for the monitoring the incidence rates and the severity of the infection and for the evaluation of the complete remission of patients with COVID-19 (18-20).

Several parameters should be considered for the selection of the optimal diagnostic test. A diagnostic test must have good sensitivity and specificity rates; however, these parameters are not the only features to be considered. Indeed, in the case of the COVID-19 pandemic, an effective test should be rapid, repeatable, based on technologies available at numerous centers and keep costs limited in order to be carried out on a large fraction of the population (21).

Therefore, it is evident that the selection of the diagnostic test should be performed taking into account the clinical or surveillance purpose of the investigation as well as the possibility to repeat the test several times until the patients are no longer positive.

For all these clinical and epidemiological needs, three main types of tests for COVID-19 diagnosis are available: i) Molecular RT-PCR swab tests; ii) serological tests; iii) rapid antigen or antibody tests.

Although the molecular and structural characteristics of SARS-CoV-2 were initially unknown, in a very short period of time, research laboratories and biomedical companies studied the main features of the virus, thus assisting researchers world-wide in developing various diagnostic solutions for a correct diagnosis of COVID-19 (22-24). Among such solutions, the most commonly used and validated methodologies are rapid antigen or antibody tests, immunoenzymatic serological tests and RT-PCR-based molecular tests. Each of these three types of diagnostic tests can be applied at a precise moment of infection. Of note, only kits, reagents and molecular probes validated by the Centers for Diseases Control and Prevention (CDC) and the World Health Organization (WHO) and approved at the American and European level by the FDA and by the EMA can be used for diagnostic purposes (25).

Although immunoenzymatic (either classic or rapid methods) and molecular methods are the most widely used techniques for the diagnosis of COVID-19, other approaches were used in the early phase of the pandemic to identify positive patients and the etiological agent of infection. Among these approaches, viral culture and next-generation sequencing (NGS) methods were crucial for the identification of the novel coronavirus and for the characterization of its molecular structure. These two techniques have made it possible to fully characterize the genome and the viral protein structure, allowing the understanding of the viral mechanisms of action, the mode of transmission, the clinical impact and the development of any therapeutic strategies and diagnostic tools (26,27).

Apart from these conventional strategies, other diagnostic methods are under development and validation or are currently finding application in research contexts. Among these methods, digital PCR, isothermal amplification techniques, clusters of regularly interspaced short palindromic repeats/Cas (CRISPR/Cas)-based diagnostic methods, biosensors, electron microscopy-based methods, etc., represent the armamentarium to effectively diagnose COVID-19 infection and to effectively assess the epidemiological spreading of the pandemic (15,28). Notably, clinical investigations and radiological imaging have represented valid diagnostic alternatives, particularly during the early stages of the pandemic when there were still no validated molecular and serological tests available (29) (Fig. 1).

Despite the availability of all these diagnostic techniques, a correct diagnosis of COVID-19 infection can only be established considering the test to be used, the type of sample to be analyzed and the timing of the test itself. Therefore, it is necessary to perform the correct test, at the correct time in the correct biological sample (30,31).

In particular, it is important to take into account the moment of the suspected infection, the patient's medical history, the symptoms and general clinical picture for a successful outcome of the diagnostic test. Furthermore, the positivity of a diagnostic test strongly depends on the moment at which it is carried out (Fig. 2). Both serological and molecular tests are not useful during the first week of the supposed infection because the virus is still in its incubation period and there are not yet sufficient copies of viral RNA in circulation neither antibodies nor viral proteins identifiable by serological tests (32,33). Therefore, before the onset of symptoms, the probability of correctly determining the presence of the virus, particularly using molecular tests, remains low (32).

At two weeks after the presumed infection, and in parallel with the onset of symptoms, molecular tests carried out on nasopharyngeal swabs or bronchoalveolar lavage fluid (BALF) samples will have a greater chance of being positive, as the virus is actively replicating. However, this probability is reduced over time, as a result of the elimination of the virus and the remission of the disease. In the case of molecular tests, the probability of a positive or negative result should be considered, as no commercial diagnostic tools have a sensitivity of 100%, particularly in cases of patients with low viral load as asymptomatic or paucisymptomatic individuals (32,34).

As regards serological tests, the search for IgM and IgG antibodies begins to yield positivity at approximately one month after the presumed infection and the levels of these immunoglobulins remain high for long periods of time. Of note, the IgM serum levels decrease significantly after six weeks from the onset of symptoms (35). The detection of viral proteins and mucosal IgA antibodies is also very important. Indeed, these markers increase in the early stages of infection and can be used for the early diagnosis of COVID-19 infection (36). In summary, all these data suggest that for a correct diagnosis of COVID-19 infection to be made, it is necessary to use the right test, at the right time.

As already mentioned, the selection and the correct handling of samples for both molecular and serological analyses are fundamental for the diagnosis of COVID-19 infection. In particular, the collection of samples and the pre-analytical phases are crucial for the positive outcome of the diagnostic procedure (37,38). Different studies have demonstrated that the positivity rate of molecular tests significantly depends on the quality of the starting sample influenced by sample collection, poor quality material, wrong transport or storage, the presence of inhibitors, etc. (38). Technical troubleshooting will be discussed in the following chapters.

The chief samples used for molecular analyses are obtained from the respiratory tract. In particular, both oropharyngeal and nasopharyngeal swabs represent good samples for viral RNA extraction and amplification through RT-PCR; however, previous studies have highlighted that BALF is the most appropriate sample for SARS-CoV-2 molecular detection (39,40). Apart from these commonly analyzed respiratory tract samples, other specimens may also be collected, including nasal mid-turbinate swabs and nasal or nasopharyngeal wash/aspirate (41). Of note, the permanence of the virus in the respiratory tract is only temporary, with a peak of positivity in the first three days of infection, followed by a constant decrement of the rate of positivity for molecular analyses until the 10th week following symptom onset (42). To ascertain SARS-CoV-2 positivity after a long period of time, other biological samples are used. Among these, fecal samples represent the optimal material to evaluate the permanence of viral RNA in patients with negative nasopharyngeal swabs, but with clinical symptoms attributable to COVID-19 infection (43). Indeed, a greater persistence of SARS-CoV-2 in the gastrointestinal tract has been demonstrated, allowing the detection of SARS-CoV-2 RNA even after more than a month from infection (44).

Other studies have proposed diagnostic screening based on the analysis of saliva and urine samples through serological and molecular tests. However, the collection of these samples is generally carried out by the patients themselves without the surveillance of a healthcare professional, resulting in a possible non-representative sample. In addition, the presence of interfering substances or substances that degrade viral RNA or human antibodies represents a considerable bias that significantly limits the use of these samples in clinical practice (30).

All the types of biological samples used for diagnostic purposes are presented in Table I. A detailed description of the timing of use and the technique for which specific samples are collected is also provided (Table I).

Overall, the type of sample should be collected considering the timing of the infection. Collecting different samples from different sites may be useful to avoid misdiagnosis of asymptomatic patients negative for molecular tests. Otherwise, it is possible to carry out repeated nasopharyngeal swabs in two or three consecutive days in order to overcome the window period of SARS-CoV-2 incubation thus being able to correctly diagnose a patient infected with SARS-CoV-2 (45).

RT-PCR-based molecular methods represent the gold standard techniques used worldwide to make a confirmatory diagnosis of COVID-19 infection (46). Since the complete sequencing of the SARS-CoV-2 genome (26), researchers of different countries have begun to design molecular primers and probes specific to SARS-CoV-2 RNA sequences in order to perform differential diagnosis between COVID-19 infections and other pathologies with similar symptoms, such as seasonal flu or bacterial infections (47,48). Of note, the whole genome sequence of SARS-CoV-2 is 29,903 bp in length containing the following functional elements: A poly-A cap of 50 bases, the open reading frame 1/ab (ORF1/ab) containing the coding sequences for the RNA-dependent RNA polymerase (RdRP), the spike proteins, the envelope proteins, the membrane and nucleocapsid proteins, and a poly-A tail of 30 bp (49).

At present, several portions of the SARS-CoV-2 genome are used for the design of specific primers and probes, including the genomic portions encoding for the RdRP gene, for proteins constituting the nucleocapsid (N gene) and spike molecules (S gene), for proteins of the envelope (E gene), for the membrane, etc. (Fig. 3).

The CDC has made available a list of approved and validated kits and reagents to clinically diagnose COVID-19 infection (50). Similarly, the CDC has also made available a list of three primer pairs that can be used for research purposes only, each working with two different molecular probes, specific for two portions of the N gene and the gene coding for viral RdRP, respectively (51).

RT-PCR-based molecular tests are considered the optimal diagnostic option for wide surveillance strategies due to the relatively low costs of the entire viral RNA extraction, reverse transcription and amplification procedure, and the availability of RT-PCR thermal cyclers in hospitals, research institutes and private laboratories (52). Other advantages of RT-PCR methods compared to other diagnostic techniques are the timesaving of the procedure, the easy execution of the technique and the non-necessity of highly trained personnel (53). In addition, a number of the available RT-PCR kits are based on one-step amplification methods, where the buffer of the nasopharyngeal swab is inserted into the plate and the machine autonomously provides for the extraction, reverse transcription, amplification and analysis of the samples. These procedures ensure fast results guaranteeing excellent reproducibility and standardization of the data obtained which are less influenced by operator bias (54).

As already mentioned, the EMA has approved 192 PCR-based methods while the FDA has approved 235 different molecular tests for both RT-PCR and the rapid detection of SARS-CoV-2 RNA (17,55).

All these approved molecular tests detect two or three fragments of SARS-CoV-2 RNA mainly using multiplex assays based on standard RT-PCR protocols. Briefly, following the collection of nasopharyngeal swabs from individuals with suspected COVID-19 infection, the viral RNA is extracted using commercial kits and lysing solutions with manual or automated extraction protocols. Subsequently, the obtained RNA, containing both human and viral RNA, can be directly analyzed by RT-PCR in the one-step protocols or must be retrotranscribed into complementary DNA (cDNA) before RT-PCR amplification. After obtaining the cDNA, the SARS-CoV-2 targets are amplified through RT-PCR using TaqMan probes specific for the two-three viral targets. In particular, the exonuclease activity of the Taq polymerase (5′→3′ direction) cleaves the probes annealed with the viral targets allowing the emission and real-time detection of fluorescent signals. In this manner, the intensity of the fluorescent signal is proportional to the total amount of the amplified targets. However, all the existing RT-PCR protocols are only qualitative and not quantitative (Fig. 4) (56,57).

The entire analytical procedure is completed in 4-8 h, based on the type of RT-PCR protocol used (one-step or two-steps). This makes it possible to establish the positivity of an individual in a relatively short amount of time, allowing the initiation of quarantine protocols that limit the spread of infections (58,59). In addition, the majority of the automated or semi-automated systems available are based on 48- or 96-well platforms in order to process a series of samples, thus reducing the execution times, the costs of plastics and consumables and procedural errors that may occur by analyzing the samples individually (60).

RT-PCR-based methods ensure also a low limit of detection (LoD) of SARS-CoV-2 RNA (61). Specifically, during the early stages of the pandemic, when diagnostic techniques had not yet been optimized and standardized, a significant fraction of COVID-19-positive patients were identified as false-negative due to the low sensitivity of the primers and probes used or the inaccuracy of the whole RT-PCR procedure (false-negative rates ranging from 38% at the day of symptoms onset to 67% before one day from the onset of symptoms or to 66% after 16 days from the onset of symptoms) (62). As will be better described at the end of this chapter, in the case of asymptomatic or paucisymptomatic patients, generally characterized by a low viral load, the RT-PCR investigations could mistakenly yield negative results. Currently, the commercially available RT-PCR kits have partially solved the problem of the low sensitivity of RT-PCR. Indeed, the methods available today have a theoretical LoD that varies from 0.3 copies/µl to 100 copies/µl, depending on the diagnostic system used (63). However, it should be considered that this limit is only theoretical; therefore, in clinical practice, procedural errors or reaction interferers raise the LoD by 10-fold (64).

Although RT-PCR represents the gold standard method for the diagnosis of COVID-19 infection, this method is subject to several limitations and criticisms that can lead to false-positive or false-negative results, thus affecting the correct management of the pandemic. As already mentioned, one of the main limits of RT-PCR is its low sensitivity in correctly diagnosing samples with low viral load, including swabs taken incorrectly or obtained from asymptomatic or paucisymptomatic individuals (65). In addition, RT-PCR is affected by contaminants and interferers contained in the sample or introduced by the operator capable of inhibiting the reaction (65). Other limits also concern the execution time of the analysis, which in non-automated systems, can take up to 24 h to obtain a result that can be communicated to the patient. Finally, RT-PCR is a method that is profoundly affected by pre-analytical and analytical bias regarding the collection, storage and handling of samples, therefore, careful attention should be paid during the collection and management of samples (66).

In order to reduce the errors in the interpretation of molecular tests, the WHO has published specific recommendations useful to make a correct diagnosis of COVID-19. In particular, a molecular test must be conducted in two different biological matrices, for example, a nasopharyngeal swab and a fecal swab, or be performed on two consecutive nasopharyngeal swabs in order to obtain reliable results. The approved tests include the analysis of three different viral genes or two viral genes and a human control gene and the use of specific negative and positive controls useful to ascertain the presence of contamination or evaluate the inhibition of the reaction (67).

Overall, RT-PCR represents the gold standard and most widely used method worldwide to make an accurate diagnosis of COVID-19 infection due to the rapidity of the method and the availability of instruments in public and private hospitals and laboratories. In addition, the RT-PCR tests currently available on the market ensure good sensitivity and specificity rates for the diagnosis of COVID-19 infection. The main RT-PCR diagnostic systems currently available on the market, illustrating their main technical features are presented in Table SI.

With the increase in the number of individuals with a suspected COVID-19 infection, it became necessary to adopt more rapid and low-cost diagnostic strategies to carry out extensive surveillance campaigns (68-70). To cope with this emergency, various rapid tests have been developed to detect viral antigens or anti-SARS-CoV-2 human antibodies in salivary, nasal or oropharyngeal swabs and blood samples. These tests are currently adopted for the frequent monitoring of personnel operating in at-risk environments such as schools or hospitals or to carry out extensive screening strategies on populations where a new outbreak of infection is suspected (70,71).

Compared to RT-PCR-based methods, rapid antigenic and rapid antibody tests are characterized by more rapid execution times of ~15-30 min, a lower cost and an easier procedure that does not require the presence of highly trained personnel (72). These tests are mainly built on platforms based on the principle of lateral flow immunoassay (LFIA) for the direct detection of viral proteins (rapid antigen tests) or human antibodies against SARS-CoV-2 antigens (rapid antibody tests). As regards rapid antigen tests, these allow the identification of COVID-19-positive individuals through the detection of SARS-CoV-2 nucleocapsid or Spike proteins (viral antigens) in swabs collected from the upper airways of the subject with suspected infection. The principle behind these LFIA rapid antigen tests is very simple and based on the bond between antigens and antibodies that occurs on the surface of a porous membrane where the swab buffer flows by capillarity. Briefly, the swab buffer is loaded in the sample well of the cartridge and flows by capillarity at the level of the conjugation pad containing control rabbit antibodies and specific antibodies against SARS-CoV-2 antigens both linked to detector molecules (conjugated antibodies). In the case of a positive sample, the link between viral antigens and conjugated antibodies take place at the level of the conjugation pad. Subsequently, the buffer containing conjugated control rabbit antibodies and the antigen-conjugated antibody complex flows to the test line where other antibodies specific for viral antigens are immobilized. In the case of a positive sample, the binding between antibodies immobilized in the test line and the antigen-conjugated antibody complex gives rise to a colorimetric reaction indicating the positivity of the sample. Finally, the buffer flows further to the control line where anti-antibodies specific for the conjugated control rabbit antibodies are immobilized; if the test is performed correctly, the reaction between these two molecules also gives rise to a colorimetric reaction. To be trusted, the test control line must always be positive (73,74) (Fig. 5A).

Similarly, rapid antibody tests use the same LFIA principle; however, human IgA or IgG and IgM against SARS-CoV-2 antigens are searched for. In particular, the blood (or saliva) sample is loaded inside the sample well and flows to the conjugated pad containing gold-tagged viral antigens and gold-tagged control antibodies. In this test, anti-SARS-CoV-2 human antibodies (IgA, IgG or IgM) bind the gold-tagged antigens and the sample flows to the test line where anti-human IgA, IgG or IgM antibodies are pre-attached to the membrane. In the case of a positive sample, the IgA, IgG or IgM-gold-tagged antigen complex binds to the human anti-antibodies immobilized in the test line determining a colorimetric reaction which indicates the positivity of the sample. Finally, the gold-tagged control rabbit antibodies flow to the control line binding anti-rabbit antibodies thus giving rise to a confirmatory colorimetric reaction (75,76) (Fig. 5B).

Both rapid antigen and rapid antibody tests yield results readable to the naked eye in a very short period of time and can be performed at the 'point-of-care' without any specific instruments or sample processing. Of note, these tests yield qualitative, but not quantitative results; therefore, it is only possible to establish whether the individual is positive or not for COVID-19 infection without assessing the viral load. In addition, in the case of rapid antibody tests, it is possible to establish whether the patient is carrying anti-SARS-CoV-2 antibodies; however, it is not possible to establish whether the patient has an active SARS-CoV-2 infection or already resolved disease (77,78).

Overall, among the main advantages of rapid antigen and rapid antibody tests are the low cost, the possibility of carrying out the test directly at the point-of-care and the high speed and easy execution that ensure a positive or negative result in ~30 min. However, these tests suffer from important limitations mainly related to a low sensitivity and specificity of 56.2 and 99.5%, respectively (79). Indeed, a recent review of the literature demonstrated that rapid antigen tests had a false negative rate of 27.9%, while no false-positive results were observed (68). As regards rapid antibody tests, a recent study comparing three different kits demonstrated a false-positive rate ranging from 51.6 to 28.1%, and a false-negative rate ranging from 0 to 4.2% (80).

As regards rapid antibody tests, these are very straight-forward and can be quickly performed, and provide useful information about the stage of infection. Indeed, in the case of negativity, only the control band is colored. In the case of an ongoing infection, the IgM and control bands are stained. In the case of a recent infection, both the IgM and IgG bands are stained, since both antibodies are present in the bloodstream. Following the remission of the disease, for a period of time that varies from patient to patient, only the IgG band is positive together with the control one, and in the case of re-infection, all the three bands are positive again (Fig. S1). The test is considered invalid when none of the lines stain or when the test lines are stained but the control line is not (81).

The limited sensitivity and the related high false-negative results are mainly related to the low viral load and the low antibody response observed in some patients; however, as already mentioned, the probability of obtaining a positive test depends also on the time of the presumed infection and on the test execution time. Indeed, although the viral antigens are found in the samples after a short time from infection, their permanence and stability in the biological sample are limited; therefore, it is not always possible to correctly identify these proteins (82). Similarly, rapid antibody tests are mainly designed to identify IgM and IgG antibodies, which are not produced by the body immediately, but begin to be found in the bloodstream after the third week of the suspected infection. Therefore, it is important to use the most appropriate test considering the time of the presumed infection. More recently, the use of rapid tests for the detection of IgA to accelerate the diagnosis of COVID-19 infection has been proposed (83).

Overall, rapid antigen and rapid antibody tests are widely used for screening strategies on large portions of the population (70,84); however, they do not ensure a precise diagnosis of COVID-19. Therefore, these tests should be always confirmed by RT-PCR analyses. Details of rapid antigen and rapid anti-body tests currently approved by international agencies are reported in Table SII.

Most of the immunoenzymatic serological tests used for COVID-19 investigations are based on the principle of indirect enzyme-linked immunosorbent assay (ELISA). Of note, ELISA is a colorimetric, chemiluminescent or fluorescent microwell plate-based assay used for the quantitation and detection of human proteins, immunoglobulins, antigens and other peptides through the binding between the target protein and a specific antibody that results in a detectable signal (85). This technique allows researchers to obtain highly specific and sensitive results in a relatively short time ranging from 1 to 5 h (85).

Briefly, 96-well commercial COVID-19 ELISA indirect tests contain immobilized viral antigens at the bottom of each well that are recognized and bound by anti-SARS-CoV-2 antibodies present in the serum of patients following proper dilutions. A series of washes is then made to remove the serum and unbounded antibodies; subsequently, a conjugated anti-body specific for human immunoglobulins is added to each well. Following further washing, a chromogenic (or fluorescent or chemiluminescent) substrate is finally added and metabolized in the presence of antibodies against SARS-CoV-2 and enzyme-conjugated anti-human antibodies. This process gives rise to a colorimetric reaction easily detectable by optical densitometry (or fluorescence or chemiluminescence) whose intensity is indicative of the quantity of IgG, IgM or IgA anti-body presents in the sample (Fig. 6A).

Similarly, 96-well commercial COVID-19 ELISA sandwich tests contain immobilized antibodies against SARS-CoV-2 antigens at the bottom of each well able to bind antigens contained in the serum samples of patients. Subsequently, a series of washes is performed to remove the serum and unbounded antigens and a conjugated antibody specific for SARS-CoV-2 antigens is added to each well. Finally, following further washing, a chromogenic substrate is added and metabolized in the case of positive samples giving rise to a colorimetric reaction (Fig. 6B).

As regards the management of the COVID-19 pandemic, ELISAs are currently used for the detection of IgM and IgG antibodies (86) specific for SARS-CoV-2 antigens or for the identification of viral Spike proteins (87). More recently, other ELISAs have been developed for the detection of human IgA antibodies whose determination is of fundamental importance as they are the first antibodies to be produced following exposure to the virus (86).

In a very short period of time, several COVID-19 ELISAs have been developed and some of these have been approved by international agencies (Table SIII). These tests are mainly adopted to monitor the immunological status of patients or for the immunosurveillance of specific work categories, such as healthcare workers or school personnel (88,89). In particular, ELISA tests for the detection of IgG and IgM are often performed on COVID-19 patients who obtained a negative result in molecular tests conducted on nasopharyngeal swabs in order to ascertain the seroconversion of patients and the acquisition of immunocompetence against COVID-19 infection (90). Contrariwise, ELISAs for the detection of viral proteins and anti-SARS-CoV-2 human IgA can be used for diagnostic purposes or for large screening strategies, as these molecules are rapidly founded in clinical samples (83,91). Of note, ELISAs for the detection of viral antigens are based on sandwich ELISA instead of indirect ELISA (92).

Several ELISAs have been approved for the management of COVID-19 infection, allowing the identification of individuals exposed to the SARS-CoV-2 virus (Table SIII); however, these tests are not able to confirm the infectious status of the subject. Indeed, in the case of an ELISA-positive result (either IgA, IgM, IgG or viral antigens), it is necessary to perform molecular confirmatory analyses on nasopharyngeal swabs (93). However, ELISA-based serological tests are much more reliable than rapid antigen or antibody tests; the sensitivity and specificity values of these tests range from 75.6 to 100%, and 85.7 to 100%, respectively, albeit important variations in these values may be related to both the manufacturer or the human antibody or viral antigen tested (94). Indeed, generally, the search for IgG is more accurate compared to that of IgM or IgA (this latter is less sensitive) (94). Of note, both sensitivity and specificity depend on the timing of infection and the timing of test execution.

Overall, SARS-CoV-2 ELISAs represent a good clinical option for large screening and surveillance campaigns mainly adopted for specific work categories due to the rapidity of this method, the possibility of analyzing multiple samples in one round and the availability of automated or semi-automated systems that allow a precise quantitation of viral antigens or human antibodies.

The COVID-19 pandemic has prompted research groups worldwide to develop novel technologies for the diagnosis of COVID-19 infection or to readapt existing diagnostic systems according to the characteristics of the new SARS-CoV-2 virus.

When the infection had not yet assumed the dimension of a global pandemic, traditional culture methods, clinical investigations and NGS techniques were among the first methods used to diagnose COVID-19 infection. Apart from these approaches, other methods have been developed, including biosensors, CRISPR/Cas-based tests, nucleic acid isothermal amplification methods, electron microscopy, etc. (25,95).

The COVID-19 pandemic has highlighted the importance of laboratory diagnostics in the management of a health emergency that has deeply affected the social, economic and health fabric of the entire world. At present, nucleic acid amplification methods represent the gold standard for the diagnosis of COVID-19 infection with several RT-PCR-based tests approved by different national and international regulatory agencies. These tests, together with clinical and radiological investigations, have significantly improved the ability to correctly and rapidly diagnose a COVID-19 infection compared to the first diagnostic methods mainly represented by viral culture analysis and de novo sequencing of the SARS-CoV-2 genome. However, despite the high sensitivity of RT-PCR, patients with a low viral load are not often correctly diagnosed. In addition, this technique requires confirmatory analyses, trained personnel and expensive instruments and reagents that limit its application, especially in low-income countries. For these purposes, other methods have been developed to overcome the limitations of RT-PCR. Among these methods, CRISPR/Cas-based assays and isothermal amplification methods represent promising low-cost diagnostic strategies that can be used for the effective diagnosis of COVID-19 infection in middle- and low-income countries. Moreover, more sensitive molecular methods, including ddPCR and biosensors, are beginning to be approved by international agencies and used for the diagnosis of COVID-19 infection, as well as for the monitoring of viral load in hospitalized patients or quarantined individuals.

Apart from these widely used diagnostic strategies, other techniques have been developed for the screening of SARS-CoV-2 infection in the population or in high-risk environments, such as hospitals and schools. Among these, rapid antigen and rapid antibody tests and immunoenzymatic serological tests represent the most used techniques for the monitoring of COVID-19 infection spreading. Of note, despite the low-cost of such techniques, the low sensitivity and specificity of LFIAs and ELISAs do not allow clinicians to formulate a precise diagnosis of COVID-19 infection that should be confirmed with other techniques. However, the use of such point-of-care testing made it possible to implement effective health surveillance systems that allowed the effective management of the COVID-19 pandemic thus limiting the number of infections.

At present, other techniques are essential for the genomic surveillance of the pandemic. Indeed, NGS techniques and SARS-CoV-2 whole-genome sequencing are fundamental for the precise characterization of novel genetic variants of SARS-CoV-2. However, the high costs of these methods limit their application in clinical practice.

Overall, in a short period of time, the scientific community has developed several methods useful for correctly diagnosing a suspected case of COVID-19 infection. However, in order to perform a correct diagnosis of COVID-19, it is necessary to take into account, not only the test to be used, but also the patient's medical history, the time of the suspicious SARS-CoV-2 exposure, the type of sample to be collected and analyzed and how to interpret the result. Only by integrating all these elements, it will be possible to formulate a correct diagnosis of COVID-19 infection and effectively manage the COVID-19 pandemic.