Corbett EL, Bandason T, Duong T, et al. Comparison of two active case-finding strategies for community-based diagnosis of symptomatic smear-positive tuberculosis and control of infectious tuberculosis in Harare, Zimbabwe (EDTECTB): a cluster-randomised trial. Lancet. 2010;376:1244-53.
Africa accounts for nearly 80% of the world's tuberculosis (TB) burden.(1) TB disease in endemic regions is primarily due to recent infection and is acquired through casual contact.(2) Delays in diagnosis result in a long period of infectiousness in the community. Efforts to identify and treat persons early in the course of disease are likely to reduce transmission and improve individual health outcomes.
To measure the impact of two methods of active case-finding on the prevalence of TB.
Residential communities in suburban Harare, Zimbabwe; Adult household residents in sampled communities.
Forty-six clusters of residential neighborhoods in suburban Harare were randomly assigned to either active case-finding for at six-month intervals by mobile van or to door-to-door visits. Randomization was done by selecting discs (with colors associated with group assignment) from an opaque bag at a public meeting. Case-finding was done by trained health workers who screened a single household member for chronic cough (> 2 weeks). Mobile van and door-to-door teams rotated through community clusters over between January 2006 and November 2008. The mobile van remained in each cluster for five days from 9 am to 4 pm Monday through Saturday and advertised their presence through loudspeaker announcements and distribution of leaflets describing benefits of TB screening and study services. Symptomatic persons were provided with specimen collection containers for themselves and any others they knew of with symptoms. Door-to-door inquiries were made between 9 am and 4 pm and included one weekend visit. Health workers and participants could not be blind to group assignment. Participants with cough produced two sputum samples per adult, one collected immediately and the other in the early morning. Sputum was tested for TB using fluorescence microscopy at a central laboratory where personnel were blinded to the study group allocation. Positive results were returned to the participant in his/her home within 14 days of specimen collection. Analysis comparing case-finding using a mobile van and with a door-to-door enquiry was done using an intention-to-treat approach.
To identify TB cases that were diagnosed outside of the intervention, municipal electronic records were reviewed and patients with known TB who resided at the selected study addresses were classified as having been diagnosed through routine methods or as part of active case-finding.
To assess the broad effects of the intervention, two cross-sectional surveys were conducted. During 2005 and 2006 and again in 2008 study, clusters were visited and the number of households and number of residents per household were enumerated. This provided the sampling frame for the cross-sectional surveys which consisted of a 12% random sample of households. Adults aged 16 years and above provided informed consent and produced two sputum specimens for TB culture and venous blood for HIV testing. Participants with symptoms of TB disease had sputum tested by microscopy and had a chest radiograph. These prevalence surveys were conducted prior to and after the active case-finding interventions.
There were 110,432 adults in the 46 clusters at baseline and this increased to 124,244 after the fifth intervention round. In the 12% random sample of households, 10,092 (81%) of the adults provided sputum before the intervention and 11,211 (77%) provided sputum after the intervention. Participation was higher among women than men. There were 20,700 household enumerated in the mobile van group and 20,719 in the door-to-door group. The characteristics of households and adults assessed at baseline during the household enumeration were similar. HIV prevalence was 21% and prevalence of culture-positive TB was 6.5 per 1000 adults. Previous treatment for TB was reported by three percent. In the six months prior to the intervention, smear-positive TB diagnoses were made in 2.8 per 1000 adults.
There were 5,466 persons in the mobile van group and 4,711 in the door-to-door group. There were 255 smear positive TB cases in the mobile van group and 137 cases in the door-to-door group and the mean cumulative rate per 1000 adults was 4.22 in the mobile van group and 2.46 in the door-to-door group (risk ratio (RR)=1.48, 95% CI 1.11-1.96, after adjustment for cluster variables). The smear positive yield was higher in the mobile van group for all survey rounds. The cumulative TB yield increased with increasing cluster of HIV prevalence in the mobile group but not the door-to-door group.
From the start of round one of active case-finding through the end of round six, a total of 472 smear-positive persons were diagnosed (44 in the mobile van group and 36 in the door-to-door group). Routine health services diagnosed an additional 670 adults in the study clusters (367 in the mobile van group and 303 in the door-to-door group). That is, active case-finding contributed 41% of the smear-positive diagnoses.
The overall prevalence of culture-positive TB declined by 44% (95% CI 17-62). The reduction was greater in participants without HIV compared to those with HIV infection (58% reduction and 25% reduction respectively).
The population in the study area increased by 13% and the overall HIV prevalence declined from 21% to 19%.
Active case-finding for TB, particularly with a mobile van may reduce TB transmission and rates of disease.
This study was a randomized clinical trial using a truly random method of group assignment. Complete blinding was not possible but was done when feasible. Loss to follow-up was not addressed. Intention-to-treat analysis was done when appropriate.
The use of community-based active case-finding for TB, especially when done using a mobile van, holds the promise of an effective method to reduce TB transmission. There has been an increased focus on facility-based diagnosis and treatment and while this has improved treatment outcomes, its effect on overall incidence and prevalence of TB has been disappointing.(3) Using chronic cough as the key symptom to trigger screening with collection of sputum specimens may miss subclinical TB cases that could be identified using screening chest radiographs. Radiologic screening is not practical at the residential level and identification and treatment of persons with cough is likely to have a greater impact on reducing transmission that would occur with radiographic screening.(4) The methods used in this study, when applied outside of research may be an important adjunct to current TB control efforts.
- Global Tuberculosis Programme, World Health Organization. Global tuberculosis control: epidemiology, strategy, financing. WHO/HTM/TB/2009.411. Geneva: World Health Organization; 2009.
- Verver S, Warren RM, Munch Z, et al. Proportion of tuberculosis transmission that takes place in households in a high-incidence area. Lancet 2004; 363: 212-14.
- Lönnroth K, Castro KG, Chakaya JM, et al. Tuberculosis control and elimination 2010-50: cure, care, and social development. Lancet 2010; 375: 1814-29.
- Corbett EL, Bandason T, Cheung YB, et al. Epidemiology of tuberculosis in a high HIV prevalence population provided with enhanced diagnosis of symptomatic disease. PLoS Med 2007;4: e22.