Study setting
Brazilian guidelines for TB control and prevention recommend isoniazid 300mg/day for 6 or 9 months in HHCs of all ages who have a positive tuberculin skin test (TST) or positive Interferon-Gamma Release Assay (IGRA). Rifampicin may be used in cases were isoniazid is not recommended, and is preferred for children under 10 years of age, adults over 50 years of age, and those with liver diseases. All treatments for LTBI are self-administered and LTBI is not a compulsory notifiable condition. (10)
This study was carried out in two clinics where the intervention trial had previously been implemented. Both clinics are run by the Department of Health of Rio de Janeiro and were set in urban areas. (9)
Study periods
The intervention trial was conducted from May to October of 2018. The follow-up study consisted of two components. Firstly, we conducted a retrospective analysis of the cascade of LTBI care of HHC of index-TB patients diagnosed between November 1, 2018 to March 31, 2019. Secondly, questionnaires were administered to TB patients, HHCs and HCWs between May 7, 2019 to July 4, 2019.
Intervention trial
The intervention trial consisted of a rapid public health evaluation to identify barriers to LTBI treatment for HHC of index-TB patients, followed by site specific selection of strengthening activities and implementation of interventions to the barriers. (9)
In Rio de Janeiro, Brazil, LTBI program strengthening interventions consisted of: (i) initial training of HCWs by a TB physician covering all steps of the LTBI cascade of care, (ii) intensified in-service training of HCWs provided by an infectious disease physician (weekly visits for the first two months, then every two weeks for two months, then once a month), (iii) development and use of a contact registry to facilitate a cascade analysis to support the in-service training of HCWs, (iv) leaflets with educational information for index-TB patients and their contacts and (v) educational material developed for health care workers (TB booklet). (9)
After October 2018, all trial interventions were stopped. This meant cessation of in-service training, provision of leaflets to HHC and index patients, and no further on-site visits by research staff.
Follow-up study
For the cascade of LTBI care, information on HHCs was obtained from the TB registry at each clinic. After diagnosis of index-TB patients, information on the number of HHCs recorded at each of the following steps was abstracted: HHC identification, initial assessment (TST application and measurement), medical evaluation (consultation with a doctor or nurse, chest radiograph, sputum testing), and HHC initiating treatment (Supplementary Figure 1). Only HHC of new microbiologically confirmed pulmonary index-TB patients (using either AFB, TB culture, Xpert MTB/RIF or a combination thereof) diagnosed between November 1, 2018 and March 31, 2019 were included in this cascade analysis. Information on the LTBI regimen participants received was not recorded.
In the follow-up study, interviewer-administered open-ended structured questions were applied to TB patients, HHCs and HCWs. TB patients were eligible to be interviewed if they had confirmed (as defined above), or clinical pulmonary TB (defined as a TB diagnosis based on chest X-ray abnormalities and suggestive signs and symptoms, with negative or absent microbiological test results). A HHC was defined as someone who slept in the same house at least one night per week, or spent more than one hour in the house at least five days per week, on average, with an index-TB patient, over the preceding 3 months, and in whom active TB has been ruled out. This includes child HHCs, in which case the parents or legal guardians were interviewed. HCWs were defined as doctor, nurse, auxiliary nurse or community health agent that assisted TB patients in either of the two clinics. All HCWs were employed by the Municipal Department of Health of Rio de Janeiro and were not funded or received any incentive from the research funds.
For the questionnaires on barriers and facilitators, a consecutive sample of all TB patients and their HHCs presenting to both clinics from May 7, 2019 date to July 4, 2019 were invited to participate. The health clinic directors identified HCW from the TB programs; all were approached and accepted to participate.
For the questionnaires, semi-structured knowledge, attitudes and practices questionnaires were adapted from those used in the ACT4 trial (Supplementary Methods 1 to 3). For HHCs and TB patients, questions focused on perspectives and perceptions about the identification of contacts and reasons why linkage to LTBI care was or was not achieved. For HCWs, open-ended questions related to motivation for contact tracing and continuation of activities implemented in the intervention trial were used. All questionnaires were interviewer-administered in the participating health clinics. All interviewees were 18 years of age or older. Written informed consent was provided by all participants prior to data gathering. For participants who could not read or write, the informed consent form was read to them and verbal consent as well as their fingerprint in place of signing was provided. If a family member was accompanying the interviewee, then they were asked to read the informed consent to the patient prior to agreeing to participate.
Outcomes
The two primary outcomes were the number of HHCs identified and the number of HHCs initiating LTBI treatment within three months of diagnosis of the index-TB patient. Both outcomes from the follow-up study (November 2018 to March 2019) were compared to: (i) the pre-trial intervention period (July-December 2017), and (ii) the intervention period of the trial (May-October 2018). Outcomes were presented per 100 index-TB patients.
Secondary outcomes included: current barriers and facilitators to LTBI linkage to care, and acceptance and initiation of treatment identified by index-TB patients, HHCs and HCWs; and identifying the acceptability of study interventions from a HCW perspective.
Analysis
A quasi-Poisson regression model (accounting for over dispersion), with identity link, was used to compare the cascade of care data (number of HHC identified and number of HHC starting LTBI treatment) between the three time periods.
For questionnaire responses, open-ended questions were transcribed from audiotapes and coded by two independent reviewers (MYL, MLB) into common themes. Disagreements were resolved by consensus. The results from these common themes are presented as frequencies and proportions. We compared the responses of HHCs and TB patients from the intervention trial questionnaires to the follow up study questionnaires using the Mantel-Haenszel method for adjusted odds ratio and Wald method for the confidence intervals.
Data analysis was performed using the statistical package R version 3.5.1 (R Foundation for Statistical Computing, Vienna, Austria) and Microsoft Office Excel 2016 (Microsoft Corporation; Redmond, WA).
Ethical approval
The study was approved in Rio de Janeiro by the Municipal Health Ministry (CAAE 38278214.3.1001.5279) and by the McGill University Health Centre ethical review board (15-291-MUHC).
Role of funding source
This study was funded by the Canadian Institute of Health Research, grant number FDN-143350. The funder had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.