Impact of tuberculosis screening and isoniazid preventive therapy on incidence of TB and death in the TB/HIV in Rio de Janeiro (THRio) study B. Durovni1,2, V. Saraceni1, A. Pacheco3, S. Cavalcante1,3, S. Cohn4, B. King4, L. Moulton4, R. Chaisson4, J. Golub4, THRio study group 1Rio de Janeiro City Health Secretariat, Rio de Janeiro, Brazil, 2Federal University of Rio de Janeiro, Rio de Janeiro, Brazil, 3Fiocruz, Rio de Janeiro, Brazil, 4Johns Hopkins University, Baltimore, United States
THRio Objectives To determine if implementation of a policy of widespread use of isoniazid preventive therapy (IPT) in HIV-infected patients with access to ARV therapy reduces the incidence of active TB and death in the HIV clinic population Scale up of tuberculin skin testing (TST) and IPT among HIV-infected patients in public primary health units in Rio de Janeiro within the current HIV clinic infrastructure In an effort to reduce: TB Incidence Mortality
Setting 29 Public Health units managed by the Health Department of Rio de Janeiro City AP5.3 AP5.2 AP4 AP5.1 AP3.3 AP3.2 AP1 AP3.1 AP 2.1 AP2.2 3
THRio Study Design and Timeline Cluster-Randomized, Step-Wedge Trial Sep 05 Jan 08 48 60 Aug 09 Intervention and Follow-up Period (for all clinics) 4
Intervention Training clinics to properly implement TB screening and adhere to TST/IPT policy for all HIV-infected patients TST to be done for all eligible clinic patients No prior TB history No prior IPT No prior +TST IPT x 6 months for all TST+ without active TB and all contacts of active TB cases 5
TST and IPT 69% had at least one TST placed and read 83% started IPT 0.84% had an adverse event 84% completed IPT
Time to TST and Time to IPT Before and After THRio Intervention Time to TST and time to IPT are both markedly improved post-intervention Durovni et al., AIDS 2010, 24 (suppl 5):S49–S56
Methods for Current Analysis Primary endpoints: Incidence of TB and TB or death at the clinic level before and after the intervention Eligible patients who made > 1 visit after 1 Sept 2003 ‘Eligible’ = no prior TB or IPT ‘Prevalent’ TB and deaths (within 60 days of enrollment) excluded Patients remain in the denominator until TB or Death Intent-to-treat Analysis – includes all eligibles “Stayers” Analysis -- among those remaining in clinic contact, censoring those missing for >1 year (mITT) Statistical analysis: Crude hazard ratios (HR) obtained from frailty-adjusted Cox models are presented
CONSORT Diagram (modified for stepped wedge cluster-randomized trial) Clinics Eligible for Inclusion (n=29) Clinics Receiving Intervention (n=29) Patients in Clinics Eligible for Intervention (n=12,815) Eligible patients contributing to control phase (n=9,853 ) Eligible patients contributing to intervention phase (n= 10,840) Patients in Clinics Ineligible for Intervention (n=4,480)
THRio cohort characteristics (n=12,815) Median age: 37 years old Male: 61% Median years since HIV diagnosis: 2.4 years HAART at entry: 60% Median CD4 cell count at entry: 403 cells/mm3
Incidence of TB Over Time X: start of intervention | : control case | : intervention case
THRio Results TB cases, total contribution time, incidence per 100pyrs Control Phase Intervention Phase Cases 221 254 Person years 16,834 23,126 Rate/100pyrs 1.31 1.10 TB/Death cases, total contribution time, incidence per 100pyrs Control Phase Intervention Phase Cases 617 696 Person years 16,834 23,126 Rate/100pyrs 3.67 3.01
THRio Results: Unadjusted Cox Models Outcome Cases HR (95% CI) p-value Intent To Treat TB 475 0.87 (0.68-1.10) 0.233 TB or Death 1313 0.72 (0.62-0.82) <0.001 Intent-to-treat – Among all eligibles
THRio Results: Unadjusted Cox Models Outcome Cases HR (95% CI) p-value Intent To Treat TB 475 0.87 (0.68-1.10) 0.233 TB or Death 1313 0.72 (0.62-0.82) <0.001 Modified To Treat (Stayers) 403 0.57 (0.44-0.76) TB or Death 1073 0.56 (0.47-0.66) Intent-to-treat – Among all eligibles Stayers – mITT - Among those remaining in clinic contact (Patients censored at the moment they go one year without a clinic contact)
Conclusions Overall, the THRio intervention had a modest impact (13% reduction) on TB, but showed an important and statistically significant impact on TB and death (28% reduction) in the primary intent-to-treat analysis; The “Stayers” analysis, including those who were more likely to be exposed to the intervention (mITT), showed an even larger and highly significant impact on reduction of TB incidence (43%) and TB incidence and death (44%) TST screening and provision of IPT to HIV-infected patients with access to highly active antiretroviral therapy significantly reduces the risk of TB and death and should be widely implemented
THRio Study Team Rio de Janeiro Betina Durovni Solange Cavalcante Valeria Saraceni Antonio Pacheco Giselle Israel Vitoria Vellozo Rita Ferreira Lilian Lauria THRio Study Team JHU Richard Chaisson Jonathan Golub Larry Moulton Silvia Cohn Bonnie King Anne Efron Susan Dorman THRio Study Team Funding: Bill and Melinda Gates Foundation