Food Incentives for TB Treatment Compliance in East Timor (FITTCET)

We propose to conduct a randomised controlled trial of a food intervention to improve compliance in East Timor. Patients will be diagnosed in the usual way and given the opportunity to enrol in the study. Consenting individuals will be randomised to receive the intervention (meal at the clinic daily during the first two months of treatment and a food parcel fortnightly during the other six months of treatment) or standard care (nutritional advice only). The primary outcome will be TB cure rates and the secondary outcomes will include response to treatment (weight gain, decreased symptoms, sputum clearance) and percentage of days attended at the clinic. In conjunction with this study of TB treatment compliance we will collect blood samples to measure micro-nutrient changes in response to the food intervention. A cost-effectiveness analysis of the intervention will also be performed.

We plan to do a pilot study in three urban clinics in Dili to establish feasibility and provide pilot data for other funding applications so that, if the intervention does improve compliance, and funding applications are successful, we can expand the project to include rural districts in conjunction with the National TB Control Program. Ultimately, the health outcome addressed by this project is improved control of tuberculosis in a high burden, low income setting. Better compliance with tuberculosis treatment at the community level will contribute to this process.

TRIAL OBJECTIVES AND PURPOSES

The specific objectives of the project are:

1. Design a locally appropriate food intervention to improve adherence to TB treatment and TB treatment outcomes in East Timor; and
2. Conduct a study of the intervention in three clinics in Dili. If this pilot study is successful, we will use our experience and pilot data to obtain funds to generalise the intervention to rural clinics as well.
3. Food incentives are an achievable, effective method of encouraging full adherence to DOTS and thus improved cure rates in East Timor;
4. Locally available, whole food supplements are a cost-effective intervention to improve cure rates in East Timor; and
5. Lessons learned in East Timor are adaptable to other settings in other parts of the world.

TRIAL DESIGN

Expected Outcomes and its measurements :

1. To show statistically significant differences in treatment success (cure rate) between the intervention and the control arm. We will assess the sputum clearance, treatment compliance and treatment completion rates and compare the two groups.
2. To show the effectiveness of food supplementation in improving compliance with treatment. In the intensive phase of treatment (usually the first two months), we will compare the observed versus expected daily visits to the clinic, compliance with daily medications and the default rate between the intervention and control arms. In the continuation phase of treatment (usually 6 months), we will compare the observed versus expected fortnightly visits to the clinic, daily medication diaries and default rates between the two groups.
3. To show a statistically significant difference in nutritional status between the intervention arm and the control arm. We will measure clinical improvement, sputum clearance (in sputum smear positive patients), gain in BMI, changes in bio-impedence and improvement in levels of micronutrients in both groups.
4. To demonstrate the cost-effectiveness of the intervention to improve TB control in East Timor.

Randomisation Patients will be randomly assigned to receive one of two interventions: food supplements or nutritional advice (standard care). The random assignment to intervention group will be provided by the statistician at the Menzies School of Health Research. The random allocation sequence will be computer-generated (Stata Version 8.0) and concealed from all investigators throughout the study. Allocation will be stratified by community health clinic and by the TB diagnosis (sputum smear positive and other). Block randomisation will be used to maintain similar numbers of participants in both intervention groups and to minimise the potential influence of time of enrolment. The random allocation will be in sealed envelopes in two boxes (smear positive and other) in each clinic. Prior to opening the envelope, the research assistant will complete the randomisation form to check if the patient is eligible for enrolment, obtain informed consent and complete the enrolment questionnaire. The research assistant will then inform the local investigator of the name of the patient, the time of randomisation and the randomisation number.

Outcome Assessment

1. The primary outcome measure will be the proportion of patients who successfully complete the eight month course of TB treatment and achieve cure in each group (treatment success) and the proportion of patients who do not complete treatment (defaulters) or have persistent disease (treatment failure). The primary analysis will be by intention to treat. All participants in the study will contribute an outcome for analysis.
2. Secondary analyses will be conducted to evaluate the proportion of clinic visits compared with expected and response to treatment measures: symptoms (cough, sputum, fever); changes in weight, BMI and bio-impedence; sputum clearance; and micronutrient measurements.
3. The primary outcome (treatment success) will be determined by an independent observer (based in Darwin) who will be blinded to the intervention received by the patients.