Advancing the public health applications of Chlamydia trachomatis serology

Abstract

Genital Chlamydia trachomatis infection is the most commonly diagnosed sexually transmitted infection. Trachoma is caused by ocular infection with C trachomatis and is the leading infectious cause of blindness worldwide. New serological assays for C trachomatis could facilitate improved understanding of C trachomatis epidemiology and prevention. C trachomatis serology offers a means of investigating the incidence of chlamydia infection and might be developed as a biomarker of scarring sequelae, such as pelvic inflammatory disease. Therefore, serological assays have potential as epidemiological tools to quantify unmet need, inform service planning, evaluate interventions including screening and treatment, and to assess new vaccine candidates. However, questions about the performance characteristics and interpretation of C trachomatis serological assays remain, which must be addressed to advance development within this field. In this Personal View, we explore the available information about C trachomatis serology and propose several priority actions. These actions involve development of target product profiles to guide assay selection and assessment across multiple applications and populations, establishment of a serum bank to facilitate assay development and evaluation, and development of technical and statistical methods for assay evaluation and analysis of serological findings. The field of C trachomatis serology will benefit from collaboration across the public health community to align technological developments with their potential applications.

Introduction

Chlamydia trachomatis is an obligate intracellular bacterium that can cause both genital and ocular infections. Genital C trachomatis infection is the most common, curable STI worldwide, with an estimated 131 million new cases each year.¹ Most infections are asymptomatic, but if left untreated can cause scarring of the upper reproductive tract in women and lead to serious complications in women, including pelvic inflammatory disease, ectopic pregnancy, and tubal factor infertility.² Trachoma is caused by ocular infection with C trachomatis and is the leading infectious cause of blindness worldwide.³ WHO estimates that around 190 million people are at risk ofblindness from trachoma, with most of these people living in Africa. There has been substantial investment in the past decade in public health programmes to control both genital and ocular infection due to C trachomatis. Control efforts for genital chlamydia have largely focused on enhanced detection and treatment of C trachomatis among young adults, and for trachoma control efforts have focused on improved access to water, sanitation, and hygiene, and mass drug administration with azithromycin in affected communities.³ However, several important questions remain about C trachomatis epidemiology, the most effective means of control, and optimal models of surveillance.^4–6 Given the ongoing control efforts and the promise of new C trachomatis vaccines,⁷ robust methods are needed to allow monitoring of and insight into the prevalence and incidence of chlamydia infection,⁸ and the progression to scarring sequelae. Measures of current infection based on DNA or RNA detection (ie, using nucleic acid amplification tests) provide an inadequate understanding of these features of C trachomatis infection. Therefore, alternative approaches are required, and in recent years interest in the use of serological assays within the fields of genital chlamydia and trachoma has been increasing.

Methods to detect C trachomatis antibodies in serum have been available for several decades.⁹ However, use of serological assays has been limited by C trachomatis’ crossreactivity with other Chlamydia spp,¹⁰ suboptimal sensitivity of many assays,^11,12 an incomplete understanding about the longevity and clinical implications of C trachomatis antibodies, and the relationship between C trachomatis infection and antibody response.¹¹ Consequently, the widespread use of chlamydia seroepidemiology decreased among researchers for several years, as did funding. Following the development of novel, sensitive, and more specific C trachomatis serological assays,^13–17 there is now growing interest in the use of these assays as an epidemiological tool. For example, assays have been developed with the capability to detect antibodies against a range of C trachomatis antigens, lateral flow assays are being evaluated for field use, and dried blood spots have been used to facilitate specimen collection, transport, and storage ( table 1).

Table 1:

Chlamydia trachomatis serological assays and examples of their public health applications to date

	Platform or format	Antigens detected (antibody class or subclass)	Examples of public health applications
			Measure seroprevalence or estimate incidence	Investigate association with disease	Evaluate control interventions
					Genital Chlamydia trachomatis	Trachoma
ELISAs
Wills et al13	Indirect ELISA	Pgp3 (IgG)	18	··	18	··
Geisler et al14	Indirect ELISA	Elementary bodies from C trachomatis serovars D/UW-3, F/ICCal-13, and J/UW-36 (IgG1, IgG3)	14,19	20,21	··	··
Horner et al15	Double-antigen ELISA	Pgp3 (IgG, IgA, IgM)	15,18,22	··	18	··
Winstanley et al16	Indirect ELISA	pgp3 (IgG)	··	··	··	··
Albritton et al23	Indirect ELISA	Elementary bodies from C trachomatis serovars D/UW3/Cx and E/UW5/Cx (IgG, IgA)	··	23	··	··
Migchelsen et al24	Indirect ELISA on dried blood spots	Pgp3 (IgG)	24, 25	··	··	··
Menon et al26	Multi-peptide indirect ELISA	12-mer peptides derived from HtrA, hsp60, and Ct44327 (IgG)	··	26,28	··	··
Commercially available*	ELISA or EIA	MOMP and hsp60 (IgG, IgA)	29–31	21,32–39	··	··
Multiplex bead arrays
Goodhew et al17	Multiplex bead array	Pgp3 and CT694 (IgG, IgA)	17	··	··	40–43
Willhauck-Fleckenstein et al44	Multiplex bead array	MOMP A/D/L2, PorB, TARP, hsp60–1, and Pgp3 (IgG, or IgG, IgA, IgM)	··	··	··	··
Near-patient testing
Gwyn et al45	Lateral flow	Pgp3 (IgG, IgA, IgM)	45,46	··	··	··
Whole proteome microarray
Budrys et al47	ELISA-based proteome array	Representing 908 proteins of C trachomatis strain D/UW-3/Cx (IgG)	··	47	··	··
Lu et al48	Whole proteome microarray	Representing 908 genomic and plasmid open reading frames of C trachomatis strain D/UW3 (IgG, IgA, IgM)	··	48,49	··	··
Hufnagel et al50	Whole proteome microarray	Representing 895 proteins of C trachomatis strain D/ UW-3/Cx (IgG, or IgG, IgA, IgM)	··	50	··	··
Immunofluorescence
Chernesky et al51	Whole-cell inclusion immunofluorescence	L2 serovar (IgG, IgA, IgM)	··	52,53	··	··
Commercially available†	Micro-immunofluorescent assay	Elementary bodies from C trachomatis serovars D-K (IgG, IgA, IgM)	··	54	··	··
Wang55	Modified micro-immunofluorescent assay protocol	Elementary bodies from C trachomatis (IgG)	56	57	··	··

This table is not intended to be a comprehensive review. HtrA=high temperature requirement A protease. hsp60=heat shock protein 60. EIA=enzyme immunoassay. MOMP=major outer membrane protein. TARP=translocated actin-recruiting phosphoprotein.

^*Chlamydien IgG rELISA medac assay, medac, Wedel, Germany; Chlamydia trachomatis EIA assay kit, Pathtech, Melbourne, Australia; and SeroCT RT kit, Savyon Diagnostics, Ashdod, Israel.

^†Chlamydia MIF IgG, Focus Diagnostics, Cypress, USA and SeroFIA, Savyon Diagnostics, Ashdod, Israel.

The current understanding of mucosal immunity and C trachomatis immunology suggests that urogenital and ocular infections with C trachomatis lead to detectable IgG response with use of appropriate serological assays in most confirmed infections.^13–15,23 Several factors affect the magnitude of IgG response and the ability of serological assays to detect a previous C trachomatis infection, including the target antibody, assay used, time since infection, and patient characteristics such as age, sex, race, and the number of previous C trachomatis infections.^13,14,16,58 In a UK-based study^13,15 that compared several assays in the same population, the sensitivity to detect a previous known chlamydia infection ranged from 46% (IgG pELISA plus medac assay, medac, Wedel, Germany) to 83% (Pgp3 double-antigen) in women and 40% (SeroCT; Savyon Diagnostics, Ashdod, Israel) to 54% (Pgp3 double-antigen) in men when compared with a previous C trachomatis diagnosis by nucleic acid amplification tests. Seroreversion (ie, loss of detectable antibodies) has been shown in some cases but varies by infection history and assay,⁵⁸ with minimal loss of detectable antibody reported in one study using a Pgp3 double-antigen ELISA.¹⁵ In a study⁵⁹ of C trachomatis seroprevalence in the context of mass azithromycin treatment for trachoma prevention in a high prevalence area, no instances of seroreversion were observed after 6 months. C trachomatis serological assays can therefore be used to measure age-specific cumulative incidence,^11,18 despite representing a lower bound estimate due to potentially incomplete seroconversion and loss of detectable antibodies over time. C trachomatis antibody response has also been found to correlate with a known history of scarring sequelae. For example, titres of C trachomatis antibodies have been found to be higher in women with tubal factor infertility than in women without this disease and some specific antibodies have been found to be more common in women with known disease than in those without disease.⁴⁷ Thus, serological assays might be used as a potential biomarker of disease.

Because the sensitivity of serological assays is inadequate, C trachomatis serology has limited diagnostic value; in the absence of genetic diagnostic methods it is an accepted tool for presumptive diagnosis of lymphogranuloma venereum,⁶⁰ but it is not used for diagnosis of other biovars. Instead, C trachomatis serological assays have a potential role in monitoring and surveillance by providing a measure of history of C trachomatis exposure among individuals tested. When applied to appropriate samples, such measures might be useful to inform resource allocation and possible clinical need by indicating the size of the population affected, and the effect of population-based interventions such as screening or mass drug administration.

Therefore in this Personal View, we explore the potential public health applications of C trachomatis serology, discuss key challenges of its use, and finally propose priorities for research and development that might help advance the field of C trachomatis control. This work grew out of an expert meeting convened by Public Health England in September, 2016 (The Public Health applications of Chlamydia trachomatis serology; London, UK), and subsequent discussions of studies relating to different assays presented at the 2016 European Society for Chlamydia Research meeting (Oxford, Sept 6–9, 2016).

Public health applications of C trachomatis serology

C trachomatis serology provides a means of quantifying the prevalence and incidence of chlamydia infection. A thorough understanding ofpopulation-level C trachomatis prevalence and incidence is crucial to identify unmet needs for screening and treatment services and to evaluate the impact of C trachomatis control interventions. However, obtaining reliable estimates of C trachomatis prevalence and incidence is challenging. In the case of genital chlamydia, most C trachomatis infections are asymptomatic, so an increase in screening would lead to an increase in reported diagnoses.⁶¹ As a result, surveillance is often centred on only case-based reporting, with scarce information about numbers of patients tested, resulting in an underestimate of the true incidence of C trachomatis infection.⁶¹ Furthermore, comparability of surveillance data between countries is limited by differences in testing recommendations, performance characteristics of diagnostic tests, and reporting policies and practices.⁶² Even when the total number of people tested in a given country is known, interpreting estimates of the proportion of people who test positive for chlamydia infection is difficult, because the tested population has a different underlying risk from the general population. Therefore, the proportion testing positive will often present a biased measure of prevalence.^61,63 Few countries have undertaken surveys of prevalence in samples of the general population, and when they have been done,^64–66 the surveys were resource intensive. Such studies are unlikely to be feasible in many countries, because of the costs involved.

C trachomatis seroprevalence as a marker of cumulative incidence has been used in several countries^15,18,22,29_‘³⁰_‘⁶⁷ to assess C trachomatis epidemiology, and in some cases to investigate population effect of control interventions. In the field of trachoma, mass drug administration programmes have been successful in reducing C trachomatis infection and C trachomatis-related ocular disease.^24,68,69 Longitudinal C trachomatis serology monitoring has strong potential as a tool for post-elimination surveillance,^40–42 and so provides an opportunity to evaluate programme effectiveness and possibly a further understanding of the public health response needed in countries where trachoma has not been eliminated.⁷⁰ Subject to certain assumptions about the population sampled, C trachomatis incidence can be estimated from repeated, cross-sectional C trachomatis seroprevalence surveys,⁷¹ although it would be necessary to adjust for imperfect assay sensitivity and specificity (AE Ades, University ofBristol, personal communication). C trachomatis serology can also be used to detect step changes in exposures by birth cohort, which would be expected in the context of control measures.^40,72 Distinguishing between recent and past or long-standing infections would also help determine the incidence of disease, and methods are already being developed to enable C trachomatis serology to be used for this purpose.⁷³

Another potential application of C trachomatis serology is as a measure of diseases related to chlamydia infections, such as pelvic inflammatory disease or ectopic pregnancy. Because the end goal of C trachomatis control is to reduce the incidence of disease, monitoring of disease biomarkers and not just of infection might improve understanding of whether C trachomatis control is leading to a reduction in reproductive sequelae, even in the absence of substantial reductions in transmission. A C trachomatis-specific biomarker of disease would be particularly useful, because C trachomatis-related diseases might occur many years after the causative infection and C trachomatis is not the only cause of long-term reproductive complications such as pelvic inflammatory disease, ectopic pregnancy, and tubal factor infertility.² Measures ofthe proportion of long-term sequelae that are attributable to C trachomatis infection (the so-called population excess fraction) are also essential to determine the need for and cost-effectiveness of control interventions.

Serological methods have been used to investigate the relationship between C trachomatis infection and sequelae as well as to estimate the proportion of long-term sequelae attributable to genital C trachomatis infection.²⁰_‘²¹_‘³²_‘⁵²_‘⁷⁴ Novel approaches also offer some promise in this area; for example, Ades and colleagues⁵² have developed a method using finite mixture modelling of antibody titre to estimate the population excess fraction of tubal factor infertility caused by chlamydia infection. Additionally, proteomic arrays are also being assessed as a means of identifying serological fingerprints to indicate the presence of disease related to genital C trachomatis infection⁵⁰ and scarring following ocular infection.⁴⁹

Another potential application of C trachomatis serology is development and evaluation of C trachomatis vaccines. The joint WHO and National Institutes of Health (NIH) sexually transmitted infection vaccine roadmap⁷⁵ has set out the need for an effective C trachomatis vaccine. Substantial progress towards this vaccine has been made in recent years, with candidate vaccines now in the preclinical and clinical testing phases.⁷ Several priority action areas set out in the WHO and NIH roadmap might be addressed through the development and application of serological assays. This includes using serological assays to obtain better epidemiological data, improve understanding of the natural history of C trachomatis and burden of sequelae, and expedite clinical development and evaluation of candidate vaccines, thereby encouraging investment in C trachomatis vaccine development. Spe-cifically, if acting as biomarkers for disease C trachomatis serological assays could be used to obtain more complete and precise estimates of the global burden of C trachomatis-associated sequelae, which are important for establishing the public health rationale for vaccination and for potential investors to assess the possible effect of investing in any successful vaccine candidate.⁷

When a safe vaccine candidate does enter a phase 3 clinical trial, C trachomatis serology could help identify C trachomatis-naive participants for recruitment, and help develop vaccination strategies through an understanding of age-specific exposure. Vaccine evaluation would also benefit from a biomarker of tubal damage for use as part of a clinical endpoint for assessing vaccine efficacy, because of the current diagnostic inaccuracy for C trachomatis-related outcomes such as pelvic inflammatory disease.² The time and resources needed for a clinical trial of candidate C trachomatis vaccines with pelvic inflammatory disease or tubal factor infertility as outcomes might also be prohibitive. It is unclear whether serology will be able to provide such a biomarker of tubal damage, and serological methods might need to be used in combinaton with cellular markers or radiological findings. However, serology is an important area of research given the need for such measures in any future vaccine evaluation. Because C trachomatis infections (with the exception of lymphogranuloma venereum) are localised in the columnar epithelium, detection of antibodies from genital secretions has been proposed as a means of investigating correlates of immune protection against C trachomatis,²³ which might complement serological investigations. Assessment of vaccine-induced immune responses will depend on the vaccine’s mechanism of action. Assays that distinguish between the natural and vaccine-induced antibody response will therefore be needed.

Key challenges to the use of C trachomatis serology in public health

Although progress has been made in recent years, some important challenges remain within the field of C trachomatis serology that need to be addressed to improve the use and value of new serological assays in a public health context. Interpretation of C trachomatis seroprevalence is difficult; several complexities exist and include the following: not everyone exposed to C trachomatis will become infected, some individuals with this infection might not develop antibodies, women are more likely to develop detectable antibodies than men following urogenital C trachomatis in-fection,^13,18,76 C trachomatis antibodies are not specific to the infection site (ie, ocular or urogenital), and seroprevalence can vary with the number of previous infections and time since infection as antibodies develop or wane.⁵⁸ These complexities require careful consideration when planning studies and doing statistical analyses.

Determining assay sensitivity and specificity in the absence of universal guidelines is also challenging. The population that will be tested should be carefully considered when selecting positive and negative controls (ie, those with or without infection or disease) and setting assay thresholds. For example, a study aimed at investigating C trachomatis infection might need a different definition of positive and negative controls than a study aimed at investigating C trachomatis-associated disease. A further challenge involves the choice of thresholds that define C trachomatis antibody response in different populations, because serological assays might be affected by differences in cross-reactivity and background antibody concentrations, which can vary for example by country or ethnicity, or both.

The relative performance of different tests cannot easily be determined without evaluation against the same reference sera. Some laboratory-developed assays have been compared with commercial assays or other laboratory-developed assays,^13,15,45,46 but there are few data available to show how different assays perform within the same population. To establish performance characteristics of C trachomatis serological assays for different applications and populations, large numbers of serum samples linked to well characterised clinical and demographical information are needed. Serum samples from previous studies (eg, human papillomavirus vaccine trials⁷⁷ and HIV unlinked anonymous testing⁷⁸) or residual samples from clinical testing⁶⁷ could be used, but they often have limited clinical or demographical information, and varying access arrangements mean that assays have not been evaluated on comparable samples.

Optimal test characteristics might vary between different applications of C trachomatis serology,suggesting that different characteristics might be prioritised. For example, a test to measure whether someone has had C trachomatis infection will need to detect antibodies that persist over time at relatively low concentrations with high specificity. However, a test that is used to estimate the population excess fraction would ideally be able to distinguish between complicated and uncomplicated C trachomatis infections (eg, by identifying high versus low concentrations of antibody in serum,⁵² the antibody subclass,¹⁴ or antibodies specifically associated with complications). ⁴⁷

Similarly, the context in which an assay is to be deployed will influence prioritisation. For example, in a research setting, tests could be more operator-intensive and less cost-effective than tests used for ongoing surveillance given limited government budgets. Furthermore, a test requiring high volume of sera might be acceptable in a setting where additional blood can be collected from consenting patients, whereas surveillance systems relying on leftover sera from routine testing might have an inadequate volume available. Applications of C trachomatis serological assays in a surveillance context might be more tolerant to some reduced precision than when used within a vaccine trial, where previous infection needs to be ruled out to precisely define populations for inclusion in any efficacy analysis.

Priority actions for research and development

To address these challenges, we suggest three priority actions for research and development (panel).

Generating target product profiles

Target product profiles originated in the field of drug development to focus discussions between regulatory authorities and research sponsors. They allow the drug development process to be directed with the end goal in mind, so that both patient and market needs are met.⁷⁹ The process of establishing target product profiles is now commonly used in drug and vaccine development and their use has also extended into the field of diagnostics— eg, for tuberculosis⁸⁰ and point-of-care diagnostics for sexually transmitted infections.⁸¹ Target product profiles for C trachomatis serological assays should establish the minimal and optimal assay requirements for the different applications previously described. Table 2 sets out some of the initial considerations that can be used to inform target product profiles. A target product profile requires broad technical consultation across clinical, microbiological, and epidemiological fields, as well as representation from vaccine and diagnostic development companies, research groups, public health agencies, and funders.

Table 2:

Initial considerations that can be used to inform target product profiles for Chlamydia trachomatis serological assays

	Measuring infection	Measuring disease	For use in vaccine evaluation
Performance requirements	High sensitivity and specificity; ability to distinguish first from repeat infection; ability to measure recent infection; and lower precision than in vaccine studies would be sufficient for monitoring and surveillance applications	Ability to distinguish between complicated and uncomplicated infections; ability to identify disease-specific antigens or combinations of antigens or the magnitude of response associated with the disease; and high specificity for sequelae to prevent over investment in resource-poor environments arising from over estimation of the incidence of C trachomatis sequelae	Ability to quantify burden of infection and disease; high precision and high sensitivity for determining C trachomatis-naive status for trials and distinguishing between exposure and infection; to measure vaccine efficacy, availability of markers of tubal involvement, potentially in combination with other measures (eg, cellular markers, radiological measures); and markers of vaccine-induced immune response will depend on the mechanism of action of vaccine and will need to distinguish between vaccine-induced and natural responses
Dependencies	Appropriate collections of population-based sera	Appropriate collections of population-based sera to estimate disease incidence; and availability of reliable cases and controls with clear case definition to estimate population excess fraction	Vaccine trial design; and mechanism of action of vaccine candidate
Statistical methods	C trachomatis serological assay thresholds appropriate to the application and population; and relation between seroprevalence and cumulative or annual incidence should consider the impact of time since infection and repeat infections on Chlamydia trachomatis antibodies	C trachomatis serological assay thresholds appropriate to the application and population	C trachomatis serological assay thresholds appropriate to the application and population
Technical requirements	High throughput, low volume, and low resource use methods would be valued	High throughput, low volume, and low resource use methods for monitoring and surveillance applications; and should tolerate methods requiring higher specimen volume and operator intensive methods	Should tolerate methods requiring higher specimen volume, and operator intensive methods

Establishing a serum bank

A well defined serum bank focused on the evaluation of C trachomatis serology will be an invaluable resource. The value of serum banks in research of infectious diseases was recently set out in The Lancet,⁸² which highlighted the role serological studies could play in understanding worldwide distribution of disease and argued for the establishment of a World Serology Bank. The development of a C trachomatis-specific serum bank would enable clear and fair access to specimens and relevant epidemiological and clinical data (eg, age, sex, and a history of C trachomatis diagnosis).A serum bank would have the potential to support the development and evaluation of serological assays, and to facilitate identification of potential vaccine targets and correlates of protection. The bank should include sera from women and men of a variety of countries, ages, and ethnicities. These individuals would also have a range of characteristics that include varying histories of C trachomatis diagnosis, incorporating different times since treatment and different numbers of known infections; those with and without reproductive complications; and those exposed to potentially cross-reactive pathogens such as Chlamydia pneumoniae. Some applications of C trachomatis serology, such as the identification of biomarkers of scarring sequelae or developing serological assays to distinguish between infection and exposure, might also benefit from simultaneous assessment of cellular immunity. Therefore, collections that incorporate both serum and whole blood specimens would be particularly valuable, although it would require different regulatory permissions in some settings and would incur additional expenses for collection and storage arrangements.

Developing methods for assay evaluation and analysis

Shared protocols to guide assay evaluation would allow for comparability of estimates with use of C trachomatis serological methods across assays and increase consistency of reporting. Evaluation protocols should incorporate a consensus position on optimum methods of estimating sensitivity and specificity of C trachomatis serological assays, and recognise the need for selection of controls and assay thresholds to be determined according to the intended application while also considering the potential for cross-reactivity. Future efforts should also focus on development and application of statistical methods to appropriately analyse C trachomatis serological findings.

Conclusions

As the reported incidence of cases of C trachomatis infection remains high or continues to increase in many developed countries, the use of C trachomatis serology in several areas of public health is likely to be effective, and has already provided further insight into C trachomatis epidemiology and natural history. We explored the available information about C trachomatis serology and identified three priority actions that we believe would directly benefit public health and advance knowledge within the C trachomatis field.

As public health agencies continue to address the high rates of C trachomatis infection and the considerable morbidity that arises as a result, a more data-driven approach to programmatic decision making at the country, state, and municipality level is essential. Promising interventions, including vaccines, do and will need robust measures for estimating the population at risk and for determining the potential effect of prevention measures. C trachomatis serology is a promising additional tool for public health that could help improve the understanding of the populations at risk and could support the development of novel and effective interventions.

Panel: Priority actions to further develop the public health applications of Chlamydia trachomatis serological assays.

Generate target product profiles for C trachomatis serological assays

• What are the minimal or optimal characteristics of C trachomatis serological assays for different purposes (eg, design and evaluation of genital C trachomatis control programmes, design and evaluation of trachoma or ocular C trachomatis control programmes, or vaccine development and evaluation) and measures (eg, seroprevalence of C trachomatis antibodies as a measure of prevalence and incidence, measure of population excess fraction of disease such as pelvic inflammatory disease or tubal factor infertility, biomarker of disease either alone or in combination with other measures, measure of being C trachomatis-naive, or measure of vaccine-induced immune response)?

• What are the minimal or optimal characteristics of the aforementioned purposes and measures in different countries?

Establish a serum bank

• Adequate volumes of well characterised serum samples from individuals who have had C trachomatis infection and sera from those who have not had this infection should be included. Samples should be stored from individuals of a variety of ages and ethnicities, including children who might still have the maternal antibody, with different characteristics including the numbers of known infections, time since treatment, and presence of known reproductive tract or ocular complications.

• Standardised assessments of clinical outcome and epidemiological data should be incorporated.

• The serum bank should be established with appropriate access arrangements.

Develop methods for assay evaluation and analysis of serological findings

• Define how sensitivity and specificity should be estimated for different purposes.

• Define how positive and negative controls should be selected for different test applications.

• Better understand what thresholds should be used for each assay for different applications or test settings.

• Define how head-to-head comparison studies be done.

• Explore and assess what statistical methods should be used to measure epidemiological parameters (eg, incidence of infection and population excess fraction).