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 1,261 records · Page 70Linked to original sources
PubMed2006

The Path-A metabolic pathway prediction web server.

Pathway Analyst (Path-A) is a publicly available web server (http://path-a.cs.ualberta.ca) that predicts metabolic pathways. It takes a FASTA format file containing a set of query protein sequences from a single organism (a partial or complete proteome) and identifies those sequences that are likely to participate in any of its supported metabolic pathways (currently 10). Path-A uses a number of machine-learning and sequence analysis techniques (e.g. SVM, BLAST and HMM) to predict pathways. Each machine-learned classifier exploits similarity between sequences in the pathways of its model organisms and sequences in the query set. It predicts the pathways that are present in the query organism and annotates each predicted reaction and catalyst, using the appropriate sequences from the query set. Path-A also provides a browsable and searchable database of the pathways for the model organisms that are used to make its predictions. Path-A's predictor sets (using different classifier technologies) have been evaluated using standard cross-validation techniques on a dataset of 10 metabolic pathways across 13 model organisms--a total of 125 organism-specific pathways. The most accurate classifier technology obtained a mean precision of 78.3% and a mean recall of 92.6% in predicting all catalyst proteins, of all reactions, in all pathways present in the dataset. Although Path-A currently only supports metabolic pathways, the underlying prediction techniques are general enough for other types of pathways. Consequently, it is our intent to extend Path-A to predict other types of pathways, including signalling pathways.

Algorithms↗
PubMed2006

Protein binding site prediction using an empirical scoring function.

Most biological processes are mediated by interactions between proteins and their interacting partners including proteins, nucleic acids and small molecules. This work establishes a method called PINUP for binding site prediction of monomeric proteins. With only two weight parameters to optimize, PINUP produces not only 42.2% coverage of actual interfaces (percentage of correctly predicted interface residues in actual interface residues) but also 44.5% accuracy in predicted interfaces (percentage of correctly predicted interface residues in the predicted interface residues) in a cross validation using a 57-protein dataset. By comparison, the expected accuracy via random prediction (percentage of actual interface residues in surface residues) is only 15%. The binding sites of the 57-protein set are found to be easier to predict than that of an independent test set of 68 proteins. The average coverage and accuracy for this independent test set are 30.5 and 29.4%, respectively. The significant gain of PINUP over expected random prediction is attributed to (i) effective residue-energy score and accessible-surface-area-dependent interface-propensity, (ii) isolation of functional constraints contained in the conservation score from the structural constraints through the combination of residue-energy score (for structural constraints) and conservation score and (iii) a consensus region built on top-ranked initial patches.

Algorithms↗
PubMed2006

A set of nearest neighbor parameters for predicting the enthalpy change of RNA secondary structure formation.

A complete set of nearest neighbor parameters to predict the enthalpy change of RNA secondary structure formation was derived. These parameters can be used with available free energy nearest neighbor parameters to extend the secondary structure prediction of RNA sequences to temperatures other than 37 degrees C. The parameters were tested by predicting the secondary structures of sequences with known secondary structure that are from organisms with known optimal growth temperatures. Compared with the previous set of enthalpy nearest neighbor parameters, the sensitivity of base pair prediction improved from 65.2 to 68.9% at optimal growth temperatures ranging from 10 to 60 degrees C. Base pair probabilities were predicted with a partition function and the positive predictive value of structure prediction is 90.4% when considering the base pairs in the lowest free energy structure with pairing probability of 0.99 or above. Moreover, a strong correlation is found between the predicted melting temperatures of RNA sequences and the optimal growth temperatures of the host organism. This indicates that organisms that live at higher temperatures have evolved RNA sequences with higher melting temperatures.

Algorithms↗
PubMed1998

Total body nitrogen predicts long-term mortality in haemodialysis patients--a single-centre experience.

BACKGROUND: It has been estimated that 30-50% of adult haemodialysis patients have moderate to severe malnutrition. We have previously shown that estimation of total body nitrogen, expressed as a nitrogen index (NI) by in vivo neutron activation analysis (IVNAA) is an accurate tool for estimating total body protein in dialysis patients. It is not clear whether the nitrogen index is predictive of mortality and morbidity in dialysis patients. METHODS: We studied the long-term predictive value of nutritional assessment by IVNAA and serum albumin on mortality and morbidity (including infection episodes requiring hospital admission, ischaemic heart disease (IHD), cerebrovascular or peripheral vascular disease (PVD). Seventy-six chronic haemodialysis patients were initially studied between 1989 and 1991, with a minimum follow-up of 5 years. The mean age of the patients was 48.3 years (range 21-76). Patients were divided into two groups, group I, n = 22, had a NI < or = 0.8 (NI < or = 0.8 represents protein malnutrition) and group II, n = 54, had a NI > 0.8. RESULTS: Fifteen patients in group II died in the follow-up period compared to nine from group I (P < 0.05), but NI < or = 0.8 did not predict vascular or infective morbidity. Serum albumins < or = 35 g/day did predict over all mortality (P < 0.05) as well as infection episodes (P < 0.001). When patients above the age of 50 years were analysed, NI did predict mortality (P < 0.05) but serum albumin did not, while the age of> 50 itself was a strong predictor of mortality (P < 0.001). CONCLUSION: We conclude that NI < or = 0.8 is predictive of long-term mortality. This reinforces the view that low body protein stores are predictive of increased mortality in dialysis patients and that the serum albumin is predictive of mortality because of its reflection of protein stores.

Adult↗
PubMed1997

Secondary structure prediction using segment similarity.

We present a secondary structure prediction method based on finding similarities between sequence segments from the target sequence and segments contained in the database of proteins with known structures. The similarity definition is optimized using a genetic algorithm and is based on a 21 x 40 similarity matrix, comparing a target sequence with the sequence and burial status of the proteins from the database. The three-state secondary structure prediction accuracy reaches 72.4% on a non homologous (maximum sequence identity <25%) data set derived from PDB and is reproduced on two independent testing sets, including the set of CASP2 prediction targets and a group of newly solved PDB structures. The prediction method was developed with simplicity and open architecture in mind, allowing for an easy extension to other types of predictions and to the analysis of the contributions to the local structure formation. For instance, the design of the prediction procedure allows us to trace back segments of the database that contributed to the prediction. It can be shown that those segments came from various structural classes and that even complete exclusion of related folds from the database does not result in a significant decrease in prediction accuracy.

Databases, Factual↗
PubMed1997

Protein distance constraints predicted by neural networks and probability density functions.

We predict interatomic Calpha distances by two independent data driven methods. The first method uses statistically derived probability distributions of the pairwise distance between two amino acids, whilst the latter method consists of a neural network prediction approach equipped with windows taking the context of the two residues into account. These two methods are used to predict whether distances in independent test sets were above or below given thresholds. We investigate which distance thresholds produce the most information-rich constraints and, in turn, the optimal performance of the two methods. The predictions are based on a data set derived using a new threshold which defines when sequence similarity implies structural similarity. We show that distances in proteins are predicted more accurately by neural networks than by probability density functions. We show that the accuracy of the predictions can be further increased by using sequence profiles. A threading method based on the predicted distances is presented. A homepage with software, predictions and data related to this paper is available at http://www.cbs.dtu.dk/services/CPHmodels/.

Amino Acids↗
PubMed1995

A new approach to the evaluation of protein secondary structure predictions at the level of the elements of secondary structure.

For many purposes, such as the prediction of the class of protein folds, the existence of an element of secondary structure rather than its precise position and length must be defined correctly. However, most methods for the evaluation of secondary structure prediction consider success in terms of the percentage of individual amino acids predicted correctly. In this paper the success in predicting elements of secondary structure is discussed. The number of overlapping residues in the predicted and observed secondary structures were considered as a function of the total number of amino acids in the observed and predicted secondary structures. A matrix search procedure was used to remove the ambiguity which similar studies may have had in defining the equivalent secondary structures between predicted and observed structures. In this study a loop was treated in the same way as an alpha-helix and a beta-strand. To describe the accuracy at the level of elements of secondary structure, a set of parameters was defined, similar to those used commonly at the level of individual amino acids. This approach was used to assess the methods of Chou and Fasman (1974b, Biochemistry, 13, 222-245), Lim (1974b, J. Mol. Biol., 88, 873-894) and Garnier et al. (1978, J. Mol. Biol., 120, 97-120). It was found that these methods were much poorer at the secondary structure level than at the amino acid level. This approach can be used generally for secondary structure prediction methods.

Amino Acids↗
PubMed2003

Recursive prediction of broiler growth response to feed intake by using a time-variant parameter estimation method.

The objective of this study was to explore whether time-variant parameter estimation procedures allow modeling and predicting the dynamic growth response of broiler chickens to feed intake in real time. A recursive linear model was used that estimated the model parameters every 24 h based on a fixed number of actual and past measurements (i.e., time window). Based on 48 datasets, it was concluded that the mean relative prediction error (MRPE) of the recursive linear modeling approach had a minimum for a window size of 5 d. Weight of the birds could be predicted during the growth process 3 to 7 d ahead with a mean relative prediction error of 5% or less. In comparison with the prediction results of three static empirical growth models (one linear and two nonlinear models), the recursive modeling technique had a similar accuracy to the nonlinear empirical models (MRPE of 1.4% to 2.3% vs. 1.1% to 2.8%), but it was less accurate for larger prediction horizons (2 to 7 d). The compact recursive linear model was more accurate than the static linear growth model for prediction horizons of one up to 4 d, depending on the feeding strategy. Since such recursive modeling approach allows the prediction of broiler growth without any prior knowledge of the system and takes into account the time-variant (nonlinear) nature of the growth process based on only a small window of measured information, it is suitable for real-time integration in process management.

Animal Nutritional Physiological Phenomena↗
PubMed2005

Multiple solar particle event dose time profile predictions using Bayesian inference.

The prediction of solar particle event occurrence and the resulting effects on humans and electronics continues to be a mission and/or life-threatening concern for the National Aeronautics and Space Administration and military and commercial satellite operators. While the frequency of events generally follows the solar cycle, individual event occurrence is sporadic and the prediction of resulting effects prior to the event onset is difficult. In one approach to space weather prediction, the forecaster begins to make predictions after the onset of an event. Previous work proved the efficacy of a forecasting methodology that used Bayesian inference and dose and/or dose rate information obtained early after the onset of an event to make predictions of dose and dose rate time profiles out to 120 h beyond onset. The previous work, however, was restricted to predictions for single-event solar particle events. Some of the largest recorded events, including the October 1989 and August 1972 events, were actually multiple events. In this study, we present an analysis of nine large events, some single and some multiple. This work ties together particle flux and fluence data with dose rate and dose calculations in an effort to develop a criterion for characterising an event as multiple and thus, generalising the Bayesian methodology to allow predictions for all events. Dose time profile predictions are made for the four separate events that made up the October 1989 event.

Bayes Theorem↗
PubMed2004

Prediction of torsade-causing potential of drugs by support vector machine approach.

In an effort to facilitate drug discovery, computational methods for facilitating the prediction of various adverse drug reactions (ADRs) have been developed. So far, attention has not been sufficiently paid to the development of methods for the prediction of serious ADRs that occur less frequently. Some of these ADRs, such as torsade de pointes (TdP), are important issues in the approval of drugs for certain diseases. Thus there is a need to develop tools for facilitating the prediction of these ADRs. This work explores the use of a statistical learning method, support vector machine (SVM), for TdP prediction. TdP involves multiple mechanisms and SVM is a method suitable for such a problem. Our SVM classification system used a set of linear solvation energy relationship (LSER) descriptors and was optimized by leave-one-out cross validation procedure. Its prediction accuracy was evaluated by using an independent set of agents and by comparison with results obtained from other commonly used classification methods using the same dataset and optimization procedure. The accuracies for the SVM prediction of TdP-causing agents and non-TdP-causing agents are 97.4 and 84.6% respectively; one is substantially improved against and the other is comparable to the results obtained by other classification methods useful for multiple-mechanism prediction problems. This indicates the potential of SVM in facilitating the prediction of TdP-causing risk of small molecules and perhaps other ADRs that involve multiple mechanisms.

Algorithms↗
PubMed2004

Human respiratory tract cancer risks of inhaled formaldehyde: dose-response predictions derived from biologically-motivated computational modeling of a combined rodent and human dataset.

Formaldehyde inhalation at 6 ppm and above causes nasal squamous cell carcinoma (SCC) in F344 rats. The quantitative implications of the rat tumors for human cancer risk are of interest, since epidemiological studies have provided only equivocal evidence that formaldehyde is a human carcinogen. Conolly et al. (Toxicol. Sci. 75, 432-447, 2003) analyzed the rat tumor dose-response assuming that both DNA-reactive and cytotoxic effects of formaldehyde contribute to SCC development. The key elements of their approach were: (1) use of a three-dimensional computer reconstruction of the rat nasal passages and computational fluid dynamics (CFD) modeling to predict regional dosimetry of formaldehyde; (2) association of the flux of formaldehyde into the nasal mucosa, as predicted by the CFD model, with formation of DNA-protein cross-links (DPX) and with cytolethality/regenerative cellular proliferation (CRCP); and (3) use of a two-stage clonal growth model to link DPX and CRCP with tumor formation. With this structure, the prediction of the tumor dose response was extremely sensitive to cell kinetics. The raw dose-response data for CRCP are J-shaped, and use of these data led to a predicted J-shaped dose response for tumors, notwithstanding a concurrent low-dose-linear, directly mutagenic effect of formaldehyde mediated by DPX. In the present work the modeling approach used by Conolly et al. (ibid.) was extended to humans. Regional dosimetry predictions for the entire respiratory tract were obtained by merging a three-dimensional CFD model for the human nose with a one-dimensional typical path model for the lower respiratory tract. In other respects, the human model was structurally identical to the rat model. The predicted human dose response for DPX was obtained by scale-up of a computational model for DPX calibrated against rat and rhesus monkey data. The rat dose response for CRCP was used "as is" for the human model, since no preferable alternative was identified. Three sets of baseline parameter values for the human clonal growth model were obtained through separate calibrations against respiratory tract cancer incidence data for nonsmokers, smokers, and a mixed population of nonsmokers and smokers, respectively. Additional risks of respiratory tract cancer were predicted to be negative up to about one ppm for all three cases when the raw CRCP data from the rat were used. When a hockey-stick-shaped model was fit to the rat CRCP data and used in place of the raw data, positive maximum likelihood estimates (MLE) of additional risk were obtained. These MLE estimates were lower, for some comparisons by as much as 1,000-fold, than MLE estimates from previous cancer dose-response assessments for formaldehyde. Breathing rate variations associated with different physical activity levels did not make large changes in predicted additional risks. In summary, this analysis of the human implications of the rat SCC data indicates that (1) cancer risks associated with inhaled formaldehyde are de minimis (10(-6) or less) at relevant human exposure levels, and (2) protection from the noncancer effects of formaldehyde should be sufficient to protect from its potential carcinogenic effects.

Animals↗
PubMed2001

Auditory evoked potential index predicts the depth of sedation and movement in response to skin incision during sevoflurane anesthesia.

BACKGROUND: The auditory evoked potential (AEP) index, which is a single numerical parameter derived from the AEP in real time and which describes the underlying morphology of the AEP, has been studied as a monitor of anesthetic depth. The current study was designed to evaluate the accuracy of AEPindex for predicting depth of sedation and anesthesia during sevoflurane anesthesia. METHODS: In the first phase of the study, a single end-tidal sevoflurane concentration ranging from 0.5 to 0.9% was assigned randomly and administered to each of 50 patients. The AEPindex and the Bispectral Index (BIS) were obtained simultaneously. Sedation was assessed using the responsiveness portion of the observer's assessment of alertness-sedation scale. In the second phase of the study, 10 additional patients were included, and the 60 patients who were scheduled to have skin incisions were observed for movement in response to skin incision at the end-tidal sevoflurane concentrations between 1.6 and 2.6%. The relation among AEPindex, BIS, sevoflurane concentration, sedation score, and movement or absence of movement after skin incision was determined. Prediction probability values for AEPindex, BIS, and sevoflurane concentration to predict depth of sedation and anesthesia were also calculated. RESULTS: The AEPindex, BIS, and sevoflurane concentration correlated closely with the sedation score. The prediction probability values for AEPindex, BIS, and sevoflurane concentration for sedation score were 0.820, 0.805, and 0.870, respectively, indicating a high predictive performance for depth of sedation. AEPindex and sevoflurane concentration successfully predicted movement after skin (prediction probability = 0.910 and 0.857, respectively), whereas BIS could not (prediction probability = 0.537). CONCLUSIONS: Auditory evoked potential index can be a guide to the depth of sedation and movement in response to skin incision during sevoflurane anesthesia.

Acoustic Stimulation↗
PubMed1993

Predicting mortality risk for infants weighing 501 to 1500 grams at birth: a National Institutes of Health Neonatal Research Network report.

OBJECTIVES: To develop and evaluate a model that predicts mortality risk based on admission data for infants weighing 501 to 1500 grams at birth, and to use the model to identify neonatal ICUs where the observed mortality rate differs significantly from the predicted rate. DESIGN: Validation cohort study. SETTING: University-based, tertiary care neonatal ICUs. PATIENTS: Sample of 3,603 infants with birth weights of 501 to 1500 grams who were born at seven National Institute of Child Health and Human Development (NICHHD) Neonatal Research Network Centers, over a 2-yr period of time. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Based on logistic regression analysis, admission factors associated with mortality risk for inborn infants were: decreasing birth weight, appropriate size for gestational age, male gender, non-black race, and 1-min Apgar score of < or = 3. The mortality prediction model based on these factors had a sensitivity of 0.50, a specificity of 0.92, a correct classification rate of 0.82, and an area under the receiver operating characteristic curve of 0.82 when applied to a validation sample. Goodness-of-fit testing showed that there was a marginal degree of fit between the observations and model predictions (chi 2 = 15.4, p = .06). The observed mortality rate for 3,603 infants at the seven centers was 24.7%, ranging from 21.8% to 27.7% at individual centers. There were no statistically significant differences between observed and predicted mortality rates at any of the centers. One center had an observed mortality rate that was 2.8% lower than predicted by the model (95% confidence interval -6.0% to 0.5%), and another center had an observed rate that was 3% higher than expected (95% confidence interval -0.3% to 6.2%). CONCLUSIONS: Mortality risk for infants weighing 501 to 1500 grams can be predicted based on admission factors. However, until more accurate predictive models are developed and validated and the relationships between care practices and outcomes are better understood, such models should not be relied on for evaluating the quality of care provided in different neonatal ICUs.

Cohort Studies↗
PubMed1995

An overview of mortality risk prediction in sepsis.

OBJECTIVE: To review the evolution and development of mortality risk prediction methods as they have been applied to the management of septic patients. DATA SOURCES: Selected relevant articles from the pertinent literature. STUDY SELECTION: Theoretical and clinical data on the mortality risk identification, severity of illness scoring systems, and cytokine levels as they relate to mortality in patients with sepsis. DATA EXTRACTION: All concepts relating to mortality risk prediction, cytokines, severity of illness, and intensive care unit (ICU) mortality were explored and interrelated accordingly. DATA SYNTHESIS: In order to improve the precision of the evaluation of new therapies for the treatment of sepsis, to monitor their utilization and to refine their indications, it has been recommended that mortality risk stratification or severity of illness scoring systems be utilized in clinical trials and in practice. With the increasing influence of managed care on healthcare delivery, there will be an increased demand for techniques to stratify patients for cost-effective allocation of care. Severity of illness scoring systems are widely utilized for patient stratification in the management of cancer and heart disease. However, the use of such systems in patients with sepsis has been limited to application in clinical trial design for assurance of balance among treatment groups. Mortality risk prediction in sepsis has evolved from identification of risk factors, and simple counts of failing organs, to sophisticated techniques that mathematically transform a raw score, comprised of physiologic and/or clinical data, into a predicted risk of death. Most of the developed systems are based on global ICU populations rather than upon sepsis patient databases. A few, newer systems are derived from such databases. However, the overall discriminating ability of the various methods is similar. Mortality prediction has also been carried out from assessments of endotoxin or cytokine (interleukin-1, interleukin-6, tumor necrosis factor) plasma concentrations. While increased levels of these substances have been correlated with increased mortality, difficulties with bioassay and their sporadic appearance in the bloodstream prevent these measurements from being practically applied. The calibration of risk prediction methods comparing predicted with actual mortality across the breadth of risk for a population of patients is excellent, but overall accuracy in individual patient predictions is such that clinical judgment must remain a major part of decision-making. However, as databases of appropriate patient information increase in size and complexity, it may be possible in the future to devise a scoring system that can be relied on to assist in clinical decision-making. CONCLUSIONS: Severity of illness scoring systems are widely used in critically ill patients. However, their use in patients with sepsis has largely been limited to a means of stratification in clinical trials. As newer sepsis therapies become available, it may be possible to use such systems for refining their indications, and monitoring their utilization. Finally, as the databases supporting the systems increase in size and complexity, it may be possible to utilize them in clinical decision-making.

Cytokines↗
PubMed1995

Evaluation of predictive ability of APACHE II system and hospital outcome in Canadian intensive care unit patients.

OBJECTIVES: To evaluate the ability of the acute Physiology and Chronic Health Evaluation II (APACHE II) scoring system to predict patient outcome in two Canadian intensive care units (ICUs). To compare the severity of illness and outcome of Canadian ICU patients with existing United States data. DESIGN: Prospective data collection on 1,724 Canadian ICU patients for validation of the APACHE II system. Comparison of the outcome of Canadian ICU patients to retrospective United States data on 4,087 patients from the 1985 APACHE II multicenter study. SETTING: Canadian data from two university teaching hospital ICUs. United States data from 13 ICUs, ten of which were in university teaching hospitals. PATIENTS: Consecutive patients admitted to adult medical/surgical ICUs. Coronary care unit, neurosurgical and cardiac surgery patients were excluded. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: For each patient, demographic data, diagnosis, APACHE II score and hospital survival data were collected. The predicted risk of death was calculated for each patient using the APACHE II risk of death equation. The accuracy in outcome prediction of the APACHE II system was assessed by means of the receiver operating characteristic curve, 2 x 2 decision matrices and linear regression analysis. The severity of illness and hospital mortality for the Canadian patients was compared with that of United States patients from the 1985 APACHE II multicenter study. In 1,724 Canadian ICU patients, the mean +/- SEM APACHE II score was 16.5 +/- 0.2. The predicted death rate was 24.7% and the observed death rate was 24.8%. Using receiver operating curve analysis, good correlation was found between predicted outcome and observed outcome. The area under the curve was 0.86. From the 2 x 2 decision matrix constructed for a predicted risk of death of 0.5, 83% of patients were correctly classified. The sensitivity was 50.9% and the specificity was 93.6%. When observed death rate was plotted against predicted death rate, linear regression analysis gave an r2 of .99. Canadian patients had a higher death rate and APACHE II score than the United States patients. After controlling for severity of illness using the APACHE II score, the Canadian and United States death rates were similar. CONCLUSIONS: The ability of the APACHE II system in predicting group outcome is validated in this Canadian ICU population by receiver operating characteristic curve, 2 x 2 decision matrices and linear regression analysis. The Canadian patients had a higher overall hospital death rate than the United States patients. After controlling for severity of illness using APACHE II scores, the hospital death rate was comparable between the Canadian and United States patients.

APACHE↗
PubMed1997

Prediction of outcome from intensive care: a prospective cohort study comparing Acute Physiology and Chronic Health Evaluation II and III prognostic systems in a United Kingdom intensive care unit.

OBJECTIVE: To evaluate the ability of two prognostic systems to predict hospital mortality in adult intensive care patients. DESIGN: Prospective cohort study. SETTING: A mixed medical and surgical intensive care unit (ICU) in the United Kingdom. PATIENTS: A total of 1,144 patients consecutively admitted to the study. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Acute Physiology and Chronic Health Evaluation (APACHE) II and III prognostic systems were applied to assess probabilities of hospital mortality, which were compared with the actual outcome. The overall goodness-of-fit of both models was assessed. Hospital death rates were higher than those predicted by each system. Risk estimates showed a strong positive correlation between both systems (nonsurvivors r2 = 0.756, p < .0001; survivors r2 = 0.787, p < .0001). Calibration of APACHE II (chi 2 = 98.6, Lemeshow-Hosmer) was superior to that of APACHE III (chi 2 = 129.8, Lemeshow-Hosmer). The total correct classification rate of APACHE III was greater for all decision criteria applied; the best overall total correct classification rate was 80.6% for APACHE III and 77.9% for APACHE II (both for a decision criterion of 40%). The areas under the receiver operating characteristic curves were 0.806 and 0.847 for APACHE II and III, respectively, confirming the better discrimination of APACHE III. When patients were classified by diagnostic categories, risk predictions did not fit uniformly across the spectrum of disease groups. For both models, mortality ratios were highest for trauma patients and lowest for the group with respiratory disease. APACHE II predictions for patients with gastrointestinal disease were significantly better. Risk estimates for surgical admissions were superior with APACHE II (MR = 1.27) compared with APACHE III (MR = 1.56), but were similar for medical patients (1.22 vs. 1.28 for APACHE II and III, respectively). Bias induced by factors reflecting the clinical practice in an individual ICU (e.g., admission criteria, treatment before admission) may have considerable impact on risk estimates. The identification of such factors appears to be a prerequisite for the meaningful interpretation of observed and predicted death rates on the individual ICU level. CONCLUSIONS: Both predictive models demonstrated a similar degree of overall goodness-of-fit. APACHE II showed better calibration, but discrimination was better with APACHE III. Hospital mortality was higher than predicted by both models, but was underestimated to a greater degree by APACHE III. Risk estimates by both models showed considerable variation across the disease spectrum of ICU patients. Risk predictions for surgical patients and patients with gastrointestinal disease were better with APACHE II. Factors reflecting the clinical practice of an individual ICU are not accounted for by APACHE II and III. Overall, the performance of APACHE III was not superior to that of its predecessor for a cohort of United Kingdom ICU patients; for certain diagnostic categories, APACHE III performed worse than APACHE II despite an improved system of disease classification.

APACHE↗
PubMed2002

The prediction of breast reduction weight.

Reimbursement for reduction mammaplasty has become more stringent because many insurers require specific documentation of patient symptoms and estimated weight of planned breast resection. The purpose of this study was to develop a simple, clinically useful method for predicting weight of breast tissue to be removed, using routine, easily obtained predictors (i.e., height, weight, age, measurements from sternal notch to nipple, and measurements from sternal notch to inframammary crease). Data were available from a retrospective review of 263 women undergoing reduction mammaplasty. Analyses were performed to predict resected weights obtained both in the operating room and by a pathologist for left and right breasts separately. Regression analyses showed that the sternal notch-to-nipple measurement accounted for nearly all of the explained variance in the resected weights, with correlations around 0.80 between sternal notch to nipple and resected weight. For sternal notch-to-nipple measurements > or 28.5 cm, predicted resected weights were approximately 600 g or more, and in general, 80 percent or more patients had specimen weights >500 grams. From 25.5 to 28 cm, the predicted weights ranged from about 400 to 600 g and the prediction rate of weights >500 g was 50 percent. The senior author predicted the resected breast weight to be >500 g 94 percent of the time. The equation alone did not produce an accurate prediction in the critical range, 400 to 600 g. The experienced surgeon more accurately predicted resected weights with use of practiced spatial relationship skills.

Adult↗
PubMed1990

Models of relative reinforcing efficacy of drugs and their predictive utility.

Studies of drugs as reinforcers in animals have indicated a close agreement between the drugs that function as reinforcers in animals and those that are abused by humans. This agreement has prompted the use of results of studies in animals as a model to predict the abuse of newly synthesized drugs. While this model has widely accepted nominal predictive utility, there have been attempts to extend its predictive capability to assess the relative reinforcing efficacy of drugs. This concept implies that the reinforcing effects of drugs can be compared on ordinal, interval or ratio scales. Relative reinforcing efficacy has been assessed in studies of performances under progressive-ratio schedules and choice procedures, and it has been suggested that results of these studies might be used to predict the extent to which a drug will be abused. In order to use these data for prediction, scaling issues need to be addressed. With progressive-ratio schedules, it is unclear whether the results represent interval or ordinal data. Development and analysis of choice procedures must to occur before each study will yield more than a pair of ordered data points. In order to assess the predictive utility of models of relative reinforcing efficacy, laboratory data from assessments of known drugs of abuse need to be validated with data on the actual abuse of the compounds. There are significant issues in determining what data on "actual abuse" (e.g. epidemiologic results) should be used in attempts to validate the experimental results. Results of epidemiological surveys are influenced by non-pharmacological variables, such as social, marketing and legal factors, that can affect the degree to which results of laboratory studies agree with those from the surveys. Differential influences of these factors on particular drugs may render results of epidemiologic surveys unsuitable for validation of laboratory results. Other types of data, such as laboratory reports of subjective effects or reinforcing effects of drugs in humans subjects, may be influenced less by those factors but are not themselves reflective of "actual abuse." Currently nominal scaling may represent the best information that can be provided for predicting abuse of drugs. However, there are clear paths to take to overcome these shortcomings. More directed laboratory studies that concentrate on scaling issues as well as more pharmacological specificity of epidemiologic studies will undoubtedly increase our predictive ability.

Journal Article↗