Can Screening People Registering With Primary Care Improve the Detection of Tuberculosis?

This randomised controlled trial tests whether screening people registering with general practices cost-effectively increases the detection of tuberculosis. The study objectives are to:

1. recruit, and randomise with stratification, forty Hackney general practices to usual activity or screening for TB primarily during registration health checks (by practice nurses and GPs) and transfer-in visits for children (by health visitors)
2. implement a screening programme comprising education, resources and clinical support for intervention practices using proven behaviour change strategies
3. gather outcomes and costs
4. build research capacity within an ELENoR Primary Care Research Team and across a Primary Care Trust.

The primary outcome is the proportion of TB cases identified in primary care. The study is powered to detect a 20% increase (from 55% to 75%) in the detection of new cases of TB in primary care with 80% power at the 5% significance level. Secondary outcomes are rates of prophylaxis for latent TB infection (LTBI), BCG immunisation, smear negative (early) TB and costs. Analyses will be by intention to treat.

To date, studies of screening for tuberculosis have used relatively weak designs. A randomised trial would give stronger evidence as to whether screening is worthwhile and therefore would be welcome. We have tested the feasibility of screening in two general practices in Hackney; our results suggest that it is simple and straightforward, that the people screened were comfortable taking part, and that screening does indeed identify people with tuberculosis.

How (or indeed whether) to screen for tuberculosis constitutes a major policy issue for inner city public health care. Our work will go some way toward answering this important question and will have local, national and international relevance. There are important benefits for those taking part. First, the project is part of a local health development scheme which encourages GPs to register refugees and asylum seekers. This means that a group of people who find it difficult to get health care will be more likely to be able to find a GP. Second, people registering with general practices who are screening will all have a brief discussion about tuberculosis when they attend a health check at the surgery. This will make people more informed about tuberculosis and help to remove the stigma that attends tuberculosis. Third, screening is likely to identify people with tuberculosis, or who need BCG vaccination to prevent tuberculosis. These people will benefit from early treatment or immunisation.

We will invite all general practices in Hackney to take part. Practices that agree will then be randomly allocated (by a computer programme) into two groups: practices who will screen and those who will carry on as usual without screening. The project researcher responsible for data collection will not know to which group the practices have been assigned. Practices allocated to the screening group will be offered a programme designed to encourage screening as part of the new patient registration checks (by practice nurses and GPs) and during children's transfer-in visits (by health visitors). The programme is founded on proven ways of helping people change how they work and includes the following: an education programme run by the primary care research team, with a local chest consultant as opinion leader; Heaf guns, Heaf heads, and tuberculin (to test for presence of TB); TB screening guidelines; regular support and advice from the local TB specialist nurse; incorporation of a TB screening template into the practice computer consulting system; and an incentive fee paid to the practices for carrying out the Heaf tests. TB identication rates for both groups will be ascertained after 25 months of screening. The cost of implementing the screening process will also be measured.

Patients / public involvment (subjects and partners):

People attending for new registration checks and children seen at transfer-in visits are the study participants. They will be informed about the study (using information sheets in various languages); they can decline to take part if they like. Public involvement in the design and execution of the study comes in three forms. First, the East London Refuge and Asylum seeker worker has contributed to the study design and will be involved throught the study. Second, the work is part of an East London Health Improvement Plan, which itself has public support. Third, the study has approval from the East London and Esssex Network of Researchers’ Research Advisory Committee, which has two consumer representatives

Inclusion Criteria:

The intervention is screening for TB in general practices which were randomised to control or intervention

Exclusion Criteria:

Practices that refused to participate could not be included. We excluded the general practice where the pilot study was undertaken