Table 5.
Non-randomised studies evaluating effect of active case-finding on tuberculosis prevalence
| Country, population | Case-finding method | Diagnostic method | Co-interventions | Clusters | Tuberculosis cases among people screened at sequential prevalence surveys, n/N (cases per 100 000 population) | Reported measure of association | |
|---|---|---|---|---|---|---|---|
| Sanchez et al (2013)38 | Brazil, people in prison | Door to door and at prison entry | Chest x-ray for all, sputum smear and culture if chest x-ray abnormal | None | 1 | Baseline, 83/1374 (6040); endline, 32/1244 (2800) | Authors report p<0·001 for difference baseline to endline |
| Kolapann et al (2013)39 | India, remote rural | Door to door | Chest x-ray for all, sputum culture if chest x-ray abnormal | Change to NTP guidelines in area (DOTS introduced) | 53 | 1999–2001, 457/83 425 (607); 2001–03, 344/85 474 (454); 2004–06, 253/89 413 (309); 2006–08, 332/92 255 (388) | Significant decrease in culture-positive tuberculosis prevalence at years 2·5, 5·0, and 7·5; regression analysis showed that a linear model was inadequate to explain the variation in prevalence, with r2=0·59 |
| Chatterjee et al (2014)40 | India, remote rural | Door to door | Chest x-ray and sputum for culture if symptoms | Change to NTP guidelines in area (DOTS introduced) | 5 | June, 1999, to April, 2000, 25/5096 (490·6); year 2·5, 9/4042 (222·7); year 5, 3/3978 (75·24); year 7·5, 7/3712 (188·6) | No measure of association reported |
| Liu et al (2019)41 | China, general population | Door to door | Chest x-ray if symptoms or in high-risk group; sputum smear if symptoms or abnormal chest x-ray | None | 3 | 2013, 35/92 822 (37·7); 2014, 25/92 638 (27·0); 2015, 15/89 799 (16·7)* | Site A, 2013 vs 2015, p<0·001; site B, 2013 vs 2015, p=0·064; site C, 2013 vs 2015, p=0·20 |
| Tsegaye Sahle et al (2019)42 | Ethiopia, people in prison | Group meetings and at prison entry | Sputum tests if symptoms (mainly smear, but some Xpert and culture); chest x-ray available if symptoms | None | 1 | Baseline, 3/3024 (99·2); endline, 10/2551 (392) | Prevalence increased from 0·10% in the first screening to 0·39% in the second screening (p=0·027) |
| Rao et al (2019)43 | India, indigenous population | Door to door | Sputum smear and culture if symptoms | None | 53 | Baseline, 293/9756 (3003); endline, 195/9775 (1995) | Prevalence had decreased significantly at endline compared with baseline (trend χ2 19·97, odds ratio 1·521, p=0·000) |
NTP=national tuberculosis programme. DOTS=directly observed therapy, short course.
*
The prevalence of tuberculosis in each year was averaged across sites A–C.