Hybrid Approach to Estimation of Underreporting of Tuberculosis Case Notification in High-Burden Settings With Weak Surveillance Infrastructure: Design and Implementation of an Inventory Study
Bibliographic record
Abstract
BACKGROUND: The greatest risk of infectious disease undernotification occurs in settings with limited capacity to detect it reliably. World Health Organization guidance on the measurement of misreporting is paradoxical, requiring robust, independent systems to assess surveillance rigor. Methods are needed to estimate undernotification in settings with incomplete, flawed, or weak surveillance systems. This study attempted to design a tuberculosis (TB) inventory study that balanced rigor with feasibility for high-need settings. OBJECTIVE: This study aims to design a hybrid TB inventory study for contexts without World Health Organization preconditions. We estimated the proportion of TB cases that were not reported to the Ministry of Health in 2015. The study sought to describe TB surveillance coverage and quality at different levels of TB care provision. Finally, we aimed to identify structural-, facility-, and provider-level barriers to notification and reasons for underreporting, nonreporting, and overreporting. METHODS: Retrospective partial digitalization of paper-based surveillance and facility records preceded deterministic and probabilistic record linkage; a hybrid of health facilities and laboratory census with a stratified sampling of HFs with no capacity to notify leveraged a priori knowledge. Distinct extrapolation methods were applied to the sampled health facilities to estimate bacteriologically confirmed versus clinical TB. In-depth interviews and focus groups were used to identify causal factors responsible for undernotification and test the acceptability of remedies. RESULTS: The hybrid approach proved viable and instructive. High-specificity verification of paper-based records in the field was efficient and had minimal errors. Limiting extrapolation to clinical cases improved precision. Probabilistic record linkage is computationally intensive, and the choice of software influences estimates. Record absence, decay, and overestimation of the private sector TB treatment behavior threaten validity, meriting mitigation. Data management demands were underestimated. Treatment success was modest in all sectors (R=37.9%-72.0%) and did not align with treatment success reported by the state (6665/8770, 75.99%). One-fifth of TB providers (36/178, 20%) were doubtful that the low volume of patients with TB treated in their facility merited mastery of the extensive TB notification forms and procedures. CONCLUSIONS: Subnational inventory studies can be rigorous, relevant, and efficient in countries that need them even in the absence of World Health Organization preconditions, if precautions are taken. The use of triangulation techniques, with minimal recourse to sampling and extrapolation, and the privileging of practical information needs of local decision makers yield reasonable misreporting estimates and viable policy recommendations.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.077 | 0.109 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.003 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.003 | 0.005 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".