Reliable and Scalable SARS-CoV-2 qPCR Testing at a High Sample Throughput: Lessons Learned from the Belgian Initiative

Abstract

We present our approach to rapidly establishing a standardized, multi-site, nation-wide COVID-19 screening program in Belgium. Under auspices of a federal government Task Force responsible for upscaling the country's testing capacity, we were able to set up a national testing initiative with readily available resources, putting in place a robust, validated, high-throughput, and decentralized qPCR molecular testing platform with embedded proficiency testing. We demonstrate how during an acute scarcity of equipment, kits, reagents, personnel, protective equipment, and sterile plastic supplies, we introduced an approach to rapidly build a reliable, validated, high-volume, high-confidence workflow based on heterogeneous instrumentation and diverse assays, assay components, and protocols. The workflow was set up with continuous quality control monitoring, tied together through a clinical-grade information management platform for automated data analysis, real-time result reporting across different participating sites, qc monitoring, and making result data available to the requesting physician and the patient. In this overview, we address challenges in optimizing high-throughput cross-laboratory workflows with minimal manual intervention through software, instrument and assay validation and standardization, and a process for harmonized result reporting and nation-level infection statistics monitoring across the disparate testing methodologies and workflows, necessitated by a rapid scale-up as a response to the pandemic.

Keywords: SARS-CoV-2; data analysis; high-throughput testing; lab automation; qPCR; qc monitoring.

Figure 1

Analysis of the N gene across multiple sites and instruments. The data set used was a limited sample of validation data for a period of 10 days in July 2020. While not representative of the post-validation workflow, the analysis illustrates variation across instrumentation and site, which needs to be taken into account in data analysis and reporting.

Figure 2

Overview of logistics and sample flow shows the flow from sampling location to the testing site. The sample logistics and the related sample accession and result data flow are organized based on a single and unique sample number, which makes the data flow compliant to Global Data Privacy Requirements and facilitates sample chain of custody, traceability across the multiple sites, and centralized result reporting through a governmental health portal.

Figure 3

Data flow through the IT systems. At sample collection sites (bottom left), national identification and pre-labeled tubes are made known to the system. HL7 connectivity is used to create sample work lists for the testing sites (top left). Work lists and raw test results from cycler instrumentation are ingested by the interpretation and result generation software. Environments and clients are available per tenant, and dashboards for QC are available to the NRC. Called results go back to the requesting sites (bottom right), with restricted views per requester. Results immediately flow back to the requesting physicians and patients.

Figure 4

A layered software architecture. Assay logic and curve-calling algorithm (Assay Plugins) are separated from the execution and visualization framework, and laboratory dashboarding is decoupled from the platform database.

Figure 5

FastFinder Insights. A dashboard overview with key metrics across participating laboratories, devices, and assays.

Figure 6

Trend analysis. Time series of the total samples and positivity rates. Interactive visualization available in the FastFinder Insights software user interface.