Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Predictive”

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.

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

At least 667 records · Page 37Linked to original sources
PubMed1994

Electroencephalography improves the prediction of functional outcome in the acute stage of cerebral ischemia.

BACKGROUND AND PURPOSE: We studied the value of clinical and electroencephalographic assessment in patients with acute first-ever supratentorial ischemia in predicting functional outcome after 1 year. METHODS: In 55 consecutive patients admitted after a median interval of less than 24 hours, the degree of handicap was dichotomized as moderate (Rankin grade 1, 2, or 3) or severe (Rankin grade 4 or 5). Clinical deficits were categorized according to signs of a lacunar or a cortical syndrome. Without knowledge of clinical data, electroencephalograms (EEGs) were classified according to findings predicting good or poor prognosis. The outcome after 1 year was assessed as good (Rankin grade 3 or less) or poor (Rankin grade 4 or 5 or death from stroke) and was correlated to clinical data and to EEG findings in the acute stage. RESULTS: Thirty patients with a moderate handicap on admission all had a good outcome (predictive value [PV] of the initial handicap, 1.00; 95% confidence interval [CI], 0.88 to 1.00). Of the 25 patients with severe handicap on admission a poor outcome occurred in 13 (PV, 0.52; 95% CI, 0.31 to 0.72). If these patients with severe handicap at baseline were subdivided according to clinical features, a lacunar syndrome predicted good outcome in 4 of 5 patients (PV, 0.80; 95% CI, 0.28 to 1.00), but a cortical syndrome predicted poor outcome in only 12 of 20 patients (PV, 0.60; 95% CI, 0.36 to 0.81). Of the 20 patients with severe handicap and a cortical syndrome at baseline, an EEG with features predicting a good prognosis correctly predicted good outcome in 6 of 7 patients (PV, 0.86; 95% CI, 0.42 to 1.00). An EEG with features predicting poor prognosis correctly predicted poor outcome in 11 of 13 patients (PV, 0.85; 95% CI, 0.55 to 0.98). CONCLUSIONS: Electroencephalography improves the prediction of functional outcome in patients with a severe neurological deficit in the acute stage of cerebral ischemia. This may have implications for the design of future intervention trials in acute stroke.

Activities of Daily Living↗
PubMedDate not supplied

Comparison of four bedside indicators used to predict duodenal feeding tube placement with radiography.

The validity of four indicators to predict successful duodenal feeding tube placement was evaluated in a prospective trial. Data were collected on each indicator at prepyloric (< or = 65 cm) and postpyloric (> or = 75 cm) feeding tube lengths. Feeding tubes were placed in 106 patients. Eighteen feeding tubes were located in the stomach, and 88 were in the duodenum. Auscultation (progression of loudest sound locations from the left to the right abdomen) had a positive predictive value of 85% (negative predictive value, 31%). The vacuum effect (a change from 40 mL of aspirated air to < or = 10 mL after 60 mL of air instillation) had a positive predictive value of 86% (negative predictive value, 45%) and was significantly correlated with duodenal placement (p = .02). Aspirate was present at prepyloric and postpyloric lengths in 35 cases. Ten of these 35 cases had the defined change in pH from < or = 4.0 to > or = 6.0 (positive predictive value, 100%; negative predictive value, 28%). The positive predictive value of color (a change to yellow) was also 100% (n = 11); the negative predictive value was 29%. The low negative predictive values of the indicators suggest that the absence of defined changes is of no assistance in discriminating between stomach and duodenal placement. A positive auscultation or vacuum effect test is not conclusive for duodenal placement. A positive pH or color change test may obviate the need for a confirmatory radiograph.(ABSTRACT TRUNCATED AT 250 WORDS)

Aged↗
PubMed2006

Using ESTs to improve the accuracy of de novo gene prediction.

BACKGROUND: ESTs are a tremendous resource for determining the exon-intron structures of genes, but even extensive EST sequencing tends to leave many exons and genes untouched. Gene prediction systems based exclusively on EST alignments miss these exons and genes, leading to poor sensitivity. De novo gene prediction systems, which ignore ESTs in favor of genomic sequence, can predict such "untouched" exons, but they are less accurate when predicting exons to which ESTs align. TWINSCAN is the most accurate de novo gene finder available for nematodes and N-SCAN is the most accurate for mammals, as measured by exact CDS gene prediction and exact exon prediction. RESULTS: TWINSCAN_EST is a new system that successfully combines EST alignments with TWINSCAN. On the whole C. elegans genome TWINSCAN_EST shows 14% improvement in sensitivity and 13% in specificity in predicting exact gene structures compared to TWINSCAN without EST alignments. Not only are the structures revealed by EST alignments predicted correctly, but these also constrain the predictions without alignments, improving their accuracy. For the human genome, we used the same approach with N-SCAN, creating N-SCAN_EST. On the whole genome, N-SCAN_EST produced a 6% improvement in sensitivity and 1% in specificity of exact gene structure predictions compared to N-SCAN. CONCLUSION: TWINSCAN_EST and N-SCAN_EST are more accurate than TWINSCAN and N-SCAN, while retaining their ability to discover novel genes to which no ESTs align. Thus, we recommend using the EST versions of these programs to annotate any genome for which EST information is available.TWINSCAN_EST and N-SCAN_EST are part of the TWINSCAN open source software package http://genes.cse.wustl.edu/distribution/download_TS.html.

Algorithms↗
PubMed2005

Ambulatory teaching: do approaches to learning predict the site and preceptor characteristics valued by clerks and residents in the ambulatory setting?

BACKGROUND: In a study to determine the site and preceptor characteristics most valued by clerks and residents in the ambulatory setting we wished to confirm whether these would support effective learning. The deep approach to learning is thought to be more effective for learning than surface approaches. In this study we determined how the approaches to learning of clerks and residents predicted the valued site and preceptor characteristics in the ambulatory setting. METHODS: Postal survey of all medical residents and clerks in training in Ontario determining the site and preceptor characteristics most valued in the ambulatory setting. Participants also completed the Workplace Learning questionnaire that includes 3 approaches to learning scales and 3 workplace climate scales. Multiple regression analysis was used to predict the preferred site and preceptor characteristics as the dependent variables by the average scores of the approaches to learning and perception of workplace climate scales as the independent variables. RESULTS: There were 1642 respondents, yielding a 47.3% response rate. Factor analysis revealed 7 preceptor characteristics and 6 site characteristics valued in the ambulatory setting. The Deep approach to learning scale predicted all of the learners' preferred preceptor characteristics (beta = 0.076 to beta = 0.234, p < .001). Valuing preceptor Direction was more strongly associated with the Surface Rational approach (beta = .252, p < .001) and with the Surface Disorganized approach to learning (beta = .154, p < 001) than with the Deep approach. The Deep approach to learning scale predicted valued site characteristics of Office Management, Patient Logistics, Objectives and Preceptor Interaction (p < .001). The Surface Rational approach to learning predicted valuing Learning Resources and Clinic Set-up (beta = .09, p = .001; beta = .197, p < .001). The Surface Disorganized approach to learning weakly negatively predicted Patient Logistics (beta = -.082, p = .003) and positively the Learning Resources (beta = .088, p = .003). Climate factors were not strongly predictive for any studied characteristics. Role Modeling and Patient Logistics were predicted by Supportive Receptive climate (beta = .135, p < .001, beta = .118, p < .001). CONCLUSION: Most site and preceptor characteristics valued by clerks and residents were predicted by their Deep approach to learning scores. Some characteristics reflecting the need for good organization and clear direction are predicted by learners' scores on less effective approaches to learning.

Adult↗
PubMed1988

Predicting serum lithium concentration using Bayesian method: a comparison with other methods.

Two pharmacokinetic approaches (single-point Bayesian and two fixed volume of distribution-iterative methods) for predicting serum lithium concentrations in patients treated with lithium carbonate for manic-depressive illness or cyclic neutropenia in Kyushu University Hospital were evaluated and compared retrospectively. Prior to these analyses, three methods (prediction using mean parameters reported by Mason et al., the Pepin method, and the Zetin method) without measuring serum concentrations were also compared. In the Bayesian analysis, the effect of population mean parameters (reported by Mason et al. and Pepin et al.), which were used as initial estimates in a fitting process, on predictive performance was also studied. Forty five patients (21 male, 24 female) were included in this study. The average number of determinations per patient was 6.3, and the sampling times ranged from 2 to 18 h after the last dose. Serum lithium concentrations were measured by atomic absorption spectrometer. The Bayesian method used a computer program (PEDA) developed previously by one of us. The prediction using the population mean values from Mason's report gave the least root mean squared error (RMSE; a composite measure for bias and precision of prediction), and was considered to be the most precise among the methods without measuring serum concentrations. Among the methods using a single measured concentration, the Bayesian prediction was less biased and more precise than that by the two fixed volume of distribution-iterative methods. The Bayesian method reduced prediction error in serum concentration prediction compared with those obtained from population mean parameters in both cases: A high reduction of RMSE was observed when the values from Pepin method were used as initial estimates (from 0.320 to 0.219 meq/l), while, Mason's values gave less reduction (from 0.219 to 0.213 meq/l). In the Bayesian prediction of serum lithium concentration, the selection of population-based initial estimates gave no effect on predictive ability of the Bayesian method in terms of RMSE. In conclusion, the Bayesian method was robust and flexible with regard to dosing schedule, sampling time and number of blood samples, and gave the most clinically acceptable precision among the methods evaluated.

Adolescent↗
PubMed1998

Predicting the effects of optical defocus on human contrast sensitivity.

We used diffraction modulation transfer functions and model eyes to predict the effect of defocus on the contrast sensitivity function (CSF) and compared these predictions with previously published experimental data. Using the principle that optically induced changes in the modulation transfer function should be paralleled by identical changes in the CSF, we used the modulation transfer function calculations with the best-focus CSF measurements to predict the defocused CSF. An aberration-free model predicted the effects of defocus well when the CSF was measured with small pupils (e.g., 2 mm) but not with larger pupils (6-8 mm). When the model included average aberrations, prediction of the defocused CSF with large pupils was better but remained inaccurate, failing, in particular, to reflect differences between individual subjects. Inclusion of measured aberrations for individual subjects provided accurate predictions in the shape of the monochromatic CSF of two of three subjects with hyperopic defocus and good predictions of the polychromatic CSF of two subjects with hyperopic defocus. Prediction of the effects of myopic defocus by use of measured individual aberrations of one subject were less successful. Hence a diffraction optics model can provide good predictions of the effects of defocus on the human CSF, given that one has knowledge of the individual ocular aberrations. These predictions are dependent on the quality of the aberration measurements.

Contrast Sensitivity↗
PubMed2004

Effects of nonlinearities and uncorrelated or correlated errors in realistic simulated data on the prediction abilities of augmented classical least squares and partial least squares.

Comparisons of prediction models from the new augmented classical least squares (ACLS) and partial least squares (PLS) multivariate spectral analysis methods were conducted using simulated data containing deviations from the idealized model. The simulated data were based on pure spectral components derived from real near-infrared spectra of multicomponent dilute aqueous solutions. Simulated uncorrelated concentration errors, uncorrelated and correlated spectral noise, and nonlinear spectral responses were included to evaluate the methods on situations representative of experimental data. The statistical significance of differences in prediction ability was evaluated using the Wilcoxon signed rank test. The prediction differences were found to be dependent on the type of noise added, the numbers of calibration samples, and the component being predicted. For analyses applied to simulated spectra with noise-free nonlinear response, PLS was shown to be statistically superior to ACLS for most of the cases. With added uncorrelated spectral noise, both methods performed comparably. Using 50 calibration samples with simulated correlated spectral noise, PLS showed an advantage in 3 out of 9 cases, but the advantage dropped to 1 out of 9 cases with 25 calibration samples. For cases with different noise distributions between calibration and validation, ACLS predictions were statistically better than PLS for two of the four components. Also, when experimentally derived correlated spectral error was added, ACLS gave better predictions that were statistically significant in 15 out of 24 cases simulated. On data sets with nonuniform noise, neither method was statistically better, although ACLS usually had smaller standard errors of prediction (SEPs). The varying results emphasize the need to use realistic simulations when making comparisons between various multivariate calibration methods. Even when the differences between the standard error of predictions were statistically significant, in most cases the differences in SEP were small. This study demonstrated that unlike CLS, ACLS is competitive with PLS in modeling nonlinearities in spectra without knowledge of all the component concentrations. This competitiveness is important when maintaining and transferring models for system drift, spectrometer differences, and unmodeled components, since ACLS models can be rapidly updated during prediction when used in conjunction with the prediction augmented classical least squares (PACLS) method, while PLS requires full recalibration.

Algorithms↗
PubMed1998

Early prediction of poor response in acute asthma patients in the emergency department.

STUDY OBJECTIVES: The aim of this study was to develop an acute asthma index for utilization in the early differentiation between patients with poor and good therapeutic response in the emergency department (ED) setting. SETTING: The ED of a large tertiary-care hospital in Montevideo, Uruguay. PATIENTS AND DESIGN: The study included 145 consecutive adult patients (mean age [+/- SEM], 33.4+/-0.97) who presented to an ED (analysis sample). The inclusion criteria were: (1) age between 18 and 50 years; (2) a peak expiratory flow rate (PEFR) or FEV1 below 35% of predicted; and (3) no history of chronic cough or cardiac, hepatic, renal, or other medical disease. INTERVENTIONS: All patients were treated with salbutamol delivered by metered-dose inhaler into a spacer device in four puffs actuated at 10-min intervals. The protocol involved 3 h of this treatment. After that time, patients with poor response received hydrocortisone, 500 mg IV. The outcome was defined as the FEV1 after 3 h of treatment in a dichotomized form: < or =45% of predicted = poor response, and >45% of predicted = good response. RESULTS: Biserial correlations between different variables and the outcome showed that PEFR as percent of predicted and PEFR variation over baseline, both measured at 30 min, were the most important predictors of a good or poor response after 3 h of treatment. Next, we developed an acute asthma index using these predictive measures. A comparison of index sensitivity, specificity, predictive values, and the area under the receiver operating characteristic (ROC) curve across different cutoff scores indicates that a score of 4 results in the least error of classification (sensitivity = 0.79; specificity = 0.96; area under the ROC curve = 0.87; positive predictive value = 0.94; and negative predictive value = 0.86). To validate the developed index, we prospectively studied a second sample of 77 consecutive patients (mean age 32.6+/-1.22 years) who presented for treatment of acute asthma (validation sample). The area under the ROC for the analysis sample was not greater than the validation sample area (p = 0.24). Thus, the validation sample showed similar levels of sensitivity and specificity, positive and negative predictive values, and area under the ROC curve (0.80, 0.88, 0.85, 0.84, and 0.89, respectively), indicating the stability of the model. CONCLUSIONS: The study suggested the predictive accuracy of a two-item bedside index. This acute asthma index provides a tool for assessing acute asthma severity using objective criteria easily accessible to the ED physician.

Acute Disease↗
PubMed2000

Predicting the therapeutic alliance in alcoholism treatment.

OBJECTIVE: Prediction of the therapeutic alliance in alcoholism treatment (as rated by the client and by the therapist) was examined in light of a range of potentially relevant factors, including client demographics, drinking history, current drinking, current psychosocial functioning and therapist demographics. METHOD: The data were gathered in Project MATCH. The present analyses were based on data from 707 outpatients and 480 aftercare clients assigned to one of the three Project MATCH treatments. Potential predictor variables were evaluated by first examining bivariate linear relationships between the variables and ratings of the alliance, and then entering blocks of these predictors into multiple linear regression equations with alliance ratings as the dependent variables. All analysis incorporated adjustments for the nonindependence of ratings pertaining to clients seen by the same therapist. RESULTS: In simple regressions evaluating bivariate relationships, outpatients' ratings of the alliance were positively predicted by client age, motivational readiness to change, socialization, level of perceived social support and therapist age, and were negatively predicted by client educational level, level of depression, and meaning seeking. Therapist ratings in the outpatient sample were positively predicted by the client being female and by level of overall alcohol involvement, severity of alcohol dependence, negative consequences of alcohol use, and readiness to change. Among aftercare clients, ratings of the alliance were positively predicted by readiness to change, socialization and social support, and were negatively predicted by level of depression. Therapist ratings of the alliance in the aftercare sample were positively predicted by the client being female and therapist educational level, and were negatively predicted by pretreatment drinks per drinking day. Of the variables having significant bivariate relationships with alliance scores, only a few were identified as significant predictors in multiple regression equations. Among outpatients, client age and motivational readiness to change remained positive predictors and client education a negative predictor of client ratings of the alliance, while client gender remained a significant predictor of therapist ratings. Among aftercare clients, readiness to change and level of depression remained significant predictors of client ratings, while none of the variables remained a significant predictor of therapist ratings. CONCLUSIONS: While the data indicate that several client variables predict the nature of both the client's and therapist's perception of the therapeutic alliance, the significant relationships are of modest magnitude, and few variables remain predictive after controlling for causally prior variables. The strongest relationship identified in both the outpatient and aftercare samples is that between clients' motivational readiness to change and their ratings of the alliance.

Adult↗
PubMed2000

Simulated predictions of insect phenological events made by using mean and median functional lower developmental thresholds.

A computer-simulated study was conducted to determine whether mean or median functional lower developmental thresholds and required degree-days were superior for predicting the dates on which insect phenological events occurred. In addition, these simulations allowed us to determine if the type of year (weatherwise) influenced those predictions. Results indicated that when median functional lower developmental thresholds and required degree-days were used their predictions were closer to the dates on which the phenological events occurred than were predictions that were made using mean thresholds and required degree-days. Also, the predictions of phenological events made when using median functional lower developmental thresholds and required degree-days were not strongly influenced by the type of year. However, the influence of type of year was quite strong when predictions were made when using mean thresholds and required degree-days. The variability in predictions that were made when using median functional lower developmental thresholds and required degree-days was greater than the variability in predictions that were made when using mean thresholds and required degree days. However, the increased variability was caused by many predictions being closer to, rather than farther from, the actual dates on which the phenological events occurred. Based on these findings, we suggest that median functional lower developmental thresholds, along with median required degree-days, be considered for use when predicting insect phenological events in the field.

Animals↗
PubMed2000

Computational methods for protein secondary structure prediction using multiple sequence alignments.

Efforts to use computers in predicting the secondary structure of proteins based only on primary structure information started over a quarter century ago [1-3]. Although the results were encouraging initially, the accuracy of the pioneering methods generally did not attain the level required for using predictions of secondary structures reliably in modelling the three-dimensional topology of proteins. During the last decade, however, the introduction of new computational techniques as well as the use of multiple sequence information has lead to a dramatic increase in the success rate of prediction methods, such that successful 3D modelling based on predicted secondary structure has become feasible [e.g., Ref 4]. This review is aimed at presenting an overview of the scale of the secondary structure prediction problem and associated pitfalls, as well as the history of the development of computational prediction methods. As recent successful strategies for secondary structure prediction all rely on multiple sequence information, some methods for accurate protein multiple sequence alignments will also be described. While the main focus is on prediction methods for globular proteins, also the prediction of trans-membrane segments within membrane proteins will be briefly summarised. Finally, an integrated iterative approach tying secondary structure prediction and multiple alignment will be introduced [5].

Algorithms↗
PubMed1994

Technical note: detection of bias in genetic predictions.

The theoretical development of a procedure to detect bias in genetic predictions is presented. The procedure is based on the expectation of three statistics. These statistics detect bias by identifying systematic, unexpected change in subsequent analyses. Expectations of the following statistics were obtained: linear correlation coefficient between subsequent predictions, linear regression of recent (more accurate) on previous (less accurate) genetic prediction, and variance of the genetic prediction difference (recent minus previous genetic prediction). Deviations from these expectations can be used to indicate bias. The covariance between subsequent BLUP of genetic value is shown to equal the variance of the early estimate, implying that the expected value of the regression of recent on previous genetic prediction equals 1 regardless of the distribution of the observations and predictions. Also, the expected value of the linear correlation coefficient between subsequent genetic predictions equals the square root of the ratio of the means of the square of accuracy values. The expected value of the variance of the genetic prediction difference was shown to be equal to the difference between prediction error variances.

Analysis of Variance↗
PubMed2006

Examination of potential methods to predict pulmonary arterial pressure score in yearling beef cattle.

Susceptibility of beef cattle to high altitude disease (HAD) is of major importance to economic and genetic selection on high elevation ranches. However, currently the best indicator of HAD susceptibility is the pulmonary arterial pressure (PAP) test, a test with high cost and invasive nature. Therefore, 2 experiments were undertaken to determine whether emerging technologies that predict blood components could be used to predict the PAP score in yearling Angus cattle. In Exp. 1, 39 yearling Angus bulls were used to determine if a relationship existed between PAP score and 10 blood components provided by a hemogram using whole blood or oxygen saturation as predicted by pulse oximetry in nonanesthetized cattle measured rectally or orally. Three of the hemogram values (packed cell volume, hemoglobin concentration, and red cell distribution width) were correlated (P < 0.10) with the PAP score. Prediction equations for PAP score were generated using the hemogram values and resulted in R2 values of 0.375 and 0.305 for the regression model using all of values and the best 2-variable model, respectively. Pulse oximetry was able to provide oxygen saturation predictions rectally or orally; however, the predicted values were not correlated with the PAP score (P > 0.10) or with each other (P > 0.10). In Exp. 2, 84 yearling Angus cattle (62 bulls, 22 heifers) were used to evaluate the ability of a portable clinical analyzer to predict the PAP score using 11 blood components from a sample of whole blood evaluated at the processing chute. The portable clinical analyzer was able to provide values for all of the 11 blood components; however, none of the predicted values were correlated with the PAP score (P > 0.10). In these preliminary experiments, 3 blood component values provided via the hemogram were the only variables both correlated with the PAP score and able to contribute to the development of a useful PAP prediction equation that could reduce the cost of traditional measures of HAD susceptibility. Future research is needed to determine whether additional blood components or emerging blood analysis technologies are able to accurately predict the PAP score in beef cattle.

Altitude Sickness↗
PubMed1997

Evaluation of alternative equations for prediction of intake for Holstein dairy cows.

Six prediction equations for dry matter intake (DMI) were evaluated for accuracy with independent data. The equations were selected based on ease of parameter measurement and practical on-farm use. The equations were assessed for accuracy of predicting individual weekly DMI for primiparous (n = 105) and multiparous (n = 136) cows; three-fourths of these cows were supplemented with a sustained-release form of bovine somatotropin (bST). Large variations in accuracy were identified across the six prediction equations for effects of parity and bST. Prediction accuracy of all equations for cows in wk 1 to 24 of lactation was better for primiparous cows than for multiparous cows. Precision of prediction equations was poor for cows in wk 8 through 12 of lactation and for cows in > 40 wk of lactation. The equation for DMI with the best accuracy measured by a low total lactation mean square prediction error was the modified equation of the National Research Council: DMI (kilograms per day) = -0.293 + 0.372 x fat-corrected milk (kilograms per day) + 0.0968 x body weight 0.75 (kilograms). However, the overall mean bias (predicted minus observed) of the prediction of weekly DMI of a single cow was high for all equations, including the modified equation of the National Research Council. For wk 2, 4, 8, 10, and 20 of lactation, the mean bias for the modified equation was +6, +3.4, -1.3, -2.1, and -2.8 kg/d. The accuracy of prediction was lower for cows treated biweekly with bST. High yielding cows and cows treated biweekly with bST had higher milk yields in relation to body weight, and standardized prediction equations for DMI were less accurate.

Animal Nutritional Physiological Phenomena↗
PubMed1981

Improved confidence of outcome prediction in severe head injury. A comparative analysis of the clinical examination, multimodality evoked potentials, CT scanning, and intracranial pressure.

An analysis of clinical signs, singly or in combination, multimodality evoked potentials (MEP's), computerized tomography scans, and intracranial pressure (ICP) data was undertaken prospectively in 133 severely head-injured patients to ascertain the accuracy, reliability, and relative value of these indicants individually, or in various combinations, in predicting one of two categories of outcome. Erroneous predictions, either falsely optimistic (FO) or falsely pessimistic (FP), were analyzed to gain pathophysiological insights into the disease process. Falsely optimistic predictions occurred because of unpredictable complications, whereas FP predictions were due to intrinsic weakness of the indicants as prognosticators. A combination of clinical data, including age, Glasgow Coma Scale (GCS) score, pupillary response, presence of surgical mass lesions, extraocular motility, and motor posturing predicted outcome with 82% accuracy, 43% with over 90% confidence. Nine percent of predictions were FO and 9% FP. The GCS score alone was accurate in 80% of predictions, but at a lower level of confidence (25% at the over-90% level), with 7% FO and 13% FP. Computerized tomography and ICP data in isolation proved to be poor prognostic indicants. When combined individually with clinical data, however, they increased the number of predictions made with over 90% confidence to 52% and 55%, respectively. Data from MEP's represented the most accurate single prognostic indicant, with 91% correct predictions, 25% at the over-90% confidence level. There were no FP errors associated with this indicant. Supplementation of the clinical examination with MEP data yielded optimal prognostic power, an 89% accuracy rate, with 64% over the 90% confidence level and only 4% FP errors. The clinical examination remains the strongest basis for prognosticating outcome in severe head injury, but additional studies enhance the reliability of such predictions.

Adult↗
PubMed1990

Predicting the course of AIDS in Australia.

There have been urgent demands for knowledge about the epidemic of the acquired immunodeficiency syndrome (AIDS) in Australia. Accurate predictions are important for efficient allocation and planning of limited health-care resources. Ideal data for this purpose would be reliable knowledge of the past and present incidence of human immunodeficiency virus (HIV) infection. However, since the incidence of the infection is unknown predictions can only be based on historical data of the incidence of AIDS. In this article, we show the limitations of such predictions by examining a broad range of mathematical models that successfully track the observed data (1187 cases diagnosed to December 31, 1988). In addition, we describe a simple method for prediction in subgroups where the numbers of cases observed so far are small. Four models representing different forms of departure from the simple exponential model provide the best fits to the Australian AIDS data. Regional variability and a possible effect resulting from the introduction of zidovudine were incorporated into the models. Significant regional variability in the course of the epidemic was observed between New South Wales, Victoria and the rest of the country. For Australia as a whole, the doubling time changed from less than one year before mid 1987 to more than two years after this time. Model fits were improved by fitting the models to just the four years of data from 1985. The models give comparable predictions for the first year (1989) of around 600 new cases. However, by 1993 the predictions vary considerably, ranging from 500 to 2300 new cases. It is predicted that between 3100 and 6700 cases are likely to be diagnosed in Australia between 1989 and 1993. The results from the subgroup prediction demonstrate that when the observed number of cases is small, then the range of predictions for a future time interval is very wide. For reliable long-term predictions that are necessary for public health planning, basic information on the past and present incidence of HIV infection is urgently needed.

Acquired Immunodeficiency Syndrome↗
PubMed1999

Case-based prediction of survival in colorectal cancer patients.

OBJECTIVE: To develop an approach to the prediction of survival in patients with colorectal cancer using nearest neighbor analysis and case-based reasoning. STUDY DESIGN: A total of 216 patients with full clinicopathologic records and five-year follow-up were the subjects of this study. They were divided into a core database of 162 cases and a test group of 54 cases, with follow-up on all patients. When the patient was still alive at the end of the follow-up period, censored survival time was used. For each of the test cases, the four closest neighbors from the database were retrieved and their median survival time recorded and used as the predicted estimate of survival. Case matching was based on a Euclidean multivariate distance measure for the three best predictor variables: patient age, Dukes stage and tubule configuration. Cases with the smallest distance from the test case were considered to be the most similar. The predicted survival times for the test cases were compared with the actual, observed survival in the test cases to determine the success of this approach. RESULTS: The results showed reasonable concordance between observed and predicted survival figures, although there was a large degree of spread. Classification of cases into < or = 60 and > 60 months' survival showed a correct classification rate of 63%. For the prediction of survival time, the distribution of differences between observed and predicted survival times for the uncensored test cases had a median value of--5 months but also showed a wide dispersion of values. Correlation of observed and predicted survival times, while not reaching statistical significance at P < .05, did show a strong positive association. CONCLUSION: Case-based approaches to the prediction of survival times in cancer patients are important. The results of the current study illustrate the difficulties in applying this approach to survival data and highlight the complexity of patient information and the inability to accurately predict patient outcome on a small subset of clinicopathologic features. While extensive work needs to be carried out to improve prediction power, this study illustrates the potential for case-based analyses. The ability to retrieve feature-matched cases from hospital patient databases has clear, independent advantages in patient management, but the ability to provide reliable, targeted prognostic estimates on individual cases should be a common goal in medical research.

Adult↗
PubMed2000

Clinical versus mechanical prediction: a meta-analysis.

The process of making judgments and decisions requires a method for combining data. To compare the accuracy of clinical and mechanical (formal, statistical) data-combination techniques, we performed a meta-analysis on studies of human health and behavior. On average, mechanical-prediction techniques were about 10% more accurate than clinical predictions. Depending on the specific analysis, mechanical prediction substantially outperformed clinical prediction in 33%-47% of studies examined. Although clinical predictions were often as accurate as mechanical predictions, in only a few studies (6%-16%) were they substantially more accurate. Superiority for mechanical-prediction techniques was consistent, regardless of the judgment task, type of judges, judges' amounts of experience, or the types of data being combined. Clinical predictions performed relatively less well when predictors included clinical interview data. These data indicate that mechanical predictions of human behaviors are equal or superior to clinical prediction methods for a wide range of circumstances.

Decision Making, Computer-Assisted↗