key: cord-0276256-7haq2dy0
authors: Martinson, N. A.; Lebina, L.; Webb, E. L.; Ratsela, A.; Varavia, E.; Kinghorn, A.; Lala, S. G.; Golub, J. E.; Bosch, Z.; Motsomi, K. P.; MacPherson, P.
title: Household contact tracing with intensified tuberculosis and HIV screening in South Africa: a cluster randomised trial
date: 2021-10-25
journal: nan
DOI: 10.1101/2021.10.21.21265356
sha: d1844efd7789b53c3f605474296dacf4485e38d5
doc_id: 276256
cord_uid: 7haq2dy0

Background Household contact tracing for tuberculosis (TB) may facilitate TB diagnosis and identify individuals who may benefit from TB preventive therapy (TPT). In this cluster-randomised trial, we investigated whether household contact tracing and intensive TB/HIV screening would improve TB-free survival. Methods Household contacts of index TB patients in two Provinces of South Africa were randomised to home tracing and intensive HIV/TB screening (sputum Xpert and culture; HIV testing with treatment linkage; and TPT, if eligible), or standard of care (SOC, clinic referral letters). The primary outcome was incident TB or death at 15-months. Secondary outcomes included tuberculin skin test (TST) positivity in children [≤]14 years and undiagnosed HIV. (ISRCTN16006202). Results From December 2016-March 2019, 1,032 index patients (4,459 contacts) and 1,030 (4,129 contacts) were randomised to the intervention and SOC arms. 3.2% (69/2166) of intervention arm contacts had prevalent microbiologically-confirmed TB. At 15-months, the cumulative incidence of TB or death did not differ between the intensive screening (93/3230, 2.9%) and SOC (80/2600, 3.1%) arms (hazard ratio: 0.90, 95% confidence interval (CI): 0.66-1.24). TST positivity was higher in the intensive screening arm (38/845, 4.5%) compared to the SOC arm (15/800, 1.9%, odds ratio: 2.25, 95% CI: 1.07-4.72). Undiagnosed HIV was similar between arms (41/3185, 1.3% vs. 32/2543, 1.3%; odds ratio: 1.02, 95% CI: 0.64-1.64). Conclusions Household contact tracing with intensive screening and referral did not reduce incident TB or death. Providing referral letters to household contacts of index patients is an alternative strategy to home visits in high TB/HIV-prevalence settings.

Contact tracing of people with tuberculosis (TB) has been advocated as part of TB control for many 79 years [1] [2] [3] because it facilitates early diagnosis and treatment of infectious individuals and identifies 80 those who could benefit from TB preventive treatment (TPT) [4] . Although World Health 81 Organization (WHO) and numerous national guidelines recommend household TB contact tracing, 82 these have not been widely implemented in high TB burden countries because of limited effectiveness 83 data [5, 6] . The COVID-19 pandemic has severely impacted TB care and prevention programmes in is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted October 25, 2021. ; https://doi.org/10.1101/2021.10.21.21265356 doi: medRxiv preprint

Study design and participants 102 We conducted an open two-arm cluster randomised trial of household contact tracing and intensive 103 TB/HIV screening in South Africa (ISRCTN16006202). Methods have been described previously (S1 104 Protocol) [19] . In brief, we recruited index TB patients diagnosed at two South African sites with 105 large differences in annual TB incidence and HIV prevalence ( is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

This study is reported following CONSORT guidelines for cluster randomised trials (S3 Checklist).

We summarised baseline index and household characteristics by trial arm. For the primary outcome, 175 follow-up time began one month after randomisation (to avoid counting prevalent TB cases) and 

. CC-BY 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted October 25, 2021. (Table 1) . There were 4,459 household contacts identified in the intervention arm and 186 4, 192 in the SOC arm (Table 2) .

A total of 974 (94%) and 977 (95%) of households randomised to the intervention and SOC arms, is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

Protocol-specified sensitivity analyses for the primary outcome, including those who had entered the 214 household after the baseline census, and based on only bacteriologically-confirmed cases of TB, 215 showed similar results (S4 Table) . In protocol-specified subgroup analysis (S5 Table) , there was no is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Table) , TST positivity was higher in the intervention arm compared to the SoC arm among is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

There was weak evidence that household interventions reduced TB prevalence and childhood TB 268 transmission. In Uganda, intervention households received HIV testing and linkage to care, and TB is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

Our findings suggest that household contact tracing with home visits and intensive screening for TB 306 and HIV is unlikely to be considered for implementation by National TB Programmes in low-307 resourced, high TB burden settings. Although household contact tracing of index TB patients is 308 widely recommended, implementation is often poor due to the substantial resource requirements. Cost 309 studies will be reported separately, but we anticipate resource implications of household visits to be 310 substantial. Conversely, the strategy of providing referral letters for household contacts to take to their 311 . CC-BY 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted October 25, 2021. ; https://doi.org/10.1101/2021.10.21.21265356 doi: medRxiv preprint local clinics to prompt facility attendance for TB/HIV screening and care is affordable and 312 implementable at scale, but requires further implementation research and evaluation.

We found that, in the intervention arm, prevalence of latent TB (defined by TST) was 13%, 315 comparable to previous household contact tracing studies from the region [26], and with strong age-316 and site-specific dependency [28] . At 15-months, TST positivity in children was higher in the 317 intervention arm than the SOC arm. One possible explanation is differential rates of acceptance of 318 TST between the intervention and SOC arms by site: in the intervention arm in Mangaung, 319 completion of TST was 77% compared to 55% in Capricorn; in the SOC arm, completion was 84% at 320 both sites. We did not record data on reasons for refusal of the TST but it may be that those with a 321 strong response previously were reluctant to be retested: in Capricorn, children who had baseline TST 

The study had several limitations. The planned sample size was reduced due to budget constraints; is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

Supplemental material S1 Protocol: Study protocol (version 5) S2 Questionnaires: Study questionnaires S3 Checklist: CONSORT extension for cluster randomised trials checklist S4 Table: Effect of intervention versus standard of care on primary trial outcome, sensitivity analyses S5 Table: Subgroup analyses of primary and secondary outcomes by study site S6 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted October 25, 2021. ; https://doi.org/10.1101/2021.10.21.21265356 doi: medRxiv preprint

Contacts of cases of active pulmonary tuberculosis

Contacts of tuberculosis patients in high-incidence countries

Systematic screening for active tuberculosis: principles and 359 recommendations

The effectiveness of contact investigation among contacts of 362 tuberculosis patients: a systematic review and meta-analysis

Yield of household contact 364 tracing for tuberculosis in rural South Africa

High prevalence of tuberculosis and insufficient 366 case detection in two communities in the Western Cape, South Africa

Household-Contact Investigation for Detection of 379 Tuberculosis in Vietnam

Effect of household and community interventions on 381 the burden of tuberculosis in southern Africa: the ZAMSTAR community-randomised trial

Home-based tuberculosis contact investigation in 384 Uganda: a household randomised trial

Community-randomized trial of enhanced DOTS 387 for tuberculosis control in Rio de Janeiro, Brazil

Subclinical tuberculosis disease -a review and 389 analysis of prevalence surveys to inform definitions, burden, associations and screening 390 methodology

Is passive diagnosis enough? The impact of subclinical 393 disease on diagnostic strategies for tuberculosis

TEMPRANO ANRS 12136 Study Group

Antiretrovirals and Isoniazid Preventive Therapy in Africa