Table 1S—[Section 1.0] Question Definition and Eligibility Criteria
Section / Informal Questions / PICO Question / Study MethodsPopulations / Intervention (s) / Comparator / Outcome
Candidates for curative treatment / Nodule characterization / Solid nodule(s), ≥8 to 10 mm; solid nodules <8 to 10 mm; sub-solid nodules <8 to 10 mm / Chest CT
Functional imaging: dynamic CT, dynamic MRI, SPECT, PET / Any or none / Yield
Accuracy
Dx thinking
Treatment plan
Complications
Unnecessary or
futile surgery
Delayed surgery
Survival
Mortality
Resource use
Cost-effectiveness / RCT
Cohort
Case-control
Case series
Tissue diagnosis / TTNB
Bronchoscopy
Radial EBUS EMN
Virtual bronchoscopy / Any or none / Yield
Accuracy
Dx thinking
Treatment plan
Complications
Unnecessary or
futile surgery
Delayed surgery
Survival
Mortality
Resource use
Cost-effectiveness
Management or follow-up / VATS / Any / Complications
Unnecessary or
futile surgery
Survival
Mortality
Resource use
Cost-effectiveness
CT surveillance: frequency, duration, radiation dose, volumetric measurement / Any or none / Timeliness
Dx thinking
Treatment plan
Complications
Delayed surgery
Survival Mortality
Resource use
Cost-effectiveness
Consequences of delayed surgery
Table 2S: Studies of chest radiography for lung nodule detection and characterization
In the Excel file
Copp, 20061 / Inclusion: Solid organ transplant recipients, University of Washington, 1990-2005 / 55 / 53.5
(range 33-73) / 62 / 26
(range 6-60) / 56 (infection)
25 (PTLD)
18 (carcinoma) / Available pathologic, microbiologic and clinical data
Exclusion: Opacity >6 cm; medical records or imaging studies not available
Gould, 20072 / Inclusion: U.S. veterans with new nodule on CXR, 1998-2001 / 375 / 66±11 / 97 (benign),
98.5 (malignant) / 15 (benign), 19 (malignant) / 54 / Pathology or 2-year follow-up
Exclusion: age<21, pregnancy or lactation, weight>350 to 400 lbs, pulmonary infection, chest surgery <6 months, life expectancy <1 year, did not have qualifying CT or PET
Yonemori, 20073 / Inclusion: SPN<30 mm and underwent lung resection, National Cancer Center, Tokyo. Japan, 1998-2004 / 452 / 62 / 55 / 19 / 75 / Pathology
Exclusion: Metastasis from extrapulmonary primary
May, 20114 / Inclusion: Consecutive patients with resected lung nodule and pathology diagnostic of carcioma or granuloma / 93 patients with 96 nodules / 67±9.8 / 35 / 16 ± 6 / 68 / Pathology
Exclusion: Diffuse nodular disease
Author year / Predictors / OR / 95% CI / Goodness of Fit / AUC (Development) / Calibration / Validation / AUC (Validation)
Copp, 20061 / Infection: non-lung vs. lung transplant / 20.7 / 1.6 to 27 / NS / NS / NS / NS / NS
Infection: consolidation / 20.2 / 2.6 to 941
Aspergillus: Dx<90 days post transplant / 13.7* / 2.9 to 64
Aspergillus: heart transplant vs. other / 4.4* / 1.1 to 19
PTLD: EBV seronegativity / 11.3* / 1.2 to ∞
PTLD: lung transplant / 14.6* / 2.5 to 86
Gould, 20072 / For malignant SPN: Smoking
Age (per 10 years) Diameter (per mm) Time since quitting (per 10 years) / 7.9
2.2
1.1
0.6 / 2.6 to 23.6
1.7 to 2.8
1.1 to 1.2
0.4 to 0.7 / p=0.61 / 0.79
(0.74 to 0.84) / Good / Cross validation by resampling without replacement (jackknife) / 0.78
(0.73 to 0.83)
Yonemori, 20073 / For benign SPN:
serum CRP (square root)
serum CEA (square root)
calcification
speculation
CT bronchus sign / 0.2
0.3
21.6
0.05
0.3 / NS / p=0.17 / 0.97 / NS / Independent sample of 148 patients with nodules enrolled prospectively at same institution / 0.84
Author year / Predictors / OR / 95% CI / Goodness of Fit / AUC (Development) / Calibration / Internal Validation / AUC (Validation)
May, 20112 / Smooth or lobulated borders, irregular shape and solid attenuation / NPV for cancer was 0.86 for this combination of findings / NS / NS / NS / NS / NS / NS
Author, year / Inclusion/Exclusion criteria / Subjects, N / Age / Men, % / Nodule size, mm / Prevalence of malignancy, %
Herder, 20055 / Inclusion: Indeterminate SPN referred for PET in academic center in Netherlands, 1997-2001
Exclusion: Prior malignancy within 5 years, unknown history of malignancy or unknown final diagnosis / 106 / 64 (mean)
range 32 to 85 / 58 / ≤10 (n=33)
11-20 (n=42)
21-30 (n=31) / 58
Schultz, 20086 / Inclusion: Veterans with 7-30 mm nodule referred for PET
Exclusion: Lung mass, atelectasis, pneumonia, pleural effusion, thoracic lymphadenopathy or bony abnormality on CXR or chest CT; no chest CT; unknown final diagnosis or diagnosis made prior to PET scan / 151 / 67±10
(mean±SD) / 96 / median 15 (range 7-30) / 44
Isbell, 20117 / Inclusion: Focal pulmonary lesion treated by surgical resection
Exclusion: Known metastatic cancer, unknown nodule size (1) / 189 / 63 (median)
range 54-72 / 50 / median 24 (range 15-40) / 73
Author, year / Model validated / Subjects, N / Reference standard / Goodness of Fit / Accuracy / Calibration
Herder, 20055 / Mayo / 106 / Pathology or CXR follow-up at least 1 year (n=6 malignant and 40 benign) / p=0.46 / 0.79
(0.70 to 0.87) / Predicted probability lower than observed frequency, especially at lower predicted values
Schultz, 20086 / Mayo
VA / 118
124 / Pathology or 2-year follow-up / NS
NS / 0.80
(0.72 to 0.88)
0.73
(0.64 to 0.82) / Predicted probability lower than observed frequency
Predicted probability higher than observed frequency
Isbell, 20117 / Mayo / 189 / Surgical pathology / NS / 0.78
(0.70-0.85) / Predicted probability lower than observed frequency at low to moderate predicted values
Gurney / 189 / NS / 0.80
(0.73-0.87) / Predicted probability lower than observed frequency at low predicted values and higher than observed at high predicted values
Tables 7, 8, 9 in Excel file
Author, year / Prospective? / Consecutive enrollment? / Blinded interpretation?Chang, 20088 / yes / Yes / No
De Felippo, 20089 / no / ?yes / no
Gupta, 200510 / no / yes / no
Hiraki, 201011 / no / yes / no
Khan, 200812 / no / yes / no
Kothary, 200913 / no / yes / no
Ng, 200814 / no / yes / no
Rizzo, 201115 / no / yes - with some clearly defined exclusions / no
Satoh, 200516 / yes / yes / no
Schaefer, 200717 / yes / yes / no
Yoshimatsu, 200818 / no / yes / no
Author, year / Inclusion criteria / Subjects / Age / % men / Nodules / Mean (SD), Median (IQR) Size (mm) / Overall prevalence of malignancy (%) / Reference standard
Chang, 20088 / All biopsy of non-vascular pulmonary lesion >4 mm with 19-gauge coaxial guiding needles and 20-gauge biopsy needles under CT guidance; excluded if scanty specimen / 582 / Mean (SD) = 62.8 (13.5) / 57.60% / 65 15 mm; 243 15-30 mm / 36 (20) / confirmed histopathological diagnosis from surgical resection, microbiology, or clinical course after follow-up >3 years
De Felippo, 20089 / 84 patients who underwent transthoracic needle biopsy for suspected lung neoplasm between 9/06 - 1/08 / 84 / mean = 65 / 65.5% / 73 (87%) <20 mm; 11 20-160 cm / 83.3% / "definitive cytological diagnosis"
Gupta, 200510 / all adults who underwent CT-guided biopsy of pulmonary nodules between 11/1/00-12/31/02, with lesion within 1 cm of pleura with max diameter 2 cm / 176 / mean =62 / 44.9% / 42 with nodule 0-1 cm, 134 nodule 1-2 cm / 22/48 direct (46%); 87/128 indirect (68%) / True positive = surgical confirmation, when histology c/w patient's known primary malignancy, when biopsy of another site had same histology, when lesion grew and mets found, or clinical dx based on response to rx
Author, year / Inclusion criteria / Subjects / Age / % men / Nodules / Mean (SD), Median (IQR) Size (mm) / Overall prevalence of malignancy (%) / Reference standard
Hiraki, 201011 / all CT fluoroscopy guided lung biopsies with 20G coaxial cutting needle from 4/00-8/08; EXCLUDED patients in whom biopsy diagnosis could not be confirmed by other means / 901 / mean 66.9 / 63.0% / 795 with nodule<30 mm / mean (SD): 23 (15) mm / 87% / surgical pathology, histology c/w known malignancy, clinical course
Khan, 200812 / all percutaneous transthoracic needle biopsies performed from 11/98-4/02 / 133 / 88 with nodule <4 cm
Kothary, 200913 / all adults who underwent CT-guided percutaneous fine needle aspiration biopsy or 20-gauge core biopsy of a pulmonary nodule / 139 / mean = 62.5 / 43.20% / 37 (26.6%) with nodule<1.5 cm / 70.50% / tissue diagnosis, follow-up imaging, or clinical course
Ng, 200814 / all CT-guided FNAB of lung nodules 10 mm or less between 1/03-2/06 / 54 / mean 63.3 / 51.9% / 54 / mean 9 mm (SD 1.1 mm) / 66.0% / surgical confirmation, histology c/w patient's known primary malignancy, or clinical course c/w malignancy
Author, year / Inclusion criteria / Subjects / Age / % men / Nodules / Mean (SD), Median (IQR) Size (mm) / Overall prevalence of malignancy (%) / Reference standard
Rizzo, 201115 / all CT-guided lung FNA or core biopsy between 1/07-7/08; EXCLUDED: prior lung resection, presence of pleural effusion, pleural-based lesions / 157 / not broken down by size / overall: mean (SD) 40.1 (25.6) mm
Satoh, 200516 / all automated cutting needle biopsies of pulmonary nodules under CT guidance between 2/93-9/98 / 57 / mean 59 (range 6-84) / 57.89% / n=57: 21 <20 mm / mean 34 mm (range 7-120 mm) / 61.4% / pathologic evaluation of surgical resected specimen, TBBx specimens, autopsy, or biopsy of other sites; OR culture results or clinical f/u.
Schaefer, 200717 / all patients with suspicious lesions of lung <1.5 cm from 11/05-10/06; excluded those unable to follow instructions; nodules affixed to chest wall; INR>1.5; PTT>40; platelet<50 / 41 / mean (SD) = 64.3 (10.5) / 54.0% / 41 / mean 12.5 mm (SD 2.0 mm) / 63.4% / diagnostic biopsy result considered definitive
Yoshimatsu, 200818 / all adults who underwent TTNB of lung between 3/03-4/04 using I-I device under fluoroscopic guidance / 126 / nodule30mm: 105 / mean (SD) 23.3 (16.9) mm; range 3-110 mm / 77% / surgical pathology or clinical course
range 46% to 83.3%
median 68%
Author, year / TTNA or interpretation method / True positive malignant, N (%) / True positive benign, N (%) / Non-diagnostic, N (%)
Chang, 20088 / imprint cytology vs. histology vs. both together / 15 mm: imprint cyotology: sens 82.4%, spec 100%; PPV 100%; NPV 83.8%; accuracy 90.8%; histopathology: sens 82.4%, spec 100%, PPV 100%, 83.8%; accuracy 90.8%. 15-30 mm: imprint cytology: sens 93.8%, spec 98.5%, PPV 99.4%; NPV 85.5%; accuracy 95.1%; histopathology: sens: 91.0%, spec 100%, PPV 100%, NPV 80.5%, acc 93.4%
De Felippo, 20089 / 22 gauge Chiba point needle / 70/70 (100%) / 10/10 (100%) / 4/84 (4.8%)
Gupta, 200510 / coaxial biopsy with 18 gauge guide, 22 gauge inner needle; direct path=<10 mm through aerated lung; indirect = >10 mm transpulmonary needle path; on-site cytopathologist present / 21/22 direct (95%); 83/87 indirect (95%) / 4/4 direct (100%), 30/30 indirect (100%) / 14/48 direct (29%); 7/128 indirect (6%)
Hiraki, 201011 / CT fluoroscopy guidance, 19G introducer, 20G cutting biopsy needle coaxial system / 579/607 (95%) / 183/183 (100%) / 5/795 (0.6%)
Khan, 200812 / 18 G cutting needle
Kothary, 200913 / coaxial technique with 19-g introducer needle and 22 g aspiration needle; cytotechnologist present. All patients underwent FNA, some cutting biopsy as well with 20G needle / for nodule1.5 cm: 16/23 (69.6%) / for any size: 17/37 (45.9%) / for nodule1.5 cm: 18/37 (48.6%)
Ng, 200814 / coaxial technique with outer 19G and inner 22 G; on-site cytotechnologist / 21/31 (67.7%) / 1/16 (6.3%) / 30/55 (54.5%)
Rizzo, 201115 / FNA with 20-25 G needle; core with 14-20 G; no on-site pathologist
Satoh, 200516 / tandem system with 20 gauge and 18 gauge automated cutting needle with throw length of 23 mm / 10/11 (91%) for nodule20mm / 9/10 (90%) for nodule<20mm / 2/21 (10%) for nodule <20mm
Schaefer, 200717 / CT-fluoroscopic -guided biopsy / ? / ? / 2/43 (4.7%)
Yoshimatsu, 200818 / FNA with 21 G needle, 18-20G for core biopsy; used I-I device to position and guide needle; no on-site cytopathologist / overall: 75/80 (93.8%) / 24/24 (100%) / 5/102 (5%)
Author, year / Pneumothorax, % / Chest tube, % / Hemorrhage, %
Chang, 20088 / 38.0% overall (by imaging) / 1% / 60% for nodules15 mm
De Felippo, 20089 / 14.3% / 1.2% / 3.6%
Gupta, 200510 / 33/48 (69%) direct; 88/128 indirect (69%) / 8/48 (17%) direct; 49/128 (38%) indirect / 17/48 (35%) direct; 97/128 (76%) indirect
Hiraki, 201011 / 42%
Khan, 200812 / for nodule<4 cm: 15/88 (17%) / for nodule<4 cm: 3/88 (3.4%) / for nodule<4 cm: 32/88 (36%)
Kothary, 200913 / for nodule1.5 cm: 35% / 5% / not reported
Ng, 200814 / 52.70% / 9.10%
Rizzo, 201115 / for nodule<20 mm: 61% / overall: 8% / for nodule<20 mm: 55%
Satoh, 200516 / overall: 32% / overall: 2% / overall: 28%
Schaefer, 200717 / 7.3% / 0% / 26.8%
Author, year / Pneumothorax, % / Chest tube, % / Hemorrhage, %
Yoshimatsu, 200818 / 34% / 3% / 28.2% parenchymal hemorrhage; 9.9% hemoptysis
range 7% to 69% / range 0% to 38% / range 3.6% to 76%
median 35% / median 3.4% / median 35%
Bibliographic Citation / Sources of Funding and Competing Interest / Setting / Objective(s) of the study / Questions addressed / Study Design (cited by author or actual)
Disayabutr et al, 201019 / Not disclosed / Mahidol University, Bangkok, Thailand; August 2007-March 2009 / To evaluate the diagnostic yield of EBUS-TBB for diagnosis of pulmonary lesions / PICO 1 - and did analyze nodules < 2 cm separately / cross-sectional
Roth et al, 201120 / Authors report no conflicts / Haukeland University Hospital and Aalesund Hospital, Norway; June 2005-January 2008 / To evaluate the use of EBUS for peripheral lesions in a clinical practice where the bronchoscopies are performed by pulmonologists with various levels of experience / PICO 1 - did not analyze nodules < 2 cm separately / RCT
Eberhardt et al, 200721 / Ernst and Herth disclose financial relationships with manufacturer, other authors disclose no conflict / University of Heidelberg and Harvard; January 2003- August 2006 / To determine if combining EBUS and ENB improves the diagnostic yield of flexible bronchoscopy in peripheral lung lesions without compromising safety / PICO 1 - and did analyze nodules < 2 cm separately / RCT
Paone et al, 200522 / Not disclosed / University of Rome; January 2001-September 2003 / To compare the diagnostic sensitivity, specificity, accuracy and predictive values of EBUS TBB vs TBB in peripheral lung cancers, particularly for lesions <3cm. / PICO 1 - and did analyze nodules < 2 cm separately / RCT
Steinfort et al, 201123 / Authors report no conflicts / Royal Melbourne Hospital, Australia; February 7, 2008- January 22, 2010 / To compare the diagnostic accuracy and complications of EBUS-TBB to CT-PNB for the investigation of solitary PPL in patients who had no clinical features to give preference to either procedure / PICO 1 - did not analyze nodules < 2 cm separately / RCT
Bibliographic Citation / Reference standard test / Diagnostic test(s) evaluated / Investigator(s) and assessor(s) training / Study population expected
Disayabutr et al, 201019 / none / EBUS / not described / Patients with pulmonary lesions detected by imaging that were beyond the segmental bronchus and had no evidence of endobronchial lesion
Roth et al, 201120 / Compared EBUS to TBB without EBUS guided by fluoroscopy. All patients were followed to September 2009 unless the diagnosis was confirmed by operation or autopsy prior to that time / EBUS - TBB vs TBB without EBUS guidance / all physicians in bronchoscopy lab had a training session in EBUS prior to use and few had prior experience with EBUS. Experience with bronchoscopy ranged from 30 years to less than one year / All patients referred for any lung lesion suspicious of malignancy and not visible by bronchoscopy.
Eberhardt et al, 200721 / Compared EBUS to ENB and to ENB plus EBUS. If transbronchial lung biopsy failed to yield a definitive histological dx, than patients were referred for surgical biopsy / EBUS, ENB, and the combined procedures without fluoroscopy / not described / > 18 years, with informed consent, and candidates for elective bronchoscopy with evidence of PPL or SPN on CT
Paone et al, 200522 / Compared EBUS-TBB to TBB obtained with flexible bronchoscopy without EBUS guidance. If transbronchial lung biopsy failed to yield a definitive histological dx, than patients were referred for other procedures or clinical/radiologic follow-up for confirmation / EBUS - TBB vs. TBB without EBUS guidance / 2 experienced, independent bronchoscopist randomly performed all procedures / > 18 years, with informed consent and referred for diagnostic bronchoscopy. 506 patients were deemed ineligible because they were outpatients, and the authors expected low compliance with follow-up in this population. It is therefore likely that the population enrolled may have been a subset with greater complexity.
Steinfort et al, 201123 / Compared EBUS - TBB to CT PNB. In patients who were non-diagnostic, were determined either on basis of subsequent invasive biopsy or 12 month follow-up / EBUS - TBB vs CT PNB / one author performed all EBUS, however 12 CT-PNB were performed by consultant radiologists, and 4 were performed by fellows / All consecutive patients referred for initial evaluation of PPL suspicious for malignancy and deemed to be eligible for diagnosis by either CT-PNB or EBUS
Bibliographic Citation / Numbers / Patients and disease characteristics / Adverse effects / Authors conclusion / Assessment of Quality--overall quality from QUADAS tool
Disayabutr et al, 201019 / overall diagnostic yield = 66.4% for EBUS / 85 men, 67 women with average age of 60.68 +/-13.45 years of age and 65% of the lesions were 3 cm or larger / no pneumothorax observed / The authors note that fluoroscopic guidance was required in most cases to direct the operator to the target bronchus and that diagnostic accuracy was less when EBUS probe was located adjacent compared to when the probe was located within the lesion. / Poor
Roth et al, 201120 / detection rate for cancer for EBUS (31/86) = 36% and for non-EBUS (45/103) = 43.7% / For EBUS arm = 87 men, 37 women with 27 <59 years old, 41 between 59-67, 28 between 68-74 and 28 over 70 years of age. For non-EBUS arm = 83 men, 57 women with 25 < 59 years, 32between 59-67, 40 between 68-74 and 43 over 70 years of age / pneumothorax in 10/99 = 10.1% for EBUS, 6/111 = 5.4% for non-EBUS / The authors found that EBUS did not increase the overall detection / Fair
Eberhardt et al, 200721 / overall diagnostic yield = 69% for EBUS, 59% for ENB, and 88% for combined; yield for lesions <or= to 2 cm = 75% for ENB (3/4), 78% for EBUS (7/9) and 90% for combined (9/10) / For EBUS arm = 23 men, 16 women, ages 30-79) Nine had lesions < 2 cm, 23 had lesions 2-3 cm and 7 had lesions >3 cm. For ENB arm = 20 men, 19 women, ages 19-81. Four had lesions < 2cm, 22 had lesions 2-3 cm and 13 had lesions >3 cm / pneumothorax in 2/39 = 5% for ENB, 2/39=5% for EBUS and 3/40=8% for combined / The authors hypothesize that the lower diagnostic yield for ENB in their study than in prior studies may be related to the surgical confirmation of diagnosis / Fair
Paone et al, 200522 / overall diagnostic yield = 75.8% for EBUS (66/87) and 52.1% for TBB (62/119) Diagnostic yields were higher for malignant lesions in both groups. Diagnostic accuracy for lesions < or = 2 cm with EBUS was 84% (x/25) and for TBB was 58% (x/31) / For EBUS 62 men, 25 women with mean age 65 +/- 12 years and 25 had lesions up to 2 cm. For TBB 81 men, 38 women with mean age 68 +/- 10 years and 31 had lesions up to 2 cm. / no pneumothorax in EBUS, 2.5% (3/119) TBB had pneumothorax / The authors summarize their limitations including the higher drop out rate for EBUS, but believe it suggests that in a selected subset of patients EBUS is safe and useful. / Fair
Bibliographic Citation / Numbers / Patients and disease characteristics / Adverse effects / Authors conclusion / Assessment of Quality--overall quality from QUADAS tool
Steinfort et al, 201123 / overall diagnostic yield = 87.5% (71-96) for EBUS and 93.3% (68-99) for CT-PNB / For EBUS 16 men, 16 women with mean age 71 +/- 11 years, actual distribution of the 16 ultimately receiving CT- PNB not provided / 3% (1/32) for EBUS and 18.7% (3/16) for CT PNB / The authors acknowledge potential for bias in subject selection because they allowed for clinical acumen to select potential subjects out of the study (n=28) and had 20 additional patients with unstated reasons for failure to randomize. There was significant difference in lesion size between the 2 groups, with larger lesions in the CT PNB group. One author performed all EBUS, while fellows were allowed to perform some of the CT PNB and a high rate of complications was noted in this subset, potentially biasing the complication analysis. / Fair
Bibliographic Citation / Sources of Funding and Competing Interest / Setting / Objective(s) of the study / Questions addressed / Study Design (cited by author or actual)
Becker et al, 200524 / Author disclosure = no conflict of interest / University of Heidelberg, Germany; July to December 2003 / Determine the use and safety of ENB guidance system in obtaining biopsies from 30 consecutive adults with isolated, peripheral lung lesions / PICO 2- but did not analyze nodules < 2 cm separately / case series
Eberhardt et al, 201025 / Not stated / University of Heidelberg, Germany; September 2005-July 2006 / Assess the efficiency of ENB / PICO 2 - but did not analyze nodules < 2 cm separately / case series
Eberhardt et al, Chest 200726 / Ernst has financial conflicts of interest with manufacturer, other authors disclose no conflict / University of Heidelberg and Harvard; February 2005- August 2006 / To determine the yield of ENB without fluoroscopy in the diagnosis of peripheral lung lesions and solitary pulmonary nodules / PICO 2 - and did analyze nodules < 2 cm separately / case series
Eberhardt et al, AJRCCM 200721 / Ernst and Herth disclose financial relationships with manufacturer, other authors disclose no conflict / University of Heidelberg and Harvard; January 2003- August 2006 / To determine if combining EBUS and ENB improves the diagnostic yield of flexible bronchoscopy in peripheral lung lesions without compromising safety / PICO 2 - and did analyze nodules < 2 cm separately / RCT