CV
PUBLICATIONS · PEER-REVIEWED · 2025 – 2026

Evidence, not opinions.

Journal articles, conference presentations, and the WHO guideline update that the Cameroon evidence streams I lead data for contributed to.

Peer-reviewed
3
Preprint
1
Years
2025 – 26
WHO citations
1
2026
CO-AUTHOR (9TH)JOURNAL

Diagnostic accuracy and inter-lot precision of the FujiLAM II point-of-care assay to detect tuberculosis in stored urine specimens from people living with HIV

Microbiology Spectrum (American Society for Microbiology), vol. 14, issue 8

Diagnostic accuracy and inter-lot variability evaluation of the redesigned Fujifilm SILVAMP TB LAM II urine lateral-flow assay, on biobanked specimens from two prospective hospital studies in Cameroon. 798 people living with HIV tested January–March 2025, 128 (16%) with microbiologically confirmed TB. Against a reference standard of two sputum liquid cultures plus sputum, urine, and blood Xpert Ultra, FujiLAM II reached 49% sensitivity (95% CI 40–58%) against 35% for AlereLAM, and 96% specificity (95% CI 94–97%) against 92%. Across four manufacturing lots, positivity ranged 37–42% with no significant difference (Cochran's Q = 4.2, P = 0.2), resolving the inter-lot variability that affected the original assay.

798
people living with HIV
49%
sensitivity (vs 35% AlereLAM)
96%
specificity
4
manufacturing lots, no difference
TBUrine LAMPoint-of-careHIVDiagnostic accuracyMicrobiology Spectrum
FIRST & CORRESPONDING AUTHORPREPRINT

Implementation of pooled testing of sputum specimens from multiple individuals for rapid molecular detection of tuberculosis in Cameroon

openRxiv — preprint (posted 23 March 2026)

First large-scale programmatic evaluation of TB pooled testing. 71,328 sputum specimens across 16 GeneXpert laboratories and 6 regions of Cameroon (Oct 2023 – Mar 2025). 59,164 specimens tested in pools of 2–8 consumed only 20,838 Ultra cartridges (0.35 cartridges per result), enabling 38,326 additional people to be tested. Time to result fell from 66 min (individual) to 45→10 min (pools of 2→8). Cost per result: $2.81 pooled vs $7.97 individual. Core evidence stream behind the WHO 9 March 2026 recommendation on sputum pooling.

71,328
specimens
16
GeneXpert labs
38,326
additional people tested
$2.81
per result (vs $7.97)
TBPooled testingImplementation scienceFirst authorCorresponding authorWHO evidence
POLICY · GUIDELINEPOLICY · GUIDELINEWHO · 9 MAR 2026

WHO recommendations on near-point-of-care TB tests, tongue swabs, and sputum pooling

World Health Organization — Guideline update

Recommendations issued 9 March 2026 — the first time WHO formally endorsed near-point-of-care molecular TB tests, tongue swabs as an alternative to sputum, and sputum pooling to expand molecular testing capacity. Cameroon evidence streams from Start4All and TB Reach Wave 10 — for which I served as Country Data Manager / Data Manager — contributed to the underlying evidence base. LSTM's 27 February 2026 commentary explicitly cited the programme.

WHOPolicyGuidelineTB diagnostics9 March 2026
2025
CO-AUTHOR (6TH)JOURNAL

Diagnostic Performance of the Pluslife MiniDock MTB and Molbio MTB Ultima Assays to Detect Tuberculosis From Tongue and Sputum Swabs Among Outpatients and in Active Case Finding in Cameroon

Clinical Infectious Diseases (Oxford University Press)

Head-to-head diagnostic accuracy evaluation of two low-cost near-point-of-care molecular TB assays (Pluslife MiniDock MTB and Molbio MTB Ultima) against Xpert MTB/RIF Ultra, using tongue and sputum swabs. 1,097 participants across 5 health facilities and active case-finding events in Cameroon. Sputum-swab sensitivity 79–91% (MiniDock) and 74–89% (MTB Ultima); specificity > 95%. MiniDock operated in the field at 35–40 °C ambient with solar power, < $200 instrument, < $4 test cost. Feeds into WHO's 9 March 2026 recommendations on near-POC TB tests and tongue swabs.

1,097
participants
79–91%
MiniDock sputum sensitivity
< $200
instrument cost
35–40 °C
field operating range
TBDiagnosticsTongue swabsNear-POCCIDActive case finding
CO-AUTHOR (3RD OF 14)JOURNAL

Implementation of pooled testing to increase access to routine viral load monitoring for people living with HIV on antiretroviral therapy

Scientific Reports (Springer Nature), vol. 15, 14713

Peer-reviewed implementation evaluation at the TB Reference Laboratory Bamenda using the Abbott RealTime HIV-1 viral load assay in pools of three. 12,396 plasma specimens processed with only 6,797 assays (0.55 assays per result) — a 45% saving that enabled an additional 5,409 people (+80% capacity) to receive routine viral load results. Bland–Altman mean bias only –0.51 log copies/mL; estimated misclassification of unsuppressed specimens at 0.01% under the implemented ≥ 40 copies/mL cutoff.

12,396
plasma specimens
+80%
testing capacity
5,409
additional people served
0.01%
misclassification rate
HIVViral loadPooled testingImplementation scienceScientific ReportsMethod validation
PRESENTING AUTHORCONFERENCE

Chest X-ray AI scores guide pooled testing to increase access to rapid molecular testing for TB in communities in Cameroon

Union World Conference on Lung Health 2025 · E-Poster EP10-793-19 · Copenhagen

Implementation-science evidence that CXR-AI-guided pooled testing cuts cartridge consumption by 45% across 2,183 specimens in community case-finding. Track K2 TB diagnostics, Enhanced TB detection session, 19 November 2025.

2,183
specimens
45%
cartridge savings
0.55
tests per person
TBAI triagePooled testingCommunity case findingUnion 2025