A Prediction Model to Help with Oncologic Mediastinal Evaluation for Radiation: HOMER.

Rationale: When stereotactic ablative radiotherapy is an option for patients with non-small cell lung cancer (NSCLC), distinguishing between N0, N1, and N2 or N3 (N2|3) disease is important.Objectives: To develop a prediction model for estimating the probability of N0, N1, and N2|3 disease.Methods: Consecutive patients with clinical-radiographic stage T1 to T3, N0 to N3, and M0 NSCLC who underwent endobronchial ultrasound-guided staging from a single center were included. Multivariate ordinal logistic regression analysis was used to predict the presence of N0, N1, or N2|3 disease. Temporal validation used consecutive patients from 3 years later at the same center. External validation used three other hospitals.Measurements and Main Results: In the model development cohort (n = 633), younger age, central location, adenocarcinoma, and higher positron emission tomography-computed tomography nodal stage were associated with a higher probability of having advanced nodal disease. Areas under the receiver operating characteristic curve (AUCs) were 0.84 and 0.86 for predicting N1 or higher (vs. N0) disease and N2|3 (vs. N0 or N1) disease, respectively. Model fit was acceptable (Hosmer-Lemeshow, P = 0.960; Brier score, 0.36). In the temporal validation cohort (n = 473), AUCs were 0.86 and 0.88. Model fit was acceptable (Hosmer-Lemeshow, P = 0.172; Brier score, 0.30). In the external validation cohort (n = 722), AUCs were 0.86 and 0.88 but required calibration (Hosmer-Lemeshow, P < 0.001; Brier score, 0.38). Calibration using the general calibration method resulted in acceptable model fit (Hosmer-Lemeshow, P = 0.094; Brier score, 0.34).Conclusions: This prediction model can estimate the probability of N0, N1, and N2|3 disease in patients with NSCLC. The model has the potential to facilitate decision-making in patients with NSCLC when stereotactic ablative radiotherapy is an option.

Complications following symptom-limited thoracentesis using suction.

BACKGROUND: Thoracentesis using suction is perceived to have increased risk of complications, including pneumothorax and re-expansion pulmonary oedema (REPO). Current guidelines recommend limiting drainage to 1.5 L to avoid REPO. Our purpose was to examine the incidence of complications with symptom-limited drainage of pleural fluid using suction and identify risk factors for REPO. METHODS: A retrospective cohort study of all adult patients who underwent symptom-limited thoracentesis using suction at our institution between January 1, 2004 and August 31, 2018 was performed, and a total of 10 344 thoracenteses were included. RESULTS: Pleural fluid ≥1.5 L was removed in 19% of the procedures. Thoracentesis was stopped due to chest discomfort (39%), complete drainage of fluid (37%) and persistent cough (13%). Pneumothorax based on chest radiography was detected in 3.98%, but only 0.28% required intervention. The incidence of REPO was 0.08%. The incidence of REPO increased with Eastern Cooperative Oncology Group performance status (ECOG PS) ≥3 compounded with ≥1.5 L (0.04-0.54%; 95% CI 0.13-2.06 L). Thoracentesis in those with ipsilateral mediastinal shift did not increase complications, but less fluid was removed (p<0.01). CONCLUSIONS: Symptom-limited thoracentesis using suction is safe even with large volumes. Pneumothorax requiring intervention and REPO are both rare. There were no increased procedural complications in those with ipsilateral mediastinal shift. REPO increased with poor ECOG PS and drainage ≥1.5 L. Symptom-limited drainage using suction without pleural manometry is safe.

Long-term quality-adjusted survival following therapeutic bronchoscopy for malignant central airway obstruction.

BACKGROUND: While therapeutic bronchoscopy has been used to treat malignant central (CAO) airway obstruction for >25 years, there are no studies quantifying the impact of therapeutic bronchoscopy on long-term quality-adjusted survival. METHODS: We conducted a prospective observational study of consecutive patients undergoing therapeutic bronchoscopy for CAO. Patients had follow-up at 1 week and monthly thereafter until death. Outcomes included technical success (ie, relief of anatomic obstruction), dyspnoea, health-related quality of life (HRQOL) and quality-adjusted survival. RESULTS: Therapeutic bronchoscopy was performed on 102 patients with malignant CAO. Partial or complete technical success was achieved in 90% of patients. At 7 days postbronchoscopy, dyspnoea improved (mean ∆Borg-day-7=-1.8, 95% CI -2.2 to -1.3, p<0.0001) and HRQOL improved (median prebronchoscopy 0.618 utiles, 25%-75% IQR 0.569 to 0.699, mean ∆utility-day-7+0.047 utiles, 95% CI +0.023 to 0.071, p=0.0002). Improvements in dyspnoea and HRQOL were maintained long-term. Compared with the prebronchoscopy baseline, HRQOL per day of life postbronchoscopy improved (mean ∆utility-long-term+0.036 utiles, 95% CI +0.014 to 0.057, p=0.002). Median quality-adjusted survival was 109 quality-adjusted life-days (QALDs) (95% CI 74 to 201 QALDs). Factors associated with longer quality-adjusted survival included better functional status, treatment-naïve tumour, endobronchial disease, less dyspnoea, shorter time from diagnosis to bronchoscopy, absence of cardiac disease, bronchoscopic dilation and receiving chemotherapy. CONCLUSIONS: Therapeutic bronchoscopy improves HRQOL as compared with baseline, resulting in approximately a 5.8% improvement in HRQOL per day of life. The risk-benefit profile in these carefully selected patients was very favourable. TRIAL REGISTRATION NUMBER: Results; NCT03326570.

Centrally located lung cancer and risk of occult nodal disease: an objective evaluation of multiple definitions of tumour centrality with dedicated imaging software.

INTRODUCTION: Current guidelines recommend invasive mediastinal staging in patients with centrally located radiographic stage T1N0M0 nonsmall cell lung cancer (NSCLC). The lack of a specific definition of a central tumour has resulted in discrepancies among guidelines and heterogeneity in practice patterns. METHODS: Our objective was to study specific definitions of tumour centrality and their association with occult nodal disease. Pre-operative chest computed tomography scans from patients with clinical (c) T1N0M0 NSCLC were processed with a dedicated software system that divides the lungs in thirds following vertical and concentric lines. This software accurately assigns tumours to a specific third based both on the location of the centre of the tumour and its most medial aspect, creating eight possible definitions of central tumours. RESULTS: 607 patients were included in our study. Surgery was performed for 596 tumours (98%). The overall pathological (p) N disease was: 504 (83%) N0, 56 (9%) N1, 47 (8%) N2 and no N3. The prevalence of N2 disease remained relatively low regardless of tumour location. Central tumours were associated with upstaging from cN0 to any N (pN1/pN2). Two definitions were associated with upstaging to any N: concentric lines, inner one-third, centre of the tumour (OR 3.91, 95% CI 1.85-8.26; p<0.001) and concentric lines, inner two-thirds, most medial aspect of the tumour (OR 1.91, 95% CI 1.23-2.97; p=0.004). CONCLUSIONS: We objectively identified two specific definitions of central tumours. While the rate of occult mediastinal disease was relatively low regardless of tumour location, central tumours were associated with upstaging from cN0 to any N.

Lung Cancer Screening, Version 3.2018, NCCN Clinical Practice Guidelines in Oncology.

Lung cancer is the leading cause of cancer-related mortality in the United States and worldwide. Early detection of lung cancer is an important opportunity for decreasing mortality. Data support using low-dose computed tomography (LDCT) of the chest to screen select patients who are at high risk for lung cancer. Lung screening is covered under the Affordable Care Act for individuals with high-risk factors. The Centers for Medicare & Medicaid Services (CMS) covers annual screening LDCT for appropriate Medicare beneficiaries at high risk for lung cancer if they also receive counseling and participate in shared decision-making before screening. The complete version of the NCCN Guidelines for Lung Cancer Screening provides recommendations for initial and subsequent LDCT screening and provides more detail about LDCT screening. This manuscript focuses on identifying patients at high risk for lung cancer who are candidates for LDCT of the chest and on evaluating initial screening findings.

Diagnostic performance of endobronchial ultrasound-guided mediastinal lymph node sampling in early stage non-small cell lung cancer: A prospective study.

BACKGROUND AND OBJECTIVE: Standard nodal staging of lung cancer consists of positron emission tomography/computed tomography (PET/CT), followed by endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) if PET/CT shows mediastinal lymphadenopathy. Sensitivity of EBUS-TBNA in patients with N0/N1 disease by PET/CT is unclear and largely based on retrospective studies. We assessed the sensitivity of EBUS-TBNA in this setting. METHODS: We enrolled patients with proven or suspected lung cancer staged as N0/N1 by PET/CT and without metastatic disease (M0), who underwent staging EBUS-TBNA. Primary outcome was sensitivity of EBUS-TBNA compared with a composite reference standard of surgical stage or EBUS-TBNA stage if EBUS demonstrated N2/N3 disease. RESULTS: Seventy-five patients were included in the analysis. Mean tumour size was 3.52 cm (±1.63). Fifteen of 75 patients (20%) had N2 disease. EBUS-TBNA identified six while nine were only identified at surgery. Sensitivity of EBUS-TBNA for N2 disease was 40% (95% CI: 16.3-67.7%). CONCLUSION: A significant proportion of patients with N0/N1 disease by PET/CT had N2 disease (20%) and EBUS-TBNA identified a substantial fraction of these patients, thus improving diagnostic accuracy compared with PET/CT alone. Sensitivity of EBUS-TBNA however appears lower compared with historical data from patients with larger volume mediastinal disease. Therefore, strategies to improve EBUS-TBNA accuracy in this population should be further explored.

A Prediction Model to Help with the Assessment of Adenopathy in Lung Cancer: HAL.

RATIONALE: Estimating the probability of finding N2 or N3 (prN2/3) malignant nodal disease on endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) in patients with non-small cell lung cancer (NSCLC) can facilitate the selection of subsequent management strategies. OBJECTIVES: To develop a clinical prediction model for estimating the prN2/3. METHODS: We used the AQuIRE (American College of Chest Physicians Quality Improvement Registry, Evaluation, and Education) registry to identify patients with NSCLC with clinical radiographic stage T1-3, N0-3, M0 disease that had EBUS-TBNA for staging. The dependent variable was the presence of N2 or N3 disease (vs. N0 or N1) as assessed by EBUS-TBNA. Univariate followed by multivariable logistic regression analysis was used to develop a parsimonious clinical prediction model to estimate prN2/3. External validation was performed using data from three other hospitals. MEASUREMENTS AND MAIN RESULTS: The model derivation cohort (n = 633) had a 25% prevalence of malignant N2 or N3 disease. Younger age, central location, adenocarcinoma histology, and higher positron emission tomography-computed tomography N stage were associated with a higher prN2/3. Area under the receiver operating characteristic curve was 0.85 (95% confidence interval, 0.82-0.89), model fit was acceptable (Hosmer-Lemeshow, P = 0.62; Brier score, 0.125). We externally validated the model in 722 patients. Area under the receiver operating characteristic curve was 0.88 (95% confidence interval, 0.85-0.90). Calibration using the general calibration model method resulted in acceptable goodness of fit (Hosmer-Lemeshow test, P = 0.54; Brier score, 0.132). CONCLUSIONS: Our prediction rule can be used to estimate prN2/3 in patients with NSCLC. The model has the potential to facilitate clinical decision making in the staging of NSCLC.

Diagnostic Yield and Complications of Bronchoscopy for Peripheral Lung Lesions. Results of the AQuIRE Registry.

RATIONALE: Advanced bronchoscopy techniques such as electromagnetic navigation (EMN) have been studied in clinical trials, but there are no randomized studies comparing EMN with standard bronchoscopy. OBJECTIVES: To measure and identify the determinants of diagnostic yield for bronchoscopy in patients with peripheral lung lesions. Secondary outcomes included diagnostic yield of different sampling techniques, complications, and practice pattern variations. METHODS: We used the AQuIRE (ACCP Quality Improvement Registry, Evaluation, and Education) registry to conduct a multicenter study of consecutive patients who underwent transbronchial biopsy (TBBx) for evaluation of peripheral lesions. MEASUREMENTS AND MAIN RESULTS: Fifteen centers with 22 physicians enrolled 581 patients. Of the 581 patients, 312 (53.7%) had a diagnostic bronchoscopy. Unadjusted for other factors, the diagnostic yield was 63.7% when no radial endobronchial ultrasound (r-EBUS) and no EMN were used, 57.0% with r-EBUS alone, 38.5% with EMN alone, and 47.1% with EMN combined with r-EBUS. In multivariate analysis, peripheral transbronchial needle aspiration (TBNA), larger lesion size, nonupper lobe location, and tobacco use were associated with increased diagnostic yield, whereas EMN was associated with lower diagnostic yield. Peripheral TBNA was used in 16.4% of cases. TBNA was diagnostic, whereas TBBx was nondiagnostic in 9.5% of cases in which both were performed. Complications occurred in 13 (2.2%) patients, and pneumothorax occurred in 10 (1.7%) patients. There were significant differences between centers and physicians in terms of case selection, sampling methods, and anesthesia. Medical center diagnostic yields ranged from 33 to 73% (P = 0.16). CONCLUSIONS: Peripheral TBNA improved diagnostic yield for peripheral lesions but was underused. The diagnostic yields of EMN and r-EBUS were lower than expected, even after adjustment.

Lung cancer screening, version 1.2015: featured updates to the NCCN guidelines.

The NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Lung Cancer Screening provide recommendations for selecting individuals for lung cancer screening, and for evaluation and follow-up of nodules found during screening, and are intended to assist with clinical and shared decision-making. These NCCN Guidelines Insights focus on the major updates to the 2015 NCCN Guidelines for Lung Cancer Screening, which include a revision to the recommendation from category 2B to 2A for one of the high-risk groups eligible for lung cancer screening. For low-dose CT of the lung, the recommended slice width was revised in the table on "Low-Dose Computed Tomography Acquisition, Storage, Interpretation, and Nodule Reporting."

Endobronchial ultrasound: A novel screening test for pulmonary hypertension prior to major pulmonary surgery.

Objectives: Pulmonary hypertension (PH) is an important physiologic variable in the assessment of patients undergoing major thoracic operations but all too often neglected because of the need for right heart catheterization (RHC) due to the inaccuracy of transthoracic echocardiography. Patients with lung cancer often require endobronchial ultrasound (EBUS) as part of the staging of the cancer. We sought to investigate whether EBUS can be used to screen these patients for PH. Methods: Patients undergoing a major thoracic operation requiring EBUS for staging were included prospectively in the study. All patients had also a RHC (gold standard). We aimed to compare the pulmonary artery pressure measurements by EBUS with the RHC values. Results: A total of 20 patients were enrolled in the study. The prevalence of abnormal pulmonary artery pressure was 65% based on RHC. All patients underwent measurement of the pulmonary vascular acceleration time (PVAT) by EBUS with no adverse events. Linear regression analysis comparing PVAT and RHC showed a correlation (r = -0.059, -0.010 to -0.018, P = .007). A receiver operator characteristic curve (area under the curve = 0.736) was used to find the optimal PVAT threshold (140 milliseconds) to predict PH; this was used to calculate a positive and negative likelihood ratio following a positive diagnosis of 2.154 and 0.538, respectively. Conclusions: EBUS interrogation of pulmonary artery hemodynamic is safe and feasible. EBUS may be used as a screening test for PH in high-risk individuals.

Gravity- vs Wall Suction-Driven Large-Volume Thoracentesis: A Randomized Controlled Study.

BACKGROUND: Prior studies have found no differences in procedural chest discomfort for patients undergoing manual syringe aspiration or drainage with gravity after thoracentesis. However, whether gravity drainage could protect against chest pain due to the larger negative-pressure gradient generated by wall suction has not been investigated. RESEARCH QUESTION: Does wall suction drainage result in more chest discomfort compared with gravity drainage in patients undergoing large-volume thoracentesis? STUDY DESIGN AND METHODS: In this multicenter, single-blinded, randomized controlled trial, patients with large free-flowing effusions of ≥ 500 mL were assigned at a 1:1 ratio to wall suction or gravity drainage. Wall suction was performed with a suction system attached to the suction tubing and with vacuum pressure adjusted to full vacuum. Gravity drainage was performed with a drainage bag placed 100 cm below the catheter insertion site and connected via straight tubing. Patients rated chest discomfort on a 100-mm visual analog scale before, during, and after drainage. The primary outcome was postprocedural chest discomfort at 5 min. Secondary outcomes included measures of postprocedure chest discomfort, breathlessness, procedure time, volume of fluid drained, and complication rates. RESULTS: Of the 228 patients initially randomized, 221 were included in the final analysis. The primary outcome of procedural chest discomfort did not differ significantly between the groups (P = .08), nor did the secondary outcomes of postprocedural discomfort and dyspnea. Similar volumes were drained in both groups, but the procedure duration was longer in the gravity arm by approximately 3 min. No differences in rate of pneumothorax or reexpansion pulmonary edema were noted between the two groups. INTERPRETATION: Thoracentesis via wall suction and gravity drainage results in similar levels of procedural discomfort and dyspnea improvement. CLINICAL TRIAL REGISTRY: ClinicalTrials.gov; No.: NCT05131945; URL: www. CLINICALTRIALS: gov.

Lung cancer screening: making the transition from research to clinical practice.

PURPOSE OF REVIEW: The purpose of this review is to examine the literature on lung cancer screening with an emphasis on the prevalence of cancer in screen-detected nodules. On the basis of the evidence, we will then develop a practical approach to screen-detected lung nodules. RECENT FINDINGS: The first large randomized controlled trial using low-dose computed tomography (LDCT) found that persons undergoing three annual screening examinations with LDCT had a 20% relative reduction in lung cancer mortality as compared with those screened with annual chest X-rays. The probability of cancer in screen-detected nodules depends on their size and whether the nodules are detected on prevalence or incidence screens. The probability of cancer in screen-detected nodules ranges from 2.4 to 5.2%. Management strategies for screen-detected nodules that have been used successfully include careful observation using serial CT imaging, CT-guided fine needle biopsy, and surgery in carefully selected cases. The most frequently used strategies involve serial CT imaging and CT-guided biopsy for larger nodules and those that demonstrate growth on follow-up. SUMMARY: There is now evidence that LDCT in carefully selected high-risk populations can lead to better outcomes but the cost effectiveness of mass screening with LDCT is still unknown. Only patients at high risk for cancer should be screened.

Ventilatory Strategy to Prevent Atelectasis During Bronchoscopy Under General Anesthesia: A Multicenter Randomized Controlled Trial (Ventilatory Strategy to Prevent Atelectasis -VESPA- Trial).

BACKGROUND: Atelectasis negatively influences peripheral bronchoscopy, increasing CT scan-body divergence, obscuring targets, and creating false-positive radial-probe endobronchial ultrasound (RP-EBUS) images. RESEARCH QUESTION: Can a ventilatory strategy reduce the incidence of atelectasis during bronchoscopy under general anesthesia? STUDY DESIGN AND METHODS: Randomized controlled study (1:1) in which patients undergoing bronchoscopy were randomized to receive standard ventilation (laryngeal mask airway, 100% Fio2, zero positive end-expiratory pressure [PEEP]) vs a ventilatory strategy to prevent atelectasis (VESPA) with endotracheal intubation followed by a recruitment maneuver, Fio2 titration (< 100%), and PEEP of 8 to 10 cm H2O. All patients underwent chest CT imaging and a survey for atelectasis with RP-EBUS bilaterally on bronchial segments 6, 9, and 10 after artificial airway insertion (time 1) and 20 to 30 min later (time 2). Chest CT scans were reviewed by a blinded chest radiologist. RP-EBUS images were assessed by three independent, blinded readers. The primary end point was the proportion of patients with any atelectasis (either unilateral or bilateral) at time 2 according to chest CT scan findings. RESULTS: Seventy-six patients were analyzed, 38 in each group. The proportion of patients with any atelectasis according to chest CT scan at time 2 was 84.2% (95% CI, 72.6%-95.8%) in the control group and 28.9% (95% CI, 15.4%-45.9%) in the VESPA group (P < .0001). The proportion of patients with bilateral atelectasis at time 2 was 71.1% (95% CI, 56.6%-85.5%) in the control group and 7.9% (95% CI, 1.7%-21.4%) in the VESPA group (P < .0001). At time 2, 3.84 ± 1.67 (mean ± SD) bronchial segments in the control group vs 1.21 ± 1.63 in the VESPA group were deemed atelectatic (P < .0001). No differences were found in the rate of complications. INTERPRETATION: VESPA significantly reduced the incidence of atelectasis, was well tolerated, and showed a sustained effect over time despite bronchoscopic nodal staging maneuvers. VESPA should be considered for bronchoscopy when atelectasis is to be avoided. TRIAL REGISTRY: ClinicalTrials.gov; No.: NCT04311723; URL: www. CLINICALTRIALS: gov.

Endobronchial ultrasound-guided transbronchial needle aspiration biopsy is useful evaluating mediastinal lymphadenopathy in a cancer center.

BACKGROUND: Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) biopsy is used to stage mediastinal lymph nodes in cancer patients to optimize treatment strategies. In this retrospective study, the authors determined the utility of EBUS-TBNA biopsy in the evaluation of mediastinal lymphadenopathy at a high-volume cancer center. MATERIALS AND METHODS: The pathology database was searched for all patients who had undergone EBUS-TBNA biopsy of mediastinal lymph nodes over a one-year period. Cytologic diagnoses were correlated with clinical histories, subsequent resection, and clinical follow-up data. RESULTS: Of 928 lymph node samples, 226 (24%) were diagnosed as malignant, 4 (0.4%) were suspicious for malignancy, 9 (1%) were atypical, 640 (69%) were benign, and 47 (5%) were insufficient for evaluation. In 89 (9.6%) cases, the patients had surgical resection. There was one false positive, in which the primary tumor contained infiltrating lymphocytes, had been sampled. There were five false-negative cases, which resulted from sampling errors, including two with micrometastases. The sensitivity, specificity, and positive and negative predictive value rates for EBUS-TBNA biopsy in the evaluation of mediastinal lymph nodes were 68.7% and 98.6% and 91.6% and 93.5%, respectively on a per lymph node basis. The overall clinical sensitivity, specificity, and positive and negative predictive value rates after one year clinical/radiological and histologic follow-up were 97%, 99.3%, 96.7% and 99.4%, respectively. CONCLUSIONS: EBUS-TBNA biopsy is a sensitive and specific method for evaluating mediastinal lymphadenopathy in patients with lung and other primary tumors.

Predicting Lymph Node Metastasis in Non-small Cell Lung Cancer: Prospective External and Temporal Validation of the HAL and HOMER Models.

BACKGROUND: Two models, the Help with the Assessment of Adenopathy in Lung cancer (HAL) and Help with Oncologic Mediastinal Evaluation for Radiation (HOMER), were recently developed to estimate the probability of nodal disease in patients with non-small cell lung cancer (NSCLC) as determined by endobronchial ultrasound-transbronchial needle aspiration (EBUS-TBNA). The objective of this study was to prospectively externally validate both models at multiple centers. RESEARCH QUESTION: Are the HAL and HOMER models valid across multiple centers? STUDY DESIGN AND METHODS: This multicenter prospective observational cohort study enrolled consecutive patients with PET-CT clinical-radiographic stages T1-3, N0-3, M0 NSCLC undergoing EBUS-TBNA staging. HOMER was used to predict the probability of N0 vs N1 vs N2 or N3 (N2|3) disease, and HAL was used to predict the probability of N2|3 (vs N0 or N1) disease. Model discrimination was assessed using the area under the receiver operating characteristics curve (ROC-AUC), and calibration was assessed using the Brier score, calibration plots, and the Hosmer-Lemeshow test. RESULTS: Thirteen centers enrolled 1,799 patients. HAL and HOMER demonstrated good discrimination: HAL ROC-AUC = 0.873 (95%CI, 0.856-0.891) and HOMER ROC-AUC = 0.837 (95%CI, 0.814-0.859) for predicting N1 disease or higher (N1|2|3) and 0.876 (95%CI, 0.855-0.897) for predicting N2|3 disease. Brier scores were 0.117 and 0.349, respectively. Calibration plots demonstrated good calibration for both models. For HAL, the difference between forecast and observed probability of N2|3 disease was +0.012; for HOMER, the difference for N1|2|3 was -0.018 and for N2|3 was +0.002. The Hosmer-Lemeshow test was significant for both models (P = .034 and .002), indicating a small but statistically significant calibration error. INTERPRETATION: HAL and HOMER demonstrated good discrimination and calibration in multiple centers. Although calibration error was present, the magnitude of the error is small, such that the models are informative.