Comparative cost analysis of community-based tuberculosis screening: Computer-Aided Detection (CAD) alone versus CAD with point-of-care C-reactive protein testing in South Africa and Lesotho

@article{harkare2026comparative,
  title = {{Comparative cost analysis of community-based tuberculosis screening: Computer-Aided Detection (CAD) alone versus CAD with point-of-care C-reactive protein testing in South Africa and Lesotho}},
  author = {Harsh Vivek Harkare and Anna Verjans and Phillip Joseph and Mashaete Kamele and Shriya Misra and Mamatlakeng Keitseng and Mojabeng Mabeleng and Thandanani Madonsela and Alastair van Heerden and Johanna Kurscheid and Bart K.M. Jacobs and Aita Signorell and Fiona Vanobberghen and Klaus Reither and Irene Ayakaka and Shannon Bosman and Fabrizio Tediosi},
  journal = {PLOS Global Public Health},
  year = {2026},
  month = sep,
  doi = {10.1371/journal.pgph.0007341},
  url = {https://journals.plos.org/globalpublichealth/article?id=10.1371/journal.pgph.0007341},
  note = {CC BY 4.0},
}

TY  - JOUR
TI  - Comparative cost analysis of community-based tuberculosis screening: Computer-Aided Detection (CAD) alone versus CAD with point-of-care C-reactive protein testing in South Africa and Lesotho
AU  - Harsh Vivek Harkare
AU  - Anna Verjans
AU  - Phillip Joseph
AU  - Mashaete Kamele
AU  - Shriya Misra
AU  - Mamatlakeng Keitseng
AU  - Mojabeng Mabeleng
AU  - Thandanani Madonsela
AU  - Alastair van Heerden
AU  - Johanna Kurscheid
AU  - Bart K.M. Jacobs
AU  - Aita Signorell
AU  - Fiona Vanobberghen
AU  - Klaus Reither
AU  - Irene Ayakaka
AU  - Shannon Bosman
AU  - Fabrizio Tediosi
PY  - 2026
DA  - 2026/09/21
JO  - PLOS Global Public Health
DO  - 10.1371/journal.pgph.0007341
UR  - https://journals.plos.org/globalpublichealth/article?id=10.1371/journal.pgph.0007341
AB  - Community-based active case finding (ACF) for tuberculosis (TB) is recommended in high-burden settings, yet evidence on implementation costs from prospective programmatic data remains limited. We estimated and compared the costs of two community-based ACF screening approaches: Computer-Aided Detection of chest X-rays alone (CAD4TBv7 approach) and CAD4TBv7 followed by point-of-care C-reactive protein (CRP) testing within a predefined score window (CAD4TBv7-CRP approach), evaluated in a prospective, paired screen-positive trial in South Africa and Lesotho. Nested within the TB TRIAGE+ trial (September 2022-September 2024), we applied a bottom-up ingredients-based approach from a programmatic perspective to estimate economic and financial costs, separately for TB-only and integrated (TB, HIV, and non-communicable disease) screening. Uncertainty was estimated via probabilistic sensitivity analysis, while scenario analyses explored higher modelled TB prevalence (1%) and expanded screening volume (100,000 participants). Among 20,023 participants screened (6,986 in South Africa; 13,037 in Lesotho), 72 TB cases were detected under the CAD4TBv7 approach and 60 under the CAD4TBv7-CRP approach (full analysis set). For TB-only screening, the average cost per TB case detected was USD 5,454 (95% uncertainty interval [UI]: 4,294–6,777) for the CAD4TBv7 approach and USD 7,486 (95% UI: 5,948–9,246) for the CAD4TBv7-CRP approach, 37% more expensive. Recurrent costs accounted for 67–72% of total costs. Personnel was the dominant cost driver in South Africa, and vehicle costs in Lesotho. Higher TB prevalence would substantially reduce the average cost per TB case detected, whereas scaling up participant numbers at the observed prevalence would not. The additional cost of CRP testing was not offset by savings from averted confirmatory tests, making the CAD4TBv7-CRP approach more expensive than CAD4TBv7 alone. CRP-based sequential screening may offer greater cost advantages in settings with lower inflammatory comorbidity burden, though this requires prospective evaluation. TB screening strategies should be tailored to local epidemiological and health-system contexts.
SN  - 2767-3375
ER  -