Summary of findings'. ''Summary of findings' table.

PET‐CT for assessing mediastinal lymph node involvement in participantswith suspected resectable non‐small cell lung cancer
Population	Participants with suspected/confirmed NSCLC who are considered potentially suitable for primary resection
Index test	PET‐CT carried out on the various available integrated PET‐CT scanners with cut‐off values for test positivity as reported in the included studies
Target condition	Resectable NSCLC defined as NSCLC that has not spread to either the ipsilateral mediastinal lymph nodes, subcarinal (N2) lymph nodes, or both
Reference standard	Pathological confirmation of PET‐CT results
Included studies	45 studies with 6095 participants available for analysis (median = 112, interquartile range (IQR) = 54 to 169), 4551 of whom were N0 or N1 and 1544 participants were N2 or N3 Different criteria for test positivity were used in the included studies: Activity > background (18 studies; N = 2823; prevalence of N2 and N3 nodes = 679/2328) SUVmax ≥ 2.5 (12 studies; N = 1656; prevalence of N2 and N3 nodes = 465/1656) Other/mixed criteria for test positivity (15 studies; N = 1616; prevalence of N2 and N3 nodes = 400/1616) None of the studies reported (any) adverse events
Test subgroup	Number of participants (studies)	Prevalence %	Summary accuracy % (95% CI)	Implications	Quality and comments
Activity > background	2823 (18)	29.2	Sensitivity: 77.4 (65.3 to 86.1) Specificity: 90.1 (85.3 to 93.5)	With the observed prevalence, there will be 66 missed cases and 70 cases who will receive futile surgery	Participant selection, index test, and flow and timing poorly reported Population spectrum narrower than in standard clinical practice in a substantial number of studies Results sensitive to selection bias, reference standard bias, and clear definition of test positivity Substantial heterogeneity was observed
SUVmax ≥ 2.5	1656 (12)	28.1	Sensitivity: 81.3 (70.2 to 88.9) Specificity: 79.4 (70.0 to 86.5)	With the observed prevalence, there will be 53 missed cases and 148 cases who will receive futile surgery	Participant selection, index test and flow, and timing poorly reported Population spectrum narrower than in standard clinical practice in a substantial number of studies Results sensitive to flow and timing bias and commercial funding bias Substantial heterogeneity was observed
All included studies	6095	25.3	Heterogeneity analyses showed significant contributions to between‐study heterogeneity from the following covariates: country of study origin, percentage of participants with adenocarcinoma, FDG dose, type of PET‐CT scanner, and study size. Study design, consecutive recruitment, attenuation correction, year of publication, and tuberculosis incidence rate per 100,000 population did not contribute significantly to the observed heterogeneity
CAUTION: The results in this table should not be interpreted in isolation from the results of the individual included studies contributing to each summary test accuracy measure. These are reported in the main body of the review

CI = confidence interval.
 FDG = (¹⁸F)‐2‐fluoro‐deoxy‐D‐glucose.
 IQR = interquartile range.
 NSCLC = non‐small cell lung cancer.
 PET‐CT = positron emission tomography–computed tomography.
 SUVmax = maximum standardised uptake value.