Step 6.1 – Develop the use of digital data and diagnostics connectivity
Step 6.2 – Develop procedures for data backup, security and confidentiality
Step 6.3 – Develop data requirements for targeted NGS tests
Step 6.1 – Develop the use of digital data and diagnostics connectivity
Many of the latest testing platforms offer the opportunity to use digital data. The implementation plan should consider software and hardware requirements, to take advantage of digital data. “Diagnostics connectivity” refers to the ability to connect diagnostic test devices that produce results in a digital format, in such a way as to transmit data reliably to a variety of users (39). Key features of the systems are the ability to monitor performance remotely, conduct QA and manage inventory. With remote monitoring, designated individuals can use any internetenabled computer to access the software, providing an overview of the facilities, devices and commodities in the network. Software can track consumption and inventory to avoid stock-outs and expiring supplies. It can also identify commodity lots or specific instruments with poor performance or abnormal error rates for QA purposes, and provide a pre-emptive service to avoid instrument failure. This approach is a highly cost-effective way to ensure that a diagnostic device network functions properly; it is also useful for reporting and connecting with treatment sites.
Data, results and information updates can also be transmitted automatically to:
- clinicians and patients, which allows for faster patient follow-up;
- laboratory information management systems or electronic registers, reducing staff time and the chance of transcription errors, and greatly facilitating monitoring and evaluation processes; and
- the NTP, to assist with surveillance of disease trends or resistance patterns and rates, and to enhance the capacity of the NTP to generate the data needed for performance indicators of the End TB Strategy.
Step 6.2 – Develop procedures for data backup, security and confidentiality
With any electronic data system, there is a risk of losing testing data. An SOP for regularly backing up data (e.g. to an external drive) is essential, as is an SOP for data retrieval. Also needed are policies and procedures to ensure the security of laboratory data and confidentiality of patient data, in line with national and international regulations. Antivirus software should be installed and kept up to date. Access restriction should be in place to safeguard confidentiality, protect personal information and prevent data breaches by unauthorized users. Data access and governance policies should be developed and enforced.
Step 6.3 – Data requirements for targeted NGS tests
This step briefly summarizes data storage requirements, data analytic tools and data-sharing protocols needed for targeted NGS tests. Technical assistance from an IT expert or department may be needed. A checklist to assess IT and data readiness is included in the site readiness checklist in Annex 3 of the WHO implementation manual (30).
Plan for data storage and computing
Adequate secure data storage is essential for successful implementation of targeted NGS. Each run of targeted NGS can generate thousands of megabytes of data (see Annex 9 of the WHO implementation manual (30)). For example, when “benchtop” NGS platforms are used, 7.5–15 gigabytes of data are generated during a 24–48 hour run. A sequencing run may include as many as 24–48 samples for WGS or 96 samples for targeted NGS.
Options for data storage include computer hard drives or external hard drives (a 1 terabyte hard drive may be able to store NGS files for up to 1500 samples); high-capacity flash drives (also known as memory sticks or thumb drives); a scalable, local computing cluster; and cloud-based systems (e.g. Illumina BaseSpace, Google Cloud and Amazon Drive).
Select data analytic tools
In recent years, many commercial and public data analytic pipelines have become available that make the analysis of MTBC NGS data accessible to both nonexperts and bioinformatic experts (29). Factors to be considered in selecting the tool or tools that are best suited for the intended use of the NGS data include compatibility with the data output by the NGS instrument, the type of information produced, the computational resources and bioinformatics expertise needed at the local site, and the time required to complete the analysis. The targeted NGS solutions included in the latest recommendations include proprietary software that analyses the user uploaded sequencing read and exports it in the form of interpretable results.
Develop procedures for data sharing and ensuring confidentiality
Because NGS generates sequences of whatever DNA is present in a sample, laboratories must be particularly careful with sequence data generated from human clinical samples (e.g. sputum specimens, which may contain considerable amounts of human DNA). Currently, two features of the NGS procedures for TB mitigate this concern: the process of isolating bacteria by culture should eliminate the risk of sequencing any human DNA; and the amplification or capture step used in the processing of clinical specimens for targeted NGS serves to minimize the relative amount of human DNA that is sequenced. Furthermore, most bioinformatics pipelines include steps to discard any sequence reads that map to the human genome before sharing TB sequencing data.