Comparing the diagnostic performance of QuantiFERON

We identified 3966 studies; 83 were selected for full-text review and 42 articles were excluded (Fig. 1 and Additional file 1: Table S14), leaving 41 studies that met our inclusion criteria. Twelve studies evaluated sensitivity [10, 21,22,23,24,25,26,27,28,29,30,31], seven evaluated specificity [21, 23, 24, 28, 31,32,33] and thirty-one evaluated positive rate [10, 22, 27, 30, 32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58].

Flow diagram for search and study selection

Twelve studies compared the sensitivity of QFT-Plus with QFT-GIT in patients with active TB (Additional file 1: Table S15). Of these, three studies compared the sensitivity of QFT-Plus with T-SPOT.TB [10, 23, 28]. The patient population that was enrolled included patients from six countries and the patient age range 0–89 years.

Seven studies compared the specificity of QFT-Plus with QFT-GIT, T-SPOT.TB and TST in populations with a very low risk of TB exposure (Additional file 1: Table S16), of which two studies compared the positive rate of QFT-Plus with QFT-GIT and QFT-Plus with T-SPOT.TB [23, 28]. The patient population that was enrolled included patients from four countries and the patient age range 2.5–75 years.

Thirty-one studies compared the positive rate of QFT-Plus with QFT-GIT, T-SPOT.TB and TST in high-risk populations (Additional file 1: Table S17), and two studies compared the positive rate of QFT-Plus with QFT-GIT and QFT-Plus with T-SPOT.TB [10, 45]. One study was included twice because it involved two populations that met the criteria for a high-risk population [10]. The patient population that was enrolled included patients from 13 countries and the patient age range 2–102 years.

We have not retrieved the original literature comparing QFT-Plus and TST in patients with active TB. Therefore, RD values were used exclusively for reporting the sensitivity of QTF-Plus versus QTF-GIT and T-SPOT.TB.

The pooled difference in sensitivity between QFT-Plus and QFT-GIT was 0.01 (95% CI − 0.02 to 0.03; Fig. 2) in 12 studies with 1004 participants. As shown in Additional file 1: Fig. S1 and Table 1, the pooled estimates of sensitivity were 0.886 (95% CI 0.812 to 0.944) and 0.879 (95% CI 0.802 to 0.939) for QFT-Plus and QFT-GIT, respectively. Subgroup analysis was conducted stratified by age of the participants, TB burden of the areas, and number of participants (Additional file 1: Fig. S9).

Pooled difference in sensitivity between QFT-Plus and QFT-GIT in 12 studies

The pooled difference in sensitivity between QFT-Plus and T-SPOT.TB was 0.09 (95% CI − 0.09 to 0.28; Fig. 3) in three studies with 317 participants. As shown in Additional file 1: Fig. S2 and Table 2, the pooled estimates of sensitivity were 0.947 (95% CI 0.873 to 0.990) and 0.872 (95% CI 0.643 to 0.991) for QFT-Plus and T-SPOT.TB, respectively. Subgroup analysis was conducted stratified by number of participants, and when the number of participants was greater than 100, QFT-Plus had a significant advantage over T-SPOT.TB (Additional file 1: Fig. S10).

Pooled difference in sensitivity between QFT-Plus and T-SPOT.TB in 3 studies

The pooled difference in specificity between QFT-Plus and QFT-GIT was 0.00 (95% CI − 0.02 to 0.01; Fig. 4) in five studies with 482 participants. As shown in Additional file 1: Fig. S3 and Table 1, the pooled estimates of specificity were 0.987 (95% CI 0.961 to 0.999) and 0.996 (95% CI 0.984 to 1.000) for QFT-Plus and QFT-GIT, respectively. Subgroup analysis was conducted stratified by TB burden of the areas and number of participants (Additional file 1: Fig. S11).

Pooled difference in specificity between QFT-Plus and QFT-GIT in 5 studies

The pooled difference in specificity between QFT-Plus and T-SPOT.TB was 0.00 (95% CI − 0.02 to 0.01; Fig. 5) in two studies with 224 participants. As shown in Additional file 1: Fig. S4 and Table 2, the pooled estimates of specificity were 0.995 (95% CI 0.959 to 1.000) and 1.000 (95% CI 0.996 to 1.000) for QFT-Plus and T-SPOT.TB, respectively.

Pooled difference in specificity between QFT-Plus and T-SPOT.TB in 2 studies

The pooled difference in specificity between QFT-Plus and TST was 0.12 (95% CI 0.02 to 0.22; Fig. 6) in two studies with 151 participants. As shown in Additional file 1: Fig. S5 and Table 3, the pooled estimates of specificity were 0.782 (95% CI 0.712 to 0.844) and 0.662 (95% CI 0.585 to 0.735) for QFT-Plus and TST, respectively.

Pooled difference in specificity between QFT-Plus and TST in 2 studies

The pooled difference in positive rate between QFT-Plus and QFT-GIT was 0.02 (95% CI 0.01 to 0.03; Fig. 7) in 18 studies with 4617 participants. As shown in Additional file 1: Fig. S6 and Table 1, the pooled estimates of the positive rate were 0.235 (95% CI 0.154 to 0.328) and 0.228 (95% CI 0.144 to 0.323) for QFT-Plus and QFT-GIT, respectively. Subgroup analysis was conducted stratified by age of the participants, TB burden of the areas, number of participants, and population (Additional file 1: Fig. S12).

Pooled difference in positive rate between QFT-Plus and QFT-GIT in 18 studies

The pooled difference in the positive rate between QFT-Plus and T-SPOT.TB was 0.01 (95% CI − 0.01 to 0.04; Fig. 8) in six studies with 2582 participants. As shown in Additional file 1: Fig. S7 and Table 2, the pooled estimates of positive rate were 0.103 (95% CI 0.047 to 0.179) and 0.069 (95% CI 0.010 to 0.174) for QFT-Plus and T-SPOT.TB, respectively. Subgroup analysis was conducted stratified by age of the participants, TB burden of the areas, number of participants, and population (Additional file 1: Fig. S13).

Pooled difference in positive rate between QFT-Plus and T-SPOT.TB in 6 studies

The pooled difference in positive rate between QFT-Plus and TST was − 0.03 (95% CI − 0.16 to 0.11; Fig. 9) in 10 studies with 1743 participants. As shown in Additional file 1: Fig. S8 and Table 3, the pooled estimates of the positive rate were 0.298 (95% CI 0.161 to 0.456) and 0.327 (95% CI 0.198 to 0.471) for QFT-Plus and TST, respectively. Subgroup analysis was conducted stratified by age of the participants, TB burden of the areas, number of participants, and population (Additional file 1: Fig. S14).

Pooled difference in positive rate between QFT-Plus and TST in 10 studies

The results of the sensitivity analysis showed that the results were stable (Additional file 1: Fig. S15–S17). The “Peters” test was set as a parameter for publication bias detection, enabling the following comparisons: QFT-PLUS versus QFT-GIT (p = 0.91) in patients with active TB; QFT-PLUS versus QFT-GIT (p = 0.19), and QFT-PLUS versus TST (p = 0.25) in high-risk populations. As a result, no evidence of publication bias was found (for details see Additional file 1: Table S18 and Figs. S18, S19).