Bulletin of the World Health Organization

Improving tuberculosis case detection rate with a lay informant questionnaire: an experience from the Lao People’s Democratic Republic

Peter Odermatt, Sayphone Nanthaphone, Hubert Barennes, Khamsay Chanthavysouk, Duc-Si Tran, Bounsou Kosanouvong, Siamphay Keola, Phetsamone Mathouchanh, Khamloune Choumlivong, Valy Keoluangkhot, Phoumindr Niranh, Sixomxay Nanthanavone, Souraxay Phrommala, Antoine Degrémont, Michel Strobel

Problem

In many countries, the tuberculosis (TB) annual case detection rate is below the target of 70%. In the Lao People’s Democratic Republic in 2005, it did not exceed 55%.

Approach

The DOTS strategy promotes passive case detection of TB. In order to increase the detection rate, we validated a questionnaire targeting lay informants at village level to notify patients with chronic cough and assessed the relevance for TB case-finding. A three-item questionnaire was sent through the district health departments to all villages in six districts in six provinces. The village headmen were asked to notify chronic cough patients. Answers were validated in a door-to-door survey (20 villages/district). In a sub-sample (four villages/district) all confirmed patients were screened for TB and paragonimiasis.

Local setting

Attapeu, Luang Namtha, Luang Prabang, Saravane, Savanakhet and Vientiane provinces in the Lao People’s Democratic Republic.

Relevant changes

Lay informant questionnaires sent from district health offices to villages are cost-effective and foster interaction between the health services and remote and underserved communities. Although the correct detection of patients is highly dependent on direct respondents, a substantial number of new TB and paragonimiasis cases were consistently diagnosed in chronic cough patients.

Lessons learned

Out of 456 questionnaires, 295 were returned (65%). Return rates were highly variable between districts (48–87%), questionnaires’ sensitivity (56–98%), positive predictive value (34–88%) and correlation between number of notified and confirmed patients (r: 0.26–0.78). In sub-sampled villages (13 541 population) 19 (5.1%) TB and 26 (7.0%) paragonimiasis cases were detected in 374 chronic cough patients. This quick questionnaire approach proved motivating for district authorities and village key informants, although no incentives were provided. The highly operator-dependent approach yielded a consistent detection rate of TB and paragonimiasis cases. This approach brings health services and populations in need in close contact, which is particularly crucial in remote and underserved areas.

Share