Send to

Choose Destination
Sci Rep. 2020 Jan 24;10(1):1135. doi: 10.1038/s41598-020-57876-0.

Quality control of multiplex antibody detection in samples from large-scale surveys: the example of malaria in Haiti.

Author information

Department of Infection Biology, London School of Hygiene & Tropical Medicine, London, UK.
Laboratoire National de Santé Publique, Port-au-Prince, Haiti.
Department of Infection Biology, London School of Hygiene & Tropical Medicine, London, UK.
Centre of Statistics and Applications, University of Lisbon, Lisbon, Portugal.
Center for Applied Malaria Research and Evaluation, Tulane University School of Public Health & Tropical Medicine, New Orleans, Louisiana, USA.
Department of Social and Preventive Medicine, University of Montreal School of Public Health, Montreal, Canada.
Malaria Branch, Division of Parasitic Diseases and Malaria, Centers for Disease Control and Prevention, Atlanta, Georgia, USA.
CDC Foundation, Atlanta, Georgia, USA.
Ministère de la santé publique et de la population, Port-au-Prince, Haiti.


Measuring antimalarial antibodies can estimate transmission in a population. To compare outputs, standardized laboratory testing is required. Here we describe the in-country establishment and quality control (QC) of a multiplex bead assay (MBA) for three sero-surveys in Haiti. Total IgG data against 21 antigens were collected for 32,758 participants. Titration curves of hyperimmune sera were included on assay plates, assay signals underwent 5-parameter regression, and inspection of the median and interquartile range (IQR) for the y-inflection point was used to determine assay precision. The medians and IQRs were similar for Surveys 1 and 2 for most antigens, while the IQRs increased for some antigens in Survey 3. Levey-Jennings charts for selected antigens provided a pass/fail criterion for each assay plate and, of 387 assay plates, 13 (3.4%) were repeated. Individual samples failed if IgG binding to the generic glutathione-S-transferase protein was observed, with 659 (2.0%) samples failing. An additional 455 (1.4%) observations failed due to low bead numbers (<20/analyte). The final dataset included 609,438 anti-malaria IgG data points from 32,099 participants; 96.6% of all potential data points if no QC failures had occurred. The MBA can be deployed with high-throughput data collection and low inter-plate variability while ensuring data quality.

Supplemental Content

Full text links

Icon for Nature Publishing Group Icon for PubMed Central
Loading ...
Support Center