SEARCH · PubMed Health

Results for “Predictive modeling”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

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

At least 559 records · Page 31Linked to original sources

PubMed2006

Predictive modelling of human walking over a complete gait cycle.

An inverse dynamics multi-segment model of the body was combined with optimisation techniques to simulate normal walking in the sagittal plane on level ground. Walking is formulated as an optimal motor task subject to multiple constraints with minimisation of mechanical energy expenditure over a complete gait cycle being the performance criterion. All segmental motions and ground reactions were predicted from only three simple gait descriptors (inputs): walking velocity, cycle period and double stance duration. Quantitative comparisons of the model predictions with gait measurements show that the model reproduced the significant characteristics of normal gait in the sagittal plane. The simulation results suggest that minimising energy expenditure is a primary control objective in normal walking. However, there is also some evidence for the existence of multiple concurrent performance objectives.

Biomechanical Phenomena↗

PubMed1998

Determination and validation of a predictive model for Clostridium difficile diarrhea in hospitalized oncology patients.

BACKGROUND: Clostridium difficile colitis in the cancer patient receiving chemotherapy is a frequent cause of morbidity which may prolong hospitalization. Techniques for identifying infection often delay the initiation of therapy. PATIENTS AND METHODS: In this retrospective case-control analysis, we identified predictors for C. difficile-associated diarrhea in 29 patients hospitalized from 1988 to 1993 on a hematologic malignancy/bone marrow transplant unit (hospital A). We then validated our model with 58 C. difficile cases and 74 controls admitted to an oncology unit from a different institution (hospital B). RESULTS: We found that low intensity of chemotherapy (P < 0.001), lack of parenteral vancomycin use (P = 0.03) and hospitalization within the past two months (P = 0.05) were independently predictive of C. difficile colitis by multivariate analysis. These variables were weighted for predictive capability using a receiver operator characteristic score; low intensity chemotherapy was assigned two points, lack of parenteral vancomycin received one point and prior hospitalization one point (P < 0.001 by chi 2 for trend). The receiver operating characteristic (ROC) curve areas were 0.78 for patients at hospital A and 0.70 at hospital B indicating moderate drop off in discrimination. Compared to hospital A patients, hospital B patients hospitalized between 1989 and 1994 were more often women (P = 0.04), received less systemic vancomycin (P = 0.01), were less frequently neutropenic (P < 0.05), and received less intense chemotherapy regimens (P < 0.05). Despite these differences in demographics in patients between these institutions, our predictive model was validated in hospital B patients (P = 0.02 by chi 2 for trend). CONCLUSIONS: The results of this study may help clinicians predict the risk of C. difficile disease in the hospitalized immunocompromised oncology patient and may help guide empiric therapy while awaiting results of stool toxin assays.

Case-Control Studies↗

PubMed1984

Cross validation of model predicting (ir)regularity of dental attendance.

A theoretical model developed for the prediction of (ir)regular dental attendance was evaluated with loglinear models (logit approach). The model found is identical to the model developed previously. The prediction of (ir)regularity, however, proved to be unsatisfactory. The disappointing result is attributed to the difficulty of measuring the predictors adequately, and to the possibility that additional variables might be needed.

Dental Care↗

PubMed2001

Prediction of hepatic metabolic clearance: comparison and assessment of prediction models.

OBJECTIVE: To perform a comparative quantitative evaluation of the prediction accuracy for human hepatic metabolic clearance of 5 different mathematical models: allometric scaling (multiple species and rat only), physiologically based direct scaling, empirical in vitro-in vivo correlation, and supervised artificial neural networks. METHODS: The mathematical prediction models were implemented with a publicly available dataset of 22 extensively metabolised compounds and compared for their prediction accuracy using 3 quality indicators: prediction error sum of squares (PRESS), r2 and the fold-error. RESULTS: Approaches such as physiologically based direct scaling, empirical in vitro-in vivo correlation and artificial neural networks, which are based on in vitro data only, yielded an average fold-error ranging from 1.64 to 2.03 and r2 values greater than 0.77, as opposed to r2 values smaller than 0.44 when using allometric scaling combining in vivo and in vitro preclinical data. The percentage of successful predictions (less than 2-fold error) ranged from 55% (rat allometric scaling) to between 64 and 68% with the other approaches. CONCLUSIONS: On the basis of a diverse set of 22 metabolised drug molecules, these studies showed that the most cost-effective and accurate approaches, such as physiologically based direct scaling and empirical in vitro-in vivo correlation, are based on in vitro data alone. Inclusion of in vivo preclinical data did not significantly improve prediction accuracy; the prediction accuracy of the allometric approaches was at the lower end of all methods compared.

Animals↗

PubMedDate not supplied

The importance of the prediction model in the validation of alternative tests.

An overview is presented of the validation process adopted by the European Centre for the Validation of Alternative Methods, with particular emphasis on the central role of the prediction model (PM). The development of an adequate PM is considered to be just as important as the development of an adequate test system, since the validity of an alternative test can only be established when both components (the test system and the PM) have successfully undergone validation. It is argued, however, that alternative tests and their associated PMs do not necessarily need to undergo validation at the same time, and that retrospective validation may be appropriate when a test system is found to be reliable, but the case for its relevance remains to be demonstrated. For an alternative test to be considered "scientifically valid", it is necessary for three conditions to be fulfilled, referred to here as the criteria for scientific relevance, predictive relevance, and reliability. A minimal set of criteria for the acceptance of any PM is defined, but it should be noted that required levels of predictive ability need to be established on a case-by-case basis, taking into account the inherent variability of the alternative and in vivo test data. Finally, in view of the growing shift in emphasis from the use of stand-alone alternative tests to alternative testing strategies, the importance of making the PM an integral part of the testing strategy is discussed.

Animal Testing Alternatives↗

PubMed2006

Risk assessment of human myelotoxicity of anticancer drugs: a predictive model and the in vitro colony forming unit granulocyte/macrophage (CFU-GM) assay.

Myelotoxicity is one of the major limitations to the use of anticancer drugs. It is desirable to evaluate human myelotoxicity before a Phase I study, however, this is difficult because of the differences in susceptibility between humans and animals. The purpose of this study was to establish a reliable method to predict the human maximum tolerated dose (MTD) of five camptothecin derivatives: SN-38, DX-8951f, topotecan (TPT), 9-aminocamptothecin (9-AC), and camptothecin (CAM). The myelotoxicity of camptothecin derivatives was evaluated on bone marrow from mice, dogs, and humans using a 14-day colony-forming unit-granulocyte/macrophage (CFU-GM) assay to determine the 50%, 75%, and 90% inhibitory concentration values (IC50, IC75, and IC90, respectively). Then, using human and murine IC90 values for myelotoxicity of these compounds, in vivo toxicological data, and pharmacokinetic parameters (data referred to the literature), human MTDs were predicted retrospectively. The mechanism-based prediction model which is proposed uses the in vitro CFU-GM assay and in vivo parameters on the basis of free fraction of area under the concentration-curve (AUC) at the MTD (r2 = 0.887) and suggests that the human MTDs were well predicted for the five camptothecin derivatives by this model rather than by other models. The application of this model for in vitro hematotoxicology could be very useful in the development of new anticancer agents.

Animals↗

PubMed2002

Staphylococcus aureus bacteremia: comparison of two periods and a predictive model of mortality.

Staphylococcus aureus is an important pathogen causing bacteremia, primarily affecting hospitalized patients. We studied the epidemiology of S. aureus bacteremia, comparing two periods (early and mid 1990s) and developed a predictive model of mortality. A nested case-control was done. All 251 patients over 14 years old with positive blood cultures for S. aureus were selected. MRSA (methicillin resistant S. aureus) was isolated in 63% of the cases. When comparing the two periods MRSA community-acquired bacteremia increased from 4% to 16% (p=0.01). There was no significant difference in the mortality rate between the two periods (39% and 33%, p=0.40). Intravascular catheters provoked 24% of the cases of bacteremia and were associated with the lowest rate of mortality. In a logistic regression analysis, three variables were associated with death: septic shock, source of bacteraemia and resistance to methicillin. The probability of dying among patients with MRSA and those with methicillin sensitive S. aureus bacteraemia ranged from 10% to 90% and from 4% to 76%, respectively, depending on the source of the bacteraemia and the occurrence of septic shock. The MRSA found in Brazil may be a particularly virulent strain.

Adult↗

PubMed1990

The health belief model: predicting compliance and dropout in cardiac rehabilitation.

We investigated the health belief model and the health locus of control constructs as predictors of group membership (compliers or dropouts) with cardiac rehabilitation and whether they added predictive utility to routinely assessed patient demographics and health behaviors. Questionnaires were completed on entry into the study by 120 patients with coronary artery disease, and by the end of the 6 month program there were 58 compliers and 62 dropouts. Discriminant function analyses were carried out to determine prediction of group membership. The health belief model predicted group membership 64.6% of the time, explaining 5.2% of the variance. Demographics, health behaviors, and health belief model factors accounted for 21.1% of the variance between compliers and total dropouts with group membership correctly predicted 74.4% of the time; avoidable and unavoidable dropout was correctly predicted 84.2% of the time with 56.9% of the variance explained. Health locus of control did not distinguish between compliers and dropouts. The addition of the health belief model provided additional information about compliance with cardiac rehabilitation beyond that explained by demographic and health behavior variables alone, particularly when predicting avoidable/unavoidable dropout.

Attitude to Health↗

PubMed2002

Basal glycogenolysis in mouse skeletal muscle: in vitro model predicts in vivo fluxes.

A previously published mammalian kinetic model of skeletal muscle glycogenolysis, consisting of literature in vitro parameters, was modified by substituting mouse specific Vmax values. The model demonstrates that glycogen breakdown to lactate is under ATPase control. Our criteria to test whether in vitro parameters could reproduce in vivo dynamics was the ability of the model to fit phosphocreatine (PCr) and inorganic phosphate (Pi) dynamic NMR data from ischemic basal mouse hindlimbs and predict biochemically-assayed lactate concentrations. Fitting was accomplished by optimizing four parameters--the ATPase rate coefficient, fraction of activated glycogen phosphorylase, and the equilibrium constants of creatine kinase and adenylate kinase (due to the absence of pH in the model). The optimized parameter values were physiologically reasonable, the resultant model fit the [PCr] and [Pi] timecourses well, and the model predicted the final measured lactate concentration. This result demonstrates that additional features of in vivo enzyme binding are not necessary for quantitative description of glycogenolytic dynamics.

Animals↗

PubMed2005

Modeling, identification and nonlinear model predictive control of type I diabetic patient.

Patients with type I diabetes nearly always need therapy with insulin. The most desirable treatment would be to mimic the operation of a normal pancreas. In this work a patient affected with this pathology is modeled and identified with a neural network, and a control strategy known as Nonlinear Model Predictive Control is evaluated as an approach to command an insulin pump using the subcutaneous route. A method for dealing with the problems related with the multiple insulin injections simulation and a multilayer neural network identification of the patient model is presented. The controller performance of the proposed strategy is tested under charge and reference disturbances (setpoint). Simulating an initial blood glucose concentration of 250 mg/dl a stable value of 97.0 mg/dl was reached, with a minimum level of 76.1 mg/dl. The results of a simulated 50 g oral glucose tolerance test show a maximum glucose concentration of 142.6 mg/dl with an undershoot of 76.0 mg/dl. According to the simulation results, stable close-loop control is achieved and physiological levels are reached with reasonable delays, avoiding the undesirable low glucose levels. Further studies are needed in order to deal with noise and robustness aspects, issues which are out of the scope of this work.

Blood Glucose↗

PubMed1991

Drug resistance-reversal strategies: comparison of experimental data with model predictions.

We previously developed a mathematical model to describe the emergence and dynamic growth of a drug-resistant subpopulation in a tumor. In the present study, our objective was to test the model's ability to mimic two strategies for reversal of drug resistance. We present data from one in vitro cell proliferation assay with drug-resistant LS174T human colon carcinoma variants and one in vivo assay of survival after treatment of female (C57BL/6 x DBA/2)F1 mice inoculated with doxorubicin-resistant P388/ADR leukemia cells. The in vitro assay examined the effects of inhibiting the biosynthesis of glutathione in cells resistant to alkylating agents or cisplatin. The in vivo assay compared the effects on cell survival of low-level continuous infusion versus high-intensity bolus dosing, with or without coadministration of the drug efflux pump blocker verapamil. Results in vitro and in vivo were comparable for qualitative accuracy and predictability to results with the model. Both the in vitro study and the model showed that, for resistant cells with high levels of glutathione, short-term cell survival was dose dependent and that even high doses of drug did not eliminate all of these cells. Addition of an inhibitor of glutathione biosynthesis did, however, augment elimination of the resistant cells. Resistant cells with low levels of glutathione could be eliminated with high drug doses or coadministration of drug and a glutathione synthesis inhibitor. In vivo, coadministration of doxorubicin with verapamil increased animal survival when either continuous infusion or bolus dosing regimens were used. The effectiveness of the blocker is crucial; when a partially (50%) effective blocker is used, continuous infusion achieves better elimination of resistant cells, but a completely (100%) effective blocker is efficacious in both dosing scenarios. Careful interpretation of these findings is necessary because the pharmacokinetics of drug in the small populations of cells in the model are not easily extrapolated to those in large tumors. This model may be useful in determining resistance mechanisms, their levels of effectiveness, and concentrations of compounds required at target sites to overcome them.

Animals↗

PubMedDate not supplied

Forecasting medical work at mass-gathering events: predictive model versus retrospective review.

INTRODUCTION: Mass-gathering events are dynamic and challenge traditional medical management systems. To improve the system for the provision of first aid at mass-gathering events, an evaluation of two models that assist in forecasting the number of patients presenting for first-aid services was conducted. METHOD: A prospective evaluation of a recurrent, mass-gathering event was undertaken comparing predicted patient presentations and ambulance transfers generated by a predictive model developed by Arbon et al and a retrospective review of seven years of historical, event data as described by Zeitz et al. RESULTS: Patient presentation rate (per 1,000 patrons) for this event was 1.6 and the transport to hospital rate (per 1,000 patrons) was 0.07. The retrospective review closely predicted the actual overall attendance. Both methods forecast the number of patients presenting on a daily basis. The prediction proved to be more accurate, on a day-by-day basis, using the Zeitz method. CONCLUSION: The Arbon method is particularly useful for events where there is no or limited information about previous medical work. Retrospective review of data generated from specific events (Zeitz method) considers the unique and individual variability that can occur from event to event and is more accurate at predicting patient presentations when the data are available. Both methods have the potential to be used more frequently to adequately and efficiently plan for the resources required for specific events.

Ambulances↗

PubMed2002

Reactive transport in porous media: a comparison of model prediction with laboratory visualization.

Groundwater transport models that accurately describe spreading of nonreactive solutes in an aquifer can poorly predict concentrations of reactive solutes. The dispersive term in the advection-dispersion equation can overpredict pore-scale mixing, and thereby overpredict homogeneous chemical reaction. We quantified this experimentally by imaging instantaneous colorimetric reactions between solutions of aqueous CuSO4 and EDTA4- within a 30-cm long translucent chamber packed with cryolite sand that closely matched the optical index of refraction of water. A charge-coupled device camera was used to quantify concentrations of blue CuEDTA2- within the chamber as it was produced by mixing of the two reactants at different flow rates. We compared these experimental results with a new analytic solution for instantaneous bimolecular reaction coupled with advection and dispersion of the product and reactants. For all flow rates, the concentrations of CuEDTA2- recorded in the experiments were about 20% less than predicted by the analytic solution, thereby demonstrating that models assuming complete mixing at the pore scale can overpredict reaction during transport.

Chelating Agents↗

PubMed1999

A porcine model predicts that a can-opener capsulotomy can be done safely in pediatric patients.

PURPOSE: The purpose of this study was to compare the strength and safety of a continuous circular capsulorhexis (CCC) with a can-opener capsulotomy (COC) in a porcine model that closely resembles the high elasticity of the human pediatric lens capsule. METHODS: COCs (N = 47) and CCCs (N = 102) were performed inside the anterior chamber of fresh pig eyes, and any uncontrolled tears were noted. The circumference of the initial opening was measured in 18 COCs and 13 CCCs. After the opening was stretched to the point of rupture, the circumference was measured again. The ratio of the circumference at rupture to the initial circumference, minus one, was used as a measure of the maximal capsular strain. RESULTS: The can-opener technique produced a smooth round opening. One of the COCs (2.1%) and 23 of the CCCs (22.5%) had uncontrolled tears (chi2, P<.001). The mean maximal strain for COCs was 46.7% (SE, 8.3%) and for the CCCs, 47.7% (SE, 9.9%). This difference was statistically not significant (P = .93 by Student's t test). CONCLUSIONS: The porcine capsule is more reliably opened with fewer uncontrolled tears by a COC than by a CCC. The porcine model predicts that pediatric capsules can be opened safely with a COC.

Animals↗

PubMed2006

A multivariate prediction model for microarray cross-hybridization.

BACKGROUND: Expression microarray analysis is one of the most popular molecular diagnostic techniques in the post-genomic era. However, this technique faces the fundamental problem of potential cross-hybridization. This is a pervasive problem for both oligonucleotide and cDNA microarrays; it is considered particularly problematic for the latter. No comprehensive multivariate predictive modeling has been performed to understand how multiple variables contribute to (cross-) hybridization. RESULTS: We propose a systematic search strategy using multiple multivariate models [multiple linear regressions, regression trees, and artificial neural network analyses (ANNs)] to select an effective set of predictors for hybridization. We validate this approach on a set of DNA microarrays with cytochrome p450 family genes. The performance of our multiple multivariate models is compared with that of a recently proposed third-order polynomial regression method that uses percent identity as the sole predictor. All multivariate models agree that the 'most contiguous base pairs between probe and target sequences,' rather than percent identity, is the best univariate predictor. The predictive power is improved by inclusion of additional nonlinear effects, in particular target GC content, when regression trees or ANNs are used. CONCLUSION: A systematic multivariate approach is provided to assess the importance of multiple sequence features for hybridization and of relationships among these features. This approach can easily be applied to larger datasets. This will allow future developments of generalized hybridization models that will be able to correct for false-positive cross-hybridization signals in expression experiments.

Algorithms↗

PubMed1998

Predictive model of weapon carrying among urban high school students: results and validation.

OBJECTIVES: The purpose of this study was to identify the behavioral, psychosocial, and demographic predictors of self-reported weapon carrying among secondary school students who attend urban public schools. METHODS: Self-reported weapon carrying was measured in a schoolwide anonymous health survey conducted in two demographically comparable high schools in 1992, in Boston, Massachusetts. Indicators of self-perception, depression, stressful life events, and adolescent risk behaviors of substance use and sexual behavior, along with self-reported weapon carrying, were measured. The students in both schools were racially heterogeneous, with the majority of about 80% from black or Hispanic backgrounds. A predictive model was developed using a forward stepwise logistic regression model in one inner-city high school, and tested in a second high school. RESULTS: Self-reported lifetime weapon carrying was 32% overall. The major predictors of weapon carrying among urban secondary school students are a combination of demographic, psychosocial, behavioral, and school-related factors. This analysis indicates consistency in eight markers predictive of weapon carrying: lower age, male gender, regular marijuana use, sexual experience, having witnessed a crime, having skipped school, suicidal ideation, and having hit or "beat up" someone. Race parental education, and family composition were not significant predictors. Significant predictors of weapon carrying were marijuana use and sexual experience, each of which was consistently high in both schools. CONCLUSIONS: The model-building and validation presented in this study provide empirical evidence for three important conclusions. First, weapon carrying is associated with multiple and interrelated factors which include demographic, psychosocial, behavioral, and school-related characteristics of high school-age adolescents. Second, students with more risk factors are more likely to carry a weapon, suggesting that the variables are independent markers. Third, this study identified marijuana use and being sexually experienced as both highly predictive of weapon carrying. Implications of this study for prevention point to the need for comprehensive multidisciplinary services in high school that include mental health counseling as well as health education efforts aimed at behavior change.

Adolescent↗

PubMed1999

Comparison of viral load and human leukocyte antigen statistical and neural network predictive models for the rate of HIV-1 disease progression across two cohorts of homosexual men.

We compared the performance of HIV-1 RNA and models based on human leukocyte antigen (HLA) in predicting the rate of HIV-1 disease progression using both linear regression and neural network models across two different cohorts of homosexual men. In all, 139 seroconverters from the Multicenter AIDS Cohort Study were used as the training set and 97 seroconverters from the District of Columbia Gay (DCG) cohort were used for validation to assess the generalizability of trained predictive models. Both viral load and HLA markers were strongly predictive of disease progression (p < .0001 and p = .001, respectively), with viral load superior to HLA (change in -2 log likelihood [-2LL] 26.7 and 10.2, respectively, in proportional hazards models). Consideration of both HLA markers and viral load offered no significant predictive advantage over viral load alone in most cases; however, HLA-based predictions obtained from neural networks modeling improved the discrimination among patients with high viral load (p = .02). Viral load, HLA scores, and rapid disease progression were moderately correlated (p < .01 for all three pairs of these variables). The median viral load was 10(3.70) copies/ml among DCG patients who had more favorable than unfavorable HLA markers and 10(4.66) copies/ml among patients with more unfavorable than favorable HLA markers. Viral load is a simpler, stronger predictor of disease progression than early developed HLA models, but neural network methods and further refined HLA models may offer additional prognostic information, especially for rapid progressors. The correlation between viral load and HLA markers suggests a possible HLA effect on setting viral load levels.

Cohort Studies↗

PubMed2004

The peak atrioventricular pressure gradient to transmitral flow relation: kinematic model prediction with in vivo validation.

Physiologists and cardiologists estimate peak transvalvular pressure gradients (DeltaP) by Doppler echocardiographic imaging of peak flow velocities using the simplified Bernoulli relationship: DeltaP (mm Hg) = 4V(2) (m/s). Because left ventricular filling is initiated by mechanical suction, V can be predicted by the motion of a simple harmonic oscillator by the parametrized diastolic filling formalism that characterizes E-wave contours by 3 unique simple harmonic oscillator parameters: initial displacement (x(o) cm); spring constant (k g/s(2)); and damping constant (c g/s). Parametrized diastolic filling predicts peak atrioventricular pressure gradient as kx(o), the peak simple harmonic oscillator force. For validation, simultaneous (micromanometric) left ventricular pressure and E-wave data from 19 patients were analyzed. Model-predicted peak gradient (kx(o)) was compared with actual gradient (DeltaP(cath)) and with 4V(2). Multiple linear regression results for all patients yielded highly significant relation between kx(o) and DeltaP(cath) (kx(o) = m(1)DeltaP(cath) + b(1), where m(1) = 40.7 +/- 8.0 dyne/mm Hg, b(1) = 1540 +/- 116 dyne, r(2) = 0.97, P <.001). Regression analysis showed no significant correlation between 4V(2) and DeltaP(cath) (4V(2) = m(2)DeltaP(cath) + b(2), where m(2) = 0.01 +/- 0.03, m(2)/s(2)/mm Hg and b(2) = 2.07 +/- 0.44 m(2)/s(2), P = nonsignificant). We conclude that E-wave analysis by parametrized diastolic filling predicts peak atrioventricular gradients reliably and more accurately than 4V(2).

Adult↗