Comparing disease screening tests when true disease status is ascertained only for screen positives

Diagnostic yield of contrast-enhanced abdominal staging CT in patients with initially diagnosed breast cancer

PURPOSE

To estimate the diagnostic yield (DY) of abdominal staging CT for detecting breast cancer liver metastasis (BCLM) in patients with initially diagnosed breast cancer and to determine the indications for abdominal staging CT.

METHODS

Patients with newly diagnosed breast cancer who underwent abdominal CT as an initial staging work-up between January 2019 and December 2020 were retrospectively analyzed. DY was calculated and analyzed according to patient age, type of treatments, histologic type, histologic grade, lymphovascular invasion, Ki-67 status, hormone receptor status, subtype, and the American Joint Committee on Cancer anatomical staging.

RESULTS

A total of 2056 patients (mean age, 51 ± 11 years) were included. The DY of abdominal staging CT for detecting BCLM was 1.1 % (22 of 2056). DY was significantly higher in stage III than in stage I or II cancers (3.9 % [18 of 467] vs. 0 % [0 of 412] or 0.4 % [4 of 1158], respectively, p < .001), and in human epidermal growth factor receptor-2 (HER2)-enriched cancers than in luminal or triple negative cancers (2.9 % [16 of 560] vs. 0.4 % [4 of 1090] or 0.5 % [2 of 406], respectively, p < .001).

CONCLUSIONS

The DY of abdominal staging CT for detecting BCLM was low among all patients with initially diagnosed breast cancer. However, although abdominal staging CT for detecting BCLM is probably unnecessary in all patients, it can be clinically useful in patients with stage III or human epidermal growth factor receptor-2-enriched breast cancers.

Artificial intelligence for breast cancer detection in screening mammography in Sweden: a prospective, population-based, paired-reader, non-inferiority study

BACKGROUND

Artificial intelligence (AI) as an independent reader of screening mammograms has shown promise, but there are few prospective studies. Our aim was to conduct a prospective clinical trial to examine how AI affects cancer detection and false positive findings in a real-world setting.

METHODS

ScreenTrustCAD was a prospective, population-based, paired-reader, non-inferiority study done at the Capio Sankt Göran Hospital in Stockholm, Sweden. Consecutive women without breast implants aged 40-74 years participating in population-based screening in the geographical uptake area of the study hospital were included. The primary outcome was screen-detected breast cancer within 3 months of mammography, and the primary analysis was to assess non-inferiority (non-inferiority margin of 0·15 relative reduction in breast cancer diagnoses) of double reading by one radiologist plus AI compared with standard-of-care double reading by two radiologists. We also assessed single reading by AI alone and triple reading by two radiologists plus AI compared with standard-of-care double reading by two radiologists. This study is registered with ClinicalTrials.gov, NCT04778670.

FINDINGS

From April 1, 2021, to June 9, 2022, 58 344 women aged 40-74 years underwent regular mammography screening, of whom 55 581 were included in the study. 269 (0·5%) women were diagnosed with screen-detected breast cancer based on an initial positive read: double reading by one radiologist plus AI was non-inferior for cancer detection compared with double reading by two radiologists (261 [0·5%] vs 250 [0·4%] detected cases; relative proportion 1·04 [95% CI 1·00-1·09]). Single reading by AI (246 [0·4%] vs 250 [0·4%] detected cases; relative proportion 0·98 [0·93-1·04]) and triple reading by two radiologists plus AI (269 [0·5%] vs 250 [0·4%] detected cases; relative proportion 1·08 [1·04-1·11]) were also non-inferior to double reading by two radiologists.

INTERPRETATION

Replacing one radiologist with AI for independent reading of screening mammograms resulted in a 4% higher non-inferior cancer detection rate compared with radiologist double reading. Our study suggests that AI in the study setting has potential for controlled implementation, which would include risk management and real-world follow-up of performance.

FUNDING

Swedish Research Council, Swedish Cancer Society, Region Stockholm, and Lunit.

Diagnostic yield of diffusion-weighted brain MR imaging in patients with cognitive impairment: Large cohort study with 3,298 patients

Objective

There is a paucity of large cohort-based evidence regarding the need and added value of diffusion-weighted imaging (DWI) in patients attending outpatient clinic for cognitive impairment. We aimed to evaluate the diagnostic yield of DWI in patients attending outpatient clinic for cognitive impairment.

Materials and methods

This retrospective, observational, single-institution study included 3,298 consecutive patients (mean age ± SD, 71 years ± 10; 1,976 women) attending outpatient clinic for cognitive impairment with clinical dementia rating ≥ 0.5 who underwent brain MRI with DWI from January 2010 to February 2020. Diagnostic yield was defined as the proportion of patients in whom DWI supported the diagnosis that underlies cognitive impairment among all patients. Subgroup analyses were performed by age group and sex, and the Chi-square test was performed to compare the diagnostic yields between groups.

Results

The overall diagnostic yield of DWI in patients with cognitive impairment was 3.2% (106/3,298; 95% CI, 2.6–3.9%). The diagnostic yield was 2.5% (83/3,298) for acute or subacute infarct, which included recent small subcortical infarct for which the diagnostic yield was 1.6% (54/3,298). The diagnostic yield was 0.33% (11/3,298) for Creutzfeldt-Jakob disease (CJD), 0.15% (5/3,298) for transient global amnesia (TGA), 0.12% (4/3,298) for encephalitis and 0.09% (3/3,298) for lymphoma. There was a trend towards a higher diagnostic yield in the older age group with age ≥ 70 years old (3.6% vs 2.6%, P = .12). There was an incremental increase in the diagnostic yield from the age group 60–69 years (2.6%; 20/773) to 90–99 years (8.0%; 2/25).

Conclusion

Despite its low overall diagnostic yield, DWI supported the diagnosis of acute or subacute infarct, CJD, TGA, encephalitis and lymphoma that underlie cognitive impairment, and there was a trend towards a higher diagnostic yield in the older age group.

Estimating Cancer Screening Sensitivity and Specificity Using Healthcare Utilization Data: Defining the Accuracy Assessment Interval

The effectiveness and efficiency of cancer screening in real-world settings depend on many factors, including test sensitivity and specificity. Outside of select experimental studies, not everyone receives a gold standard test that can serve as a comparator in estimating screening test accuracy. Thus, many studies of screening test accuracy use the passage of time to infer whether or not cancer was present at the time of the screening test, particularly for patients with a negative screening test. We define the accuracy assessment interval as the period of time after a screening test that is used to estimate the test's accuracy. We describe how the length of this interval may bias sensitivity and specificity estimates. We call for future research to quantify bias and uncertainty in accuracy estimates and to provide guidance on setting accuracy assessment interval lengths for different cancers and screening modalities.

Extending Hui-Walter framework to correlated outcomes with application to diagnosis tests of an eye disease among premature infants

Diagnostic accuracy, a measure of diagnostic tests for correctly identifying patients with or without a target disease, plays an important role in evidence-based medicine. Diagnostic accuracy of a new test ideally should be evaluated by comparing to a gold standard; however, in many medical applications it may be invasive, costly, or even unethical to obtain a gold standard for particular diseases. When the accuracy of a new candidate test under evaluation is assessed by comparison to an imperfect reference test, bias is expected to occur and result in either overestimates or underestimates of its true accuracy. In addition, diagnostic test studies often involve repeated measurements of the same patient, such as the paired eyes or multiple teeth, and generally lead to correlated and clustered data. Using the conventional statistical methods to estimate diagnostic accuracy can be biased by ignoring the within-cluster correlations. Despite numerous statistical approaches have been proposed to tackle this problem, the methodology to deal with correlated and clustered data in the absence of a gold standard is limited. In this article, we propose a method based on the composite likelihood function to derive simple and intuitive closed-form solutions for estimates of diagnostic accuracy, in terms of sensitivity and specificity. Through simulation studies, we illustrate the relative advantages of the proposed method over the existing methods that simply treat an imperfect reference test as a gold standard in correlated and clustered data. Compared with the existing methods, the proposed method can reduce not only substantial bias, but also the computational burden. Moreover, to demonstrate the utility of this approach, we apply the proposed method to the study of National-Eye-Institute-funded Telemedicine Approaches to Evaluating of Acute-Phase Retinopathy of Prematurity (e-ROP), for estimating accuracies of both the ophthalmologist examination and the image evaluation.

Body CT and PET/CT Detection of Extracranial Lymphoma in Patients with Newly Diagnosed Central Nervous System Lymphoma

BACKGROUND

We aimed to investigate the detection rate of body CT or PET/CT for sites of extracranial disease in patients with a new pathological diagnosis of CNS DLBCL and to identify factors associated with sites of extracranial disease.

METHODS

An international multicenter cohort study of consecutive immunocompetent patients with a new diagnosis of CNS DLBCL confirmed by brain biopsy who underwent CT and/or PET/CT to evaluate for sites of extracranial disease between 1998 and 2019. The primary outcome was the detection rate of extracranial lymphoma by CT or PET/CT. Subgroup analyses according to age and EBV status were also performed. Logistic regression analyses were performed to determine factors related to sites of extracranial disease. Detection rates of CT and PET/CT were compared.

RESULTS

1043 patients were included. The overall detection rate of CT or PET/CT was 2.6% (27/1043). The treatment approach was adjusted in 74% of these patients. Multivariable analysis demonstrated that age>61-years (OR, 3.10; P=.016) and EBV positivity (OR, 3.78; P=.045) were associated with greater odds of extracranial lymphoma. There was no statistically significant difference in detection rate between CT and PET/CT (P=.802). In patients≤61 years old, the false-referral rates were significantly higher than the detection rates (P<.001).

CONCLUSION

Our results showed increased odds of extracranial lymphoma in patients with older age or EBV-positive lymphoma. Treatment was adjusted in a majority of patients diagnosed with extracranial lymphoma, thereby supporting the current guidelines for the use contrast-enhanced body CT or PET/CT in patients with newly diagnosed CNS DLBCL.

MRI-Targeted or Standard Biopsy in Prostate Cancer Screening

BACKGROUND

High rates of overdiagnosis are a critical barrier to organized prostate cancer screening. Magnetic resonance imaging (MRI) with targeted biopsy has shown the potential to address this challenge, but the implications of its use in the context of organized prostate cancer screening are unknown.

METHODS

We conducted a population-based noninferiority trial of prostate cancer screening in which men 50 to 74 years of age from the general population were invited by mail to participate; participants with prostate-specific antigen (PSA) levels of 3 ng per milliliter or higher were randomly assigned, in a 2:3 ratio, to undergo a standard biopsy (standard biopsy group) or to undergo MRI, with targeted and standard biopsy if the MRI results suggested prostate cancer (experimental biopsy group). The primary outcome was the proportion of men in the intention-to-treat population in whom clinically significant cancer (Gleason score ≥7) was diagnosed. A key secondary outcome was the detection of clinically insignificant cancers (Gleason score 6).

RESULTS

Of 12,750 men enrolled, 1532 had PSA levels of 3 ng per milliliter or higher and were randomly assigned to undergo biopsy: 603 were assigned to the standard biopsy group and 929 to the experimental biopsy group. In the intention-to-treat analysis, clinically significant cancer was diagnosed in 192 men (21%) in the experimental biopsy group, as compared with 106 men (18%) in the standard biopsy group (difference, 3 percentage points; 95% confidence interval [CI], -1 to 7; P<0.001 for noninferiority). The percentage of clinically insignificant cancers was lower in the experimental biopsy group than in the standard biopsy group (4% [41 participants] vs. 12% [73 participants]; difference, -8 percentage points; 95% CI, -11 to -5).

CONCLUSIONS

MRI with targeted and standard biopsy in men with MRI results suggestive of prostate cancer was noninferior to standard biopsy for detecting clinically significant prostate cancer in a population-based screening-by-invitation trial and resulted in less detection of clinically insignificant cancer. (Funded by the Swedish Research Council and others; STHLM3-MRI ClinicalTrials.gov number, NCT03377881.).

Final diagnosis and patient disposition following equivocal results on 2-mSv CT vs. conventional-dose CT in adolescents and young adults with suspected appendicitis: a post hoc analysis of large pragmatic randomized trial data

OBJECTIVE

To compare 2-mSv CT and conventional-dose CT (CDCT, typically 7-8 mSv) regarding final diagnosis and patient disposition following equivocal CT results in adolescents and young adults with suspected appendicitis.

METHODS

In total, 3074 patients of 15-44 years (28 ± 9 years, 1672 women) from 20 hospitals were randomized to undergo contrast-enhanced 2-mSv CT (n = 1535) or CDCT (n = 1539) from December 2013 through August 2016. One hundred sixty-one radiologists prospectively rated the likelihood of appendicitis in a Likert scale (i.e., grades 1-5). The final diagnosis was based on CT image, surgical, pathologic, and clinical findings. Post hoc analysis was performed for final diagnosis, surgical procedure, and delay in patient management following equivocal results (i.e., grade 3).

RESULTS

The 2-mSv CT and CDCT groups were comparable for final diagnosis following equivocal results, including confirmed appendicitis (1.2% [18 patients] vs. 1.2% [19], p > 0.99), negative appendectomy (0.1% [2] vs. 0.3% [4], p = 0.53), and perforated appendicitis (0.1% [1] vs. 0.2% [3], p = 0.53). More patients were confirmed as not having appendicitis following equivocal results in the CDCT group than in the 2-mSv CT group (2.2% [34] vs. 1.0% [16], p = 0.016). The two groups were comparable for the need of appendectomy (1.4% [22] vs. 1.5% [23], p > 0.99), need of additional imaging tests (0.7% [11] vs. 1.1% [17], p = 0.35), and delay in patient management following equivocal results.

CONCLUSION

2-mSv CT is comparable to CDCT regarding final diagnosis and patient disposition following equivocal CT results.

KEY POINTS

• Our results strengthen evidence justifying the use of low-dose CT instead of conventional-dose CT (CDCT) in adolescents and young adults with suspected appendicitis. • The 2-mSv CT and CDCT groups were comparable for final diagnosis following equivocal CT results, including confirmed appendicitis (1.2% vs. 1.2%, p > 0.99), negative appendectomy (0.1% vs. 0.3%, p = 0.53), and perforated appendicitis (0.1% vs. 0.2%, p = 0.53). • The two groups were comparable for the need for appendectomy (1.4% vs. 1.5%, p > 0.99), need for additional imaging tests (0.7% vs. 1.1%, p = 0.35), and delay in patient management, following equivocal CT results.

Usefulness of Staging Chest CT in Breast Cancer: Evaluating Diagnostic Yield of Chest CT According to the Molecular Subtype and Clinical Stage

The aim of this study is to investigate the clinical utility of staging chest CT in breast cancer by evaluating diagnostic yield (DY) of chest CT in detection of metastasis, according to the molecular subtype and clinical stage. This retrospective study included 840 patients with 855 breast cancers from January 2017 to December 2018. The number of patients in clinical stage 0/I, II, III and IV were 457 (53.5%), 298 (34.9%), 92 (10.8%) and 8 (0.9%), respectively. Molecular subtype was identified in 841 cancers and there were 709 (84.3%) luminal type, 55 (6.5%) human epidermal growth factor receptor 2 (HER2)-enriched type and 77 (9.2%) triple-negative (TN) type. The DYs in clinical stage 0/I, cII, cIII and cIV were 0.2% (1/457), 1.7% (5/298), 4.3% (4/92) and 100.0% (8/8), respectively. The DYs in luminal type, HER2-enriched type and TN type were 1.7% (12/709), 3.6% (2/55) and 2.6% (2/77), respectively. Clinical stage was associated with the DY (p = 0.000). However, molecular subtype was not related to the DY (p = 0.343). Molecular subtype could not provide useful information to determine whether staging chest CT should be performed in early-stage breast cancer. However, chest CT should be considered in advanced breast cancer.

Diagnostic Yield of Staging Brain MRI in Patients with Newly Diagnosed Non-Small Cell Lung Cancer

Background Existing guidelines are inconsistent regarding the indications for staging brain MRI in patients with newly diagnosed, early-stage non-small cell lung cancer (NSCLC). Purpose To evaluate the diagnostic yield of staging brain MRI in the initial evaluation of lung cancer. Materials and Methods This retrospective, observational, single-institution study included patients with newly diagnosed NSCLC who underwent staging chest CT and staging brain MRI from November 2017 to October 2018. Diagnostic yield was defined as the proportion of patients with brain metastases among all patients. Yield was stratified into clinical stage groups per the eighth edition of the American Joint Committee on Cancer staging guidelines, based on staging chest CT and in adenocarcinoma with epidermal growth factor receptor (EGFR) gene mutation and anaplastic lymphoma kinase (ALK) gene rearrangement. Subgroup analyses were performed on the basis of cell types and molecular markers. The χ² test was performed to compare the diagnostic yields, and Bonferroni correction was used to account for multiple testing between stage groups. Results A total of 1712 patients (mean age, 64 years ± 10 [standard deviation]; 1035 men) were included. The diagnostic yield of staging brain MRI in newly diagnosed NSCLC was 11.9% (203 of 1712; 95% confidence interval [CI]: 10.4%, 13.5%). In clinical stage IA, IB, and II disease, the diagnostic yields were 0.3% (two of 615; 95% CI: 0.0%, 1.2%), 3.8% (seven of 186; 95% CI: 1.5%, 7.6%), and 4.7% (eight of 171; 95% CI: 2.0%, 9.0%), respectively. The diagnostic yield was higher in patients with adenocarcinoma (13.6%; 176 of 1297; 95% CI: 11.8%, 15.6%) than squamous cell carcinoma (5.9%; 21 of 354; 95% CI: 3.7%, 8.9%) and in patients with EGFR mutation-positive adenocarcinoma (17.5%; 85 of 487; 95% CI: 14.2%, 21.1%) than with EGFR mutation-negative adenocarcinoma (10.6%; 68 of 639; 95% CI: 8.4%, 13.3%) (P < .001 for both). Conclusion The diagnostic yield of staging brain MRI in clinical stage IA non-small cell lung cancer was low, but staging brain MRI had a higher diagnostic yield in clinical stage IB and epidermal growth factor receptor mutation-positive adenocarcinoma. © RSNA, 2020 Online supplemental material is available for this article.

Diagnostic Yield of Body CT and Whole-Body FDG PET/CT for Initial Systemic Staging in Patients With Suspected Primary CNS Lymphoma: A Systematic Review and Meta-Analysis

BACKGROUND. Several guidelines recommend body imaging for the initial work-up of patients with suspected primary CNS lymphoma, to exclude subclinical systemic involvement. However, to our knowledge, the diagnostic yield of body CT (contrast-enhanced CT of the chest, abdomen, and pelvis) and whole-body FDG PET/CT for the evaluation of subclinical systemic lymphoma has not yet been systematically evaluated. OBJECTIVE. The purpose of this study was to investigate and compare the diagnostic yield of body CT and whole-body FDG PET/CT in detecting subclinical systemic lymphoma in patients with suspected primary CNS lymphoma. EVIDENCE ACQUISITION. A systematic search of the MEDLINE and EMBASE databases through July 5, 2020, was conducted to identify studies evaluating the diagnostic yield of body CT or whole-body FDG PET/CT in detecting subclinical systemic lymphoma in patients with suspected primary CNS lymphoma. Pooled estimates of the diagnostic yield of both imaging modalities were calculated using the DerSimonian and Laird random-effects model. The false referral rate and the rate of incidental secondary malignancy were also pooled. EVIDENCE SYNTHESIS. Nine original articles on studies evaluating a total of 1040 patients were included. In detecting subclinical systemic lymphoma, the pooled diagnostic yields of body CT and whole-body FDG PET/CT were 2.5% (95% CI, 1.5-3.9%) and 4.9% (95% CI, 2.8-8.5%), respectively. In the subgroup analysis, the diagnostic yield of whole-body FDG PET/CT was significantly higher than that of body CT (p = .03). Four studies reported changes in the management plan: the R-CHOP (rituximab, cyclophosphamide, hydroxydaunorubicin, vincristine, and prednisone) regimen with or without radiation therapy was added if extracranial lymphoma involvement was detected by body CT or whole-body FDG PET/CT. The pooled false referral rate of whole-body FDG PET/CT was 5.3% (95% CI, 2.2-12.0%). The pooled rate of incidental secondary malignancy detected on whole-body FDG PET/CT was 3.1% (95% CI, 1.7-5.6%). CONCLUSION. Body imaging should be used in the initial workup of patients with suspected primary CNS lymphoma, to exclude systemic involvement. Whole-body FDG PET/CT may be a better alternative to body CT. CLINICAL IMPACT. Our results support current National Comprehensive Cancer Network guidelines for the use of body imaging to exclude subclinical systemic involvement in patients with suspected primary CNS lymphoma.

Cerebellar Hemangioblastoma: Diagnostic Yield of Contrast-Enhanced Abdominal CT and Whole-Spine MRI as Initial Screening Imaging

OBJECTIVE. The existing literature lacks research into the benefits of initial screening imaging for patients with cerebellar hemangioblastoma. We aimed to evaluate the diagnostic yield of initial screening imaging using abdominal CT and whole-spine MRI in patients with cerebellar hemangioblastoma. MATERIALS AND METHODS. This retrospective study included 117 consecutive patients with histopathologically confirmed, newly diagnosed cerebellar hemangioblastomas at a single tertiary hospital between January 2006 and October 2018. Patients underwent contrast-enhanced abdominal CT, whole-spine MRI, or both to detect abdominal and spinal lesions of von Hippel-Lindau disease. Diagnostic yields and false referral rates for initial screening imaging were determined. RESULTS. After exclusion of six patients who forewent any initial imaging, 111 patients were included (53 men [mean age ± SD, 51 ± 13 years] and 58 women [mean age, 43 ± 16 years]). The diagnostic yield of abdominal CT was 3.8% (4 of 105; 95% CI, 1.1-9.3%), whereas the false referral rate was 1.0% (1 of 105; 95% CI, 0.0-5.2%). For whole-spine MRI, the corresponding values were 5.6% (4 of 71; 95% CI, 1.6-13.8%) and 2.8% (2 of 71; 95% CI, 0.3-9.8%), respectively. The respective diagnostic yields in patients with a single cerebellar hemangioblastoma were both 0% (0 of 98 and 66, respectively). CONCLUSION. For patients with a single cerebellar hemangioblastoma, screening examinations with abdominal CT and whole-spine MRI are unnecessary before the results of genetic testing are available.

Statistical considerations for testing an AI algorithm used for prescreening lung CT images

Artificial intelligence, as applied to medical images to detect, rule out, diagnose, and stage disease, has seen enormous growth over the last few years. There are multiple use cases of AI algorithms in medical imaging: first-reader (or concurrent) mode, second-reader mode, triage mode, and more recently prescreening mode as when an AI algorithm is applied to the worklist of images to identify obvious negative cases so that human readers do not need to review them and can focus on interpreting the remaining cases. In this paper we describe the statistical considerations for designing a study to test a new AI prescreening algorithm for identifying normal lung cancer screening CTs. We contrast agreement vs. accuracy studies, and retrospective vs. prospective designs. We evaluate various test performance metrics with respect to their sensitivity to changes in the AI algorithm's performance, as well as to shifts in reader behavior to a revised worklist. We consider sample size requirements for testing the AI prescreening algorithm.

Contrast-enhanced US with Perfluorobutane for Hepatocellular Carcinoma Surveillance: A Multicenter Diagnostic Trial (SCAN)

Background US has served as a standard surveillance tool for hepatocellular carcinoma (HCC); however, the detection rate and false referral rate with this modality are suboptimal. Purpose To evaluate the added value of perfluorobutane-enhanced US when combined with conventional B-mode US as an HCC surveillance tool in participants with liver cirrhosis. Materials and Methods This prospective multi-institution diagnostic trial (https://ClinicalTrials.gov, NCT02188901) used an intraindividual comparison design in a single arm of study participants and was conducted at five referral hospitals. Eligible participants who had liver cirrhosis related to viral hepatitis and were undergoing US for HCC surveillance were enrolled from October 2014 to August 2016. Immediately after completion of B-mode US but before performance of perfluorobutane-enhanced US, operating radiologists entered the results of B-mode US. After completion of subsequent perfluorobutane-enhanced US (Kupffer phase with or without vascular-phase US), the radiologists recorded the results. The presence of HCC was confirmed either with pathologic analysis or radiologically by using dynamic contrast material-enhanced CT or gadoxetic acid-enhanced MRI. The primary end points were the detection rate of early-stage HCC (Barcelona Clinic Liver Cancer staging system stage 0 or A) and false referral rate. The primary end points were compared in a per-participant manner by using the McNemar test. Results A total of 524 participants (mean age, 54 years ± 9 [standard deviation]) were included. Of these, 493 (94.1%) had liver cirrhosis related to the hepatitis B virus. Ten HCCs were confirmed in eight participants. The detection rate of early-stage HCC was not significantly improved by adding perfluorobutane-enhanced US to conventional B-mode US (difference, 0.4% [95% confidence interval: -0.3%, 1.1%]; P = .16). The false referral rate was significantly reduced (difference, -3.2% [95% confidence interval: -5.0%, -1.4%]; P < .001). Conclusion The addition of perfluorobutane-enhanced US to conventional B-mode US reduced the false referral rate without a significant improvement in the detection rate of early-stage hepatocellular carcinoma for surveillance in a population in which the hepatitis B virus predominated. © RSNA, 2019 Online supplemental material is available for this article.

Diagnostic Yield and False-Referral Rate of Staging Chest CT in Patients with Colon Cancer

Purpose To measure the diagnostic yield and false-referral rate (FRR) of staging contrast material-enhanced chest CT based on the clinical stage from contrast-enhanced abdominal CT in patients with colon cancer. Materials and Methods This retrospective study included 1743 patients (mean age, 63.4 years; range, 18-96 years) with a diagnosis of colon cancer. The primary outcomes were diagnostic yield and FRR of contrast-enhanced chest CT in the detection of thoracic metastasis. The proportions of patients with occult thoracic metastasis and those undergoing pulmonary metastasectomy for true-positive metastases were key secondary outcomes. The outcomes were stratified according to clinical stage at contrast-enhanced abdominal CT. Results The diagnostic yields in clinical stage 0/I, cII, cIII, and cIV were 0% (95% confidence interval [CI]: 0%, 0.8%), 1.3% (95% CI: 0.4%, 3.3%), 4.4% (95% CI: 3.0%, 6.1%), and 43.3% (95% CI: 36.8%, 49.9%), respectively. The corresponding FRRs were 5.7% (95% CI: 3.8%, 8.2%), 2.9% (95% CI: 1.3%, 5.5%), 6.7% (95% CI: 5.0%, 8.8%), and 6.1% (95% CI: 3.4%, 10.0%), respectively. The proportions of patients with occult metastasis were 0% (95% CI: 0%, 0.8%), 3.3% (95% CI: 1.6%, 5.9%), 1.5% (95% CI: 0.8%, 2.7%), and 6.1% (95% CI: 3.4%, 10.0%), respectively. The proportion of patients who underwent pulmonary metastasectomy was 0% (none of 474; 95% CI: 0%, 0.8%) for clinical stage 0/I tumors. Conclusion In clinical stages 0 and I, the diagnostic yield of staging contrast-enhanced chest CT in detecting thoracic metastasis was zero. For clinical stages II, III, and IV, contrast-enhanced chest CT as a baseline examination was helpful for the detection of thoracic metastasis and allowed for the possibility of a curative metastasectomy. There was no significant association between clinical stage and false-referral rate. © RSNA, 2018 Online supplemental material is available for this article.

Using Relative Statistics and Approximate Disease Prevalence to Compare Screening Tests

Schatzkin et al. and other authors demonstrated that the ratios of some conditional statistics such as the true positive fraction are equal to the ratios of unconditional statistics, such as disease detection rates, and therefore we can calculate these ratios between two screening tests on the same population even if negative test patients are not followed with a reference procedure and the true and false negative rates are unknown. We demonstrate that this same property applies to an expected utility metric. We also demonstrate that while simple estimates of relative specificities and relative areas under ROC curves (AUC) do depend on the unknown negative rates, we can write these ratios in terms of disease prevalence, and the dependence of these ratios on a posited prevalence is often weak particularly if that prevalence is small or the performance of the two screening tests is similar. Therefore we can estimate relative specificity or AUC with little loss of accuracy, if we use an approximate value of disease prevalence.

Clinical Validation of a Predictive Model for the Presence of Cervical Lymph Node Metastasis in Papillary Thyroid Cancer

BACKGROUND AND PURPOSE

Ultrasound is a standard technique to detect lymph node metastasis in papillary thyroid cancer. Cystic changes and microcalcifications are the most specific features of metastasis, but with low sensitivity. This prospective study compared the diagnostic accuracy of a predictive model for sonographic evaluation of lymph nodes relative to the radiologist's standard assessment in detecting papillary thyroid cancer metastasis in patients after thyroidectomy.

MATERIALS AND METHODS

Cervical lymph node sonographic images were reported by a radiologist (R method) per standard practice. The same images were independently evaluated by another radiologist using a sonographic predictive model (M method). A test was considered positive for metastasis if the R or M method suggested lymph node biopsy. The result of lymph node biopsy or surgical pathology was used as the reference standard. We estimated relative true-positive fraction and relative false-positive fraction using log-linear models for correlated binary data for the M method compared with the R method.

RESULTS

A total of 237 lymph nodes in 103 patients were evaluated. Our analysis of relative true-positive fraction and relative false-positive fraction included 54 nodes with pathologic results in which at least 1 method (R or M) was positive. The M method had a higher relative true-positive fraction of 1.46 (95% CI, 1.12-1.91; P = .006) and a lower relative false-positive fraction of 0.58 (95% CI, 0.36-0.92; P = .02) compared with the R method.

CONCLUSIONS

The sonographic predictive model outperformed the standard assessment to detect lymph node metastasis in patients with papillary thyroid cancer and may reduce unnecessary biopsies.

Comparing sensitivity and specificity of medical imaging tests when verification bias is present: The concept of relative diagnostic accuracy

Medical imaging plays a key role in all stages of cancer management. In evaluating a new imaging modality, the optimal design involves a comparison with standard test results as well as a gold standard, such as a pathological evaluation to determine disease status. However, when both the standard and experimental test results are negative, a gold standard may not always be performed, especially if it involves an invasive and/or costly procedure. In this situation, true disease status cannot be verified, which creates an estimation problem for sensitivity and specificity. The aim of this article is to present the concept of relative accuracy which permits to remove the bias when only patients with at least one positive test receive the gold standard.

Use of NCCN Guidelines, Other Guidelines, and Biomarkers for Colorectal Cancer Screening

Colorectal cancer (CRC) remains a common cancer and significant public health burden. CRC-related mortality is declining, partly due to the early detection of CRC through robust screening. NCCN has established the NCCN Guidelines for CRC Screening to help healthcare providers make appropriate screening recommendations according to the patient's risk of developing CRC. This review describes the evolution of CRC screening guidelines for average-risk individuals, discusses the role of NCCN Guidelines for CRC Screening in cancer prevention, and comments on the current and emerging use of biomarkers for CRC screening.

Contrast-enhanced US with Perfluorobutane(Sonazoid) used as a surveillance test for Hepatocellular Carcinoma (HCC) in Cirrhosis (SCAN): an exploratory cross-sectional study for a diagnostic trial

Background

Ultrasonography (US) is widely used as a standard surveillance tool for patients who are at a high risk of having hepatocellular carcinoma (HCC); however, conventional B-mode US appears to be insufficient in order to ensure the early detection of HCC. Perfluorobutane allows very stable Kupffer phase imaging for at least 60 min, which is tolerable for examinations of the entire liver. The purpose of our study is to evaluate the added value of contrast-enhanced US using perfluorobutane to that of conventional B-mode US as an HCC surveillance tool for patients with liver cirrhosis.

Methods/Design

SCAN (Sonazoid-US for surveillance of hepatoCellulArcarciNoma) is a prospective, multi-institutional, diagnostic trial using an intra-individual comparison design in a single arm of patients. This study was approved by our five institutional review board and informed consent was obtained from all participating. We obtained consent for publication of these data (contrast enhanced US images, CT or MRI images, laboratory findings, age, sex) from all participating patients. All patients will undergo conventional B-mode US immediately followed by contrast-enhanced US. The standardized case report forms will be completed by operating radiologists after B-mode US and contrast-enhanced US, respectively. If any lesion(s) is detected, the likelihood of HCC will be recorded. The primary endpoints are a detection rate of early-stage HCC and a false referral rate of HCC. Intra-individual comparison using Mcnemar’s test will be performed between B-mode US and contrast-enhanced US. The study will include 523 patients under HCC surveillance in five medical institutions in Korea.

Discussion

SCAN is the first study to investigate the efficacy of contrast-enhanced US in surveillance using two reciprocal endpoints specialized for the evaluation of a surveillance test. SCAN will provide evidence regarding whether patients can truly benefit from contrast-enhanced US in terms of the detection of early stage HCC while avoiding additional unnecessary examinations. In addition to the study protocol, we elaborate on potentially debatable components of SCAN, including the design of an intra-individual comparison study, study endpoints, composite reference standards, and indefinite imaging criteria regarding the likelihood of HCC.

Trial registration

The date of trial registration (ClincalTrials.gov: NCT02188901) in this study is July 3, 2014. The last patient enrolled in August 30, 2016 and follow up to see the primary end point is still ongoing. All authors have no other relationships/conditions/circumstances that present a potential conflict of interest of relationships. Our study protocol has undergone peer-review by the funding body (GE Healthcare). No other relationships/conditions/circumstances that present a potential conflict of interest. Also, we clearly stated in the 'competing interests' section of my manuscript.

Corrected ROC analysis for misclassified binary outcomes

Creating accurate risk prediction models from Big Data resources such as Electronic Health Records (EHRs) is a critical step toward achieving precision medicine. A major challenge in developing these tools is accounting for imperfect aspects of EHR data, particularly the potential for misclassified outcomes. Misclassification, the swapping of case and control outcome labels, is well known to bias effect size estimates for regression prediction models. In this paper, we study the effect of misclassification on accuracy assessment for risk prediction models and find that it leads to bias in the area under the curve (AUC) metric from standard ROC analysis. The extent of the bias is determined by the false positive and false negative misclassification rates as well as disease prevalence. Notably, we show that simply correcting for misclassification while building the prediction model is not sufficient to remove the bias in AUC. We therefore introduce an intuitive misclassification-adjusted ROC procedure that accounts for uncertainty in observed outcomes and produces bias-corrected estimates of the true AUC. The method requires that misclassification rates are either known or can be estimated, quantities typically required for the modeling step. The computational simplicity of our method is a key advantage, making it ideal for efficiently comparing multiple prediction models on very large datasets. Finally, we apply the correction method to a hospitalization prediction model from a cohort of over 1 million patients from the Veterans Health Administrations EHR. Implementations of the ROC correction are provided for Stata and R. Published 2017. This article is a U.S. Government work and is in the public domain in the USA.

Inverse probability weighting estimation of the volume under the ROC surface in the presence of verification bias

In diagnostic medicine, the volume under the receiver operating characteristic (ROC) surface (VUS) is a commonly used index to quantify the ability of a continuous diagnostic test to discriminate between three disease states. In practice, verification of the true disease status may be performed only for a subset of subjects under study since the verification procedure is invasive, risky, or expensive. The selection for disease examination might depend on the results of the diagnostic test and other clinical characteristics of the patients, which in turn can cause bias in estimates of the VUS. This bias is referred to as verification bias. Existing verification bias correction in three-way ROC analysis focuses on ordinal tests. We propose verification bias-correction methods to construct ROC surface and estimate the VUS for a continuous diagnostic test, based on inverse probability weighting. By applying U-statistics theory, we develop asymptotic properties for the estimator. A Jackknife estimator of variance is also derived. Extensive simulation studies are performed to evaluate the performance of the new estimators in terms of bias correction and variance. The proposed methods are used to assess the ability of a biomarker to accurately identify stages of Alzheimer's disease.

Epigenetic Alterations in Colorectal Cancer: Emerging Biomarkers

Colorectal cancer (CRC) is a leading cause of cancer deaths worldwide. One of the fundamental processes driving the initiation and progression of CRC is the accumulation of a variety of genetic and epigenetic changes in colonic epithelial cells. Over the past decade, major advances have been made in our understanding of cancer epigenetics, particularly regarding aberrant DNA methylation, microRNA (miRNA) and noncoding RNA deregulation, and alterations in histone modification states. Assessment of the colon cancer "epigenome" has revealed that virtually all CRCs have aberrantly methylated genes and altered miRNA expression. The average CRC methylome has hundreds to thousands of abnormally methylated genes and dozens of altered miRNAs. As with gene mutations in the cancer genome, a subset of these epigenetic alterations, called driver events, are presumed to have a functional role in CRC. In addition, the advances in our understanding of epigenetic alterations in CRC have led to these alterations being developed as clinical biomarkers for diagnostic, prognostic, and therapeutic applications. Progress in this field suggests that these epigenetic alterations will be commonly used in the near future to direct the prevention and treatment of CRC.

The role of PET and PET-CT scanning in assessing response to neoadjuvant therapy in esophageal carcinoma

BACKGROUND

The response to neoadjuvant (radio-)chemotherapy for esophageal carcinoma is often assessed with the aid of positron-emission tomography (PET), either alone or in combination with computed tomography (PET-CT). In this review, we discuss the diagnostic validity and clinical benefit of these imaging techniques.

METHODS

We systematically searched the Medline, Embase, and Cochrane Library databases for randomized controlled trials (RCTs) and controlled clinical trials (CCTs) comparing PET-CT with conventional techniques such as endosonography and CT. We then determined the diagnostic validity of these methods on the basis of information from published systematic reviews, updated with further information from more recent primary studies.

RESULTS

We did not find any RCTs that addressed the question of the patient-relevant benefit of PET-CT. We found 20 studies of diagnostic methods, carried out on a total of 854 patients, of whom 82.2% were male. These studies had a high potential for bias. In two of them, PET-CT was directly compared with endosonography or CT. Estimates of sensitivity and specificity varied widely across studies. 54% of all patients (median value across studies) had no histopathological response to therapy at the end of treatment. Taking a reduction of the standard uptake value (SUV) by at least 35% as a threshold criterion, we found that the median negative predictive value of PET across all studies was 86.5.

CONCLUSION

There is no robust evidence for a patient-relevant benefit of PET and PET-CT in patients with esophageal carcinoma. PET could potentially be used to distinguish treatment responders from non-responders after the first cycle of treatment. RCTs with patient-relevant endpoints will be needed in order to determine whether this is useful.

Test equality between two binary screening tests with a confirmatory procedure restricted on screen positives

In studies of screening accuracy, we may commonly encounter the data in which a confirmatory procedure is administered to only those subjects with screen positives for ethical concerns. We focus our discussion on simultaneously testing equality of sensitivity and specificity between two binary screening tests when only subjects with screen positives receive the confirmatory procedure. We develop four asymptotic test procedures and one exact test procedure. We derive sample size calculation formula for a desired power of detecting a difference at a given nominal [Formula: see text]-level. We employ Monte Carlo simulation to evaluate the performance of these test procedures and the accuracy of the sample size calculation formula developed here in a variety of situations. Finally, we use the data obtained from a study of the prostate-specific-antigen test and digital rectal examination test on 949 Black men to illustrate the practical use of these test procedures and the sample size calculation formula.

A general latent class model for performance evaluation of diagnostic tests in the absence of a gold standard: an application to Chagas disease

We propose a new general Bayesian latent class model for evaluation of the performance of multiple diagnostic tests in situations in which no gold standard test exists based on a computationally intensive approach. The modeling represents an interesting and suitable alternative to models with complex structures that involve the general case of several conditionally independent diagnostic tests, covariates, and strata with different disease prevalences. The technique of stratifying the population according to different disease prevalence rates does not add further marked complexity to the modeling, but it makes the model more flexible and interpretable. To illustrate the general model proposed, we evaluate the performance of six diagnostic screening tests for Chagas disease considering some epidemiological variables. Serology at the time of donation (negative, positive, inconclusive) was considered as a factor of stratification in the model. The general model with stratification of the population performed better in comparison with its concurrents without stratification. The group formed by the testing laboratory Biomanguinhos FIOCRUZ-kit (c-ELISA and rec-ELISA) is the best option in the confirmation process by presenting false-negative rate of 0.0002% from the serial scheme. We are 100% sure that the donor is healthy when these two tests have negative results and he is chagasic when they have positive results.

Notes on testing non-inferiority under the partial verification design with a confirmatory procedure limited to screen positives

When a new test with fewer invasions or less expenses to administer than the traditional test is developed, we may be interested in testing whether the former is non-inferior to the latter with respect to test accuracy. We define non-inferiority via both the odds ratio (OR) of correctly identifying a case and the OR of correctly identifying a non-case between two tests under comparison. We focus our discussion on testing the non-inferiority of a new screening test to a traditional screening test when a confirmatory procedure is performed only on patients with screen positives. On the basis of well-established methods for paired-sample data, we derive an asymptotic test procedure and an exact test procedure with respect to the two ORs defined here. Using Monte Carlo simulation, we evaluate the performance of these test procedures in a variety of situations. We note that the test procedures proposed here can also be applicable if we are interested in testing non-inferiority with respect to the ratio of sensitivities and the ratio of specificities. We discuss interval estimation of these ORs and sample size calculation based on the asymptotic test procedure considered here. We use the data taken from a study of the prostate-specific-antigen (PSA) test and the digital rectal examination (DRE) test to illustrate the practical use of these test procedures, interval estimators and sample size calculation formula.

A two-stage estimation for screening studies using two diagnostic tests with binary disease status verified in test positives only

This article considers the statistical estimation and inference for screening studies in which two binary tests are used for screening with a binary disease status verified only for those subjects with at least one positive test result. The challenge encountered in these studies is the non-identifiability because the disease rate is not identifiable for subjects with negative results from both tests without additional assumptions. Different homogeneous association models have been proposed in the literature to circumvent the non-identifiability problem, which were solved using numerical methods. We propose to formulate the problem as a constrained maximum likelihood estimation (MLE) problem. The MLE has a closed-form in general, which can be solved using a unified two-stage estimation approach. We demonstrate the application of the proposed method on a set of homogeneous association models. The homogeneous association assumptions are generally not testable as all models are saturated. Therefore, we propose an association-ratio plot as a visualization tool for model comparisons. The methods are illustrated through three examples.

Comparison of Hybrid capture 2 testing at different thresholds with cytology as primary cervical screening test

Background:

We evaluated the performance of primary high-risk human papillomavirus (hrHPV) testing by hybrid capture 2 (HC2) with different thresholds for positivity, in comparison with conventional cytology.

Methods:

We used data of 25 871 women (aged 30–60 years) from the intervention group of the VUSA-Screen study (VU University Medical Center and Saltro laboratory population-based cervical screening study), who were screened by cytology and hrHPV. Primary outcome measure was the number of cervical intraepithelial neoplasia grade 3 or higher (CIN3+), detected within 3 years. We compared baseline cytology testing with three possible hrHPV screening strategies at different relative light unit/cutoff (RLU/CO) thresholds.

Results:

Compared with baseline cytology testing, hrHPV DNA testing as a sole primary screening instrument did not yield a superior sensitivity, as well as lower colposcopy referral rate and lower false positivity rate at any RLU/CO threshold. The hrHPV screening at 1 RLU/CO threshold with cytology triage at baseline and at 12 months revealed the highest sensitivity for CIN3+ (relative sensitivity of 1.32), although still displaying a lower colposcopy referral rate than cytology testing (relative colposcopy rate of 0.94). Higher thresholds (>1RLU/CO) yielded lower colposcopy rates, but resulted in substantial loss in sensitivity.

Conclusions:

The hrHPV testing at the commonly used threshold of 1 RLU/CO with cytology triage at baseline and at 12 months showed a much higher sensitivity with a lower colposcopy referral rate compared with cytology testing.

On the estimation of disease prevalence by latent class models for screening studies using two screening tests with categorical disease status verified in test positives only

To evaluate the probabilities of a disease state, ideally all subjects in a study should be diagnosed by a definitive diagnostic or gold standard test. However, since definitive diagnostic tests are often invasive and expensive, it is generally unethical to apply them to subjects whose screening tests are negative. In this article, we consider latent class models for screening studies with two imperfect binary diagnostic tests and a definitive categorical disease status measured only for those with at least one positive screening test. Specifically, we discuss a conditional-independent and three homogeneous conditional-dependent latent class models and assess the impact of misspecification of the dependence structure on the estimation of disease category probabilities using frequentist and Bayesian approaches. Interestingly, the three homogeneous-dependent models can provide identical goodness-of-fit but substantively different estimates for a given study. However, the parametric form of the assumed dependence structure itself is not 'testable' from the data, and thus the dependence structure modeling considered here can only be viewed as a sensitivity analysis concerning a more complicated non-identifiable model potentially involving a heterogeneous dependence structure. Furthermore, we discuss Bayesian model averaging together with its limitations as an alternative way to partially address this particularly challenging problem. The methods are applied to two cancer screening studies, and simulations are conducted to evaluate the performance of these methods. In summary, further research is needed to reduce the impact of model misspecification on the estimation of disease prevalence in such settings.

Comparing accuracy in an unpaired post-market device study with incomplete disease assessment

The sensitivity and specificity of a new medical device are often compared relative to that of an existing device by calculating ratios of sensitivities and specificities. Although it would be ideal for all study subjects to receive the gold standard so true disease status was known for all subjects, it is often not feasible or ethical to obtain disease status for everyone. This paper proposes two unpaired designs where each subject is only administered one of the devices and device results dictate which subjects are to receive disease verification. Estimators of the ratio of accuracy and corresponding confidence intervals are proposed for these designs as well as sample size formulae. Simulation studies are performed to investigate the small sample bias of the estimators and the performance of the variance estimators and sample size formulae. The sample size formulae are applied to the design of a cervical cancer study to compare the accuracy of a new device with the conventional Pap smear.

Comparison of a guaiac and an immunochemical faecal occult blood test for the detection of colonic lesions according to lesion type and location

We investigated variations in sensitivity of an immunochemical (I-FOBT) and a guaiac (G-FOBT) faecal occult blood test according to type and location of lesions in an average-risk 50- to 74-year-old population. Screening for colorectal cancer by both non-rehydrated Haemoccult II G-FOBT and Magstream I-FOBT was proposed to a sample of 20 322 subjects. Of the 1615 subjects with at least one positive test, colonoscopy results were available for 1277. A total of 43 invasive cancers and 270 high-risk adenomas were detected. The gain in sensitivity associated with the I-FOBT was calculated using the ratio of sensitivities (RSN) according to type and location of lesions, and amount of bleeding. The gain in sensitivity by using I-FOBT increased from invasive cancers (RSN=1.48 (1.16–4.59)) to high-risk adenomas (RSN=3.32 (2.70–4.07)), and was inversely related to the amount of bleeding. Among cancers, the gain in sensitivity was confined to rectal cancer (RSN=2.09 (1.36–3.20)) and concerned good prognosis cancers, because they involve less bleeding. Among high-risk adenomas, the gain in sensitivity was similar whatever the location. This study suggests that the gain in sensitivity by using an I-FOBT instead of a G-FOBT greatly depends on the location of lesions and the amount of bleeding. Concerning cancer, the gain seems to be confined to rectal cancer.

Methods for Assessing Improvement in Specificity when a Biomarker is Combined with a Standard Screening Test

Biomarkers that can be used in combination with established screening tests to reduce false positive rates are in considerable demand. In this article, we present methods for evaluating the diagnostic performance of combination tests that require positivity on a biomarker test in addition to a standard screening test. These methods rely on relative true and false positive rates to measure the loss in sensitivity and gain in specificity associated with the combination relative to the standard test. Inference about the relative rates follows from noting their interpretation as conditional probabilities. These methods are extended to evaluate combinations with continuous biomarker tests by introducing a new statistical entity, the relative receiver operating characteristic (rROC) curve. The rROC curve plots the relative true positive rate versus the relative false positive rate as the biomarker threshold for positivity varies. Inference can be made by applying existing ROC methodology. We illustrate the methods with two examples: a breast cancer biomarker study proposed by the Early Detection Research Network (EDRN) and a prostate cancer case-control study examining the ability of free prostate-specific antigen (PSA) to improve the specificity of the standard PSA test.

Triagem neonatal para infecção chagásica congênita: aplicação de análise de classe latente para avaliação dos testes diagnósticos

O presente estudo tem como objetivo avaliar os testes sorológicos convencionais usados na triagem neonatal para doença de Chagas, discutindo métodos estatísticos disponíveis. Estudou-se uma amostra aleatória dentre 23.308 recém-nascidos triados para doença de Chagas congênita por meio de três testes: imunoensaioenzimático, imunofluorescência indireta e hemoaglutinação indireta. Os dados obtidos foram analisados por diferentes metodologias estatísticas: a análise de classe latente, o teste Kappa e a análise de sensibilidade relativa. Utilizando a análise de classe latente, a maior sensibilidade foi do imunoensaioenzimático (48,6%), seguido pela imunofluorescência indireta (39,8%) e pela hemoaglutinação indireta (23,2%). O valor Kappa foi 0,496. A razão entre as sensibilidades dos testes imunoensaioenzimático e imunofluorescência indireta foi de 92% [0,74;1,13]. A análise de classe latente não se mostrou adequada para determinação de sensibilidade e especificidade, mas forneceu dados importantes sobre a equivalência dos testes, corroborados pela análise de sensibilidade relativa. Os resultados mostraram que o teste imunoensaioenzimático em sangue-seco pode ser utilizado com a mesma segurança do teste imunofluorescência indireta.

Comparison of accuracy measures of two screening tests for gestational diabetes mellitus

OBJECTIVE

To compare the accuracy measures of the random glucose test and the 50-g glucose challenge test as screening tests for gestational diabetes mellitus (GDM).

RESEARCH DESIGN AND METHODS

In this prospective cohort study, pregnant women without preexisting diabetes in two perinatal centers in the Netherlands underwent a random glucose test and a 50-g glucose challenge test between 24 and 28 weeks of gestation. If one of the screening tests exceeded predefined threshold values, the 75-g oral glucose tolerance test (OGTT) was performed within 1 week. Furthermore, the OGTT was performed in a random sample of women in whom both screening tests were normal. GDM was considered present when the OGTT (reference test) exceeded predefined threshold values. Receiver operating characteristic (ROC) analysis was used to evaluate the performance of the two screening tests. The results were corrected for verification bias.

RESULTS

We included 1,301 women. The OGTT was performed in 322 women. After correction for verification bias, the random glucose test showed an area under the ROC curve of 0.69 (95% CI 0.61-0.78), whereas the glucose challenge test had an area under the curve of 0.88 (0.83-0.93). There was a significant difference in area under the curve of the two tests of 0.19 (0.11-0.27) in favor of the 50-g glucose challenge test.

CONCLUSIONS

In screening for GDM, the 50-g glucose challenge test is more useful than the random glucose test.

A novel design for estimating relative accuracy of screening tests when complete disease verification is not feasible

The accuracy (sensitivity and specificity) of a new screening test can be compared with that of a standard test by applying both tests to a group of subjects in which disease status can be determined by a gold standard (GS) test. However, it is not always feasible to administer a GS test to all study subjects. For example, a study is planned to determine whether a new screening test for cervical cancer ("ThinPrep") is better than the standard test ("Pap"), and in this setting it is not feasible (or ethical) to determine disease status by biopsy in order to identify women with and without disease for participation in a study. When determination of disease status is not possible for all study subjects, the relative accuracy of two screening tests can still be estimated by using a paired screen-positive (PSP) design in which all subjects receive both screening tests, but only have the GS test if one of the screening tests is positive. Unfortunately in the cervical cancer example, the PSP design is also infeasible because it is not technically possible to administer both the ThinPrep and Pap at the same time. In this article, we describe a randomized paired screen-positive (RPSP) design in which subjects are randomized to receive one of the two screening tests initially, and only receive the other screening test and GS if the first screening test is positive. We derive maximum likelihood estimators and confidence intervals for the relative accuracy of the two screening tests, and assess the small sample behavior of these estimators using simulation studies. Sample size formulae are derived and applied to the cervical cancer screening trial example, and the efficiency of the RPSP design is compared with other designs.

The role of proteomics in the diagnosis and treatment of ovarian cancer

Ovarian cancer is the leading cause of gynecologic cancer death in the Western world and more than 70% of patients are diagnosed with advanced stage disease. The high mortality rate is due to the difficulty in the early detection of ovarian cancer. Current screening strategies lack the necessary sensitivity and specificity to reliably and accurately diagnose affected women, prompting investigators to seek alternative means of analysis found in protein pathways and networks. Proteomics seeks to advance the understanding of how proteins interact in cancer and may provide a mechanism for early stage diagnosis. The proteomic techniques of laser capture microdissection, mass spectrometry and tissue lysate arrays have led to the discovery of new biomarkers and the identification, development and approval of a number of targeted therapeutic agents. Following validation through clinical trials, the application of these techniques will contribute to the changing paradigm of cancer detection and treatment toward personalized medicine.

Bayesian estimation of diagnostic tests accuracy for semi-latent data with covariates

The performance of a diagnostic test is usually summarized by its sensitivity and specificity. Sensitivity is the probability of a positive result, once the individual is truly ill, and specificity is the probability of a negative result, regarding a healthy individual. These measures are obtained by comparing the test outcome and the results of a reference test generically denominated gold standard. However, in many applied problems considering two diagnostic tests, the gold standard is not available for those individuals with negative results on both tests. In addition, not all diagnostic tests have the same performance across different populations. In this context, we present a Bayesian inference approach for performance measures estimation and we propose an extension of this procedure involving the inclusion of covariates. This Bayesian approach is based on Markov Chain Monte Carlo methods. The conditional dependence between the diagnostic tests was considered. As an example, we applied the proposed methodology to a real data set obtained from the medical literature.

Verification bias-corrected estimators of the relative true and false positive rates of two binary screening tests

The relative accuracy of two binary screening tests can be quantified by estimating the relative true positive rate (rTPR) and relative false positive rate (rFPR) between the two tests. Ideally all study subjects are administered both screening tests as well as a gold standard to determine disease status. In practice, however, often the gold standard is so invasive or costly that only a percentage of study subjects receive disease verification and the percentage differs depending on the results of the two screening tests. This is known as verification-biased sampling and may be by design or due to differential patient dropout or refusal to have the gold standard test administered. In this paper, maximum likelihood estimators of rTPR and rFPR and corresponding confidence intervals are developed for studies with verification-biased sampling assuming that disease status is missing at random (MAR). Simulation studies are used to show that if the MAR assumption holds, then the verification bias-corrected point estimators have little small sample bias and the confidence intervals have good coverage probabilities. Simulation studies also demonstrate that the verification bias-corrected point estimators may not be robust to violation of the MAR assumption. The proposed methods are illustrated using data from a study comparing the accuracy of Papanicolaou and human papillomavirus tests for detecting cervical cancer.

Small sample estimation of relative accuracy for binary screening tests

Comparing the accuracy of two screening tests is ideally achieved by administering both tests as well as a gold standard test to all study subjects. In practice, a more ethical screen positive study design is often used, one that requires gold standard determination only for subjects that screen positive on either test under investigation. Although it is not possible to quantify the absolute accuracy of each test with such a design,the relative accuracy of the tests can be estimated. Since relative accuracy estimation has poor small sample properties, adjusted estimators based on adding constants to the observed data have been proposed. The adjusted estimators have the advantage that they yield point and variance estimates of relative accuracy in all settings. However, we show through both theory and numerical examples that adding constants to the data can be beneficial or detrimental to small sample performance. Furthermore, the performance of the adjusted estimator depends not only on the magnitude of the constant but also on parameters that cannot be estimated with data from a screen positive study, making selection of an optimal constant difficult in practice. We also examine the performance of the adjusted estimator using data from a study comparing the accuracy of two screening tests for cervical cancer.

Sample size calculations for comparative studies of medical tests for detecting presence of disease

Technologic advances give rise to new tests for detecting disease in many fields, including cancer and sexually transmitted disease. Before a new disease screening test is approved for public use, its accuracy should be shown to be better than or at least not inferior to an existing test. Standards do not yet exist for designing and analysing studies to address this issue. Established principles for the design of therapeutic studies can be adapted for studies of screening tests. In particular, drawing upon methods for superiority and non-inferiority studies of therapeutic agents, we propose that confidence intervals for the relative accuracy of dichotomous tests drive the design of comparative studies of disease screening tests. We derive sample size formulae for a variety of designs, including studies where patients undergo several tests and studies where patients receive only one of the tests under evaluation. Both cohort and case-control study designs are considered. Modifications to the confidence intervals and sample size formulae are discussed to accommodate studies where, because of the invasive nature of definitive testing, true disease status can only be obtained for subjects who are positive on one or more of the screening tests. The methods proposed are applied to a study comparing a modified pap test to the conventional pap for cervical cancer screening. The impact of error in the gold standard reference test on the design and evaluation of comparative screening test studies is also discussed.

Testing the independence of two diagnostic tests

Consider two diagnostic procedures having binary outcomes. If one of the tests results in a positive finding, a more definitive diagnostic procedure will be administered to establish the presence or absence of a disease. The use of both tests will improve the overall screening sensitivity when the two tests are independent, compared with employing two tests that are positively correlated. We estimate the correlation coefficient of the two tests and derive statistical methods for testing the independence of the two diagnostic procedures conditional on disease status. The statistical tests are used to investigate the independence of mammography and clinical breast exams aimed at establishing the benefit of early detection of breast cancer. The data used in the analysis are obtained from periodic screening examinations of three randomized clinical trials of breast cancer screening. Analysis of each of these trials confirms the independence of the clinical breast and mammography examinations. Based on these three large clinical trials, we conclude that a clinical breast exam considerably increases the overall sensitivity relative to screening with mammography alone and should be routinely included in early breast cancer detection programs.