Sample sizes of studies on diagnostic accuracy: literature survey

OBJECTIVES: To determine sample sizes in studies on diagnostic accuracy and the proportion of studies that report calculations of sample size. DESIGN: Literature survey. DATA SOURCES: All issues of eight leading journals published in 2002. METHODS: Sample sizes, number of subgroup analyses, and how often studies reported calculations of sample size were extracted. RESULTS: 43 of 8999 articles were non-screening studies on diagnostic accuracy. The median sample size was 118 (interquartile range 71-350) and the median prevalence of the target condition was 43% (27-61%). The median number of patients with the target condition--needed to calculate a test's sensitivity--was 49 (28-91). The median number of patients without the target condition--needed to determine a test's specificity--was 76 (27-209). Two of the 43 studies (5%) reported a priori calculations of sample size. Twenty articles (47%) reported results for patient subgroups. The number of subgroups ranged from two to 19 (median four). No studies reported that sample size was calculated on the basis of preplanned analyses of subgroups. CONCLUSION: Few studies on diagnostic accuracy report considerations of sample size. The number of participants in most studies on diagnostic accuracy is probably too small to analyse variability of measures of accuracy across patient subgroups.

Accuracy of magnetic resonance imaging for the diagnosis of multiple sclerosis: systematic review

OBJECTIVE: To determine the accuracy of magnetic resonance imaging criteria for the early diagnosis of multiple sclerosis in patients with suspected disease. DESIGN: Systematic review. DATA SOURCES: 12 electronic databases, citation searches, and reference lists of included studies. Review methods Studies on accuracy of diagnosis that compared magnetic resonance imaging, or diagnostic criteria incorporating such imaging, to a reference standard for the diagnosis of multiple sclerosis. RESULTS: 29 studies (18 cohort studies, 11 other designs) were included. On average, studies of other designs (mainly diagnostic case-control studies) produced higher estimated diagnostic odds ratios than did cohort studies. Among 15 studies of higher methodological quality (cohort design, clinical follow-up as reference standard), those with longer follow-up produced higher estimates of specificity and lower estimates of sensitivity. Only two such studies followed patients for more than 10 years. Even in the presence of many lesions (> 10 or > 8), magnetic resonance imaging could not accurately rule multiple sclerosis in (likelihood ratio of a positive test result 3.0 and 2.0, respectively). Similarly, the absence of lesions was of limited utility in ruling out a diagnosis of multiple sclerosis (likelihood ratio of a negative test result 0.1 and 0.5). CONCLUSIONS: Many evaluations of the accuracy of magnetic resonance imaging for the early detection of multiple sclerosis have produced inflated estimates of test performance owing to methodological weaknesses. Use of magnetic resonance imaging to confirm multiple sclerosis on the basis of a single attack of neurological dysfunction may lead to over-diagnosis and over-treatment.

Relative sensitivity factors of inorganic cations in frozen-hydrated standards in secondary ion MS analysis

We describe the measurement, at 100 K, of the SIMS relative sensitivity factors (RSFs) of the main physiological cations Na+, K+, Mg2+, and Ca2+ in frozen-hydrated (F-H) ionic solutions. Freezing was performed by either plunge freezing or high-pressure freezing. We also report the measurement of the RSFs in flax fibers, which are a model for ions in the plant cell wall, and in F-H ionic samples, which are a model for ions in the vacuole. RSFs were determined under bombardment with neutral oxygen (FAB) for both the fibers and the F-H samples. We show that referencing to ice-characteristic secondary ions is of little value in determining RSFs and that referencing to K is preferable. The RSFs of Na relative to K and of Ca relative to Mg in F-H samples are similar to their respective values in fiber samples, whereas the RSFs of both Ca and Mg relative to K are lower in fibers than in F-H samples. Our data show that the physical factors important for the determination of the RSFs are not the same in F-H samples and in homogeneous matrixes. Our data show that it is possible to perform a SIMS relative quantification of the cations in frozen-hydrated samples with an accuracy on the order of 15%. Referencing to K permits the quantification of the ionic ratios, even when the absolute concentration of the referencing ion is unknown. This is essential for physiological studies of F-H biological samples.

Electron energy-loss spectroscopy as a tool for elemental analysis in biological speciments.

A transmission electron microscope (TEM) accessory, the energy filter, enables the establishment of a method for elemental microanalysis, the electron energy-loss spectroscopy (EELS). In conventional TEM, unscattered, elastic, and inelastic scattered electrons contribute to image information. Energy-filtering TEM (EFTEM) allows elemental analysis at the ultrastructural level by using selected inelastic scattered electrons. EELS is an excellent method for elemental microanalysis and nanoanalysis with good sensitivity and accuracy. However, it is a complex method whose potential is seldom completely exploited, especially for biological specimens. In addition to spectral analysis, parallel-EELS, we present two different imaging techniques in this chapter, namely electron spectroscopic imaging (ESI) and image-EELS. We aim to introduce these techniques in this chapter with the elemental microanalysis of titanium. Ultrafine, 22-nm titanium dioxide particles are used in an inhalation study in rats to investigate the distribution of nanoparticles in lung tissue.

Accuracy of Preoperative Endoscopic Ultrasound (EUS) to Distinguish Between Early and Locally Advanced Esophageal Cancer in Daily Clinical Practice

Introduction: Preoperative chemoradiotherapy is generally recommended for locally advanced esophageal cancer (clinical stage T3 or T4 or nodal positive disease) but not for early cancer (clinical stage T0 to T2, N0). EUS has been described as the most accurate method to distinguish between early and locally advanced stage in several studies. Recently however, the high accuracy of EUS (90% or higher) was questioned by some investigators. This raises the issue whether the results of studies focused on EUS accuracy may be directly translated into daily clinical practice. Aim & Methods: The aim of this retrospective analysis was to assess the accuracy of preoperative EUS to distinguish between early and locally advanced esophageal cancer in daily clinical practice outside a study setting. EUS was performed by several investigators, including trainees in one university hospital. For this purpose, EUS reports and patient files (medical and surgical) including histological reports of 300 consecutive pts with esophageal tumors were reviewed. In pts with adenocarcinoma or squamous cell cancer and surgical resection without previous radio-/chemotherapy, EUS tumor staging was compared with histological diagnosis. Results: Out of the 300 consecutive pts with esophageal tumor and EUS 102 pts had esophageal surgery after EUS-staging without any radio-/chemotherapy. In 93 pts oesophageal cancer was confirmed, whereas 9 had other tumors. The mean age was 65 years (range 27-89), sex ratio female:male was 1:3.2. To distinguish between early and late tumor stage, the accuracy was 85%. The sensitivity and specificity for early cancer was 59%, and 93%, respectively. The diagnostic accuracy for local tumor spread was 90%, 90%, 68%, 69%, 89% for pT0, pT1, pT2, pT3 and pT4 lesions, respectively. The overall accuracy for T-stage was 74%. For pN-positive staging the accuracy of EUS was 73%. Conclusion: In daily clinical practice, the accuracy of EUS in assessing esophageal tumor staging is lower than in specific studies focusing on EUS accuracy. Mainly early esophageal cancer stages were overstaged. Thus, the implementation of recommendations for diagnostic work-up of esophageal cancer patients resulting from highly specific studies should consider the appropriate clinical setting.

Diagnostic accuracy of coronary in-stent restenosis using 64-slice computed tomography: comparison with invasive coronary angiography

OBJECTIVES: This study sought to evaluate the diagnostic accuracy of coronary binary in-stent restenosis (ISR) with angiography using 64-slice multislice computed tomography coronary angiography (CTCA) compared with invasive coronary angiography (ICA). BACKGROUND: A noninvasive detection of ISR would result in an easier and safer way to conduct patient follow-up. METHODS: We performed CTCA in 81 patients after stent implantation, and 125 stented lesions were scanned. Two sets of images were reconstructed with different types of convolution kernels. On CTCA, neointimal proliferation was visually evaluated according to luminal contrast attenuation inside the stent. Lesions were graded as follows: grade 1, none or slight neointimal proliferation; grade 2, neointimal proliferation with no significant stenosis (<50%); grade 3, neointimal proliferation with moderate stenosis (> or =50%); and grade 4, neointimal proliferation with severe stenosis (> or =75%). Grades 3 and 4 were considered binary ISR. The diagnostic accuracy of CTCA compared with ICA was evaluated. RESULTS: By ICA, 24 ISRs were diagnosed. Sensitivity, specificity, positive predictive value, and negative predictive value were 92%, 81%, 54%, and 98% for the overall population, whereas values were 91%, 93%, 77%, and 98% when excluding unassessable segments (15 segments, 12%). For assessable segments, CTCA correctly diagnosed 20 of the 22 ISRs detected by ICA. Six lesions without ISR were overestimated as ISR by CTCA. As the grade of neointimal proliferation by CTCA increases, the median value of percent diameter stenosis increased linearly. CONCLUSIONS: Binary ISR can be excluded with high probability by CTCA, with a moderate rate of false-positive results.

Accuracy and consistency of the visual diagnosis of exposed dentine on worn occlusal/incisal surfaces

Most indices for the assessment of wear of various aetiologies include the distinction between 'enamel still present' and 'dentine exposed' for grading. Since the visual diagnosis of exposed dentine has not yet been validated, the present study is a first attempt to investigate its accuracy and consistency. Sixty-one examiners (23 scientists, 18 university dentists and 20 dental students) were asked to diagnose 49 tooth areas with different grades of wear and to decide whether dentine was exposed (positive test) or not (negative test). Afterwards, the teeth were histologically evaluated. In 44 areas, dentine (also in all cases with minor wear) was exposed, and in 5 areas enamel was present. Overall sensitivity was 0.65, specificity 0.88 and the proportion of correct diagnoses was 0.67. The diagnosis 'dentine is exposed' was about 5 times as likely and the diagnosis 'dentine is not exposed' half as likely to come from an area with exposed dentine than from an enamel-covered area. The closeness of the visual diagnosis to the histological findings was only fair (kappa=0.27), no significant impact of professional experience was found. For inter- and intra-examiner agreement, kappa was 0.28 and 0.55, respectively. It was concluded that the diagnosis of exposed dentine is difficult.

Diagnostic accuracy of 64-slice computed tomography for detecting angiographically significant coronary artery stenosis in an unselected consecutive patient population: comparison with conventional invasive angiography

BACKGROUND: Multislice computed tomography (MSCT) is a promising noninvasive method of detecting coronary artery disease (CAD). However, most data have been obtained in selected series of patients. The purpose of the present study was to investigate the accuracy of 64-slice MSCT (64 MSCT) in daily practice, without any patient selection. METHODS AND RESULTS: Using 64-slice MSCT coronary angiography (CTA), 69 consecutive patients, 39 (57%) of whom had previously undergone stent implantation, were evaluated. The mean heart rate during scan was 72 beats/min, scan time 13.6 s and the amount of contrast media 72 mL. The mean time span between invasive coronary angiography (ICAG) and CTA was 6 days. Significant stenosis was defined as a diameter reduction of > 50%. Of 966 segments, 884 (92%) were assessable. Compared with ICAG, the sensitivity of CTA to diagnose significant stenosis was 90%, specificity 94%, positive predictive value (PPV) 89% and negative predictive value (NPV) 95%. With regard to 58 stented lesions, the sensitivity, specificity, PPV and NPV were 93%, 96%, 87% and 98%, respectively. On the patient-based analysis, the sensitivity, specificity, PPV and NPV of CTA to detect CAD were 98%, 86%, 98% and 86%, respectively. Eighty-two (8%) segments were not assessable because of irregular rhythm, calcification or tachycardia. CONCLUSION: Sixty-four-MSCT has a high accuracy for the detection of significant CAD in an unselected patient population and therefore can be considered as a valuable noninvasive technique.

An interfering Go/No-go task does not affect accuracy in a Concealed Information Test

Following the idea that response inhibition processes play a central role in concealing information, the present study investigated the influence of a Go/No-go task as an interfering mental activity, performed parallel to the Concealed Information Test (CIT), on the detectability of concealed information. 40 undergraduate students participated in a mock-crime experiment and simultaneously performed a CIT and a Go/No-go task. Electrodermal activity (EDA), respiration line length (RLL), heart rate (HR) and finger pulse waveform length (FPWL) were registered. Reaction times were recorded as behavioral measures in the Go/No-go task as well as in the CIT. As a within-subject control condition, the CIT was also applied without an additional task. The parallel task did not influence the mean differences of the physiological measures of the mock-crime-related probe and the irrelevant items. This finding might possibly be due to the fact that the applied parallel task induced a tonic rather than a phasic mental activity, which did not influence differential responding to CIT items. No physiological evidence for an interaction between the parallel task and sub-processes of deception (e.g. inhibition) was found. Subjects' performance in the Go/No-go parallel task did not contribute to the detection of concealed information. Generalizability needs further investigations of different variations of the parallel task.

Evaluation of magnetic resonance colonography at 3.0 Tesla regarding diagnostic accuracy and image quality

OBJECTIVES: To assess magnetic resonance (MR)-colonography (MRC) for detection of colorectal lesions using two different T1w three-dimensional (3D)-gradient-recalled echo (GRE)-sequences and integrated parallel data acquisition (iPAT) at a 3.0 Tesla MR-unit. MATERIALS AND METHODS: In this prospective study, 34 symptomatic patients underwent dark lumen MRC at a 3.0 Tesla unit before conventional colonoscopy (CC). After colon distension with tap water, 2 high-resolution T1w 3D-GRE [3-dimensional fast low angle shot (3D-FLASH), iPAT factor 2 and 3D-volumetric interpolated breathhold examination (VIBE), iPAT 3] sequences were acquired without and after bolus injection of gadolinium. Prospective evaluation of MRC was performed. Image quality of the different sequences was assessed qualitatively and quantitatively. The findings of the same day CC served as standard of reference. RESULTS: MRC identified all polyps >5 mm (16 of 16) in size and all carcinomas (4 of 4) correctly. Fifty percent of the small polyps 0.6). CONCLUSIONS: MRC using 3D-GRE-sequences and iPAT is feasible at 3.0 T-systems. The high-resolution 3D-FLASH was slightly preferred over the 3D-VIBE because of better image quality, although both used sequences showed no statistical significant difference.

Links between body mass index, total body fat, cholesterol, high-density lipoprotein, and insulin sensitivity in patients with obesity related to depression, anger, and anxiety

OBJECTIVE: Define links between psychosocial parameters and metabolic variables in obese females before and after a low-calorie diet. METHOD: Nine female obese patients (age 36.1 +/- 7.1 years, body mass index [BMI] > 30 kg/m2) were investigated before and after a 6-week low-calorie diet accompanied by behavior therapy. Blood lipids, insulin sensitivity (Bergman protocol), fat distribution (by dual-energy X-ray absorptiometry [DEXA]), as well as psychological parameters such as depression, anger, anxiety, symptom load, and well-being, were assessed before and after the dieting period. RESULTS: The females lost 9.6 +/- 2.8 kg (p < .0001) of body weight, their BMI was reduced by 3.5 +/- 0.3 kg/m2 (p < .0001), and insulin sensitivity increased from 3.0 +/- 1.8 to 4.3 +/- 1.5 mg/kg (p = .05). Their abdominal fat content decreased from 22.3 +/- 5.5 to 18.9 +/- 4.5 kg (p < .0001). In parallel, psychological parameters such as irritability (p < .05) and cognitive control (p < .0001) increased, whereas feelings of hunger (p < .05), externality (p < .05), interpersonal sensitivity (p < .01), paranoid ideation (p < .05), psychoticism (p < .01), and global severity index (p < .01) decreased. Prospectively, differences in body fat (percent) were correlated to nervousness (p < .05). Waist-to-hip ratio (WHR) differences were significantly correlated to sociability (p < .05) and inversely to emotional instability (p < .05), whereas emotional instability was inversely correlated to differences in insulin sensitivity (p < .01). DISCUSSION: Weight reduction may lead to better somatic risk factor control. Women with more nervousness and better sociability at the beginning of a diet period may lose more weight than others.

Diagnostic accuracy of exercise electrocardiogram in patients with left anterior hemi-block

BACKGROUND: Left anterior hemiblock (LAHB) is the most frequent conduction abnormality, but its impact on the diagnostic accuracy of the exercise ECG has not been studied. The aim of our study was to determine the diagnostic accuracy of ST depression for predicting ischaemia in the presence of LAHB. PATIENTS: Consecutive patients with known or suspected coronary heart disease undergoing exercise ECG and 99mTc-sestamibi single photon emission computed tomography (SPECT) were included in the analysis. Patients with left bundle branch block, with changes in QRS morphology related to myocardial infarction, and patients who had undergone pharmacological stress testing were excluded. RESULTS: Of 1532 patients assessed, 567 patients qualified for the analysis. In 69 patients with LAHB, ECG stress testing had lower sensitivity (38% vs 86%) and lower negative predictive value (82% vs 92%) than in patients with normal baseline ECG. The reduction of sensitivity appeared to be similar in patients with isolated LAHB (n=43), in patients with right bundle branch block (n=39), and with bifascicular block (n=26). In contrast, the positive predictive value of the test was excellent. CONCLUSION: The diagnostic accuracy of the exercise ECG for prediction of ischaemia is reduced in patients with LAHB.

Clinical presentation of celiac disease and the diagnostic accuracy of serologic markers in children

There has been growing recognition of a changing clinical presentation of celiac disease (CD), with the manifestation of milder symptoms. Serologic testing is widely used to screen patients with suspected CD and populations at risk. The aim of this retrospective analysis was to evaluate the clinical presentation of CD in childhood, assess the diagnostic value of serologic tests, and investigate the impact of IgA deficiency on diagnostic accuracy. We evaluated 206 consecutive children with suspected CD on the basis of clinical symptoms and positive serology results. Ninety-four (46%) had biopsy-proven CD. The median age at diagnosis of CD was 6.8 years; 15% of the children were <2 years of age. There was a higher incidence of CD in girls (p = 0.003). Iron deficiency and intestinal complaints were more frequent in children with CD than those without CD (61% vs. 33%, p = 0.0001 and 71% vs. 55%, p = 0.02, respectively), while failure to thrive was less common (35% vs. 53%, p = 0.02). The sensitivity of IgA tissue transglutaminase (IgA-tTG) was 0.98 when including all children and 1.00 after excluding children with selective IgA deficiency. The specificity of IgA-tTG was 0.73 using the recommended cut-off value of 20 IU, and this improved to 0.94 when using a higher cut-off value of 100 IU. All children with CD and relative IgA deficiency (IgA levels that are measurable but below the age reference [n = 8]) had elevated IgA-tTG. In conclusion, CD is frequently diagnosed in school-age children with relatively mild symptoms. The absence of intestinal symptoms does not preclude the diagnosis of CD; many children with CD do not report intestinal symptoms. While the sensitivity of IgA-tTG is excellent, its specificity is insufficient for the diagnostic confirmation of a disease requiring life-long dietary restrictions. Children with negative IgA-tTG and decreased but measurable IgA values are unlikely to have CD.

The accuracy of FAST in relation to grade of solid organ injuries: a retrospective analysis of 226 trauma patients with liver or splenic lesion

BACKGROUND: This study investigated the role of a negative FAST in the diagnostic and therapeutic algorithm of multiply injured patients with liver or splenic lesions. METHODS: A retrospective analysis of 226 multiply injured patients with liver or splenic lesions treated at Bern University Hospital, Switzerland. RESULTS: FAST failed to detect free fluid or organ lesions in 45 of 226 patients with spleen or liver injuries (sensitivity 80.1%). Overall specificity was 99.5%. The positive and negative predictive values were 99.4% and 83.3%. The overall likelihood ratios for a positive and negative FAST were 160.2 and 0.2. Grade III-V organ lesions were detected more frequently than grade I and II lesions. Without the additional diagnostic accuracy of a CT scan, the mean ISS of the FAST-false-negative patients would be significantly underestimated and 7 previously unsuspected intra-abdominal injuries would have been missed. CONCLUSION: FAST is an expedient tool for the primary assessment of polytraumatized patients to rule out high grade intra-abdominal injuries. However, the low overall diagnostic sensitivity of FAST may lead to underestimated injury patterns and delayed complications may occur. Hence, in hemodynamically stable patients with abdominal trauma, an early CT scan should be considered and one must be aware of the potential shortcomings of a "negative FAST".

An empirical comparison of methods for meta-analysis of diagnostic accuracy showed hierarchical models are necessary

OBJECTIVE: Meta-analysis of studies of the accuracy of diagnostic tests currently uses a variety of methods. Statistically rigorous hierarchical models require expertise and sophisticated software. We assessed whether any of the simpler methods can in practice give adequately accurate and reliable results. STUDY DESIGN AND SETTING: We reviewed six methods for meta-analysis of diagnostic accuracy: four simple commonly used methods (simple pooling, separate random-effects meta-analyses of sensitivity and specificity, separate meta-analyses of positive and negative likelihood ratios, and the Littenberg-Moses summary receiver operating characteristic [ROC] curve) and two more statistically rigorous approaches using hierarchical models (bivariate random-effects meta-analysis and hierarchical summary ROC curve analysis). We applied the methods to data from a sample of eight systematic reviews chosen to illustrate a variety of patterns of results. RESULTS: In each meta-analysis, there was substantial heterogeneity between the results of different studies. Simple pooling of results gave misleading summary estimates of sensitivity and specificity in some meta-analyses, and the Littenberg-Moses method produced summary ROC curves that diverged from those produced by more rigorous methods in some situations. CONCLUSION: The closely related hierarchical summary ROC curve or bivariate models should be used as the standard method for meta-analysis of diagnostic accuracy.