Skip to main content
. 2017 Nov 6;216(Suppl 7):S702–S713. doi: 10.1093/infdis/jix335

Table 1.

Estimation of the Number of Cases for Each Step of the Tuberculosis Care Cascade

Case Type, Cascade Step Cases, No. (Range) Estimation Method Calculation
Tuberculosis (all cases)
Annual burden 532005 (333760–764480) WHO 2016 time-series analysis of tuberculosis incidence in 2013 plus 50% of the number of undetected cases from 2012 Tuberculosis incidence, 2013 (all): 459000 (range, 327000–614000) [4] Tuberculosis incidence, 2012 (all): 471000 (range, 338000–627000) [4] Case detection rate, 2012: 69% (range, 52%–96%) [4] Estimated undetected cases 2012: 146000 (range, 13520–300960) 50% of undetected cases who have not died/self-cured: 73005 (range, 6760–150480)
Accessed tests 504514 (449505–589689) Add DS tuberculosis and RIF-R cases tested (see “DS tuberculosis, Accessed tests” and “RIF-R tuberculosis, Accessed tests” below) DS tuberculosis cases tested: 483912 (430534–566274) RIF-R tuberculosis cases tested: 20602 (18971–23415)
Diagnosed 435483 (411066–461766) Add DS tuberculosis and RIF-R cases diagnosed (see “DS tuberculosis, Diagnosed” and “RIF-R tuberculosis, Diagnosed” below) DS tuberculosis cases diagnosed: 417277 (393420–442963) RIF-R tuberculosis cases diagnosed: 18206 (17646–18803)
Notified and treated 372577 (354162–368051) Add DS tuberculosis and RIF-R cases notified and treated (see “DS tuberculosis, Notified and treated” and “RIF-R tuberculosis, Notified and treated” below) DS tuberculosis cases notified and treated: 361107 (354162–368051) RIF-R tuberculosis cases notified and treated: 11470
Successfully treated 279816 (279688–279945) Add DS tuberculosis and RIF-R cases successfully treated (see “DS tuberculosis, Successfully treated” and “RIF-R tuberculosis, Successfully treated” below) DS tuberculosis cases successfully treated: 274441 (274333–274549) RIF-R tuberculosis cases successfully treated: 5375 (5355–5395)
DS Tuberculosis
Annual burden 507533 (322078–720905) Annual tuberculosis burden (all cases; see “Tuberculosis, Annual burden” above) less RIF-R cases Annual tuberculosis burden: 532005 (333760–764480) RIF-R tuberculosis: 4.6% (range, 3.5%–5.7%) [2] = 24472 (11682–43575)
Accessed tests 483912 (430534–566274) Among bacteriologically diagnosed DS tuberculosis cases, missed cases estimated based on test type; test sensitivity to identify false-negative cases; proportion of missed cases diagnosed by culture tests and proportion treated empirically Add missed cases to total no. of DS tuberculosis cases diagnosed Total no. of bacteriologically diagnosed DS tuberculosis cases: 289537 (272625–308278; see “DS tuberculosis cases, Diagnosed” below) No. of HIV-positive bacteriologically diagnosed cases: 178544 (168115–190101) (see “DS tuberculosis cases, Diagnosed” below) Proportion of cases tested by Xpert: 58.7% (NHLS) Proportion of cases tested by smear: 41.3% (NHLS) Xpert sensitivity: HIV-negative, 89% (81%–94%); HIV-positive, 80% (67%–88%) [12] Smear sensitivity: HIV-negative, 76% (70%–80%); HIV-positive, 50% (42%–57%) [13, 14] Proportion with culture test: HIV-positive/Xpert-negative, 14%; HIV-positive/smear-negative, 32% [16] Proportion cases treated empirically: Xpert, 21.6%; smear, 35.1% [17]; we assumed that this was 10% higher in HIV-positive cases. Estimated missed DS tuberculosis cases: 66635 (37114–123311) Calculation for midrange shown below:
Variable HIV-Negative, No. HIV-Positive, No. Overall, No.
Underwent bacteriological tests 110992 178544 289537
Underwent Xpert 65192 104869 170060
Underwent smear 45801 73676 119476
Cases missed by Xpert 8057aa 26217 34275
Cases missed by smear 14543 75164 89707
Xpert-negative/HIV- positive with culture 3670 3670
Smear-negative/HIV-positive with culture 24053 24053
Underwent Xpert/received empirical treatment 5357 6610
Underwent smear/received empirical treatment 19707 23013
Missed cases 48594 66635
aaExample calculation of HIV-negative cases missed by Xpert = [no. tested by Xpert/Xpert sensitivity] – [no. tested by Xpert] = [65192/0.89] – 65192 = 8057
DS tuberculosis
Diagnosed 417277 (393420–442963) Pooled estimate of ILTF from published studies No. with ILTF back calculated from bacteriologically confirmed DS tuberculosis cases notified and treated No. with ILTF added to no. of DS tuberculosis cases notified and treated No. of bacteriologically diagnosed DS tuberculosis cases notified and treated (ETR): 233367bb (bbETR fields were only changed to accommodate Xpert results during 2013. Using estimates from NHLS test data, the no. with positive bacteriological tests recorded in ETR was increased by 30% in pulmonary tuberculosis in patients >7 years of age, to account for underreporting of Xpert results in early 2013) ILTF estimated from pooled analysis of studies undertaken in South Africa [17–32] (see Annexure A in Supplementary Materials) Rate of ILTF for bacteriologically confirmed tuberculosis cases: 19.4% (14.4%–24.3%) [17–32] Total no. of bacteriologically diagnosed cases: 289537 (272625–308278) No. of bacteriologically diagnosed cases with ILTF: 56170 (39258–74912) (Example of midrange calculation: no. of bacteriologically confirmed cases diagnosed = 233367/80.6% = 289537; no. with ILTF = 289537 – 233367 = 56170) Add to cases notified: 361107 (354162–368051)
Notified and treated 361107 (354162–368051) No. of cases recorded as newly registered tuberculosis cases in ETR, adjusted for underreporting of this category Cleaned, deduplicated data from ETR: 347218 newly registered patients with tuberculosis Adjusted to account for patients not previously newly registered and reported as moved in to district (estimated increase, 4%; range, 2%–6%)
Successfully treated 274441 (274333–274549) Based on mean success rate and 95% CI among adjusted no. of DS tuberculosis cases notified and treated Cleaned, deduplicated data from the electronic tuberculosis register; used system-generated outcome Numerator: no. cured and completed treatment among newly registered and moved-in patients Denominator: no. of newly registered and moved-in patients, minus those with outcome of moved out (reported at site where they moved in) Weighted mean and 95% CI calculated at district level on the basis of definition above used by South African National Department of Health: 76.00% (range, 75.97%–76.03%)
HIV-coinfected DS tuberculosis
Annual burden
HIV-positive tuberculosis 329655 (185860–509920) WHO 2016 time-series analysis of HIV-positive tuberculosis incidence in 2013 plus 50% of the number of undetected HIV-positive tuberculosis cases from 2012 HIV-positive tuberculosis incidence, 2013: 283000 (range, 182000–406000) HIV-positive tuberculosis incidence, 2012: 301000 (range, 193000–433000) Case detection rate, 2012: 69% (range, 52%–96%) [67] Estimated HIV-positive undetected cases, 2012: 93310 (range, 7720–207840) 50% of undetected cases who have not died/self-cured: 46655 (range, 3860–103920)
HIV-positive DS tuberculosis 314491 (179355–480855) Annual tuberculosis burden for HIV- positive cases less RIF-R cases Annual HIV-positive tuberculosis burden: 329655 (185860–509920) RIF-R rate assumed to be same for HIV-positive patients as for all patients: 4.6% (range, 3.5%–5.7%) [2], or 15164 (6505–29065)
Accessed tests 305910 (273121– 354848) Among bacteriologically diagnosed HIV-positive DS tuberculosis cases, missed cases estimated on the basis of test type, test sensitivity to identify false-negative cases, proportion of missed cases diagnosed by culture, and proportion treated empirically Add to total no. of HIV-positive DS tuberculosis cases diagnosed Total no. of HIV-positive bacteriologically diagnosed DS tuberculosis cases: 178544 (168115–190101; from “DS tuberculosis cases, Diagnosed” above) Proportion of cases tested by Xpert: 58.7% (NHLS) Proportion of cases tested by smear: 41.3% (NHLS) Xpert sensitivity in HIV positive: 80% (67%–88%) [12] Smear sensitivity in HIV-positive: 50% (42%–57%) [13, 14] Proportion with culture performed: HIV-positive/Xpert-negative, 14%; HIV-positive/smear-negative, 32% [16] Proportion of HIV-positive cases treated empirically [17] (assumed to be 10% higher in HIV-positive cases): Xpert, 23.8%; smear, 38.6 % See “DS Tuberculosis, Accessed tests” above for calculation of HIV-positive cases with missed DS Tuberculosis: 48594 (30516–81693)
HIV coinfected DS tuberculosis
Diagnosed 257316 (242605– 273155) Pooled estimate of ILTF from published studies No. with ILTF back calculated from bacteriologically confirmed HIV-positive DS Tuberculosis cases notified and treated No. with ILTF added to no. HIV-positive DS Tuberculosis cases notified and treated No. of HIV-positive patients with bacteriologically diagnosed DS tuberculosis notified and treated (ETR): 143907 ILTF estimated from pooled analysis of studies undertaken in South Africa [17–32] (Supplementary Materials); same proportion of ILTF used for HIV-positive cases because 2 studies found no association between ILTF and HIV-negative status Rate of ILTF for bacteriologically confirmed tuberculosis cases: 19.4% (14.4%–24.3%) [17–32] Total no. of HIV-positive patients with bacteriologically diagnosed tuberculosis: 178544 (168115–190101) No. of HIV-positive patients with bacteriologically diagnosed DS tuberculosis (ETR) with ILTF: 34638 (24209–46195) (Example of midrange calculation: no. of bacteriologically confirmed cases diagnosed = 143907/80.6% = 178544; no. with ILTF = 178544–143907 = 34638) Add ILTF to 222678 (218396–226961)
Notified and treated 222678 (218396– 226961) Proportion of HIV-positive cases applied to adjusted no. of DS tuberculosis cases notified (see “DS tuberculosis, “Notified and Treated” above). Cleaned, deduplicated data from the national electronic tuberculosis register (2013): HIV-positive, 192952; HIV-negative, 119949; HIV status unknown, 34317 (9.9%) Among those tested: 61.7% HIV-positive 61.7% of DS tuberculosis cases (361107; 354162–368051)
Successfully treated 164804 (161548– 168110) Based on mean success rate and 95% CI among adjusted no. of HIV- positive DS tuberculosis cases notified and treated Cleaned, deduplicated data from the electronic tuberculosis register; used system-generated outcome. Numerator: no. of HIV-positive patients who were cured and completed treatment among newly registered and moved-in patients Denominator: no. of HIV-positive newly registered and moved-in patients, minus patients with outcome of moved-out (reported at site where they moved in) Weighted mean and 95% CI calculated at district level on the basis of definition above used by National Department of Health: 74.01% (range, 73.97%–74.07%)
RIF-R tuberculosis
Annual burden 24472 (11682– 43575) Proportion of RIF-R cases in the annual tuberculosis burden (see “Tuberculosis, Annual burden” above) Annual tuberculosis burden (all cases): 532005 (333760–764480) RIF-R: 4.6% (range, 3.5%–5.7%) [2]
Accessed tests 20602 (18971– 23415) Back calculated from RIF-R tuberculosis cases diagnosed on the basis of cases bacteriologically diagnosed, by test type and test sensitivity RIF-R tuberculosis cases diagnosed: 18206 (17646–18803) Proportion of cases diagnosed by Xpert: 58.7% (NHLS); 10777 (10446–11131) Proportion of cases diagnosed by Genotype MDR-TBplus: 41.3% (NHLS); 7429 (7200–7673) Xpert sensitivity: M. tuberculosis, 88% (83%–92%); Rif-R TB, 94% (87%–97%) [12] Genotype MDR-TBplus: 98.1% (95.9%–99.1%) [15]No. of cases missed by Xpert: 2251 (1259–4284) No. of cases missed by Genotype MDR-TBplus: 144 (65–328)
Diagnosed 18206 (17646– 18803) Back calculated from RIF-R tuberculosis cases notified and treated and rate of ILTF Rate of initial loss to follow-up estimated from nationally representative DR-tuberculosis survey that reported results for the 2013 cohort [34]: 37% (35–39%) (Example of midrange calculation for RIF-R tuberculosis cases diagnosed = no. receiving treatment/[100% – ILTF rate] = 11470/[100% – 37%] = 18206) ILTF: 6736 (6176–7333)
Notified and treated 11470 Summary data from EDR Includes no. of patients receiving MDR-TB treatment (10880) and XDR-TB treatment (590)
Successfully treated 5375 (5355–5395) Proportion of RIF-R tuberculosis cases notified and treated that were successfully treated Numerator: no. of DR-tuberculosis patients who were cured and completed treatment Denominator: no. of patients with DR-tuberculosis in 2013 cohort Weighted mean and 95% CI calculated at provincial level: 46.86% (46.69%–47.04%)

Abbreviations: CI, confidence interval; DR, drug resistant; DS, drug susceptible; EDR, electronic drug-resistant tuberculosis register; ETR, electronic tuberculosis register; HIV, human immunodeficiency virus; ILTF, initial loss to follow-up; MDR-TB, multidrug-resistant tuberculosis; M. tuberculosis, Mycobacterium tuberculosis; NHLS, South African National Health Laboratory Service; RIF-R, rifampicin resistant; WHO, World Health Organization; XDR-TB, extensively drug-resistant tuberculosis; Xpert, Xpert MTB/RIF test.