SEARCH · Search PubMed

Results for “diagnostic performance”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 865 records · Page 48Linked to original sources

PubMed2005

Diagnostic work-up of rectal bleeding in general practice.

BACKGROUND: GPs have many patients with gastrointestinal discomfort. Among bowel-related complaints, the sign of rectal bleeding is of particular importance in patients aged 50 years and above, as it can be an early sign for serious bowel diseases such as colon carcinoma. Despite many guidelines offered to GPs for screening and early detection of colorectal carcinomas, there is very little information about the actual diagnostic approach to the sign of rectal bleeding. AIM: The aim of the study was to collect data concerning treatment strategies used by GPs who treat patients presenting with rectal bleeding. DESIGN OF STUDY: Prospective data collection. SETTING: General practices in Germany. METHOD: Over the course of a year, GPs recorded their treatment strategies in patients presenting with rectal bleeding and associated symptoms. Using a digital practice patient file, physicians participating in the study were able to continuously transmit data electronically to the researchers of the study about diagnostics, referrals, hospital admissions, and final diagnoses. RESULTS: During the course of 1 year, 94 participating physicians collected data on 1584 patients. Information about treating rectal bleeding was recorded for 422 patients; 60% of the patients were referred to specialists in internal medicine or gastroenterologists for further diagnostics. A colonoscopy was the most frequently performed diagnostic procedure (46.2%). Twenty-two per cent (n = 93) of the patients--54 of them aged 50 years and above--were exclusively treated by their GP without conducting a colonoscopy or cooperating with specialists. For these patients, GPs diagnosed less severe diseases like haemorrhoids or other proctologic diseases. CONCLUSION: By using a study that allows GPs to transmit electronically their findings and data, it is possible to draw a picture of treatment strategies of GPs in patients presenting with rectal bleeding. The high percentage of patients who received medical treatment in consultation with specialists underscores the significance of the sign of rectal bleeding in general practice. The need for further diagnostic measures in patients who have been treated exclusively by GPs has to be discussed.

Abdominal Pain↗

PubMed1987

Integrating management and expertise in a computerised system for hypertensive patients.

An expert system has been integrated to the data management system of the ARTEMIS programme for hypertensive patients. The patient database, which has been used since 1975, contains the medical records of about 20,000 patients. Information is interactively entered by physicians, nurses and secretaries on video display units. The computerised medical record has replaced the traditional handwritten medical record. The database management system is used to produce different summary reports (inpatient and outpatient care) and personalized recall letters which are mailed to the patients before their appointments. Suggestions provided by the expert system include additional information to be obtained (complementary patient interrogation, biological or radiological investigations, etc.), possible causes of hypertension, and medical prescriptions. The information base allows the description of both static knowledge (in the form of a semantic network) and dynamic knowledge (in the form of production rules). The inference system sequentially uses a combination of forward and backward chaining and performs both exact and approximate reasoning. The diagnostic performance of the expert system was evaluated in 100 cases of hypertension (50 of essential hypertension and 50 of secondary hypertension. Concordance between the diagnosis proposed by the expert system and the one proposed by the specialist was achieved in 92% of secondary hypertension cases and 88% of essential hypertension cases. It is suggested that the integration of data and knowledge management might enhance the overall acceptance by medical staff of a computerised system, and facilitate the validation of a knowledge base.

Database Management Systems↗

PubMed2005

Performance benchmarks for diagnostic mammography.

PURPOSE: To evaluate a range of performance parameters pertinent to the comprehensive auditing of diagnostic mammography examinations, and to derive performance benchmarks therefrom, by pooling data collected from large numbers of patients and radiologists that are likely to be representative of mammography practice in the United States. MATERIALS AND METHODS: Institutional review board approval was met, informed consent was not required, and this study was Health Insurance Portability and Accountability Act compliant. Six mammography registries contributed data to the Breast Cancer Surveillance Consortium (BCSC), providing patient demographic and clinical information, mammogram interpretation data, and biopsy results from defined population-based catchment areas. The study involved 151 mammography facilities and 646 interpreting radiologists. The study population included women 18 years of age or older who underwent at least one diagnostic mammography examination between 1996 and 2001. Collected data were used to derive mean performance parameter values, including abnormal interpretation rate, positive predictive value (for abnormal interpretation, biopsy recommended, and biopsy performed), cancer diagnosis rate, invasive cancer size, and the percentages of minimal cancers, axillary node-negative invasive cancers, and stage 0 and I cancers. Additional benchmarks were derived for these performance parameters, including 10th, 25th, 50th (median), 75th, and 90th percentile values. RESULTS: The study involved 332,926 diagnostic mammography examinations. Mean performance parameter values were abnormal interpretation rate, 8.0%; positive predictive value for abnormal interpretation, 31.4%; positive predictive value for biopsy recommended, 31.5%; positive predictive value for biopsy performed, 39.5%; cancer diagnosis rate, 25.3 per 1000 examinations; invasive cancer size, 20.2 mm; percentage of minimal cancers, 42.0%; percentage of axillary node-negative invasive cancers, 73.6%; and percentage of stage 0 and I cancers, 62.4%. CONCLUSION: The presented BCSC outcomes data and performance benchmarks may be used by mammography facilities and individual radiologists to evaluate their own performance for diagnostic mammography as determined by means of periodic comprehensive audits.

Adult↗

PubMed2005

Automatic generation and evaluation of sparse protein signatures for families of protein structural domains.

We identified key residues from the structural alignment of families of protein domains from SCOP which we represented in the form of sparse protein signatures. A signature-generating algorithm (SigGen) was developed and used to automatically identify key residues based on several structural and sequence-based criteria. The capacity of the signatures to detect related sequences from the SWISSPROT database was assessed by receiver operator characteristic (ROC) analysis and jack-knife testing. Test signatures for families from each of the main SCOP classes are described in relation to the quality of the structural alignments, the SigGen parameters used, and their diagnostic performance. We show that automatically generated signatures are potently diagnostic for their family (ROC50 scores typically >0.8), consistently outperform random signatures, and can identify sequence relationships in the "twilight zone" of protein sequence similarity (<40%). Signatures based on 15%-30% of alignment positions occurred most frequently among the best-performing signatures. When alignment quality is poor, sparser signatures perform better, whereas signatures generated from higher-quality alignments of fewer structures require more positions to be diagnostic. Our validation of signatures from the Globin family shows that when sequences from the structural alignment are removed and new signatures generated, the omitted sequences are still detected. The positions highlighted by the signature often correspond (alignment specificity >0.7) to the key positions in the original (non-jack-knifed) alignment. We discuss potential applications of sparse signatures in sequence annotation and homology modeling.

Algorithms↗

PubMed2002

Optimizing imaging time for improved performance in oncology PET studies.

PURPOSE: The potential for improving the diagnostic performance of static positron imaging tomography (PET) by judiciously choosing optimum post-injection imaging times is investigated. PROCEDURES: Dynamic and whole-body scan data, from 2-deoxy-2-[18F]fluoro-D-glucose (FDG) oncological studies, are analyzed for changing standardized uptake value (SUV) behavior with increasing post-injection times at either single- or multiple-bed positions. Model-based interpretations address d(SUV)/dt, shown to correlate with SUV, and the contrast ratio for a tumor and its surroundings. A method for correcting measurements to a standardized time is given. RESULTS: Both data and model-based equations suggest that starting data acquisition later than the average 55 +/- 15 (SD) minutes post-injection reported in the FDG literature can improve contrast ratios. Considerations for choosing an optimum time from a clinical standpoint are listed. CONCLUSIONS: It is concluded that the appropriate time for each particular protocol can be found with the aid of the information presented here. True optimization, however, remains a complex issue.

Journal Article↗

PubMed2026

The role of artificial intelligence in the diagnosis and prognosis of traumatic brain injury based on brain CT scans: a systematic review.

Traumatic brain injury (TBI) is a leading cause of emergency department visits and a major contributor to injury-related mortality and long-term neurological disability. Non-contrast computed tomography (CT) is the gold-standard imaging modality for the rapid diagnosis of TBI. Clinical outcomes depend strongly on early detection and prompt acute management. Artificial intelligence (AI)-based models may support faster automated identification of traumatic findings and early prediction of patient prognosis. A systematic literature search was conducted in PubMed/MEDLINE, Scopus, IEEE Xplore, ACM Digital Library, and the Cochrane Library in accordance with PRISMA 2020 guidelines to evaluate AI-based models for automated detection of TBI-related findings on CT and for prediction of clinical outcomes. Risk of bias and applicability were assessed using QUADAS-2 for diagnostic accuracy studies and PROBAST + AI for prediction model studies. Twenty-two studies were included. Sixteen studies evaluated diagnostic tasks and 10 evaluated prognostic outcomes, with four studies contributing to both categories. Diagnostic performance was generally high, with many studies reporting AUC values approaching or exceeding 0.90, particularly for larger lesion volumes.Prognostic performance was more variable, with moderate to high discrimination and substantial heterogeneity. Only 9 studies incorporated independent external validation, and performance was frequently lower in external cohorts. All prognostic model studies were judged to be at high overall risk of bias using PROBAST + AI, and most diagnostic accuracy studies also demonstrated high or unclear risk of bias in at least one QUADAS-2 domain, most frequently in patient selection. AI-based models applied to brain CT demonstrate strong technical performance for both diagnostic and prognostic tasks in TBI. However, most studies relied on retrospective designs and lacked independent external validation which limits models generalizability and raises concern for potential overfitting. Prospective, multicenter studies with standardized methodologies and rigorous external validation are required before widespread clinical implementation.

Humans↗

PubMed2005

Validation of Michigan neuropathy screening instrument for diabetic peripheral neuropathy.

OBJECTIVE: The reliability and accuracy of the Michigan neuropathy screening instrument (MNSI) have been discussed recently. As a result of the difficulties of performing and analyzing nerve biopsy as a standard diagnostic test, electromyography and neuronography is used as the best alternative diagnostic procedure. The objective of this study was to determine the diagnostic performance of the test characteristics and cut-off point of MNSI scoring for the diagnosis of diabetic peripheral neuropathy. METHOD: Over a 2-year period, a cross-sectional study was conducted on 176 type 2 diabetic patients. An internist carried out the MNSI and the sum of scores varying from 0 to 1 for each abnormality as revealed in foot appearance, ulceration, ankle reflexes and vibratory perception has been recorded. A neurologist, who was blind to the MNSI scores, performed all neurophysiological studies. The test performance characteristics of the MNSI procedure were measured for different cut-off values. RESULTS: MNSI scores of 1.5, 2.0, 2.5 and 3.0 were assessed as cut-off values. Sensitivities were 79%, 65%, 50% and 35% and specificities were 65%, 83%, 91% and 94%, respectively. Positive predictive values increased and negative predictive values decreased for each score. Accuracies, likelihood ratios and post-test probabilities were measured. CONCLUSION: The accuracy of MNSI scoring makes it a useful screening test for diabetic neuropathy in taking a decision regarding which patients should be referred to a neurologist for electrophysiological studies. High specificity, likelihood ratios over 5 and a moderate to good post-test probability give a high diagnostic impact for MNSI scoring. We suggest a cut-off point of 2 for the MNSI procedure. However, electrophysiological studies should be considered when the patient has signs and symptoms other than those rated by the MNSI, suggesting peripheral nerve involvement, and also because the MNSI is still just a screening test.

Adult↗

PubMed1996

What can artificial neural networks teach us about neurodegenerative disorders with extrapyramidal features?

Artificial neural networks (ANNs), computer paradigms that can learn, excel in pattern recognition tasks such as disease diagnosis. Artificial neural networks operate in two different learning modes: supervised, in which a known diagnostic outcome is presented to the ANN, and unsupervised, in which the diagnostic outcome is not presented. A supervised learning ANN could emulate human expert diagnostic performance and identify relevant predictive markers in the diagnostic task, while an unsupervised learning ANN could suggest reasonable alternative diagnostic classification criteria. In the present study, we used ANN methodology to try to overcome the neuropathological difficulties in differentiating the subtypes of progressive supranuclear palsy (PSP), and in differentiating PSP from postencephalitic parkinsonism (PEP) and corticobasal degeneration, or Pick's disease from corticobasal degeneration. First, we applied supervised learning ANN to classify 62 cases of these disorders and to identify diagnostic markers that distinguish them. In a second experiment, we used unsupervised learning ANN to investigate possible alternative nosological classifications. Artificial neural networks input data for each case consisted of values representing histological features, including neurofibrillary tangles, neuronal loss and gliosis found in multiple brain sampling areas. The supervised learning ANN achieved excellent accuracy in classifying PSP but had difficulty classifying the other disorders. This method identified a few features that might help to differentiate PEP, supported currently proposed criteria for Pick's disease, corticobasal degeneration and typical PSP, but detected no features to characterize the atypical subtype of PSP. In general, unsupervised learning ANN supported the present nosological classification for PSP, PEP, Pick's disease and corticobasal degeneration, although it overlapped some groups. Artificial neural networks methodology appears promising for studying neurodegenerative disorders.

Basal Ganglia Diseases↗

PubMed1995

Spatial features in body-surface potential maps can identify patients with a history of sustained ventricular tachycardia.

BACKGROUND: Regional disparities of ventricular primary-repolarization properties contribute to an electrophysiological substrate for arrhythmias. Such disparities can be assessed from body-surface distributions of ECG QRST areas. Our objective was to isolate and test those features of QRST-area distributions that would be suitable for identifying patients at risk for life-threatening ventricular arrhythmias. METHODS AND RESULTS: We recorded ECGs simultaneously from 120 leads during sinus rhythm for 204 patients taking no antiarrhythmic drugs: half had had sustained ventricular tachycardia (VT); the other half, a myocardial infarction but no history of VT. For each patient, we calculated the QRST area in each lead and, using Karhunen-Loeve (K-L) expansion, reduced these data to 16 coefficients (each relating to one spatial feature, an eigenvector, derived from the total set of 204 QRST-area maps). Using stepwise discriminant analysis, we selected feature subsets that best discriminated between the two groups, and we estimated by a bootstrap procedure using 1000 trials how these subsets would perform on a prospective patient population. The mean diagnostic performance of the classifier for 1000 randomly selected training sets (n = 102 in each, with both groups equally represented) increased monotonically with the number of features used for classification. The initial trend for the corresponding test sets (n = 102 in each) was the same but reversed when the number of features exceeded eight. For an optimal set of eight spatial features, the sensitivity and specificity of the classifier for detecting patients with VT in 1000 test sets were (mean +/- SD) 90.3 +/- 4.3% and 78.0 +/- 6.1%, and its positive and negative predictive accuracies were 80.7 +/- 4.2% and 89.2 +/- 4.2%, respectively. Use of QRS duration as a supplementary feature to eight K-L coefficients can, in the test sets, increase specificity to 80.9 +/- 5.4% and positive predictive accuracy to 82.8 +/- 3.9% compared with the results for the optimal number of eight K-L features alone. CONCLUSIONS: Multiple body-surface ECGs contain valuable spatial features that can identify the presence of an arrhythmogenic substrate in the myocardium of patients at risk for ventricular arrhythmias. Our results compare very favorably with those achieved by any other known test, invasive or noninvasive, for arrhythmogenicity.

Body Surface Potential Mapping↗

PubMed2005

Reflectance spectroscopy for diagnosis of epithelial precancer: model-based analysis of fiber-optic probe designs to resolve spectral information from epithelium and stroma.

Reflectance spectroscopy is a promising technology for detection of epithelial precancer. Fiber-optic probes that selectively collect scattered light from both the epithelium and the underlying stroma are likely to improve diagnostic performance of in vivo reflectance spectroscopy by revealing diagnostic features unique to each layer. We present Monte Carlo models with which to evaluate fiber-optic probe geometries with respect to sampling depth and depth resolution. We propose a probe design that utilizes half-ball lens coupled source and detector fibers to isolate epithelial scattering from stromal scattering and hence to resolve spectral information from the two layers. The probe is extremely compact and can provide easy access to different organ sites.

Computer Simulation↗

PubMed2005

Allergy testing on the IMMULITE 2000 Random-Access immunoanalyzer - a clinical evaluation study.

PURPOSE: We aimed to evaluate the diagnostic performance of the IMMULITE 2000 Allergy System from Diagnostic Products Corporation (DPC) for the detection of inhalant and food allergies, focusing on inhalant and food screens, mixes and single allergens. METHODS: Serum samples were collected from new, unselected patients who were referred to the allergist for a suspected allergy. Patients were classified as study diagnosis-positive for inhalant (food) allergy if they had both a positive clinical examination/history and a positive skin test for inhalant (food) allergy; otherwise - failing one or both of these criteria - they were classified as study diagnosis-negative. Classification and testing of the serum samples was carried out in a blinded fashion. Values greater than 0.35 kU/L were considered positive. RESULTS: Of the 118 patients included, 63 were considered study diagnosis-positive for inhalation and/or food allergy. DPC inhalation screening showed 82% total agreement (TA) and 91% sensitivity relative to the study diagnoses. The DPC food panel showed 96% TA and 98% specificity relative to the study diagnoses. Relative to specific intracutaneous testing (ICT), the DPC D1, E1 and E5 assays had sensitivity of 82-90%; tree and grass panels had sensitivity of 74% and 95%. The DPC weed panel and initial lots of DPC E5 had poor sensitivity (<40%); mold panel sensitivity was equally low for both DPC and the routinely used Pharmacia assay (36%). Relative to skin prick testing (SPT), specific food allergens had TA of 94-98% and specificity of 95-100%. CONCLUSION: In patients classified by the combination of clinical examination/history and skin test results, the DPC IMMULITE 2000 Allergy System generally demonstrated acceptable sensitivity, specificity and TA compared to the study diagnoses, both at the screening level and at the level of panels and single allergens.

Adult↗

PubMed2004

Serodiagnosis of toxoplasmosis. The impact of measurement of IgG avidity.

The development of IgG avidity assays has revolutionised serological diagnosis of Toxoplasma infections. The measurement of IgG avidity has shown its power in various clinical settings, especially in situations where timing and differentiation of primary and secondary infections is crucial. However, no laboratory test performed alone is self-sustained, whereby the diagnostic strategy of choice is sequential (or combinatorial) use of high-quality IgG, IgM, IgA and IgG-avidity assays. The impact of IgG avidity measurement will be discussed in five clinical scenarios: acute acquired infection, primary infection during pregnancy, congenital toxoplasmosis, ocular toxoplasmosis and Toxoplasma infection in immunocompromised patients. All in all in toxoplasmosis, superior diagnostic performance is achieved by appropriate combinations of serological, culture-based and PCR techniques.

Acute Disease↗

PubMed1983

Culture diagnosis of Legionella infections.

Culture diagnosis of Legionella infections is feasible in most instances. Using selective culture techniques, sputum can be used to diagnose these infections specifically, and with a degree of sensitivity equal to or greater than currently used immunological means of diagnosis. Every laboratory performing diagnostic testing for this group of infections should perform cultures to achieve optimal yield.

Culture Media↗

PubMed1983

[Provocation tests for coronary spasm. Methods, value and indications].

Five methods are currently used to demonstrate coronary spasm: they use derivatives of ergotamine, alkalinisation, cold, adenosine triphosphate and exercise stress test. The criteria used to confirm spasm are either direct visualisation at coronary angiography or indirect electrocardiographic changes. The value of these test is beginning to be established: --their specificity is excellent with all methods (a positive test is diagnostic); --their sensitivity is mediocre (a negative test does not exclude the diagnosis). This depends on three factors: the choice of criteria of positivity: angiographic appearances is the most reliable; the timing of the test: it is more likely to be positive when the test is carried out close to episodes of chest pain; the method of provocation: the most sensitive tests use either the ergotamine derivatives or alkalinisation; --reproductibility is satisfactory at short and long term. These tests have diagnostic and therapeutic indications. It is preferable to perform diagnostic tests during coronary angiography in patients with normal coronary arteries and unexplained episodes of chest pain. However, in patients with fixed coronary lesions (except triple vessel and left main stem disease) when coronary bypass surgery is planned, the results of the test can help in the choice of associated therapy, such as plexectomy and calcium antagonist drugs. The repetition of the test is useful in the evaluation of antispastic therapy: when the test becomes negative the drug used is probably effective. In conclusion, with equal patient comfort and MSR, provocation tests using the ergotamine derivatives would seem to be the method of choice: alkalinisation is a good alternative.

Adenosine Triphosphate↗

PubMed2005

Rapid scanning protocol for brain (18)F-FDG PET: a validation study.

UNLABELLED: Although (18)F-FDG PET is an established technique to assess brain glucose use, a shorter imaging time is preferable for patient convenience and increased throughput. The aim of this study was to validate a brain (18)F-FDG PET protocol more rapid than the conventional protocol. METHODS: For comparison of normalized metabolic activities, brain (18)F-FDG PET was performed on 60 healthy subjects and 25 patients with probable Alzheimer's disease (AD), and an additional 20 healthy subjects served as a control group to assess diagnostic performance between the conventional and rapid scanning protocols. Conventional scans were acquired for a total of 20 min (a 10-min emission and a 10-min transmission). Immediately after conventional scanning, rapid scanning was performed for a total of 4 min (a 3-min emission and a 1-min transmission). PET images were anatomically standardized using NEUROSTAT, with pixel values normalized to the individual global value. Two database sets, from the 2 protocols, were compared by regional values and pixel-by-pixel analysis. A receiver-operating-characteristic analysis was performed for comparison of diagnostic accuracy between the 2 protocols. A kinetic simulation study was also performed to examine the possible difference due to the time lag between the protocols. RESULTS: Although small differences in normalized activity were found in several regions in the healthy subjects between the 2 protocols, no significant difference was found in any region in the patient group. The coefficients of variation of the normalized activity were 20%-30% larger in the rapidly scanned images, but the mean z images and their coefficient-of-variation images did not differ. The kinetic simulation study suggested that the differences were caused by the time lag between the 2 protocols. No significant differences were found in area under the receiver-operating-characteristic curves, and the diagnostic accuracies for the detection of AD were virtually equal between the 2 protocols. CONCLUSION: The rapid scanning protocol used in the present study could provide results nearly equivalent to data from the conventional protocol. Thus, it is feasible to use this rapid protocol to detect AD, without losing diagnostic accuracy.

Aged↗

PubMed2001

Automatic analysis and classification of surface electromyography.

In this paper, parametric modeling of surface electromyography (EMG) algorithms that facilitates automatic SEMG feature extraction and artificial neural networks (ANN) are combined for providing an integrated system for the automatic analysis and diagnosis of myopathic disorders. Three paradigms of ANN were investigated: the multilayer backpropagation algorithm, the self-organizing feature map algorithm and a probabilistic neural network model. The performance of the three classifiers was compared with that of the old Fisher linear discriminant (FLD) classifiers. The results have shown that the three ANN models give higher performance. The percentage of correct classification reaches 90%. Poorer diagnostic performance was obtained from the FLD classifier. The system presented here indicates that surface EMG, when properly processed, can be used to provide the physician with a diagnostic assist device.

Algorithms↗

PubMed2002

Feasibility and diagnostic impact of fluorescence-based diagnostic laparoscopy in hepatocellular carcinoma: a case report.

Fluorescence-based diagnostic techniques are an evolving field in the staging of hepatologic and gastroenterologic malignancies. The method aims at improved accuracy in the detection of locoregional and distant metastases. In hepatocellular carcinoma, detection of metastases is of utmost importance, because advanced tumors preclude curative surgical therapy. We present here our first experience of fluorescence-based diagnostic laparoscopy in a patient with hepatocellular carcinoma. We performed diagnostic laparoscopy in a 76-year-old woman who presented with a tumor of the liver and increased serum levels of alpha-fetoprotein. For photosensitization, 5-aminolevulinic acid (20 mg/kg bodyweight, p.o.) was used 6 hours prior to laparoscopy, which was performed with a prograde 0 degree telescope suitable for fluorescence-based laparoscopy. The D-Light AF system was used as a light source. Laparoscopy was carried out according to common guidelines in the white-light as in the blue-light mode. The main tumor and the satellite metastases clearly showed intense fluorescence compared with unaffected liver parenchyma and other abdominal organs. Fluorescence-based laparoscopy identified several small metastases which had been missed during conventional white-light illumination. Hepatocellular carcinoma was confirmed by the histological examination of several biopsies drawn from suspicious areas. Fluorescence-based laparoscopy provided additional information in the staging of hepatocellular carcinoma. The technique can further improve diagnostic accuracy in the staging of hepatocellular carcinoma, particularly in patients potentially suitable for partial liver resection or transplantation.

Aged↗

PubMed1989

Maximizing diagnostic information from the dexamethasone suppression test. An approach to criterion selection using receiver operating characteristic analysis.

Receiver operating characteristic analysis yields indices of diagnostic performance that permit innovative mathematical descriptions and comparisons of diagnostic tests whose results are distributed over a range of possible outcomes. We employed data from seven published studies to demonstrate how ROC analysis may be used to characterize the discriminative properties of the dexamethasone suppression test (DST). We also used an intrinsic property of test results--their ability to reduce diagnostic uncertainty--to select cortisol cutoff levels with the highest information content for each of the DST studies. These cutoffs often differ from original authors' definitions of nonsuppression. The information-maximizing cortisol levels vary depending on pretest estimates of the prevalence of depression or melancholia and vary among studies. Receiver operating characteristic analytic techniques provide important tools for a full description of the DST. Future studies should use receiver operating characteristic methods to select cortisol cutoffs and to assess the test's overall performance.

Binding, Competitive↗