Main features and findings of studies reporting on tuberculosis (TB) incidence and/or risk of TB progression among individuals with select risk factors: corticosteroid use (n=1), diabetes mellitus (n=1), glomerular disease (n=1), hepatitis C (n=1), malignancies (n=1), rheumatoid arthritis (n=3), other rheumatic diseases and their treatments (n=3) and vitamin D deficiency (n=1)
| First author, year [reference] | Country and study period | Study design | Population characteristics | Main quantitative findings (95% CI) | Data sources and other notes | |
| Risk group/cases | Comparator group/controls | |||||
| Corticosteroid use | ||||||
| Lai, 2015 [19] | Taiwan January 1999 to December 2011 | CC (nested within a cohort of 1 000 000 randomly selected subjects from the NHIRD) | 6229 people aged >15 years with active TB; mean age 59.1 years; 30.4% female | 622 900 people aged >15 years without active TB; mean age 59.1 years; 30.4% female 100 controls selected for each case using risk-set sampling | Adjusted IRR of active TB according to: Corticosteroid use within previous 30 days of TB diagnosis date 2.76 (2.44–3.11) Corticosteroid use within previous 31–90 days 1.99 (1.73–2.31) Corticosteroid use within previous 91–365 days 1.17 (1.06–1.29) | NHIRD Using a time-matched CC sampling scheme, conditional odds ratios were used to estimate rate ratios Corticosteroid use could be oral or i.v. for any indication |
| Diabetes mellitus | ||||||
| Lin, 2018 [20] | Taiwan January 1998 to December 2010 | RC (subcohorts from a cohort of 1 000 000 randomly selected subjects from the NHIRD) | 49 028 adults aged 20–100 years with newly diagnosed T2DM; mean age 50.7 years; 48.9% female | 49 028 adults without T2DM, frequency matched for age, sex, region and year of T2DM diagnosis; mean age 50.6 years; 48.9% female | Adjusted HR of active TB 2.01 (1.80–2.25) | All claims data from the NHIRD |
| Glomerular disease | ||||||
| Gunning, 2021 (not peer-reviewed) [12] | Canada 2000 to 2012 | RC (province of British Columbia) | 3079 adults aged ≥18 years diagnosed on native kidney biopsy with glomerular diseases; mean age at biopsy 50.9 years; 48.6% female | Age-standardised population of the province, sample size not reported; sex data not reported | SIR of active TB 23.36 (16.76–31.68) Unadjusted HR of active TB with use of immunosuppressive agents 2.13 (1.13–4.03) (immunosuppressive agents categorised as corticosteroids, calcineurin inhibitors, antimetabolites, cyclophosphamide and rituximab) | Provincial pathology database; provincial clinical information system for patients with kidney disease; BC Vital Statistics; Population Data BC |
| Hepatitis C | ||||||
| Baliashvili, 2022 [21] | Georgia January 2015 to September 2020 | RC (among all adult residents of Georgia tested for anti-HCV antibodies) | 1) 70 341 HCV-positive adults who had not finished the HCV treatment course (untreated HCV) 2) 53 456 HCV-positive adults who had finished the HCV treatment course (treated HCV) Age and sex data not reported | 1 708 017 HCV-negative adults (uninfected HCV); age and sex data not reported | Adjusted HR of active TB in: Untreated HCV versus uninfected HCV 2.9 (2.4–3.4) Treated HCV versus uninfected HCV 1.6 (1.4–2.0) | National HCV screening registry; hepatitis C elimination programme clinical database “Elimination C” (ElimC); national TB surveillance database managed by the National Center for TB and Lung Disease |
| Malignancies | ||||||
| Cheon, 2020 [22] | Republic of Korea January 2000 to December 2014 | RC (one hospital in Ulsan province) | 34 783 adults aged ≥20 years newly diagnosed with malignancies, based on ICD-10 codes C000-C999; median age 58 years (IQR 48–68 years); 49.8% female | 1) 69 566 age- and sex-matched adults with no history of cancer who visited the hospital for health screening during the risk group enrolment period and were followed for >3 years afterwards 2) 1 151 402 people, the total population of Ulsan province, averaged over period 2000–2017; age and sex data not reported | 1) IRR of active TB in cancer patients versus comparator group 1: 10.68 (8.83–12.99) for all TB 5.82 (4.41–7.67) for clinically diagnosed TB 14.30 (11.91–17.18) for bacteriologically confirmed TB 2) IRR of active TB in cancer patients versus comparator group 2: 9.71 (8.99–10.48) for all TB | Ulsan University Hospital Information of Clinical Ecosystem; Korean Statistical Information Service |
| Rheumatoid arthritis | ||||||
| Brassard, 2006 [13] | USA September 1998 to December 2003 | CC (nested within a cohort constructed from PharMetrics database of medical claims data from >85 managed care organisations) | 386 adults aged ≥18 years with RA and TB; mean age 54 years; 77.2% female | 38 600 adults aged ≥18 years with RA but not with TB; mean age 56 years; 73.7% female | Adjusted IRR of active TB according to medication use in the previous year: Traditional DMARDs 1.2 (1.0–1.5) NSAIDs 0.9 (0.8–1.1) COX-2 inhibitors 0.9 (0.7–1.2) Corticosteroids 1.7 (1.3–2.2) | PharMetrics Patient-Centric Database |
| Brassard, 2009 [14] | Canada January 1980 to December 2003 | RC (from physician billing codes in the province of Quebec) | 24 282 people with one or more occurrences of the physician billing code for RA during an inpatient or outpatient visit; mean age 61.7 years; 70.1% female | General population of Quebec; sample size unclear; age and sex data not reported | SIR of active TB in patients with RA versus the general population 10.9 (7.9–15.0) | Provincial physician billing codes No details given on route of corticosteroid administration |
| CC (nested within aforementioned RC) | 50 people with RA and TB; mean age 65.6 years; 50% female | 1500 people with RA but not with TB; mean age 67.6 years; 70.0% female | Adjusted IRR of active TB according to medication use in the previous year: Any DMARD 3.0 (1.6–5.8) Corticosteroids 2.4 (1.1–5.4) COX-2 inhibitors 1.4 (0.5–4.4) NSAIDs 1.2 (0.6–2.3) | |||
| Carmona, 2003 [15] | Spain 1990–2000 | RC (random selection of patients from 34 clinical centres throughout the country) | 788 people aged ≥16 years with RA; mean age at baseline visit 61 years; mean age at RA onset 48 years; 72.1% female | Age- and sex-standardised general population of Spain; sample size not reported; age and sex data not reported | IRR of active TB (any TB location) 4.13 (2.59–6.83) IRR of active pulmonary TB 3.68 (2.36–5.92) | National Network of Epidemiological Surveillance; clinical registries of the participating clinics |
| Other rheumatic diseases and their treatments | ||||||
| Chen, 2013 [16] | Taiwan 1996–2008 | RC (from NHIRD) | 81 266 people with psoriasis or psoriatic arthritis; median age 43.0 years (IQR 28.7–57.8 years) | General population of Taiwan; sample size not reported; age and sex not reported | Crude IRR of active TB 1.22 (1.18–1.33) | NHIRD Antipsoriatic drugs defined as methotrexate, acitretin, cyclosporine, azathioprine and mycophenolate mofetil Corticosteroids were systemic, with no details given on route of administration |
| CC (nested within the aforementioned RC) | 497 people with psoriasis and TB; mean age 59.7 years; 17.5% female | 1988 people with psoriasis without TB; mean age 59.7 years; 17.5% female | Adjusted odds ratios of active TB according to: Antipsoriatic drugs 0.83 (0.52–1.31) Corticosteroids 3.98 (3.12–5.06) NSAIDs 2.20 (1.76–2.76) COX-2 inhibitors 1.20 (0.71–2.01) | |||
| Cho, 2020 [17] | Republic of Korea January 2012 to December 2018 | RC (from National Health Insurance Service database) | 2803 people aged ≥10 years who were administered ustekinumab for psoriasis, psoriatic arthritis and Crohn's disease; median age not reported; 32% female | “General Korean population”; sample size not reported; age and sex not reported | IRR of active TB 0.76 (0.59–2.02) | National Health Insurance service database; annual report on notified TB by Korea Centers for Disease Control and Prevention |
| Wu, 2017 [18] | Taiwan January 1999 to December 2011 | CC (nested within a cohort of 1 000 000 randomly selected subjects from the NHIRD) | 1) 123 419 person-years from adults aged ≥18 years taking traditional NSAIDs (mostly but not exclusively with arthritis or other rheumatic conditions); mean age 63.1 years; 31.4% female 2) 16 392 person-years from people aged ≥18 years taking COX-2 inhibitors (mostly but not exclusively with arthritis or other rheumatic conditions); mean age 75.5 years; 36.4% female | Adults aged ≥18 years not using traditional NSAIDs or COX-2 inhibitors (from general population, not matched on comorbidities); 340 396 person-years; mean age 55.9 years; 28.0% female 100 controls selected for each case using risk-set sampling | Adjusted IRR of active TB according to: Use of traditional NSAIDs within the previous 31–90 days 1.19 (1.05–1.35) Use of COX-2 inhibitors within the previous 31–90 days 1.07 (0.78–1.48) | NHIRD NSAID and COX-2 inhibitor use defined as having a prescription record ≥7 days Using fractional polynomial disease risk scores, odds ratios were used to estimate rate ratios |
| Vitamin D deficiency | ||||||
| Patterson, 2020 [23] | UK April 2010 to January 2019 | RC (individuals diagnosed with LTBI in one London hospital) | 1) 320 adults aged ≥18 years with moderately deficient vitamin D levels (25[OH]D 25.0–50.0 nmol·L−1) 2) 114 adults aged ≥18 years with profoundly deficient vitamin D levels (25[OH]D <25.0 nmol·L−1) Age and sex data not reported | 554 adults aged ≥18 years with sufficient vitamin D levels (25[OH]D >50.0 nmol·L−1); age and sex data not reported | Adjusted HR of active TB in: Moderately deficient versus sufficient vitamin D levels 2.14 (0.84–5.48) Profoundly deficient versus sufficient vitamin D levels 5.68 (2.18–14.82) | London North West University Healthcare electronic medical records; London TB Register, Extended Tuberculosis Surveillance Vitamin D levels within 14 days before and up to 365 days after the time of LTBI diagnosis were linked to individuals in the LTBI cohort |
CC: case–control study; NHIRD: National Health Insurance Research Database; IRR: incidence rate ratio; RC: retrospective cohort study; T2DM: type 2 diabetes mellitus; HR: hazard ratio; SIR: standardised incidence ratio; HCV: hepatitis C virus; ICD-10: International Classification of Diseases, tenth revision; IQR: interquartile range; RA: rheumatoid arthritis; DMARDs: disease-modifying antirheumatic drugs; NSAIDs: nonsteroidal anti-inflammatory drugs; COX: cyclooxygenase; LTBI: latent TB infection.