https://orcid.org/0000-0003-3702-2716
Figures
Abstract
The live attenuated tetravalent CYD-TDV vaccine (Dengvaxia) is effective but has scarcely been used due to safety concerns among seronegative recipients. Rapid diagnostic tests (RDTs) which can accurately determine individual dengue serostatus are needed for use in pre-vaccination screening. This study aimed to determine the performance of existing RDTs (which have been designed to detect levels of immunoglobulin G, IgG, associated with acute post-primary dengue) when repurposed for detection of previous dengue infection (where concentrations of IgG are typically lower). A convenience sample of four-hundred-and-six participants including 217 children were recruited during a community serosurvey. Whole blood was collected by phlebotomy and tested using Bioline Dengue IgG/IgM (Abbott) and Standard Q Dengue IgM/IgG (SD Biosensor) RDTs in the field. Serum samples from the same individuals were also tested at National Health Laboratory. The Panbio indirect IgG ELISA was used as a reference test. Reference testing determined that 370 (91.1%) participants were dengue IgG seropositive. Both assays were highly specific (100.0%) but had low sensitivity (Bioline = 21.1% and Standard Q = 4.6%) when used in the field. Sensitivity was improved when RDTs were used under laboratory conditions, and when assays were allowed to run beyond manufacturer recommendations and read at a delayed time-point, but specificity was reduced. Efforts to develop RDTs with high sensitivity and specificity for prior dengue infection which can be operationalised for pre-vaccination screening are ongoing. Performance of forthcoming candidate assays should be tested under field conditions with blood samples, as well as in the laboratory.
Author summary
Dengvaxia is effective but has scarcely been used due to safety concerns among seronegative recipients. Rapid diagnostic tests (RDTs) which can accurately determine individual dengue serostatus are therefore needed for pre-vaccination screening. This study evaluated two commercially available RDTs during a community seroprevalence survey in Timor-Leste. Field conditions were intended to replicate a pre-vaccination screening setting. Both assays were specific but had low sensitivity, which was lower than previous studies conducted in laboratories using stored serum samples. Future candidate assays should be evaluated under field conditions before they can safely be used in pre-vaccination screening.
Citation: Arkell P, Tanesi M, Gomes N, Joao JC, Oakley T, Bosco F, et al. (2022) Field evaluation of rapid diagnostic tests to determine dengue serostatus in Timor-Leste. PLoS Negl Trop Dis 16(11): e0010877. https://doi.org/10.1371/journal.pntd.0010877
Editor: Renata Rosito Tonelli, Universidade Federal de Sao Paulo, BRAZIL
Received: July 23, 2022; Accepted: October 7, 2022; Published: November 7, 2022
Copyright: © 2022 Arkell et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: This work was supported by the Department for Foreign Affairs and Trade, Australian Government [Complex Grant Agreement Number 75889, https://www.dfat.gov.au/]. Funding was received by JRF. The funders did not play any role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Dengue is a mosquito-borne disease caused by dengue viruses 1–4 which are endemic in more than 125 countries throughout tropical and sub-tropical regions [1,2]. Due to international travel, urbanisation and climate change, global incidence has increased rapidly during the last 50 years [3]. Half the world’s population are at risk of dengue, with 100–400 million infections occurring annually [1,4]. Timor-Leste is an island nation in Southeast Asia where dengue causes seasonal outbreaks across all 13 municipalities and is a major cause of morbidity and mortality among children [5].
The live attenuated tetravalent CYD-TDV vaccine (Dengvaxia) is licensed in several dengue-endemic countries for individuals 9–45 years of age. In 2017, long-term safety data showed a 1.75-fold increased risk of hospitalisation due to dengue and severe dengue among seronegative participants who received vaccination (compared with those who were seronegative but unvaccinated) [6]. The World Health Organization (WHO) therefore recommends vaccination after individuals are screened for prior infection in most settings [7]. Rapid diagnostic tests (RDTs) which can accurately determine individual dengue serostatus are needed [8,9].
Existing RDTs which detect IgG have been designed for the diagnosis of dengue in patients with acute febrile illness and/or for differentiation of primary vs. post-primary dengue. Performance is therefore likely to have been optimised based on high concentrations of IgG associated with acute infection. The potential for these devices to be repurposed for detecting previous dengue infection in healthy individuals (where typically IgG concentrations have waned over time and are lower) has been explored in previous studies. Devices have generally been shown to have high specificity but lower sensitivity when compared to traditional serological techniques, with performance varying between assays. So far, studies of commercially available RDTs have occurred in laboratories and have exclusively analysed serum samples (and not whole blood). Table 1 summarises previously published studies which have assessed the performance of various RDTs in determining dengue serostatus in healthy individuals [10–17].
The primary aim of this study was to determine the sensitivity and specificity of two RDTs when repurposed to determine dengue seropositivity in healthy individuals. Secondary aims were to compare performance of devices when used under field conditions (i.e. whole blood tested by research nurses in the community, in the presence of participants) to performance when used under laboratory conditions (i.e. serum samples tested by laboratory scientists in a diagnostic laboratory setting), explore whether any deficiencies are likely to be in-part due to assay limits-of-detection, and explore whether performance could be enhanced by allowing assays to run beyond manufacturer recommendations and read at a delayed time-point.
Methods
Ethics statement
This study received ethical approval from the Research Ethics and Technical Committee of the Instituto Nacional da Saude, Timor-Leste (Reference: 875 MS-INS/DGE/IX/2021) and the Human Research Ethics Committee of the Northern Territory Department of Health and Menzies School of Health Research, Australia (Reference: 2021–4064)
Participants
A population-representative sample of individuals aged one year or older living in Dili was identified as part of a three-stage cluster sampling design for a national seroprevalence study of vaccine-preventable diseases (including dengue) in Timor-Leste. Potential participants were visited at their homes. The first 200 adults and first 200 children (aged < 16 years) who were recruited within Dili municipality were also included in this sub-study, evaluating the performance of dengue RDTs. This sample size was considered adequate to estimate 70% sensitivity and 95% specificity of RDTs with 5% precision (alpha = 0.05) [18]. Since there has been no previous population serological surveillance of dengue in Dili, seroprevalence was assumed to be 80% (i.e. similar to other urban Southeast Asian settings with high burden of dengue) [19].
Sample collection
Research nurses collected blood from participants by phlebotomy using either a standard hypodermic or a winged butterfly needle with a syringe attached. A small volume of whole blood was applied to the index test RDTs (methods below), while most of the sample was injected into a gel serum (SST) tube, which underwent centrifugation in the field (1,500 RCF, 10 minutes). Separated serum samples were transported to the National Health Laboratory (NHL) in Dili within a portable refrigerator and then stored at 4 degrees Celsius until analysis.
Reference test
Serum samples were analysed using the Panbio indirect IgG ELISA (product code 01PE30). This assay was chosen because it has been has been designed for detection of previous dengue (as opposed to diagnosis of acute dengue infection), it has been used in serosurveillance studies previously [20,21], it has acceptable performance when compared to virus neutralisation testing (considered most accurate in determining dengue serostatus and correlates best with outcomes after Dengvaxia) [12,22], and because laboratory equipment and expertise for its use were available at the NHL. Testing was performed in accordance with manufacturers’ instructions: A cut-off value was calculated for each run by multiplying the calibration factor (batch-specific) by the mean optical densities (ODs) of calibrator material (run in triplicate). An index value (IV) was then calculated by dividing the sample OD by the cut-off value. Samples with IV < 0.9 were assigned ‘dengue IgG negative’. Samples with index value > 1.1 were assigned dengue IgG positive. Samples with IV 0.9–1.1 were repeat-tested with the second result being used. If IV remained 0.9–1.1 on repeat testing then the sample was assigned ‘dengue IgG positive’. The appropriateness of these commercial serological cut-offs was assessed visually by plotting a histogram showing ELISA IVs from all serum samples in the study, which was assessed visually (see S1 Fig). The cut-off values appeared to adequately distinguish two populations of antibody responses and were therefore adopted into the study.
Testing with RDTs
Participants were tested for the presence of dengue antibodies using two different commercially available dengue RDTs (Bioline Dengue IgG/IgM, Abbott and Standard Q Dengue IgM/IgG, SD Biosensor). These were chosen based their positive performance in previous studies (Bioline sensitivity 53.7–71.1%, specificity 96.0–99.6%) [10,12], and their availability and low cost in Timor-Leste (Standard Q). Initially, RDTs were used under field conditions (i.e. whole blood was tested by research nurses in the community, in the presence of participants). This was intended to mimic a potential pre-vaccination screening setting. Subsequently, RDTs were used under laboratory conditions (i.e. serum samples from the same individuals were tested by laboratory scientists at NHL, which is a diagnostic laboratory).
In both test settings RDTs were operated in accordance with manufacturers’ instructions. Each device was labelled with the participant ID number and placed on a flat surface. In the field they were inoculated with two drops of whole blood from the syringe. In the laboratory, they were inoculated with two drops of serum from the SST tube. A timer was set for 15 minutes, after which the IgG and IgM lines were interpreted visually by a single researcher and reported in real-time. For the laboratory (serum) experiment, the assays were also read at 60 minutes, to assess whether this would affect test performance. Research nurses and laboratory scientists attended classroom teaching sessions on the safe and appropriate use the RDTs which included practicing operating each device and being observed, which was delivered by PA.
Participant management and ethical considerations
Written consent was obtained from the participants and also from the parent/guardian (in the case of children). Participants were informed of their RDT results in the field. If dengue IgG was detected by either assay (but IgM was not detected), participants were informed that they had likely been infected with dengue at some time in their lives, but likely not recently. If IgM (+/- IgG) was detected by either assay, participants were informed that there was evidence of current or recent dengue infection. These individuals were advised to seek medical attention if they experienced fever or other acute illness during the subsequent two weeks. These interpretations were agreed by members of the research team who have backgrounds in medical microbiology, adult, and paediatric infectious disease (PA, JY and JF). They are consistent with interpretations suggested by the RDT instructions for use, with the exception of isolated IgG seropositivity which was considered significantly more likely to represent previous infection than acute dengue in this group of asymptomatic participants.
Statistical analysis
The sensitivity and specificity of each RDT for determining IgG seropositivity under each set of operating conditions was assessed against the reference test. To explore whether RDT insensitivity was likely to be in-part due to an assay limit-of-detection (LOD), ELISA IVs from reference testing were analysed as a proxy for antibody concentrations. RDT true positive samples were compared to RDT false negative samples and unpaired Students T-tests were used to determine whether there was a significant difference. Results were considered significant where p < 0.05.
Results
Four-hundred-and-six individuals were included in this study. This included 217 children whose median (interquartile range, IQR) age in years was 10 (6–13) and 189 adults whose median (IQR) age in years was 33 (24–42). Reference testing determined that 186/217 (85.7%) children and 184/189 (97.4%) adults were dengue IgG seropositive (overall seropositivity = 91.1%).
Performance of RDTs when testing blood samples under field conditions
The sensitivity and specificity of the Bioline assay when testing blood under field conditions (interpreted after 15 minutes, as per manufacturers’ instructions) were 21.1% and 100.0%, respectively. The sensitivity and specificity of the Standard Q under similar conditions were 4.6% and 100.0%, respectively.
Performance of RDTs when testing serum under laboratory conditions
The sensitivity and specificity of the Bioline assay when testing serum under laboratory conditions (interpreted after 15 minutes, as per manufacturers’ instructions) were 42.7% and 97.2%, respectively. When tests were allowed to run for 60 minutes (i.e. beyond manufacturer recommendations) sensitivity was improved to 80.3% but specificity reduced to 94.4%.
The sensitivity and specificity of the Standard Q assay when testing serum under laboratory conditions (interpreted after 15 minutes, as per manufacturers’ instructions) were 10.5% and 97.2%, respectively. When tests were allowed to run for 60 minutes (i.e. beyond manufacturer recommendations) sensitivity was improved to 38.4% and specificity remained 97.2%. Table 2 summarises the performance of each RDT under different test conditions.
Analysis of ELISA IVs as a proxy for antibody concentration
Among participants whose Bioline RDT was true positive when used under field conditions the mean (standard deviation, SD) IV of the reference ELISA test was 3.808 (0.535), compared to 3.300 (0.779) in participants whose Bioline RDT was false negative (p < 0.001). Among participants whose Standard Q test was true positive when used under field conditions the mean (SD) ELISA IV of the reference ELISA test was 3.740 (0.441), compared to 3.392 (0.772) in participants whose Standard Q RDT was false negative (p = 0.007). This indicates the RDTs were more likely to accurately identify positive samples which have higher dengue IgG concentrations and hence RDT insensitivity is at least partially due to an assay limit-of-detection problem. Fig 1 shows ELISA IVs in each of these groups. Table 3 summarises similar comparisons of ELISA IVs across different test conditions. Study data are available in the supplementary information (S1 Data).
Comparison of reference enzyme-linked immunosorbent assay (ELISA) index values (as a proxy for antibody concentration) between individuals who were rapid diagnostic test (RDT) -true positive (TP) and RDT-false negative (FN) when Bioline Dengue IgG/IgM, Abbott (A) and Standard Q Dengue IgM/IgG, SD Biosensor (B) assays were used under field conditions.
Discussion
This study evaluated the performance of two RDTs in determining dengue IgG seropositivity among healthy individuals (using serum ELISA as the reference test). It is the first study in which commercially available RDTs have been evaluated for this purpose using blood samples, and the first such study conducted under field conditions intended to replicate a pre-vaccination screening setting. When operated according to manufacturers’ instructions (i.e. read at 15 minutes), the Bioline and Standard Q assays were highly specific (both 100.0%) but had low sensitivity (21.1% and 4.6%, respectively). These results contrast with previous studies which have tested serum samples under laboratory conditions and found sensitivity of the Bioline assay to be 53.7–71.1%.[10,12] Indeed, when RDTs were operated under laboratory conditions using serum samples from the same individuals in this study, the sensitivity of both assays appeared improved (42.7% and 10.5%, respectively). When tests were allowed to run for 60 minutes (i.e. beyond manufacturer recommendations) sensitivity was further improved (80.3% and 38.4% respectively), but this appeared to be at the expense of specificity (which reduced to 97.2% and 94.4%, respectively). Similar observations were made in a recent study which aimed to optimise both commercially available and purpose-designed RDTs by varying run-time [15].
This study also found that mean IV of the reference ELISA test was consistently significantly higher in participants whose RDTs were true positive compared with those whose RDTs were false negative, indicating that RDT insensitivity is at least partially due to a limit-of-detection problem with these two RDTs. This is consistent with assays having been developed for the diagnosis of dengue in patients with acute febrile illness and differentiation of primary vs. post-primary dengue (where concentrations of IgG are usually higher compared to those seen long after convalescence). Previous studies have found that some ELISAs which detect dengue IgG can also be insensitive, particularly in individuals who have lower titres of neutralising antibodies and/or where there is a monotypic antibody profile [22,23]. Importantly, the latter group of individuals are most likely to benefit most from vaccination but would be denied this if pre-vaccination screening was falsely negative.
Additional potential sources of low RDT sensitivity include the following: First, RDTs may be interpreted too soon after inoculation. This is supported by data from the present study which showed a difference in sensitivity when time-to-interpretation was varied. Second, there may be difficulties in the visual interpretation of RDTs, particularly when operators are working under field conditions (where the level of lighting may be low). Fieldworkers in the present study received classroom training in the use of both RDTs and were observed to ensure proper technique, however there was no formal programme of quality assurance to ensure proficiency in visual interpretation. The use of lateral flow readers and/or mobile phone app readers [24], including those which verify run-time (or would not make a determination before the appropriate run-time has elapsed), may improve performance. Larger field studies which assess performance at fine temporal resolutions to determine the optimal run-time may be useful. Third, while the chosen reference test (Panbio indirect IgG ELISA) has been designed for detection of previous dengue (as opposed to diagnosis of acute infection), and has been used in serosurveillance studies previously [20,21], it may have mis-assigned some samples. This could have occurred because the commercial serological cut-off was used rather than any population-specific alternative derived from the observed antibody responses. While the appropriateness of cut-offs were assessed visually, the sample size for this study was not considered large enough to undertake mixture modelling [25,26]. Furthermore, virus neutralisation assays may have been more accurate in determining dengue serostatus (and could correlate better with outcomes after Dengvaxia) than ELISAs [12,22], however they were not available during this study which took place in Timor-Leste.
Although this study found RDTs to have high specificity, false positive results were observed. These could have been due to antibodies against other flaviviruses which may be circulating in Timor-Leste including zika and Japanese encephalitis viruses, which commonly cross-react in serological assays [27]. If used in pre-vaccination screening, RDTs which give false positive results would give rise to inappropriate vaccine administration. As such, it would be important to further characterise any false positive samples from candidate assays and perform more extensive specificity testing using panels of samples which are known to have potentially cross-reacting antibodies.
This study identified high dengue seroprevalence among both adults and children which is consistent with other studies from the Southeast Asian region. A previous study from Timor-Leste identified lower seroprevalence, but that study tested dried blood spots from individuals in rural areas and used a less sensitive ELISA [28]. Results from the ongoing population-representative serosurvey (from which this study took a convenience sample) will therefore be crucial in determining the most appropriate disease control strategy. If seroprevalence in the vaccine candidate age-group is very high (>70%), universal administration of Dengvaxia without pre-vaccination screening may be appropriate, because the number of severe cases prevented in those who were seropositive would likely be substantially greater than the excess number that occurred in those who were seronegative [7].
Dengue is an emerging infection which causes significant morbidity and mortality across tropical and sub-tropical regions. However, the only effective vaccine has scarcely been used because of safety concerns among seronegative recipients. Efforts to develop an assay with high sensitivity and specificity for prior dengue infection which can be operationalised for pre-vaccination screening are ongoing. Performance of forthcoming candidate assays should be tested under field conditions with blood samples (as well as in the laboratory).
Supporting information
S1 Fig. Histogram showing antibody responses in serum samples from all individuals with commercial cut-off values shown.
https://doi.org/10.1371/journal.pntd.0010877.s002
(PDF)
References
- 1. Bhatt S, Gething PW, Brady OJ, Messina JP, Farlow AW, Moyes CL, et al. The global distribution and burden of dengue. Nature. 2013;496: 504–507. pmid:23563266
- 2. Zeng Z, Zhan J, Chen L, Chen H, Cheng S. Global, regional, and national dengue burden from 1990 to 2017: A systematic analysis based on the global burden of disease study 2017. EClinicalMedicine. 2021;32. pmid:33681736
- 3. Messina JP, Brady OJ, Golding N, Kraemer MUG, Wint GRW, Ray SE, et al. The current and future global distribution and population at risk of dengue. Nat Microbiol. 2019;4: 1508–1515. pmid:31182801
- 4. Brady OJ, Gething PW, Bhatt S, Messina JP, Brownstein JS, Hoen AG, et al. Refining the Global Spatial Limits of Dengue Virus Transmission by Evidence-Based Consensus. PLoS Negl Trop Dis. 2012;6. pmid:22880140
- 5. Wangdi K, Clements ACA, Du T, Nery SV. Spatial and temporal patterns of dengue infections in Timor-Leste, 2005–2013. Parasites and Vectors. 2018;11: 1–9. pmid:29301546
- 6. Sridhar S, Luedtke A, Langevin E, Zhu M, Bonaparte M, Machabert T, et al. Effect of Dengue Serostatus on Dengue Vaccine Safety and Efficacy. N Engl J Med. 2018 [cited 13 Jul 2021]. pmid:29897841
- 7. Organisation WH. Dengue vaccines: WHO position paper. Wkly Epidemiol Rec. 2018; 457–76.
- 8. Luo R, Fongwen N, Kelly-Cirino C, Harris E, Wilder-Smith A, Peeling RW. Rapid diagnostic tests for determining dengue serostatus: a systematic review and key informant interviews. Clin Microbiol Infect. 2019;25: 659–666. pmid:30664935
- 9. Wilder-Smith A, Smith PG, Luo R, Kelly-Cirino C, Curry D, Larson H, et al. Pre-vaccination screening strategies for the use of the CYD-TDV dengue vaccine: A meeting report. Vaccine. 2019;37: 5137–5146. pmid:31377079
- 10. Bonaparte M, Zheng L, Garg S, Guy B, Lustig Y, Schwartz E, et al. Evaluation of rapid diagnostic tests and conventional enzyme-linked immunosorbent assays to determine prior dengue infection. J Travel Med. 2019;26. pmid:31616949
- 11. Bonaparte M, Huleatt J, Hodge S, Zheng L, Lustig Y, DiazGranados CA, et al. Evaluation of dengue serological tests available in Puerto Rico for identification of prior dengue infection for prevaccination screening. Diagn Microbiol Infect Dis. 2020;96. pmid:31839333
- 12. DiazGranados CA, Bonaparte M, Wang H, Zhu M, Lustig Y, Schwartz E, et al. Accuracy and efficacy of pre-dengue vaccination screening for previous dengue infection with five commercially available immunoassays: a retrospective analysis of phase 3 efficacy trials. Lancet Infect Dis. 2020;3099: 1–8. pmid:33212068
- 13. Chong ZL, Soe HJ, Ismail AA, Mahboob T, Chandramathi S, Sekaran SD. Evaluation of the diagnostic accuracy of a new biosensors-based rapid diagnostic test for the point-of-care diagnosis of previous and recent dengue infections in Malaysia. Biosensors. 2021;11. pmid:33921935
- 14. Daag JV, Ylade M, Adams C, Jadi R, Crisostomo MV, Alpay R, et al. Evaluation of a new point-of-care test to determine prior dengue infection for potential use in pre-vaccination screening. Clin Microbiol Infect. 2020. pmid:32866651
- 15. Echegaray F, Laing P, Hernandez S, Marquez S, Harris A, Laing I, et al. Adapting Rapid Diagnostic Tests to Detect Historical Dengue Virus Infections. Front Immunol. 2021;12. pmid:34367162
- 16. Limothai U, Tachaboon S, Dinhuzen J, Hunsawong T, Ong-Ajchaowlerd P, Thaisomboonsuk B, et al. Dengue pre-vaccination screening test evaluation for the use of dengue vaccine in an endemic area. PLoS One. 2021;16. pmid:34507347
- 17. Savarino SJ, Bonaparte M, Wang H, Dayan GH, Forrat R, Zhu M, et al. Accuracy and efficacy of pre-dengue vaccination screening for previous dengue infection with a new dengue rapid diagnostic test: a retrospective analysis of phase 3 efficacy trials. The Lancet Microbe. 2022;3: e427–e434. pmid:35659904
- 18. Hajian-Tilaki K. Sample size estimation in diagnostic test studies of biomedical informatics. J Biomed Inform. 2014;48: 193–204. pmid:24582925
- 19. Vongpunsawad S, Intharasongkroh D, Thongmee T, Poovorawan Y. Seroprevalence of antibodies to dengue and chikungunya viruses in Thailand. PLoS One. 2017;12. pmid:28662144
- 20. Garg S, Chakravarti A, Singh R, Masthi NRR, Goyal RC, Jammy GR, et al. Dengue serotype-specific seroprevalence among 5- to 10-year-old children in India: a community-based cross-sectional study. Int J Infect Dis. 2017;54: 25–30. pmid:27825949
- 21. Prayitno A, Taurel AF, Nealon J, Satari HI, Karyanti MR, Sekartini R, et al. Dengue seroprevalence and force of primary infection in a representative population of urban dwelling Indonesian children. PLoS Negl Trop Dis. 2017;11. pmid:28617803
- 22. Lopez AL, Adams C, Ylade M, Jadi R, Daag JV, Molloy CT, et al. Determining dengue virus serostatus by indirect IgG ELISA compared with focus reduction neutralisation test in children in Cebu, Philippines: a prospective population-based study. Lancet Glob Heal. 2021;9: e44–e51. pmid:33212030
- 23. Biggs JR, Sy AK, Ashall J, Santoso MS, Brady OJ, Reyes MAJ, et al. Combining rapid diagnostic tests to estimate primary and post-primary dengue immune status at the point of care. PLoS Negl Trop Dis. 2022;16: e0010365. pmid:35507552
- 24. Kadam R, White W, Banks N, Katz Z, Dittrich S, Kelly-Cirino C. Target Product Profile for a mobile app to read rapid diagnostic tests to strengthen infectious disease surveillance. PLoS One. 2020;15. pmid:31995628
- 25. Bottomley C, Otiende M, Uyoga S, Gallagher K, Kagucia EW, Etyang AO, et al. Quantifying previous SARS-CoV-2 infection through mixture modelling of antibody levels. Nat Commun. 2021;12: 1–7.
- 26. Ward D, Gomes AR, Tetteh KKA, Sepúlveda N, Gomez LF, Campino S, et al. Sero-epidemiological study of arbovirus infection following the 2015–2016 Zika virus outbreak in Cabo Verde. Sci Rep. 2022;12: 1–8. pmid:35810191
- 27. Innis BL, Nisalak A, Nimmannitya S, Kusalerdchariya S, Chongswasdi V, Suntayakorn S, et al. An enzyme-linked immunosorbent assay to characterize dengue infections where dengue and Japanese encephalitis co-circulate. Am J Trop Med Hyg. 1989;40: 418–427. pmid:2540664
- 28. Arkell P, Angelina J, do Carmo Vieira A, Wapling J, Marr I, Monteiro M, et al. Integrated serological surveillance of acute febrile illness in the context of a lymphatic filariasis survey in Timor-Leste: a pilot study using dried blood spots. Trans R Soc Trop Med Hyg. 2021 [cited 11 Feb 2022]. pmid:34850241