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 775 records · Page 43Linked to original sources
PubMed2005

Evaluation of AUC(0-4) predictive methods for cyclosporine in kidney transplant patients.

Cyclosporine (CyA) is the most commonly used immunosuppressive agent in patients who undergo kidney transplantation. Dosage adjustment of CyA is usually based on trough levels. Recently, trough levels have been replacing the area under the concentration-time curve during the first 4 h after CyA administration (AUC(0-4)). The aim of this study was to compare the predictive values obtained using three different methods of AUC(0-4) monitoring. AUC(0-4) was calculated from 0 to 4 h in early and stable renal transplant patients using the trapezoidal rule. The predicted AUC(0-4) was calculated using three different methods: the multiple regression equation reported by Uchida et al.; Bayesian estimation for modified population pharmacokinetic parameters reported by Yoshida et al.; and modified population pharmacokinetic parameters reported by Cremers et al. The predicted AUC(0-4) was assessed on the basis of predictive bias, precision, and correlation coefficient. The predicted AUC(0-4) values obtained using three methods through measurement of three blood samples showed small differences in predictive bias, precision, and correlation coefficient. In the prediction of AUC(0-4) measurement of one blood sample from stable renal transplant patients, the performance of the regression equation reported by Uchida depended on sampling time. On the other hand, the performance of Bayesian estimation with modified pharmacokinetic parameters reported by Yoshida through measurement of one blood sample, which is not dependent on sampling time, showed a small difference in the correlation coefficient. The prediction of AUC(0-4) using a regression equation required accurate sampling time. In this study, the prediction of AUC(0-4) using Bayesian estimation did not require accurate sampling time in the AUC(0-4) monitoring of CyA. Thus Bayesian estimation is assumed to be clinically useful in the dosage adjustment of CyA.

Adult↗
PubMed2004

Prediction of liver fibrosis and cirrhosis in chronic hepatitis B infection by serum proteomic fingerprinting: a pilot study.

BACKGROUND: Most noninvasive predictive models of liver fibrosis are complicated and have suboptimal sensitivity. This study was designed to identify serum proteomic signatures associated with liver fibrosis and to develop a proteome-based fingerprinting model for prediction of liver fibrosis. METHODS: Serum proteins from 46 patients with chronic hepatitis B (CHB) were profiled quantitatively on surface-enhanced laser desorption/ionization (SELDI) ProteinChip arrays. The identified liver fibrosis-associated proteomic fingerprint was used to construct an artificial neural network (ANN) model that produced a fibrosis index with a range of 0-6. The clinical value of this index was evaluated by leave-one-out cross-validation. RESULTS: Thirty SELDI proteomic features were significantly associated with the degree of fibrosis. Cross-validation showed that the ANN fibrosis indices derived from the proteomic fingerprint strongly correlated with Ishak scores (r = 0.831) and were significantly different among stages of fibrosis. ROC curve areas in predicting significant fibrosis (Ishak score >or=3) and cirrhosis (Ishak score >or=5) were 0.906 and 0.921, respectively. At 89% specificity, the sensitivity of the ANN fibrosis index in predicting fibrosis was 89%. The sensitivity for prediction increased with degree of fibrosis, achieving 100% for patients with Ishak scores >4. The accuracy for prediction of cirrhosis was also 89%. Inclusion of International Normalized Ratio, total protein, bilirubin, alanine transaminase, and hemoglobin in the ANN model improved the predictive power, giving accuracies >90% for the prediction of fibrosis and cirrhosis. CONCLUSIONS: A unique serum proteomic fingerprint is present in the sera of patients with fibrosis. An ANN fibrosis index derived from this fingerprint could differentiate between different stages of fibrosis and predict fibrosis and cirrhosis in CHB infection.

Blood Proteins↗
PubMed1994

Assessment of operative risk in patients undergoing lung resection. Importance of predicted pulmonary function.

OBJECTIVE: To evaluate the ability of preoperative variables to identify patients at increased risk for complications after lung resection and the usefulness of predicted postoperative FEV1 as a marker of risk for adverse outcomes. DESIGN: Prospective analysis of a cohort of patients undergoing pulmonary resection. Complication rates were analyzed according to preoperative pulmonary variables, demographic variables, procedure performed, and predicted postoperative FEV1. Predicted postoperative FEV1 was calculated using a formula estimating the decline in preoperative FEV1 based on the number of bronchopulmonary segments removed during surgery. SETTING: A major teaching hospital and tertiary referral center. PATIENTS: A consecutive series of patients undergoing pulmonary resection. MEASUREMENTS AND MAIN RESULTS: Medical complications were recorded as part of an ongoing clinical database. The overall complication rate was low (17 percent rate of any complication, 1 percent death rate). Univariate predictors of complications included age > or = 60, male sex, history of smoking, a pneumonectomy procedure, and a low predicted postoperative FEV1. Hypercarbia (> or = 45 mm Hg) on preoperative arterial blood gas analysis, desaturation on exercise oximetry (< or = 90 percent), and a preoperative FEV1 less than 1 L were not predictive of complications. When the effect of these variables was controlled for in a multivariate analysis, a low predicted postoperative FEV1 remained the only significant independent predictor of complications. For each 0.2 L decrease in predicted FEV1, the odds ratio for complications was 1.46 (95 percent confidence interval [CI] 1.2 to 1.8). CONCLUSIONS: A low predicted postoperative FEV1 appears to be the best indicator of patients at high risk for complications, and it was the only significant correlate of complications when the effect of other potential risk factors was controlled for in a multivariate analysis. Pulmonary resection should not be denied on the basis of traditionally cited preoperative pulmonary variables, and a prediction of postoperative pulmonary function by a technique of simple calculation may be useful to identify patients at increased risk for medical complications.

Age Factors↗
PubMed2003

The use of selective bronchography in predicting reversal of neoplastic obstructive atelectasis.

OBJECTIVES: To assess the ability of selective bronchography to predict which patients with neoplastic postobstructive atelectasis will respond to interventional therapies directed at the reexpansion of the affected lung. Furthermore, to compare the utility of selective bronchography with the current predictive standard that reversal of postobstructive atelectasis is unlikely when it is > or = 4 weeks in duration (ie, the 4-week rule). DESIGN: A prospective observational study. SETTING: A tertiary care referral center/medical school. PATIENTS: Twenty-seven consecutive patients with advanced lung cancer or other malignancy, with documented neoplastic postobstructive atelectasis involving a total of 44 lobes. INTERVENTIONS: Lobar collapse was documented radiographically. The duration of atelectasis was investigated and quantified as accurately as possible. Prior to the use of interventional therapies, selective bronchography was performed on each collapsed lobe, and the results were documented. Bronchography results did not influence the decision to proceed with interventional therapies. Patients had each of their collapsed lobes manipulated by interventional techniques that were directed at reexpansion of the lung. One week after the patient underwent the intervention, the degree of reexpansion was assessed radiographically. RESULTS: Interventional therapies leading to significant reversal of airway narrowing were completed in all 44 lobes. These were successful in reexpanding 28 of 44 collapsed lobes (64%). Selective bronchography demonstrated the following two distinct patterns: an intact bronchial tree (ie, tree pattern); or the absence of a distinguishable, distal bronchial tree (ie, blush pattern). The sensitivity of selective bronchography to predict reexpansion is 1.00 (95% confidence interval [CI], 0.90 to 1.00), and its specificity is 0.56 (95% CI, 0.30 to 0.80). There were no complications attributable to selective bronchography. The sensitivity of the 4-week rule to predict reexpansion is 0.61 (95% CI, 0.41 to 0.78), and its specificity is 0.75 (95% CI, 0.48 to 0.93). The results of selective bronchography and use of the 4-week rule were significantly different in predicting which lobes would reexpand and which would not (p = 0.0026). Using selective bronchography to predict the reversal of lobar atelectasis, the positive predictive value of the tree pattern was 0.80 and the negative predictive value of the blush pattern was 1.00. The values for the 4-week rule are 0.81 and 0.52, respectively. CONCLUSIONS: Selective bronchography is a useful tool for predicting whether patients with neoplastic postobstructive atelectasis would benefit from interventional techniques that are directed at lobar reexpansion. Selective bronchography appears to be superior to the 4-week rule in this regard.

Aged↗
PubMed1999

Predicting the need for topical anesthetic in the pediatric emergency department.

OBJECTIVE: To investigate the potential for pediatric emergency department (ED) triage nurses to apply a topical anesthetic (ie, eutectic mixture of local anesthetic) for intravenous catheter (IV) insertion. METHODS: Prospective cross-sectional survey over a 2-month period, with post hoc application of internally developed prediction rules. Eligible patients were children presenting to the ED triage area of an urban children's hospital. RESULTS: A total of 2596 (86.7% of eligible children) had a triage nursing prediction performed. Nurse prediction of IV insertion had a sensitivity of 72% (95% CI: 66,78), a specificity of 90% (88,91), and a positive predictive value (PPV) of 49% (44,54). Objective factors such as high-risk medical history (chronic neurologic, hematologic, cardiac, endocrine, or gastrointestinal illness) and high-risk chief complaint (gastrointestinal illness, skin infection, and previous seizure) were incorporated into a predictive score used to predict IV insertion independently with a sensitivity of 33% (27,39) and a PPV of 43% (44,54). Addition of the objective predictors to nursing prediction increased the sensitivity to 76% (70,81) with a PPV of 43% (38,47). Of the patients, 95% received an IV insertion </=45 minutes after triage, 89% </=60 minutes after triage. Of the IV insertions, 68% were placed in the dorsum of the hand. CONCLUSIONS: The prediction of an experienced triage nurse can identify most patients requiring an IV in a pediatric ED. Incorporation of objective criteria other than nursing prediction into this decision process can decrease the amount of wasted product at the expense of less sensitive identification. The timing of IV insertion in our ED would allow for full medication effect of the currently marketed topical anesthetics in the majority of ED patients. topical anesthetic, intravenous cannulation, children, eutectic mixture of local anesthetic.

Administration, Topical↗
PubMed2005

Assessing the predictive accuracy of QUICKI as a surrogate index for insulin sensitivity using a calibration model.

The quantitative insulin-sensitivity check index (QUICKI) has an excellent linear correlation with the glucose clamp index of insulin sensitivity (SI(Clamp)) that is better than that of many other surrogate indexes. However, correlation between a surrogate and reference standard may improve as variability between subjects in a cohort increases (i.e., with an increased range of values). Correlation may be excellent even when prediction of reference values by the surrogate is poor. Thus, it is important to evaluate the ability of QUICKI to accurately predict insulin sensitivity as determined by the reference glucose clamp method. In the present study, we used a calibration model to compare the ability of QUICKI and other simple surrogates to predict SI(Clamp). Predictive accuracy was evaluated by both root mean squared error of prediction as well as a more robust leave-one-out cross-validation-type root mean squared error of prediction (CVPE). Based on data from 116 glucose clamps obtained from nonobese, obese, type 2 diabetic, and hypertensive subjects, we found that QUICKI and log (homeostasis model assessment [HOMA]) were both excellent at predicting SI(Clamp) (CVPE = 1.45 and 1.51, respectively) and significantly better than HOMA, 1/HOMA, and fasting insulin (CVPE = 3.17, P < 0.001; 1.67, P < 0.02; and 2.85, P < 0.001, respectively). QUICKI and log(HOMA) also had the narrowest distribution of residuals (measured SI(Clamp) - predicted SI(Clamp)). In a subset of subjects (n = 78) who also underwent a frequently sampled intravenous glucose tolerance test with minimal model analysis, QUICKI was significantly better than the minimal model index of insulin sensitivity (SI(MM)) at predicting SI(Clamp) (CVPE = 1.54 vs. 1.98, P = 0.001). We conclude that QUICKI and log(HOMA) are among the most accurate surrogate indexes for determining insulin sensitivity in humans.

Adult↗
PubMed2003

Ractopamine treatment biases in the prediction of pork carcass composition.

Carcass and live measurements of 45 barrows were used to evaluate the magnitude of ractopamine (RAC) treatment prediction biases for measures of carcass composition. Barrows (body weight = 69.6 kg) were allotted by weight to three dietary treatments and fed to an average body weight of 114 kg. Treatments were: 1) 16% crude protein, 0.82% lysine control diet (CON); 2) control diet + 20 ppm RAC (RAC16); 3) a phase feeding sequence with 20 ppm RAC (RAC-P) consisting of 18% crude protein (1.08% lysine) during wk 1 and 4, 20% crude protein (1.22% lysine) during wk 2 and 3, 16% crude protein (0.94% lysine) during wk 6, and 16% crude protein (0.82% lysine) during wk 6. The four lean cuts from the right side of the carcasses (n = 15/treatment) were dissected into lean and fat tissue. The other cut soft tissue was collected from the jowl, ribs, and belly. Proximate analyses were completed on these three tissue pools and a sample of fat tissue from the other cut soft tissue. Prediction equations were developed for each of five measures of carcass composition: fat-free lean, lipid-free soft tissue, dissected lean in the four lean cuts, total carcass fat tissue, and soft-tissue lipid mass. Ractopamine treatment biases were found for equations in which midline backfat, ribbed carcass, and live ultrasonic measures were used as single technology sets of measurements. Prediction equations from live or carcass measurements underpredicted the lean mass of the RAC-P pigs and underpredicted the lean mass of the CON pigs. Only 20 to 50% of the true difference in fat-free lean mass or lipid-free soft-tissue mass between the control pigs and pigs fed RAC was predicted from equations including standard carcass measurements. The soft-tissue lipid and total carcass fat mass of RAC-P pigs was overpredicted from the carcass and live ultrasound measurements. Prediction equations including standard carcass measurements with dissected ham lean alone or with dissected loin lean reduced the residual standard deviation and magnitude of biases for the three measures of carcass leanmass. Prediction equations including the percentage of lipid of the other cut soft tissue improved residual standard deviation and reduced the magnitude of biases for total carcass fat mass and soft-tissue lipid. Prediction equations for easily obtained carcass or live ultrasound measures will only partially predict the true effect of RAC to increase carcass leanness. Accurate prediction of the carcass composition of RAC-fed pigs requires some partial dissection, chemical analysis, or alternative technologies.

Adipose Tissue↗
PubMed2002

Test day and lactation yield predictions in Italian simmental cows by ARMA methods.

Autoregressive Moving Average (ARMA) models, originally developed in the contest of time series analysis, were used to predict Test Day (TD) yields of milk production traits in dairy cows. ARMA models areable to take into account both the average lactation curve of homogeneous groups of animals and the residual individual variability that may be explained in terms of probability models, such as Autoregressive (AR) and Moving Average (MA) processes. Milk, fat, and protein yields of 6000 Italian Simmental cows with 8 TD records per lactation were analyzed. Data were grouped according to parity (1st, 2nd, and 3rd calving) and fitted to a Box-Jenkins ARMA model in order to predict TD yields in five situations of incomplete lactations. Reasonable accuracies have been obtained for a limited horizon of prediction: average correlations among actual and predicted data were 0.85, 0.72, and 0.80 for milk, fat and protein yields when the first predicted TD was one step ahead (on average 42d) of the last actual record available. Cumulative 305-d yields were calculated using all actual (actual yields) or actual plus forecasted (estimated yields) daily yields. Accuracy of lactation predictions was remarkable even when only a few actual TD records were available, with values of 0.88 for milk and protein and 0.84 for fat for the correlations between actual and estimated yields when 6 out of 8 TD records were predicted. Accuracy rapidly increases with the number of actual TD available: correlations were about 0.96 for milk and protein and 0.93 for fat when 4 out of 8 TD records were predicted. In comparison with other prediction methods, ARMA modelsare very simple and can be easily implemented in data recording software, even at the farm level.

Animals↗
PubMed2005

Development of a model to predict growth of Clostridium perfringens in cooked beef during cooling.

The objective of this work was to develop a new model to predict the growth of Clostridium perfringens in cooked meat during cooling. All data were collected under changing temperature conditions. Individual growth curves were fit using DMFit. Germination outgrowth and lag (GOL) time was modeled versus temperature at the end of GOL using conservative assumptions. Each growth curve was used to estimate a series of exponential growth rates at a series of temperatures. The squareroot model was used to describe the relationship between the square root of the average exponential growth rate and effective temperature. Predictions from the new model were in close agreement with the data used to create the model. When predictions from the model were compared with new observations, fail-dangerous predictions were made a majority of the time. When GOL time was predicted exactly, many fail-dangerous predictions shifted toward the fail-safe direction. Two important facts regarding C. perfringens should impact future modeling research with this organism and may have broader food safety policy implications: (i) the normal variability in the response of the organism from replicate to replicate may be quite large (1 log CFU) and may exceed the current U.S. Food Safety Inspection Service performance standard, and (ii) the accuracy of the GOL time model has a profound influence upon the overall prediction, with small differences in GOL time prediction (approximately 1 h) having a very large effect on the predicted final concentration of C. perfringens.

Animals↗
PubMed1998

[The use of maternal serum cytokines in the predicting of the efficacy of tocolytic therapy in case of the threat of preterm labor].

OBJECTIVES: The purpose of our study was to evaluate the use of maternal serum IL-8, IL-6, IFN-gamma levels in the predicting of the efficacy of tocolytic therapy in preterm labor. MATERIALS AND METHODS: We investigated prospectively the group of 47 women in singleton pregnancies with threatened preterm labor in less than 36 weeks gestation and administered tocolytic therapy. RESULTS: In 19 of them tocolysis failed (group II and they delivered premature newborns (the group I--successful tocolysis consisted of remaining 28 women). The incidence of clinical chorioamnionitis, histologic chorioamnionitis and inherited infection of newborns was significantly higher among women refractory to tocolytic therapy (10.2%, 36.8%, 26.3% versus 0%, 3.6%, 0%, respectively, p < 0.05). Maternal serum IL-8, IL-6, IFN-gamma (by means of ELISA technique) and CRP, WBC, ESR levels were measured at the admission to the study. The mean WBC, ESR and the median (range) IFN-gamma (0 (0-7.1) and 0.9 (0-10.4) pg/ml, respectively) didn't differ in both groups. The concentrations of serum IL-8, IL-6, CRP were significantly higher in the group of failed tocolysis (median (range): IL-8: 22.7 (6.3-83.2) vs 3.0 (0-26.0) pg/ml; IL-6: 7.4 (0-21.0) vs 0 (0-11.3) pg/ml; CRP: 1.8 (0.6-7.0) vs 0.6 (0.6-3.9) mg/dl; p < 0.05). Serum IL-8 determinations (definition of abnormal test: > 8 pg/ml) were found the most reliable in the prediction of tocolysis failure with a sensitivity 87.5%, specificity 81.8%, positive predictive value 77.8%, negative predictive value 90% and accuracy 84.2%. Also reliable were IL-6 determinations (IL-6 > 6 pg/ml had a sensitivity 75%, specificity 90.9%, positive predictive value 85.7%, negative predictive value 83.3% and accuracy 84.2%) and CRP determinations (CRP > 1.2 mg/dl had a sensitivity 75%, specificity 81.8%, positive predictive value 75%, negative predictive value 81.8% and accuracy 78.9%). The efficacy of IFN-gamma, WBC and ESR was significantly lower. CONCLUSIONS: Our data revealed that the maternal serum IL-8, IL-6 and CRP determinations are very useful in the predicting of the efficacy of tocolytic therapy in women with threatened preterm labor. The use of IFN-gamma, WBC, ESR was significantly lower.

Cytokines↗
PubMed1999

Does site specific labeling of sextant biopsy cores predict the site of extracapsular extension in radical prostatectomy surgical specimen.

PURPOSE: We determine whether site specific labeling of sextant prostate biopsy cores predicts the site of extracapsular extension in a radical prostatectomy specimen, thereby justifying increased cost of pathological evaluation. MATERIALS AND METHODS: Between January 1994 and December 1997, 407 radical prostatectomies were performed at our institution by a single surgeon (H. L.). Surgical specimens showing extracapsular extension were examined by a single pathologist (J. M.) to identify the site of extension. Several different methods of submitting transrectal ultrasound guided biopsy cores were used since the majority of cases did not undergo biopsy at our institution. In 243 cases sextant biopsies were labeled right versus left. Of these cases 103 specimen cores were individually labeled. The ability of the positive biopsy core location to predict the location of extracapsular extension in the surgical specimen was determined. Univariate and multivariate logistic regression analyses were performed to assess the ability of biopsy core characteristics, including Gleason score, percentage of cancer in the core, core location and number of positive cores in the specimen, to predict the site of extracapsular extension. A similar analysis was performed for the 243 cases with right versus left core labeling. RESULTS: The positive predictive value was 8.9+/-2.2% for a single positive core to identify the location of extracapsular extension correctly in the individually labeled core cases. The absence of cancer in a sextant biopsy had a negative predictive value of 96.9+/-1.4%. The overall sensitivity was 59.4+/-3.8% for a positive biopsy core. In the right versus left core cases the positive predictive value was 12.9+/-3.0% with a sensitivity of 85.1+/-3.2%. In an individual core Gleason score 8 or greater and/or cancer in more than 50% of tissue enhanced the positive predictive value but not to a clinically useful level. Multivariate logistic regression identified Gleason score, number of positive ipsilateral cores and base position of the positive biopsy as the most predictive variables for the site of extracapsular extension. CONCLUSIONS: When submitting biopsy specimens by individually labeled core or right versus left core, the positive predictive value of an individual positive core for the location of extracapsular extension is not sufficient to guide the surgical decision to spare or excise a neurovascular bundle. Therefore, the clinical information provided by individually labeled or right versus left core labeling does not justify the increased associated costs.

Biopsy↗
PubMed2002

Supplemental value of troponin I combined with a clinical risk model to predict in-hospital events in intermediate and high risk patients with acute coronary syndromes.

To evaluate the supplemental value of serial troponin I (Trp) measurements when combined with a clinical model composed of six clinical parameters in predicting in-hospital adverse event rates, a total of 118 consecutive patients admitted over a 23-month period with intermediate- or high-risk unstable angina or non-Q wave myocardial infarction (MI) as defined by AHCPR criteria who had coronary angiography within 72 hours of hospitalization were studied. Presenting clinical characteristics were graded using a previously validated variation of the Braunwald criteria (RUSH model). The RUSH model clinical score includes six clinical parameters: age, diabetes, intravenous nitroglycerin, pre-admission calcium-channel and beta-blocker, ST depression and post-MI angina (< 2 weeks), and creates an estimated probability of MI or death. The RUSH model was compared to serial Trp levels drawn at 6-hour intervals (0, 6 and 12 hours). An abnormal Trp value was defined as > 2.0 mg/dl. Outcome measures included death, MI, recurrent chest pain and new ST or T changes and enzyme elevation. One death, 23 MIs and 24 other adverse clinical events occurred. The event group had a RUSH score predictive of 12.7 12.4% risk and the no-event group had a score of 13.2 10.2% risk (p = 0.64). The Trp positive group had a clinical score predicting 14.2 13.2% risk and the Trp negative group had a score of 11.7 9.3% risk (p = 0.21). Patients with elevated Trp had an adverse event rate of 32/50 (64%) vs. 21/68 (31%) in patients with normal Trp (p < 0.0004). Elevated Trp had 60.4% sensitivity and 72.3% specificity, odds ratio of 3.97 (1.71 9.33), as well as 64% positive and 69.1% negative predictive values for predicting adverse events. Thus, there was significant incremental value to adding Trp to the clinical score when predicting outcomes in patients with intermediate- and high-risk clinical scores. When Trp was abnormal, it was useful when predicting higher risk; if Trp was normal, it was useful predicting lower but still elevated risk. Consequently, in a population selected for intermediate and high risk, the presence or absence of elevated Trp I is a sensitive and specific additive predictor to clinical score to predict need for revascularization and adverse in-hospital outcomes, as suggested in current guidelines.

Aged↗
PubMed2002

[ROC curve analysis of factors predictive of response to treatment with interferon plus ribavirin in patients with chronic hepatitis C relapse after previous interferon treatment].

This study aimed to identify the factors predictive of response before the initiation of treatment and throughout the treatment period in patients with chronic hepatitis C relapse after treatment with interferon-a who were retreated with a standard regimen of interferon-a plus ribavirine and followed up for 40 months. Forty-four patients (40 with genotype 1, four without genotype 1) were included in the study. Four patients (genotype 1) were excluded because of adverse effects. The rate of maintained response was 55% (50% genotype 1, 100% non-genotype 1). The stage of histological damage (>2), glutamic-pyruvic transaminase (GPT) concentration (< or = 26 UI/l) and the association between the GPT concentration and the detection of the RNA-HCV in the first and third treatment months were the variables with an area under the ROC curve and a confidence interval >0.5. The probability of predicting a maintained response (negative predictive value) if the stage of histological lesion was <2 was 62.9%, while the positive predictive value was 100%. During the treatment, the disappearance of the RNA-HCV together with GPT values < or =26 in the first treatment month were the best predictive values. In this case, the negative predictive value was 78.3% and the positive predictive value was 76.5% (OR: 11.7, 2.6-52.2). Furthermore, the GPT value with the best predictive value (<26 UI/l) was a more effective predictor of the response to treatment than the normal value of the GPT. Finally, the GPT values >26 UI/l and the detection of RNA/HCV in the first or third treatment month were certain predictors of the absence of response but with low sensitivity (10-12%). It was concluded that is possible to predict the response to the combined treatment with an acceptable level of confidence, although not unequivocally. Ninety percent of the patients would be candidates for maintaining treatment for at least 6-12 months, while approximately 10% could undergo early interruption of treatment due to the absence of response.

Adult↗
PubMed2003

Predicting the need for hospital admission in patients with intentional drug overdose.

BACKGROUND: Self-poisoned patients are often admitted to a medical unit. However, often no treatment is given. We have developed a model to predict those patients who will not be treated and how long patients should be observed before this prediction can be safely made. METHODS: In this retrospective study a model to predict treatment was developed based on cases of self-poisoning in 1996 and validated on cases between 1997 and 1999. In a teaching hospital in The Netherlands 299 adults performing 353 episodes of self-poisoning were studied. The main outcome measures were predicted versus initiated medical treatment, time to prediction and time to initiation of treatment. RESULTS: The model predicted that in 51% (156/307) of all autointoxications no treatment would be given. In 2% (6/307) of all cases, treatment was incorrectly not predicted. All but one of these were preventive treatments based on the ingested compound. 4.5 hours after admission no additional patients fulfilled the criteria for prediction of treatment and all treatments were started within 4.5 hours. CONCLUSIONS: In 51% of patients that present with an autointoxication the model accurately predicts that no treatment will be initiated. This decision can be made in the first 4.5 hours after presentation. This model can be used for a first screening of patients. It can also be used as a basis for a further prospective study to establish rational guidelines in the management of these patients.

Adult↗
PubMedDate not supplied

Predicting antigenic determinants in proteins: looking for unidimensional solutions to a three-dimensional problem?

In a recent review, Hopp (Peptide Research 6:183-190, 1993) claimed that the Hopp and Woods hydrophilicity method for locating antigenic determinants is superior to all other existing methods for predicting the B cell epitopes of proteins but that it is not useful to aid the investigator in producing peptide-protein cross-reactive antisera. In this article, we challenge both these assertions. Most investigators utilize antigenicity prediction algorithms because they wish to produce anti-peptide antibodies capable of cross-reacting with the intact protein. All prediction methods are based on propensity scales for the 20 amino acids, which describe the tendency of each residue to be associated with properties such as hydrophilicity, surface accessibility or segmental mobility. When we compared the prediction efficacy of 22 different scales, taking into account both correct and incorrect predictions, we found that none of the scales gave a level of correct prediction higher than about 50%-60%. If no antigenicity was found in a particular region of the protein, we took the view that hydrophilicity peaks located in that region amounted to wrong predictions. The much higher success rate reported by Hopp for this method stems from the way he assesses prediction efficacy, i.e., by counting the number of known epitopes located inside and outside hydrophilicity peaks. Reasons for the low success rate of antigenicity prediction are discussed. In most cases, it is unrealistic to try to reduce the complexity of discontinuous, conformational epitopes to simple, linear peptide models.

Chemical Phenomena↗
PubMed1993

Artificial neural networks for predicting failure to survive following in-hospital cardiopulmonary resuscitation.

BACKGROUND: Neural networks are an artificial intelligence technique that uses a set of nonlinear equations to mimic the neuronal connections of biological systems. They have been shown to be useful for pattern recognition and outcome prediction applications, and have the potential to bring artificial intelligence techniques to the personal computers of practicing physicians, assisting them with a variety of medical decisions. It is proposed that such an artificial neural network can be trained, using information available at the time of admission to the hospital, to predict failure to survive following in-hospital cardiopulmonary resuscitation (CPR). METHODS: The age, sex, heart rate, and 21 other clinical variables were collected on a consecutive series of 218 adult patients undergoing CPR at a 295-bed public acute-care hospital. The data set was divided into two groups. A neural network was trained to predict failure to survive to discharge following CPR, using one group as the training set and the other as the testing set. The procedure was then reversed, and the results of the two networks were combined to form an aggregate network. RESULTS: The trained aggregate neural network had a sensitivity of 52.1% and a positive predictive value of 97% for the prediction of failure to survive following CPR. The relative risk of actually failing to survive to discharge following CPR for a patient predicted not to survive was 11.3 (95% CI 3.3 to 38.2). CONCLUSIONS: Predicting failure to survive following CPR is but one possible application of neural network technology. It demonstrates how this technique can assist physicians in medical decision making. Future work should attempt to improve the positive predictive value of the neural network, to consider combining it with an expert system, and to compare it with other predictive tools. Once validated, the network can be distributed as a separate application for use by practicing physicians.

Adolescent↗
PubMed1998

Practical clinical application of predictive factors in prostate cancer. A review with an emphasis on quantitative methods in tissue specimens.

Predictive factors stratify cancer patients into homogeneous groups for treatment. There is an acute need for accurate predictive factors in patients with prostate cancer given the marked variation in treatment recommendations. These factors should be obtained prior to therapy and should include patient factors, serum factors and tissue-specific factors derived from biopsies. This review evaluates the current state of knowledge regarding quantitative methods in prostate tissue specimens, classifying predictive factors in prostate cancer into four categories. The first category, predictive factors recommended for widespread clinical use, includes Gleason grade; nontissue markers include clinical stage and serum prostate-specific antigen. The second category, predictive factors that are often collected but of unproven significance, includes DNA ploidy and volume of cancerous material in the needle biopsy. The third category, predictive factors not used for routine patient management, includes cell proliferation markers (mitotic figures, proliferating cell nuclear antigen, Ki-67 and MIB-1), apoptotic markers, microvessel density and perineural invasion. The fourth category, predictive factors under investigation, includes morphometric features, such as nuclear roundness and size, chromatin texture, silver-staining nucleolar organizer regions and nucleolar size. Standards are required for virtually every aspect of morphometric study, and these features require validation before their acceptance as clinically relevant in prostate cancer. Predictive factors in radical prostatectomies include cancer volume and extent in radical prostatectomy specimens and quantitation of number and size of lymph node metastases. Neural network models provide greater accuracy for combinations of predictive factors than traditional statistical methods of analysis, such as logistic regression and Cox models, and are expected to be incorporated into routine use in the next few years.

Humans↗
PubMed1978

Predicting hearing level from the acoustic reflex. A comparison of three methods.

Three prediction methods based on the acoustic reflex noise-tone difference (NTD) were used to predict hearing level in 370 subjects. The methods were two versions of the sensitivity prediction by acoustic reflex (SPAR) and a formula for estimating hearing threshold level. With each method, hearing loss was correctly predicted in more than one half of the subjects, while serious predictive error occurred in less than 10%. Age was found to be an important factor in hearing level prediction. Predictive accuracy for each method decreased systematically as a function of age. Predictive accuracy for the two SPAR methods decreased dramatically in subjects with minor middle ear and/or tympanogram abnormalities. Nevertheless, the value of the acoustic reflex NTD in predicting hearing level was confirmed.

Acoustic Impedance Tests↗