Missed Diagnoses, Overdiagnosis, and Improving Treatment Outcomes

(Yoseph Samodra)

Tuberculosis (TB) control requires coordinated improvements across diagnosis, prevention, and clinical management. Across these six studies, a consistent picture emerges: health systems face substantial diagnostic error at scale, TB burden continues to rise in key populations and geographies, and outcomes can be improved through stronger patient-centered supports and practical clinical tools, particularly for high-risk and drug-resistant TB patients.

See also: Lin TB Lab

---

TB diagnosis at scale is still error-prone, with major consequences

• Real-world diagnostic algorithms in LMICs show only “moderate” performance, with estimated sensitivity 82.6% and specificity 88.0%, meaning large numbers of cases are still missed or wrongly labeled as TB.
• In 2023 alone (across 111 LMICs), the model estimated approximately: ~1.00 million false-negative diagnoses (missed TB cases); ~2.05 million false-positive diagnoses (people treated for TB who likely did not have it).
• Clinical diagnosis is a key driver of overdiagnosis, contributing to: ~22% of true-positive diagnoses, but ~75% of false-positive diagnoses.
• Diagnostic performance varies widely by region, largely reflecting differences in access to rapid diagnostic tests (RDTs) and reliance on clinical judgment.
• System-level improvements could substantially reduce both missed cases and overtreatment, including wider RDT adoption, better clinical decision algorithms, and more sensitive diagnostic technologies.

See also: Jago Beasiswa

---

The TB burden is increasing, with clear demographic and geographic concentration

• In Southeast Sulawesi (Indonesia), reported TB cases increased steadily from 2,087 (2021) to 2,906 (2023), showing a sustained upward trend.
• TB burden was consistently higher among males, and the most affected age group was 45–54 years, indicating concentrated transmission and/or delayed detection in working-age adults.
• Geographic clustering was clear, with Kendari City reporting the highest number of cases, while less densely populated regions reported fewer cases.
• Although latent TB infection (ILTB/LTBI) detection increased annually, participation in tuberculosis preventive therapy (TPT) declined, signaling a critical implementation gap between identifying risk and delivering prevention.
• In Kupang City, pulmonary TB cases were mostly among productive age adults (15–50 years; 73.4%), with no cases reported in children (0–14 years) in the dataset—suggesting either lower detection in children or a surveillance/diagnostic gap.
• Education level showed a statistically significant association with pulmonary TB incidence (p = 0.048), while occupation did not (p = 0.958), reinforcing that social determinants (especially education-linked health literacy and access) may shape risk more than job category alone.

---

Patient-centered behavior and clinical risk tools strengthen TB outcomes

• In Surabaya (Indonesia), TB transmission prevention behavior was significantly influenced by three modifiable factors: Knowledge (each 1-point increase → +0.667 prevention behavior points); Supportive behavior (each 1-point increase → +0.370); Medication adherence (each 1-point increase → +0.720). This highlights adherence and education as high-impact levers for transmission control.
• In MDR-TB treatment (Indonesia), adding delamanid to bedaquiline-containing regimens produced: No significant differences in sputum/culture conversion compared with bedaquiline regimens without delamanid; No significant differences in QTc interval changes, suggesting comparable cardiac safety under the study’s exclusion criteria; Numerically faster conversion and lower QTc prolongation in the combination group, though not statistically significant.
• In Thailand, an externally validated TB mortality risk score using only three diagnosis-time predictors (CCI category + tuberculous meningitis) demonstrated good discrimination (AuROC 76.3%) and good calibration, supporting its use for risk stratification and resource targeting, even if individual prediction remains imperfect, especially in intermediate-risk patients.

---

Conclusion

Taken together, these studies indicate that TB programs must address both technical performance gaps (diagnostic accuracy and appropriate confirmation) and implementation gaps (prevention uptake and adherence support). The most actionable direction is an integrated strategy: expand high-quality rapid diagnostics to reduce false results, target rising hotspots and high-burden demographics with prevention that actually reaches completion, and strengthen outcomes through education, adherence reinforcement, and simple clinical risk tools, particularly for patients at elevated mortality risk or facing MDR-TB treatment complexity.

References:

1. van Lieshout Titan, A., Dodd, P.J., Cohen, T. and Menzies, N.A., 2026. Estimating the number of incorrect tuberculosis diagnoses in low-and middle-income countries. Nature Medicine, pp.1-8.
2. Juliasih, N.N., Sakinah, L.F., Sari, R.M., Winarso, H., Siahaan, S.C.P. and Gunawan, E.J., 2024. Determinants of transmission prevention behavior among Tuberculosis patients in Surabaya, Indonesia. Infection Prevention in Practice, 6(4), p.100404.
3. Lestari, H., 2024. Analisis Epidemiologi Kejadian Tuberkulosis Di Provinsi Sulawesi Tenggara Tahun 2021-2023. Variable Research Journal, 1(02), pp.802-810.
4. Dewi, N.P.A.N. and Susilawati, N.M., 2024. Hubungan Pekerjaan dan Pendidikan dengan Kejadian TB Paru di Kota Kupang. Inovasi Kesehatan Global, 1(4), pp.139-148.
5. Soedarsono, S., Mertaniasih, N.M., Kusmiati, T., Permatasari, A., Subay, S. and Adiono, S.H., 2024. Comparison of Individual Regimen Containing Bedaquiline with Delamanid and Bedaquiline without Delamanid on Efficacy and Safety in Multidrug-resistant Tuberculosis Patients: Implementation in Dr. Soetomo General Academic Hospital, Indonesia. The International Journal of Mycobacteriology, 13(2), pp.140-146.
6. Saisudjarit, P., Saokaew, S., Duangjai, A., Prasatkhetragarn, A., Kanchanasurakit, S. and Phisalprapa, P., 2026. External Validation of a Simple Mortality Risk Prediction Score for Tuberculosis Patients. Journal of Health Research, 40(1), p.3.

TBN 007