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 1,171 records · Page 65Linked to original sources
PubMed2006

A novel method for apoptosis protein subcellular localization prediction combining encoding based on grouped weight and support vector machine.

Apoptosis proteins have a central role in the development and homeostasis of an organism. These proteins are very important for understanding the mechanism of programmed cell death. Based on the idea of coarse-grained description and grouping in physics, a new feature extraction method with grouped weight for protein sequence is presented, and applied to apoptosis protein subcellular localization prediction associated with support vector machine. For the same training dataset and the same predictive algorithm, the overall prediction accuracy of our method in Jackknife test is 13.2% and 15.3% higher than the accuracy based on the amino acid composition and instability index. Especially for the else class apoptosis proteins, the increment of prediction accuracy is 41.7 and 33.3 percentile, respectively. The experiment results show that the new feature extraction method is efficient to extract the structure information implicated in protein sequence and the method has reached a satisfied performance despite its simplicity. The overall prediction accuracy of EBGW_SVM model on dataset ZD98 reach 92.9% in Jackknife test, which is 8.2-20.4 percentile higher than other existing models. For a new dataset ZW225, the overall prediction accuracy of EBGW_SVM achieves 83.1%. Those implied that EBGW_SVM model is a simple but efficient prediction model for apoptosis protein subcellular location prediction.

Algorithms↗
PubMed2005

Does EuroSCORE predict length of stay and specific postoperative complications after coronary artery bypass grafting?

BACKGROUND: To evaluate the performance of EuroSCORE in the prediction of in-hospital postoperative length of stay and specific major postoperative complications after coronary artery bypass grafting (CABG). METHODS: Data on 3760 consecutive patients with CABG were prospectively collected. The EuroSCORE model (standard and logistic) was used to predict in-hospital mortality, prolonged length of stay (>12 days) and major postoperative complications (stroke, myocardial infarction, sternal infection, bleeding, sepsis and/or endocarditis, gastrointestinal complications, renal and respiratory failure). A C statistic (receiver operating characteristic curve) was used to test the discrimination of the EuroSCORE. The calibration of the model was assessed by the Hosmer-Lemeshow goodness-of-fit statistic. RESULTS: In-hospital mortality was 2.7%, and 13.7% of patients had one or more major complications. EuroSCORE showed very good discriminatory ability in predicting renal failure (C statistic: 0.80) and good discriminatory ability in predicting in-hospital mortality (C statistic: 0.75), sepsis and/or endocarditis (C statistic: 0.72) and prolonged length of stay (C statistic: 0.71). There were no differences in terms of the discriminatory ability between standard and logistic EuroSCORE. Standard EuroSCORE showed good calibration (Hosmer-Lemeshow: P>0.05) in predicting these outcomes except for postoperative length of stay, while logistic EuroSCORE showed good calibration only in predicting renal failure. CONCLUSIONS: EuroSCORE can be used to predict not only in-hospital mortality, for which it was originally designed, but also prolonged length of stay and specific postoperative complications such as renal failure and sepsis and/or endocarditis after CABG. These outcomes can be predicted accurately using the standard EuroSCORE which is very simple and easy in its calculation.

Aged↗
PubMed2006

A global optimization method for prediction of muscle forces of human musculoskeletal system.

Inverse dynamic optimization is a popular method for predicting muscle and joint reaction forces within human musculoskeletal joints. However, the traditional formulation of the optimization method does not include the joint reaction moment in the moment equilibrium equation, potentially violating the equilibrium conditions of the joint. Consequently, the predicted muscle and joint reaction forces are coordinate system-dependent. This paper presents an improved optimization method for the prediction of muscle forces and joint reaction forces. In this method, the location of the rotation center of the joint is used as an optimization variable, and the moment equilibrium equation is formulated with respect to the joint rotation center to represent an accurate moment constraint condition. The predicted muscle and joint reaction forces are independent of the joint coordinate system. The new optimization method was used to predict muscle forces of an elbow joint. The results demonstrated that the joint rotation center location varied with applied loading conditions. The predicted muscle and joint reaction forces were different from those predicted by using the traditional optimization method. The results further demonstrated that the improved optimization method converged to a minimum for the objective function that is smaller than that reached by using the traditional optimization method. Therefore, the joint rotation center location should be involved as a variable in an inverse dynamic optimization method for predicting muscle and joint reaction forces within human musculoskeletal joints.

Biomechanical Phenomena↗
PubMed2004

Clinicians required very high sensitivity of a bacteremia prediction rule.

BACKGROUND AND OBJECTIVE: Efforts to improve blood culture practice have focused on developing clinical prediction rules to identify patients at risk for bacteremia. However, no such models have been accepted into general clinical use. The goal of this study was to determine physicians' criteria for acceptability of a bacteremia clinical prediction rule. METHOD: We conducted a survey of all medical and surgical house officers as well as all infectious diseases physicians at the University of Pennsylvania to identify physician requirements for the sensitivity of a bacteremia clinical prediction rule. RESULTS: Of 225 eligible physicians, 149 (66.2%) completed the survey, including 110 house officers and 39 infectious disease physicians. The median (95% confidence interval) sensitivity of a bacteremia prediction rule required by respondents was 95% (95% confidence interval, 95.9%). Furthermore, 29 (19.5%) respondents required the sensitivity of a prediction rule to be at least 99%. The median required sensitivity was significantly higher for infectious diseases physicians than for house officers (98% and 95%, respectively) (P=.04). CONCLUSION: Our survey of house staff and infectious diseases physicians demonstrates that the sensitivity of any bacteremia prediction rule must be extremely high (i.e., 99% to 100%) to be widely accepted by practicing clinicians. Elucidation of physician criteria for acceptability of clinical predictive models will be invaluable in future efforts to develop such prediction rules.

Attitude of Health Personnel↗
PubMed2005

Predicting rRNA-, RNA-, and DNA-binding proteins from primary structure with support vector machines.

In the post-genome era, the prediction of protein function is one of the most demanding tasks in the study of bioinformatics. Machine learning methods, such as the support vector machines (SVMs), greatly help to improve the classification of protein function. In this work, we integrated SVMs, protein sequence amino acid composition, and associated physicochemical properties into the study of nucleic-acid-binding proteins prediction. We developed the binary classifications for rRNA-, RNA-, DNA-binding proteins that play an important role in the control of many cell processes. Each SVM predicts whether a protein belongs to rRNA-, RNA-, or DNA-binding protein class. Self-consistency and jackknife tests were performed on the protein data sets in which the sequences identity was < 25%. Test results show that the accuracies of rRNA-, RNA-, DNA-binding SVMs predictions are approximately 84%, approximately 78%, approximately 72%, respectively. The predictions were also performed on the ambiguous and negative data set. The results demonstrate that the predicted scores of proteins in the ambiguous data set by RNA- and DNA-binding SVM models were distributed around zero, while most proteins in the negative data set were predicted as negative scores by all three SVMs. The score distributions agree well with the prior knowledge of those proteins and show the effectiveness of sequence associated physicochemical properties in the protein function prediction. The software is available from the author upon request.

Amino Acid Sequence↗
PubMed2006

Emotional modulation of spinal nociception and pain: the impact of predictable noxious stimulation.

Recent evidence suggests that emotional picture-viewing is a reliable method of engaging descending modulation of spinal nociception. The present study attempted to replicate these findings and determine the effect of noxious stimulus predictability. Participants viewed pictures from the International Affective Picture System (IAPS), during which pain and nociceptive flexion reflexes (NFR) were elicited by electric shocks delivered to the sural nerve. For half of the participants (n=25) shocks were preceded by a cue (predictable), whereas the other half received no cue (unpredictable). Results suggested emotion was successfully induced by pictures, but the effect of picture-viewing on the NFR was moderated by the predictability of the shocks. When shock was unpredictable, spinal nociception (NFR) and pain ratings were modulated in parallel. Specifically, pain and NFR magnitudes were lower during pleasant emotions and higher during unpleasant emotions. However, when shocks were predictable, only pain was modulated in this way. NFRs from predictable shocks were not altered by pictures. Further, exploratory analyses found that pain ratings, but not NFRs, were lower during predictable shocks. These data suggest emotional picture-viewing is a reliable method of engaging descending modulation of spinal nociception. However, descending modulation could not be detected in NFRs resulting from predictable noxious stimuli. Although preliminary, this study implies that separate mechanisms are responsible for emotional modulation of nociception at spinal vs. supraspinal levels, and that predictable noxious events may disengage modulation at the spinal level. The current paradigm could serve as a useful tool for studying descending modulation.

Adult↗
PubMed2006

Applying psychological theory to evidence-based clinical practice: identifying factors predictive of taking intra-oral radiographs.

This study applies psychological theory to the implementation of evidence-based clinical practice. The first objective was to see if variables from psychological frameworks (developed to understand, predict and influence behaviour) could predict an evidence-based clinical behaviour. The second objective was to develop a scientific rationale to design or choose an implementation intervention. Variables from the Theory of Planned Behaviour, Social Cognitive Theory, Self-Regulation Model, Operant Conditioning, Implementation Intentions and the Precaution Adoption Process were measured, with data collection by postal survey. The primary outcome was the number of intra-oral radiographs taken per course of treatment collected from a central fee claims database. Participants were 214 Scottish General Dental Practitioners. At the theory level, the Theory of Planned Behaviour explained 13% variance in the number of radiographs taken, Social Cognitive Theory explained 7%, Operant Conditioning explained 8%, Implementation Intentions explained 11%. Self-Regulation and Stage Theory did not predict significant variance in radiographs taken. Perceived behavioural control, action planning and risk perception explained 16% of the variance in number of radiographs taken. Knowledge did not predict the number of radiographs taken. The results suggest an intervention targeting predictive psychological variables could increase the implementation of this evidence-based practice, while influencing knowledge is unlikely to do so. Measures which predicted number of radiographs taken also predicted intention to take radiographs, and intention accounted for significant variance in behaviour (adjusted R(2)=5%: F(1,166)=10.28, p<.01), suggesting intention may be a possible proxy for behavioural data when testing an intervention prior to a service-level trial. Since psychological frameworks incorporate methodologies to measure and change component variables, taking a theory-based approach enabled the creation of a methodology that can be replicated for identifying factors predictive of clinical behaviour and for the design and choice of interventions to modify practice as new evidence emerges.

Adult↗
PubMed2005

Antigen processing is predictable: From genes to T cell epitopes.

In order to induce a cellular immune response, antigens have to be processed in intracellular compartments, transported, and presented by HLA molecules prior to recognition by specific T cells. Many of the events that contribute to antigen processing have been thoroughly investigated during the past years and are now well understood, which lead to a number of prediction programs. "Reverse immunology" has been used for about 10 years in order to identify T cell epitopes from pathogens or tumor-associated antigens. The advantages and pitfalls of T cell epitope prediction compared to classical experimental procedures such as epitope mapping and cloning experiments have been discussed many times. In this presentation, a number of internet programs that offer help in T cell epitope prediction (or prediction of antigen processing) will be discussed in the light of transfusion medicine. Some databases are listing published HLA ligands and T cell epitopes, others offer epitope prediction for many HLA class I or class II restrictions. In addition, a number of established programs will be demonstrated which are freely accessible at no cost in the world wide web for the prediction of either HLA-peptide binding, proteasomal processing of antigens, or both. Epitope prediction and processing prediction programs will be applied to minor histocompatibility antigens (miHAgs) and compared. This reflects the actual possibilities and limitations of such computer-aided work not only in cellular immunology, but also in transplantation immunology.

Animals↗
PubMed2005

Toxicogenomic analysis methods for predictive toxicology.

Toxicogenomics, the application of genomic data to elucidate or predict an organism's response to a toxicant, can inform the drug development process in important ways. It is apparent that standardized approaches to many types of toxicogenomic questions are still being formulated. Specifically, a significant body of proof of principle studies has emerged that demonstrates a range of statistical methodologies applied to predictive toxicology. These studies rely on class prediction methods--mathematical models generated using the gene expression profiles of known toxins from representative toxicological classes--to predict the toxicological effect of a compound based on the similarities between its gene expression profile and the profiles of a given toxicological class. Class prediction methods hold promise for increasing the rate at which compounds can be evaluated for toxicity early in the drug discovery process, while at the same time reducing the length of toxicological studies and their associated costs. Class prediction methods are informed by class comparison and class discovery steps, which inform, respectively, the selection of genes whose response can be used to distinguish among the toxicological classes and the number of classes distinguishable using the response of these genes. Together these steps use a variety of complementary statistical techniques to achieve a successful class prediction model. This report attempts to review some of the themes that appear to be emerging in the application of these techniques to predictive toxicology methods over toxicogenomics' short history.

Animals↗
PubMed2001

Predicting ambulatory function following lower extremity trauma using the functional capacity index.

This study evaluated the accuracy of experts' predictions of ambulatory function following lower extremity trauma using the Functional capacity index (FCI). Data from three orthopedic trauma studies designed to determine long-term function following specific types of lower extremity injuries were used to examine the extent of agreement between the reported and predicted ambulatory function of 921 subjects. Functional limitations reported by the cohort using a generalized health status measure and more detailed questions on lower extremity function were compared with those predicted by experts based on the injuries sustained. The overall agreement between predicted and self-reported FCI function for ambulation was relatively low (31%). In the majority of cases (80%), the disagreement differed by one functional level. Subjects were more likely to report worse function than predicted by the experts. Multivariate modeling identified different injuries, combinations of injuries, and patient characteristics that significantly influenced agreement. For example, subjects who sustained both a tibia and a femur fracture were three times more likely than subjects who did not sustain either fracture type to report poorer ambulatory function than predicted. Many challenges are faced in predicting long-term function following trauma. More empirical data are needed to inform the process. These data suggest that until the FCI can more accurately predict long-term ambulatory function following different lower extremity injuries, it should not be used for this purpose.

Activities of Daily Living↗
PubMed2003

Traffic sense--which factors influence the skill to predict the development of traffic scenes?

A study was conducted to evaluate the skill to predict the development of traffic situations. A stop-controlled intersection was filmed over several days, and 12 scenes with varying traffic complexity were selected. In half of the scenes, the traffic rules were violated, in half of the scenes, the rules were observed. A total of 36 participants were asked to watch the scenes and predict how the scene would most likely develop in the 2s after the film was paused. Additionally, the participants rated how certain they were about their prediction, and how complex and dangerous they assessed the scenes to be. With the method used here, experienced drivers were not found to make more correct predictions of situational development, and no difference in skill to predict could be found between genders. Nevertheless, more experienced drivers were more certain in their judgements and evaluated the situations on average as less complex and dangerous than did less experienced drivers. Scenes in which the traffic rules were violated were more difficult to predict correctly. The scenes in which the participants predicted violations were rated as more complex and dangerous. It is concluded that the low-cost method used here is more useful for examining which scenes are generally easy or difficult to predict and how they are experienced subjectively than to investigate differences in performance for different driver categories.

Accidents, Traffic↗
PubMed1999

The in-training examination in internal medicine: resident perceptions and lack of correlation between resident scores and faculty predictions of resident performance.

PURPOSE: We sought to survey residents' perceptions regarding the In-Training Examination in Internal Medicine and to assess the ability of faculty members to evaluate the knowledge base of internal medicine residents. SUBJECTS AND METHODS: Residents were asked about the perceived utility of the In-Training Examination and related self-directed educational activities. Residents predicted their own performance on the examination (into upper, middle, or lower tertile). Faculty predicted housestaffs scores, and residents predicted the scores of interns. RESULTS: Most residents (35/36; 97%) believed that the examination was useful, and 91% modified their study habits or clinical rotation schedule based on its results. Approximately half of the residents accurately predicted into which tertile they would score. Faculty predictions of resident performance on the examination were accurate 49% of the time, and resident predictions of intern scores were accurate 38% of the time. The sensitivity ofa lower-tertile prediction by faculty was 34%, with a specificity of 90%. The sensitivity of a resident prediction of a lower-tertile intern score was 15%, with a specificity of 98%. Both faculty and residents were more likely to overestimate than underestimate examination scores. CONCLUSION: Residents believe that the In-Training Examination is useful and frequently initiate educational interventions based on results. Faculty and residents lack the ability to evaluate accurately the knowledge of trainees that they supervise. In particular, both groups may be unable to identify trainees who are deficient in this element of clinical competence.

Clinical Competence↗
PubMed1996

A comparison of rapid amniotic fluid markers in the prediction of microbial invasion of the uterine cavity and preterm delivery.

OBJECTIVE: The purpose of this study was to evaluate amniotic fluid lactate dehydrogenase level in comparison with other rapid markers in prediction of microbial invasion of the uterine cavity and preterm delivery < or = 36 hours after amniocentesis. STUDY DESIGN: One hundred thirty-one women in preterm labor with intact membranes underwent transabdominal amniocentesis. Amniotic fluid was analyzed for leukocyte count, glucose level, lactate dehydrogenase level, and Gram stain. Cultures for aerobes, anaerobes, and Mycoplasma sp. were performed. Amniocentesis-to-delivery interval was calculated. The study group was divided and the findings compared according to amniotic fluid culture results and according to amniocentesis-to-delivery interval. Sensitivity, specificity, and positive and negative predictive value were calculated for lactate dehydrogenase, leukocyte count, glucose, and Gram stain in the prediction of positive amniotic fluid culture and preterm delivery < or = 36 hours after amniocentesis. Receiver-operator characteristic curve analysis, logistic regression analysis, t tests, and nonparametric tests were used. RESULTS: The prevalence of positive amniotic fluid cultures was 12% (16 of 131). The median lactate dehydrogenase level (1084 U/L) was significantly greater for women with a positive amniotic fluid culture than for those with a negative culture (median lactate dehydrogenase level 194 U/L; p < 0.0002). The critical values calculated for optimal performance in prediction of a positive amniotic fluid culture were a lactate dehydrogenase level > or = 419 U/L, leukocyte count > or = 50 cells/mm3 (50 x 10(6)/L) and glucose < or = 17 mg/dl (0.94 mmol/L). Lactate dehydrogenase, leukocyte count, glucose, and Gram stain were equally sensitive and specific in prediction of a positive amniotic fluid culture. Thirty-nine women (29.8%) gave birth < or = 36 hours after amniocentesis. The median lactate dehydrogenase level (414 U/L) was significantly greater among women giving birth < or = 36 hours after amniocentesis than among women giving birth > 36 hours after amniocentesis (median lactate dehydrogenase, 173 U/L; p < 0.001). Critical values of lactate dehydrogenase > or = 225 U/L, leukocyte count > or = 10 cells/mm3 (10 x 10(6)/L) and glucose < or = 34 mg/dl (1.9 mmol/L) were selected for optimal performance in prediction of amniocentesis-to-delivery interval < or = 36 hours. Lactate dehydrogenase level had the best sensitivity (74%) in prediction of delivery < or = 36 hours after amniocentesis in contrast to leukocyte count (49%), glucose (62%), and positive Gram stain (26%). Amniotic fluid lactate dehydrogenase values > or = 225 U/L were associated with a fivefold greater risk for delivery < or = 36 hours after amniocentesis (odds ratio 5.46, 95% confidence interval 2.00 to 14.87; p = 0.0006). CONCLUSION: Amniotic fluid lactate dehydrogenase level has diagnostic value in prediction of a positive amniotic fluid culture and delivery < or = 36 hours after amniocentesis. Lactate dehydrogenase is a readily available, inexpensive, rapid amniotic fluid marker that can be measured in any hospital laboratory.

Adult↗
PubMed1999

Accuracy of salivary estriol testing compared to traditional risk factor assessment in predicting preterm birth.

OBJECTIVE: The objective was to compare the predictive accuracy (percentage of correct vs incorrect predictions) of salivary estriol levels (SalEst; Biex, Inc, Dublin, Calif) with that of the modified Creasy score for predicting preterm labor followed by preterm delivery. STUDY DESIGN: A triple-blinded prospective trial was conducted at 8 US centers. RESULTS: Among 601 evaluable patients, serial salivary estriol testing correctly predicted the appropriate outcome 91% of the time and the Creasy scoring method correctly predicted the appropriate outcome 75% of the time (McNemar test P <. 001). Among subjects with Creasy scores >/=10 (high-risk group, n = 152), use of salivary estriol testing correctly predicted the end point 87% of the time, compared with only 7.2% correctly predicted by modified Creasy scoring (McNemar test P <.001). CONCLUSION: Salivary estriol assessment was more accurate in predicting outcome than was modified Creasy scoring.

Estriol↗
PubMed1997

Neural network modeling accurately predicts the functional outcome of stroke survivors with moderate disabilities.

OBJECTIVE: To predict the place of discharge or discharge Functional Independence Measure (FIM) score for stroke survivors with moderate disability using neural network modeling. Our previous work demonstrated that the FIM predicts the level of recovery for stroke survivors with either severe or mild disabilities. DESIGN: Neural network analysis. SETTING: Tertiary care rehabilitation program. PATIENTS: One hundred forty-seven consecutive stroke survivors admitted for rehabilitation with admission FIM scores between 37 and 96 were used as the training and internal test set. Seventeen other randomly selected stroke survivors were used as the external test set. INTERVENTION: A neural network model was developed using a small set of clinical variables and the admission FIM score. MAIN OUTCOME MEASURE: Neural network model predicting place of discharge or discharge FIM score. RESULTS: A working and accurate model was developed to predict the discharge FIM score. The model was able to predict the 17 external test cases with an accuracy = 88%, sensitivity = 83%, specificity = 91%, positive predictive value = 83%, and negative predictive value = 91%. CONCLUSION: Neural network modeling is useful in the prediction of functional recovery and helps in discharge planning and allocation of rehabilitation resources.

Aged↗
PubMed1992

Monte Carlo simulation studies on the prediction of protein folding types from amino acid composition.

In the methodology development for statistical prediction of protein structures, the founders of different methods usually selected different sets of proteins to test their predicted results. Therefore, it is hard to make a fair comparison according to the results they reported. Even if the predictions by different methods are performed for the same set of proteins, there is still such a problem: a method better that the other for one set of proteins would not necessarily remain so when applied to another set of proteins. To tackle this problem, a Monte Carlo simulation method is proposed to establish an objective criterion to measure the accuracy of prediction for the protein folding type. Such an objective accuracy is actually corresponding to the asymptotical limit genereated during the Monte Carlo simulation process. Based on that, it has been found that the average objective accuracy for predicting the all-alpha, all-beta, alpha + beta, and alpha/beta proteins by the least Euclid's distance method (Nakashima, H., K. Nishikawa, and T. Ooi. 1986. J. Biochem. 99:152-162) is 73.0% and that by the least Minkowski's distance method (Chou, P.Y. 1989. Prediction in Protein Structure and the Principles of Protein Conformation. Plenum Press. New York. 549-586) is 70.9%, indicating that the former is better than the latter. However, according to the original reports, the latter claimed a rate of correct prediction with 79.7% but the former with only 70.2%, leading to a completely opposite conclusion. This indicates the necessity of establishing an objective criterion, and a comparison is meaningful only when it is based on the objective criterion. The simulation method and the idea developed here also can be applied to examine any other statistical prediction methods.

Amino Acids↗
PubMed1992

Adenosine 5'-triphosphate consumption by smooth muscle as predicted by the coupled four-state crossbridge model.

We have proposed a four-state crossbridge model to explain contraction and the latch state in arterial smooth muscle. Ca(2+)-dependent crossbridge phosphorylation was the only postulated regulatory mechanism and the latchbridge (a dephosphorylated, attached crossbridge) was the only novel element in the model. In this study, we used the model to predict rates of ATP consumption by crossbridge phosphorylation (JPhos) and cycling (JCycle) during isometric and isotonic contractions in arterial smooth muscle; then we compared model predictions with experimental data. The model predicted that JPhos and JCycle were similar in magnitude in isometric contractions, and both increased almost linearly with myosin phosphorylation. The predicted relationship between isometric stress and ATP consumption was quasihyperbolic, but approximately linear when myosin phosphorylation was below 35%, in agreement with most of the available data. Muscle shortening increased the predicted values of JCycle up to 3.7-fold depending on shortening velocity and the level of myosin phosphorylation. The predicted maximum work output per ATP was 7.4-7.8 kJ/mol ATP and was relatively insensitive to changes in myosin phosphorylation. The predicted increase in JCycle with shortening was in agreement with available data, but the model prediction that work output per ATP was insensitive to changes in myosin phosphorylation was unexpected and remains to be tested in future experiments.

Adenosine Triphosphate↗
PubMed1997

Using fuzzy logic to predict response to citalopram in alcohol dependence.

INTRODUCTION: The prediction of patient response to new pharmacotherapies for alcohol dependence has usually not been successful with standard statistical techniques. We hypothesized that fuzzy logic, a qualitative computational approach, could predict response to 40 mg/day citalopram and 40 mg/day citalopram with a brief psychosocial intervention in alcohol-dependent patients. METHODS: Two data sets were formed with patients from our studies who received 40 mg/day citalopram alone (n = 34) or 40 mg/day citalopram and a brief psychosocial intervention (n = 28). The output variable, "response," was the percentage decrease in alcohol intake from baseline. Input variables included age, gender, baseline alcohol intake, and levels of anxiety, depression, alcohol dependence, and alcohol-related problems. RESULTS: A fuzzy rulebase was created from the data of 26 randomly chosen patients who received 40 mg/day citalopram and was used to predict the responses of the remaining eight patients. Eight rules related response with depression, anxiety, alcohol dependence, alcohol-related problems, age, and baseline alcohol intake. The average magnitude of the error in the predictions (RMSE) was 2.6 with a bias (ME) of 0.6. Predicted and actual response correlated (r = 0.99; p < 0.001). A fuzzy rulebase was created from the data of 28 randomly chosen patients who received 40 mg/day citalopram and a brief psychosocial intervention and was used to predict the responses of the remaining five patients. Six rules related response with age, anxiety, depression, alcohol dependence, and baseline alcohol intake with good predictive performance (RMSE = 6.4; ME = -1.5; r = 0.96; p < 0.01). CONCLUSIONS: This study indicates that fuzzy logic modeling can predict response to pharmacotherapies for alcohol dependence.

Alcoholism↗