Badri (7)
(2002) | Observational cohort | HIV patients attending New Somerset Hospital adult HIV clinic, University of Cape Town, between 1992 and 2001. This is a major public health-care facility. | Inclusion: aged >15 years. Exclusion: acute opportunistic infection, significant laboratory abnormalities, current evidence of active substance abuse, pregnancy or lactation, TB at baseline visit, use of IPT in the 6 months prior to baseline, and treatment with immune-modulating or systemic chemotherapeutic agents | HAART (defined as 2 NRTIs plus one of the following: (i) an NNRTI, (ii) a PI, or (iii) a third NRTI) | For patients on ART, every 2–3 months, clinical, immunological and virological information was taken, or more frequently if clinically indicated. For patients not on ART, follow-up was every 3–6 months. Mean follow-up in those exposed to ART was significantly greater than in the controls (16.8 (SD 8.3) VS13.2 (SD 15.5) months). | 2 cases during 100.1 person-years on ART with baseline CD4 counts >350 cells/mm3; 14 cases during 388.3 person-years off ART with baseline CD4 counts >350. Adjusted RR was 0.36 (95% CI 0.10 to 1.74) for this CD4 stratum.
2 cases during 121.2 person-years on ART with baseline CD4 counts 200–350; 27 cases during 225 person-years off ART with baseline CD4 counts 200–350. Adjusted RR was 0.12 (95% 0.03 to 0.53) for this CD4 stratum.
Across all CD4 strata, of the 9 TB cases on ART, 5 were definite (culture/autopsy confirmed) and 4 were probable (presence of AFB or histological finding of caseating granulomata). Of the 82 TB cases off ART, 48 were probable and 34 were definite. |
del Amo (8)
(2011) | Observational cohort | This analysis from the HIV-CAUSAL Collaboration includes 11 observational cohorts from 1996 to 2007. The cohorts are in the UK, the Netherlands, France, Spain and the USA. | HIV-infected individuals who were aged ≥ 18 years, antiretroviral naive, without AIDS, not pregnant, and in whom CD4 counts and viral load were measured within 6 months of each other at baseline. Individuals with TB during the first month of follow up were excluded (prevalent TB case). | Combined ART (defined as a regimen including >3ARTs, 2 ritonavir-boosted Pls, or 1 NNRTI and 1 boosted PI) | For each patient, follow-up ended at the earliest of: TB diagnosis, death, 12 months after the most recent laboratory measurement, pregnancy (if known), or the cohort-specific administrative end of follow-up. | The overall hazard ratio of TB for combined ART was 0.56 (95% CI 0.44–0. 72).
The hazard ratios were similar for individuals with baseline CD4 counts >350 (0.50, 95% CI 0.36–0.69) and 200–350 (0.45, 95% CI 0.27, 0.74).
TB cases were diagnosed according to standard clinical practice in Europe and the USA. |
Girardi (13)
(2005) | Observational cohort | The ART Cohort Collaboration (ART-CC) is an international collaboration of cohort studies from Europe and North America. This study included 12 cohorts. | Antiretroviral-naive subjects had to be aged ≥16 years. Data were from subjects who started ART between 1996 and 2008 (ART-CC updated the data presented in the 2005 article). Follow-up was censored at 3 years from start of ART | HAART (defined as at least 3 drugs, including Pls, NNRTIs and NRTIs) | Data on HAART regimen (PI/NNRTI/NRTI-based), CD4 count, viral load, gender, age, year of HAART initiation and risk group (MSM, injection drug user, heterosexual, blood-product recipient, other/not known) were collected | 676 cases during 124 669 person-years on HAART with baseline CD4 counts ≤350; 58 cases during 23 420 person-years on HAART with baseline CD4 counts 351–500.
The IRR was 0.46 (95% CI 0.35, 0.60) for higher vs lower baseline CD4 counts at HAART initiation.
All TB cases underwent antituberculosis treatment. Since the standard of medical care is to confirm TB prior to starting antituberculosis treatment, investigators indicate that all the TB cases were culture-confirmed. |
Golub (12)
(2009) | Observational cohort | HIV patients receiving primary HIV care at two clinics affiliated with the University of the Witwatersrand: the Perinatal HIV Research Unit (PHRU) in Soweto, a large, urban setting, and the Tintswalo Hospital, a remote, rural clinic in Mpumalanga Province. | HIV-infected adults aged >18 years. No exclusion criteria listed | HAART (not defined) and/or IPT | Time between visits is scheduled to be 4–7 months, but patients may visit the clinic at any time. Follow-up data collection focuses on clinical diagnoses, symptoms and longitudinal data including height and weight, smoking status, alcohol consumption and measures of socioeconomic status. Laboratory tests include CD4 and full blood counts at recruitment and every 6 months. | 200 cases of TB during 2815 unexposed person-years; 44 cases during 952 person-years on HAART; 1 case during 93 person-years on IPT and ART The IRR for HAART was 0.65 (0.46–0.91). The IRR for IPT and HAART was 0.15 (0.00–0.85). The adjusted HR for HAART was 0.36 (0.25, 0.51). The adjusted HR for IPT and HAART was 0.11 (0.02, 0.78). |
Golub (11)
(2007) | Observational cohort | Baseline medical records were analysed for a 2-year period (from 1 September 2003 to 1 September 2005) in 29 public clinics in Rio de Janeiro, Brazil. | Patients who had made at least one visit to one of the clinics (from 1 September 2003 to 1 September 2005) and who received their primary
care from the clinics were included. Patients who attended the clinic to collect antiretroviral medications prescribed by a private physician were excluded, as were patients who died before 1 September 2003. | ART (not defined) and/or IPT | Follow-up ended at TB diagnosis or the earlier of their last visit
to an HIV clinic, the date of administrative censoring or death. Information collected included age, sex, date of HIV diagnosis, treatment history (antiretroviral drugs, IPT), dates of opportunistic diseases including TB, and results of diagnostic tests including CD4 counts, HIV viral loads and TST. | 155 cases during 3865 unexposed person-years; 221 cases during 11627 person-years on ART; 10 cases during 1253 person-years on ART and IPT. IRR for ART was 0.48 (0.39, 0.59). IRR for ART and IPT was 0.20 (0.09, 0.91). Adjusted HR was 0.41 (0.31, 0.54) for ART. Adjusted HR was 0.24 (0.11, 0.53) for ART and IPT. |
Jones (14)
(2000) | Observational cohort | The Adult/Adolescent Spectrum of HIV Disease project is a multicentre observational cohort study including sites in Atlanta, GA; Dallas, Houston, and San Antonio, TX; Denver, CO; Detroit, MI; Los Angeles, CA; Seattle, WA; New Orleans, LA; New York, NY; and Bayamon, Puerto Rico. | HIV-infected persons are identified at their first health-care encounter regardless of the stage of their HIV infection. All patients aged >13 years who were infected with HIV and who attended participating clinics during the study period were eligible for enrolment. | HAART (defined as triple therapy with 2 NRTIs and an NNRTI or PI, or any combination including a PI) | On successive 6-month follow-up intervals, medical records are reviewed for illnesses, AIDS-defining conditions, prescriptions, laboratory tests and medical care utilization. TB incidence was assessed by examining the first prospective occurrence of TB from January 1996 to June 1998. | 45 cases during 6250 unexposed person-years; 7 cases during 3684 person-years on HAART. The IRR was 0.2 (0.1, 0.5). |
Lau (15)
(2010) | Observational cohort | The North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD) includes 22 research groups, representing more than 60 study sites. | HIV-infected patients who received medical care between January 1996 and December 2005 and had no history of an AIDS-defining illness and were ART-naive. | HAART (defined as a regimen containing at least 3 ART drugs, one of which had to be a PI, an NNRTI, abacavir/tenofovir, an integrase inhibitor, or maraviroc/enfuvirtide) | Individuals contributed person-time starting from the month of initiation of HAART up until TB diagnosis, loss to follow-up (last contributed CD4 count), or administrative censoring. | 124 cases during 149 011 person-years on HAART with baseline CD4 counts ≤ 350; 13 cases during 27 601 person-years on HAART with baseline CD4 counts 351–500. The IRR was 0.57 (95% CI 0.32, 1.00) for earlier HAART; 8 cases during follow-up for persons on HAART with baseline CD4 counts >500. Of the 145 TB cases, 98 were culture-positive, 39 were culture-negative and 8 were of unknown culture status. For the TB cases that were not positive, TB diagnosis was established by signs, symptoms and chest radiography consistent with TB, pathology with necrotizing granulomas and AFB, positive NAAT and/or clinical response to antituberculosis therapy. |
Ledergerber (16)
(2000) | Observational cohort | The Swiss HIV Cohort Study enrols HIV-infected persons aged >16 years in Basel, Bern, Geneva, Lausanne, Lugano, St Gall and Zurich. | All participants who started potent ART between September 1995 and December 1997; had a CD4 count and a viral load during the 3 months before starting; and made ≥ 1 follow- up visit >1 month after starting ART. | Potent ART (defined as combination treatment with at least 3 drugs, including at least 1 PI) | Data are collected at 6-month intervals. Medications are registered by month of initiation and discontinuation. | 7 cases during 897 unexposed person-years; 6 cases during 2727 person-years on ART. The IRR was 0.28 (95% CI 0.09, 0.84). |
Moreno (17)
(2008) | Observational cohort | CoRIS-MD is an open multicentre hospital-based cohort study within Spain's HIV Research Network of Excellence. | All HIV-infected subjects aged >18 years, with at least 6 months of follow-up, seen at any time between 1 January 1997 and 30 June 2003 in any of the participating centres. | HAART (defined as ART regimens including 2 NRTIs plus either an NNRTI or a PI, or 3 NRTIs) | Data were collected for: sex, age, transmission category, period of observation, CD4 count, viral load, AIDS diagnosis, treatment status at entry and over time, and vital status. Median follow-up was 4.1 years (IQR, 2.0–6.4) in HAART-naive and 2.9 years (1.5–4.6) in HAART. | 129 cases during 5215 unexposed person-years; 75 cases during 7825 person-years on HAART. The IRR was 0.26 (95% CI 0.16, 0.40). |
Cohen (10)
(2011) | Randomized controlled trial | The trial had global representation with 13 sites in Botswana, Brasil, India, Kenya, Malawi, South Africa, Thailand, the United States and Zimbabwe. | HIV-serodiscordant couples in which the HIV-infected partner is ART-naive and has a CD4 count between 350 and 550 cells/mm3 | The ART regimens in this study were consistent with WHO guidelines | All participants completed monthly follow-up visits throughout the study. For those on ART, laboratory and adherence measurements took place. For those not on ART laboratory and counselling took place. | 33 cases among the 877 adults initiating ART with CD4 counts <350 cells/μL; 17 cases among the 886 people initiating ART with CD4 counts >350 cells/μL This translated into an RR 0.51 (0.28–0.91).a |
Severe (9)
(2010) | Randomized controlled trial | The study was conducted at the centre of the Haitian Group for the Study of Kaposi's Sarcoma and Opportunistic Infections in Port-au-Prince, Haiti | HIV-infected participants aged ≥ 18 years with CD4 counts from 200 and 350 cells/mm3 within 45 days before enrolment were included. Participants with a history of an AIDS-defining illness or who had received ART previously were excluded | Treatment included lamivudine, zidovudine, and efavirenz | Adherence was measured every 6 months using a questionnaire. Adverse events were monitoring using DAIDS criteria. Complete blood count, liver enzyme tests and serum chemical tests were repeated every 3 months for participants on ART. CD4 results were collected 6-monthly. | 36 cases of TB among the 393 people who were randomized to defer ART until their CD4 counts were <200; 18 cases of TB among the 380 people who were randomized to initiate ART immediately (i.e. between 200 and 350). This translated into an RR of 0.52 (0.30–0.89).b |
