TB Data and connectivity for data-driven Case-Finding strategies

The integration of digital technologies in TB screening is transforming TB control. Automated TB data transmission from diagnostic devices to a central database can provide real-time quality data on device performance, continuous disease surveillance and patient care [1]. Aggregated TB data can be analysed in real-time to identify high-risk populations, monitor disease trends, and detect potential outbreaks early. Predictive data-driven analytics will enable TB programs to allocate resources more effectively.

WHO’s Digital Health for the End TB Strategy advocates that TB programmes integrate digital health solutions in their implementation of the End TB Strategy. ICT can help advance person-centred care, surveillance, programme management, staff development (e.g. eLearning), and the engagement of communities [2]. Digital imaging and AI are driving and supporting the evolution of TB case finding in particular in resource-constrained settings as illustrated below.

Evolution of TB case finding in resource-constrained settings

Digital TB screening tools, including Computer-Aided Detection (CAD), are emerging as powerful enablers to optimize, plan, execute, and control active TB case-finding programs in both health facilities and community outreach settings. By leveraging data generated from these digital tools, healthcare providers can enhance the efficiency and effectiveness of TB screening programs, resulting in improved outcomes for at-risk populations.

Role of digital TB screening

TB Data in health facility and outreach settings

• In health facilities, digital TB screening tools - using CAD score data - streamline the diagnostic workflow by quickly and accurately triaging patients with suspected TB. This leads to a more efficient allocation of resources and minimizes patient wait times, reducing the risk for “loss to follow-up”.
• In outreach settings, where TB case finding often involves screening large numbers of people in remote or underserved areas, digital tools enhance the ability to capture and analyse data on the spot. Mobile and ultra-portable digital X-ray units equipped with CAD can be deployed to screen high-risk populations such as urban slum dwellers, PDL, mining communities, or people living with HIV, allowing for immediate identification and referral of presumptive TB cases [3].

CAD & test data for Diagnostic Network Optimization

• By aggregating and analysing CAD and test data, health systems can identify patterns of TB prevalence, assess the performance of various diagnostic centres or outreach teams, and streamline patient referral pathways.
• CAD and test data permit to identify gaps in service delivery and guide decisions on where to establish new diagnostic facilities or optimize existing ones.
• Ultimately, the integration of CAD and test data into diagnostic networks facilitate early TB diagnoses, reduces delays in treatment initiation, and improves overall program efficiency and impact [4].

Example: CAD4TB Platform providing data modules

• Patient registration
• Symptom screening with customisable questionnaires
• HIV status
• CAD4TB outputs (CXR heatmap & score)
• Provisions for integrating data from other systems (e.g. lab systems)

The CAD4TB Platform generates individual reporting and insights as outputs to enable the TB programs to become more cost-effective, evidence-based and data-driven like:

• Identification of TB root causes, prioritize interventions, and optimize resource allocation through additional data for the NTLP PCF dashboard;
• Enhancing people-centric approach by more effectively addressing the needs of the most vulnerable and most at risk populations integrating digital CAD4TB data with the LMIS.

Such Platform can support insights into for instance individual risk factors for more data-driven active case finding and diagnostic network optimization.

Optimizing TB Case-Finding through data-integration & connectivity

• Automatic data transmission to a software platform via a server or a cloud, for data integration and advanced analysis. Facilitate real-time performance monitoring and optimised data management, leading to data-driven decision-making and better patient care.
• Timely, high-quality data can also expose inequities in health outcomes and help decision makers to identify children’s unmet needs [5]. The value of digital TB screening data extends beyond individual patient diagnosis; it can be leveraged to optimize the entire TB case-finding strategy.
• When screening data, including the CAD abnormality score, are integrated into a country’s Lab Management Information System (LMIS) through connectivity solutions, it can provide real-time visibility of the productivity and yield of TB screening programs.
• Monitor KPI’s, such as the number of screenings, the % of abnormal X-ray results, and the yield of confirmed TB cases, for data-driven decision-making and diagnostic network optimization.

Integrating HIV status data for more effective TB screening

• Individuals living with HIV are at a much higher risk of developing TB due to their compromised immune systems, making them a critical group for targeted interventions. By combining TB screening data with HIV status and viral load, health programs can prioritize screening and early diagnosis for those at dual risk, ensuring timely and appropriate treatment including TPT.
• This integrated approach allows for people-centred approaches, more comprehensive patient management, facilitating simultaneous screening and treatment for both TB and HIV where needed, reducing mortality rates, and improving overall health outcomes.

Data-integration TIBULIMS - Kenya

Data-integration LabXpertDS - Uganda

Performance monitoring & reporting

• The diagnostic connectivity solution LabXpertDS used in Uganda allows for a performance reward system that recognizes individual efforts in terms of daily screening throughput and case notifications.
• A user-friendly dashboard features for alerts that support adherence to equipment service and maintenance schedules through reminders about service and maintenance, while also highlighting machine performance problems [8].
• The Dashboard that aggregates data from the connected devices into a simple graph or KPIs, allowing the monitoring of test results and system utilisation/performance trends.

Geospatial mapping: visualize TB hotspots with incidence heatmaps

• Integrated digital TB screening data when connected to molecular test systems can facilitate real-time visualization of TB hotspots through incidence heatmaps.
• ACF TB data can be used as an input for sub-national TB prevalence surveys
• Allows for a more focused allocation of resources, such as deploying mobile or ultra-portable screening units to areas with high transmission rates or increasing public health messaging in identified hotspots.
• Predictive models may forecast potential TB outbreaks based on factors like population density, socio-economic status, seasonality and previous TB case data, allowing for more targeted interventions.

The map below shows near real-time disease surveillance based on weekly TB data automatically collected from reporting sites in Uganda [8]

Example Hot spot mapping with AI - Nigeria

Conclusions

Automated data transmission is feasible by networking the point of care (POC) devices to a central database at the Ministries of Health, thereby providing real-time data on diagnostic instrument performance [11]. In addition, aggregated data as an input for real-time Dashboards can be used to optimize and control TB screening initiatives and to effectively integrate data from the HIV program.