Home About us Contact
|
Home About us Contact | ||
|
|
||
Receiver Operator Characteristic (receiver + operator_characteristic)
Terms modified by Receiver Operator Characteristic Selected AbstractsMulticenter Validation of the Philadelphia EMS Admission Rule (PEAR) to Predict Hospital Admission in Adult Patients Using Out-of-hospital DataACADEMIC EMERGENCY MEDICINE, Issue 6 2009Zachary F. Meisel MD Abstract Objectives:, The objective was to validate a previously derived prediction rule for hospital admission using routinely collected out-of-hospital information. Methods:, The authors performed a multicenter retrospective cohort study of 1,500 randomly selected, adult patients transported to six separate emergency departments (EDs; three community and three academic hospitals in three separate health systems) by a city-run emergency medical services (EMS) system over a 1-year period. Patients younger than 18 years or who bypassed the ED to be evaluated by trauma, obstetric, or psychiatric teams were excluded. The score consisted of six weighted elements that generated a total score (0,14): age , 60 years (3 points); chest pain (3); shortness of breath (3); dizzy, weakness, or syncope (2); history of cancer (2); and history of diabetes (1). Receiver operator characteristic (ROC) curves for the decision rule and admission rates were calculated among individual hospitals and for the entire cohort. Results:, A total of 1,102 patients met inclusion criteria. The admission rate for the entire cohort was 40%, and individual hospital admission rates ranged from 28% to 57%. Overall, 34% had a score of ,4, and 29% had a score of ,5. Area under the ROC curve (AUC) for the combined cohort was 0.83 for all admissions and 0.72 for intensive care unit (ICU) admissions; AUCs at individual hospitals ranged from 0.72 to 0.85. The admission rate for a score of ,4 was 77%; for a score of ,5 the admission rate was 80%. Conclusions:, The ability of this EMS rule to predict the likelihood of hospital admission appears valid in this multicenter cohort. Further studies are needed to measure the impact and feasibility of using this rule to guide decision-making. [source] A PCA-based modelling technique for predicting environmental suitability for organisms from presence recordsDIVERSITY AND DISTRIBUTIONS, Issue 1-2 2001M. P. Robertson We present a correlative modelling technique that uses locality records (associated with species presence) and a set of predictor variables to produce a statistically justifiable probability response surface for a target species. The probability response surface indicates the suitability of each grid cell in a map for the target species in terms of the suite of predictor variables. The technique constructs a hyperspace for the target species using principal component axes derived from a principal components analysis performed on a training dataset. The training dataset comprises the values of the predictor variables associated with the localities where the species has been recorded as present. The origin of this hyperspace is taken to characterize the centre of the niche of the organism. All the localities (grid-cells) in the map region are then fitted into this hyperspace using the values of the predictor variables at these localities (the prediction dataset). The Euclidean distance from any locality to the origin of the hyperspace gives a measure of the ,centrality' of that locality in the hyperspace. These distances are used to derive probability values for each grid cell in the map region. The modelling technique was applied to bioclimatic data to predict bioclimatic suitability for three alien invasive plant species (Lantana camara L., Ricinus communis L. and Solanum mauritianum Scop.) in South Africa, Lesotho and Swaziland. The models were tested against independent test records by calculating area under the curve (AUC) values of receiver operator characteristic (ROC) curves and kappa statistics. There was good agreement between the models and the independent test records. The pre-processing of climatic variable data to reduce the deleterious effects of multicollinearity, and the use of stopping rules to prevent overfitting of the models are important aspects of the modelling process. [source] Long-Term Fracture Prediction by DXA and QUS: A 10-Year Prospective Study,JOURNAL OF BONE AND MINERAL RESEARCH, Issue 3 2006Alison Stewart PhD Abstract This study investigated the ability of DXA and QUS to predict fractures long term when measured around the time of the menopause. We found both DXA and QUS are able to predict both any fracture and "osteoporotic" fractures and that QUS can predict independently of BMD. Introduction: There are now many treatments available for prevention of osteoporotic fracture. To be cost-effective, we need to target those most at risk. This study examines the ability of DXA and QUS to predict fractures in an early postmenopausal population of women. Materials and Methods: We prospectively measured 3883 women who had been randomly selected from a community-based register. At baseline, they were measured using DXA of spine and hip (Norland XR-26) and QUS of the heel (Walker Sonix UBA 575). Follow-up had a mean of 9.7 ± 1.1 (SD) years. All incident fractures were identified and validated by examination of X-ray reports, and these were compared with those without fracture in a Cox-regression model to calculate hazard ratios (HRs). Results: We found adjusted HRs for any fracture per 1 SD reduction in spine BMD to be 1.61 (1.42-1.83), whereas neck of femur BMD was 1.54 (1.34-1.75). Areas under the curve (AUC) for a receiver operator characteristic (ROC) analysis were 0.62 for spine BMD and 0.59 for neck BMD. In a subgroup where QUS was also measured, the HR for a 1 SD reduction in BMD was 1.69 (1.29-2.22) for spine BMD and 1.55 (1.17-2.06) for neck BMD. The HR for a 1 SD reduction in broadband ultrasound attenuation (BUA) was 1.53 (1.19-1.96), and 1.44 (1.12-1.86) when further adjusted for neck BMD. The AUCs were 0.63 for spine BMD, 0.59 for neck BMD, and 0.62 for BUA. When only osteoporotic fractures were examined, the HRs increased in all situations. BUA showed the highest HR of 2.25 (1.51-3.34), and when further adjusted for neck BMD was 2.12 (1.38-3.28). Conclusions: In conclusion, it may be possible to scan women around the time of the menopause to predict future fractures. It seems that, for "osteoporotic" fractures, BUA may be an improved predictor of fractures in comparison with DXA, because the relative risk is highest for BUA, and independent of BMD. [source] Measuring diagnostic and predictive accuracy in disease management: an introduction to receiver operating characteristic (ROC) analysisJOURNAL OF EVALUATION IN CLINICAL PRACTICE, Issue 2 2006Ariel Linden DrPH MS Abstract Diagnostic or predictive accuracy concerns are common in all phases of a disease management (DM) programme, and ultimately play an influential role in the assessment of programme effectiveness. Areas, such as the identification of diseased patients, predictive modelling of future health status and costs and risk stratification, are just a few of the domains in which assessment of accuracy is beneficial, if not critical. The most commonly used analytical model for this purpose is the standard 2 × 2 table method in which sensitivity and specificity are calculated. However, there are several limitations to this approach, including the reliance on a single defined criterion or cut-off for determining a true-positive result, use of non-standardized measurement instruments and sensitivity to outcome prevalence. This paper introduces the receiver operator characteristic (ROC) analysis as a more appropriate and useful technique for assessing diagnostic and predictive accuracy in DM. Its advantages include; testing accuracy across the entire range of scores and thereby not requiring a predetermined cut-off point, easily examined visual and statistical comparisons across tests or scores, and independence from outcome prevalence. Therefore the implementation of ROC as an evaluation tool should be strongly considered in the various phases of a DM programme. [source] Abstracts of the 8th Meeting of the Italian Peripheral Nerve Study Group: 57JOURNAL OF THE PERIPHERAL NERVOUS SYSTEM, Issue 1 2003D Cocito BACKGROUND: Since 1991, five sets of electrophysiological criteria for CIDP have been reported. However, until now, receiver operator characteristic (ROC), such as sensitivity and specificity, of only AAN criteria were investigated, showing a high specificity, but intermediate sensitivity. The application of these criteria may be useful in clinical trials, but is inadequate in clinical practice, since they preclude immunomodulating treatment in patients who do not meet them. OBJECTIVE: 1) to evaluate the ROC and predictive value of five different electrophysiological criteria for CIDP (AAN, INCAT, Rotta et al, Nicholas et al. and Saperstein et al.); 2) to identify the most informative electrophysiological features indicative of demyelination by mean of the likelihood ratio; 3) to determine, in our series of cases, a set of minimal electrophysiological criteria (albeit aspecific) enough sensitive for CIDP diagnosis. PATIENTS AND METHODS: 20 patients with sensorimotor polyneuropathy, progressive for at least 2 months, with weakness in least two limbs and documented improvement in strength in response to immunotherapy. Other potential causes were excluded, including diabetes and IgM paraproteinemia with or without anti-MAG reactivity. Twelve patients with axonal polyneuropathy associated with diabetes and 12 patients with amyotrophic lateral sclerosis were included as controls. Nerve conduction studies (NCS) were evaluated according to all five sets of NCS criteria and ROC was calculated. Likelihood ratio for CIDP was evaluated for each set of NCS criteria and for presence of each of the following features in different number of nerves: motor conduction velocities (MCV), conduction blocks/temporal dispersion, distal motor (DML) and F-wave latencies (FWL). MCV, DML and FWL were considered indicative of demyelination if decreased or increased as suggested by AAN criteria. Results will be discussed. [source] Lung function and exhaled nitric oxide levels in infants developing chronic lung diseasePEDIATRIC PULMONOLOGY, Issue 2 2007Olivia Williams MRCPCH Abstract Chronic lung disease (CLD) is a common outcome of neonatal intensive care. To determine whether the results of serial exhaled nitric oxide (eNO) measurements during the perinatal period differed between infants who did and did not develop CLD. In addition, we wished to assess whether eNO results were more predictive of CLD development than lung function test results or readily available clinical data (gestational age and birthweight). The patients were 24 infants with a median gestational age of 27 (range 25,31) weeks. Measurements of eNO levels, functional residual capacity (FRC), and compliance of the respiratory system (CRS) were attempted on postnatal days 1, 3, 5, 7, 14, and 28 days. The 12 infants who developed CLD were of significantly lower birthweight and gestational age than the rest of the cohort; in addition, they had lower median FRC (P,<,0.02) and CRS (P,<,0.02) results, but not higher eNO levels, in the first week after birth. Construction of receiver operator characteristic (ROC) curves demonstrated that the CRS and FRC results on Day 3 were the best predictors of CLD development; the areas under the ROC curves were 0.94 and 0.91, respectively. Early lung function test results, but not eNO levels, are useful in predicting CLD development, but are not significantly better than birthweight. Pediatr Pulmonol. 2007; 42:107,113. © 2006 Wiley-Liss, Inc. [source] Automatic generation and evaluation of sparse protein signatures for families of protein structural domainsPROTEIN SCIENCE, Issue 1 2005Matthew J. Blades Abstract 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. [source] Prospective Validation of the Pediatric Appendicitis Score in a Canadian Pediatric Emergency DepartmentACADEMIC EMERGENCY MEDICINE, Issue 7 2009Maala Bhatt MD Abstract Objectives:, Clinical scoring systems attempt to improve the diagnostic accuracy of pediatric appendicitis. The Pediatric Appendicitis Score (PAS) was the first score created specifically for children and showed excellent performance in the derivation study when administered by pediatric surgeons. The objective was to validate the score in a nonreferred population by emergency physicians (EPs). Methods:, A convenience sample of children, 4,18 years old presenting to a pediatric emergency department (ED) with abdominal pain of less than 3 days' duration and in whom the treating physician suspected appendicitis, was prospectively evaluated. Children who were nonverbal, had a previous appendectomy, or had chronic abdominal pathology were excluded. Score components (right lower quadrant and hop tenderness, anorexia, pyrexia, emesis, pain migration, leukocytosis, and neutrophilia) were collected on standardized forms by EPs who were blinded to the scoring system. Interobserver assessments were completed when possible. Appendicitis was defined as appendectomy with positive histology. Outcomes were ascertained by review of the pathology reports from the surgery specimens for children undergoing surgery and by telephone follow-up for children who were discharged home. Sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) were calculated. The overall performance of the score was assessed by a receiver operator characteristic (ROC) curve. Results:, Of the enrolled children who met inclusion criteria (n = 246), 83 (34%) had pathology-proven appendicitis. Using the single cut-point suggested in the derivation study (PAS 5) resulted in an unacceptably high number of false positives (37.6%). The score's performance improved when two cut-points were used. When children with a PAS of ,4 were discharged home without further investigations, the sensitivity was 97.6% with a NPV of 97.7%. When a PAS of ,8 determined the need for appendectomy, the score's specificity was 95.1% with a PPV of 85.2%. Using this strategy, the negative appendectomy rate would have been 8.8%, the missed appendicitis rate would have been 2.4%, and 41% of imaging investigations would have been avoided. Conclusions:, The PAS is a useful tool in the evaluation of children with possible appendicitis. Scores of ,4 help rule out appendicitis, while scores of ,8 help predict appendicitis. Patients with a PAS of 5,7 may need further radiologic evaluation. [source] Correlation of Optic Nerve Sheath Diameter with Direct Measurement of Intracranial PressureACADEMIC EMERGENCY MEDICINE, Issue 2 2008Heidi Harbison Kimberly MD Abstract Background:, Measurements of the optic nerve sheath diameter (ONSD) using bedside ultrasound (US) have been shown to correlate with clinical and radiologic signs and symptoms of increased intracranial pressure (ICP). Objectives:, Previous literature has identified 5 mm as the ONSD measurement above which patients exhibit either clinical or radiologic signs of elevated ICP. The goals of this study were to evaluate the association between ONSD and ICP and to validate the commonly used ONSD threshold of 5 mm using direct measurements of ICP as measured by ventriculostomy. Methods:, A prospective blinded observational study was performed using a convenience sample of adult patients in both the emergency department (ED) and the neurologic intensive care unit (ICU) who had invasive intracranial monitors placed as part of their clinical care. Ocular USs were performed with a 10,5 MHz linear probe. Emergency physicians (EPs) with previous ocular US experience performed ONSD measurements while blinded to the contemporaneous ICP reading obtained directly from invasive monitoring. The association between ONSD and ICP was assessed with the Spearman rank correlation coefficient, and a receiver operator characteristic (ROC) curve was created to determine the optimal ONSD cutoff to detect ICP > 20 cm H2O. Results:, Thirty-eight ocular USs were performed on 15 individual patients. Spearman rank correlation coefficient of ONSD and ICP was 0.59 (p < 0.0005) demonstrating a significant positive correlation. An ROC curve was created to assess the ability of ONSD to distinguish an abnormal ICP greater than 20 cm H2O. The area under the ROC curve was 0.93 (95% confidence interval [CI] = 0.84 to 0.99). Based on inspection of the ROC curve, ONSD > 5 mm performed well to detect ICP > 20 cm H2O with a sensitivity of 88% (95% CI = 47% to 99%) and specificity of 93% (95% CI = 78% to 99%). Conclusions:, Using an ROC curve the authors systematically confirmed the commonly used threshold of ONSD > 5 mm to detect ICP > 20 cm H2O. This study directly correlates ventriculostomy measurements of ICP with US ONSD measurements and provides further support for the use of ONSD measurements as a noninvasive test for elevated ICP. [source] A CRITICAL LOOK AT PAP ADEQUECY: ARE OUR CRITERIA SATISFACTORY?CYTOPATHOLOGY, Issue 2006D.R. Bolick Liquid based Pap (LBP) specimen adequacy is a highly documented, yet poorly understood cornerstone of our GYN cytology practice. Each day, as cytology professionals, we make adequacy assessments and seldom wonder how the criteria we use were established. Are the criteria appropriate? Are they safe? What is the scientific data that support them? Were they clinically and statistically tested or refined to achieve optimal patient care? In this presentation, we will take a fresh look at what we know about Pap specimen adequacy and challenge some of the core assumptions of our daily practice. LBP tests have a consistent, well-defined surface area for screening, facilitating the quantitative estimates of slide cellularity. This provides an unprecedented opportunity to establish reproducible adequacy standards that can be subjected to scientific scrutiny and rigorous statistical analysis. Capitalizing on this opportunity, the TBS2001 took the landmark step to define specimen adequacy quantitatively, and set the threshold for a satisfactory LBP at greater than 5,000 well visualized squamous epithelial cells. To date, few published studies have attempted to evaluate the validity or receiver operator characteristics for this threshold, define an optimal threshold for clinical utility or assess risks of detection failure in ,satisfactory' but relatively hypocellular Pap specimens. Five years of cumulative adequacy and cellularity data of prospectively collected Pap samples from the author's laboratory will be presented, which will serve as a foundation for a discussion on ,Pap failure'. A relationship between cellularity and detection of HSIL will be presented. Risk levels for Pap failure will be presented for Pap samples of different cellularities. The effect of different cellularity criterion on unsatisfactory Pap rates and Pap failure rates will be demonstrated. Results from this data set raise serious questions as to the safety of current TBS2001 adequacy guidelines and suggest that the risk of Pap failure in specimens with 5,000 to 20 000 squamous cells on the slide is significantly higher than those assumed by the current criteria. TBS2001 designated all LBP to have the same adequacy criterion. Up to this point, it has been assumed that ThinPrep, SurePath, or any other LBP would be sufficiently similar that they should have the same adequacy criteria. Data for squamous cellularity and other performance characteristics of ThinPrep and SurePath from the author's laboratory will be compared. Intriguing data involving the recently approved MonoPrep Pap Test will be reviewed. MonoPrep clinical trial data show the unexpected finding of a strong correlation between abundance of endocervical component and the detection of high-grade lesions, provoking an inquiry of a potential new role for a quantitative assessment of the transition zone component. The current science of LBP adequacy criteria is underdeveloped and does not appear to be founded on statistically valid methods. This condition calls us forward as a body of practitioners and scientists to rigorously explore, clarify and define the fundamental nature of cytology adequacy. As we forge this emerging science, we will improve diagnostic performance, guide the development of future technologies, and better serve the patients who give us their trust. Reference:, Birdsong GG: Pap smear adequacy: Is our understanding satisfactory? Diagn Cytopathol. 2001 Feb; 24(2): 79,81. [source] Predictors of Long-Term Risk for Heart Failure Hospitalization after Acute Myocardial InfarctionANNALS OF NONINVASIVE ELECTROCARDIOLOGY, Issue 3 2010Juha S. Perkiömäki M.D. Background: Data on the value of baseline brain natriuretic peptide (BNP) and autonomic markers in predicting heart failure (HF) hospitalization after an acute myocardial infarction (AMI) are limited. Methods: A consecutive series of patients with AMI without a previous history of HF (n = 569) were followed up for 8 years. At baseline, the patients had a blood sample for determination of BNP, a 24-hour Holter recording for evaluating heart rate variability (HRV) and heart rate turbulence (HRT), and an assessment of baroreflex sensitivity (BRS) using phenylephrine test. Results: During the follow-up, 79 (14%) patients were hospitalized due to HF. Increased baseline BNP, decreased HRV, HRT, and BRS had a significant association with HF hospitalization in univariate comparisons (P < 0.001 for all). After adjusting with all the relevant clinical parameters, BNP, HRV, and HRT still significantly predicted HF hospitalization (P < 0.001 for BNP and for the short-term scaling exponent ,1, P < 0.01 for turbulence slope). In the receiver operator characteristics curve analysis, the area under the curve for BNP was 0.77, for the short-term scaling exponent ,1 0.69, for turbulence slope 0.71, and for BNP/standard deviation of all N-N intervals ratio 0.80. Conclusion: Baseline increased BNP and impaired autonomic function after AMI yield significant information on the long-term risk for HF hospitalization. Ann Noninvasive Electrocardiol 2010;15(3):250,258 [source] A novel algorithm to improve pathologic stage prediction of clinically organ-confined muscle-invasive bladder cancerCANCER, Issue 7 2009David Margel MD Abstract BACKGROUND: An algorithm was created to predict pathologic stage in patients with clinically organ-confined muscle-invasive bladder cancer. METHODS: The sample consisted of 133 consecutive patients scheduled to undergo cystectomy. To develop a tool to predict nonorgan-confined disease before surgery, principal component analysis (PCA) was applied. Patients were stratified into a training set (n = 89) and a validation set (n = 44), and 7 parameters were evaluated: levels of carcinoembryonic antigen, cancer antigen (CA) 125, and carbohydrate antigen (CA) 19-9; clinical stage; presence of hydronephrosis; presence of carcinoma in situ; and initial tumor size >3 cm. PCA was applied to the training set to determine the weight of each parameter. A PCA score was generated for each patient in the set, and a cutoff defining nonorgan-confined disease was established. The accuracy of the cutoff was quantified by the area under the receiver operator characteristics curve (AUC). The model was then applied to the validation set without recalculation; the AUC and the positive and negative predictive values of the validation set were calculated. RESULTS: On pathologic evaluation, 71 patients (53%) were found to have organ-confined tumors and 62 patients (47%) had extravesical disease. The AUC was 0.85 in the training group (95% confidence interval [95% CI], 0.71-0.97) and 0.84 in the validation group (95% CI, 0.75-0.93). The positive and negative predictive values in the validation group were 88% (95% CI, 71%-96%) and 94% (95% CI, 71%-99%), respectively. CONCLUSIONS: The newly devised, internally validated, algorithm was 85% accurate in predicting nonorgan-confined bladder disease before cystectomy. Further external validation in a large cohort was recommended as still necessary. Cancer 2009. © 2009 American Cancer Society. [source] | ||||||||||