SEARCH · Search PubMed
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.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
OBJECTIVE: To assess whether physicians can identify very low likelihood of survival and very low likelihood of favorable functional outcome in adult nontrauma patients before admission to the intensive care unit (ICU) from the emergency department (ED). DESIGN: Prospective survey. SETTING: University hospital ED and ICU. PARTICIPANTS AND PATIENTS: Critical care fellows and ED physicians and all adult nontrauma patients admitted to the ICU from the ED over 1 yr. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The survey compared predictions of poor outcome from three sources: critical care fellows, ED physicians, and the admission Mortality Probability Model (MPM0). All patients were followed until hospital death or hospital discharge. Six-month follow-up data were obtained for patients predicted to have a < 2% chance of surviving with favorable functional outcome. In the ED, critical care fellows and ED physicians predicted likelihood of patient survival and likelihood of favorable functional outcome. MPM0 estimates of mortality were determined. The sensitivities, specificities, and positive predictive values were calculated for the predictions of < 2% survival and the predictions of < 2% chance of favorable functional outcome made by each prediction group. Complete data were obtained on 236 (96%) of 243 eligible patients. With regard to hospital mortality rate, fellows' predictions had a sensitivity of 27%, a specificity of 99%, and a positive predictive value of 88%; ED physicians' predictions had a sensitivity of 24%, a specificity of 98%, and a positive predictive value of 81%; and MPM0 predictions had a sensitivity of 2%, a specificity of 100%, and a positive predictive value of 100%. With regard to mortality rate combined with poor functional outcome, fellows' predictions had a sensitivity of 35%, a specificity of 99%, and a positive predictive value of 96%; ED physicians' predictions had a sensitivity of 37%, a specificity of 99%, and a positive predictive value of 96%. CONCLUSIONS: If a cutoff point of < 2% predicted survival is used in the triage of patients away from the ICU, the MPM0 has too low a sensitivity to be used as an effective screen. The low sensitivities and relatively low positive predictive values with wide confidence intervals of physician predictions of < 2% survival also preclude their use in triage. The addition of functional outcome as an end point improves the sensitivity, specificity, and positive predictive value of subjective predictions, making triage of patients away from the ICU at the time of ED evaluation a realistic possibility.
A variable that predicts an outcome with sufficient accuracy is called a predictive factor. Predictive factors can be divided into three types based on the outcomes to be predicted and on the accuracy with which they can be predicted. These three types include risk factors, where the main outcome of interest is incidence and the predictive accuracy is less than 100%; diagnostic factors, where the main outcome of interest is also incidence but the predictive accuracy is almost 100%; and prognostic factors, where the main outcome of interest is death and the predictive accuracy is variable. Surrogate outcomes are predictive factors that are used for a purpose beyond the prediction of an outcome--surrogate outcomes are predictive factors that are substituted for the true outcome in order to determine the effectiveness of an intervention. Surrogate outcomes used in clinical trials are called intermediate endpoints and surrogate endpoints. Predictive factors used as surrogate outcomes have a poor accuracy rate in predicting the true outcome; aggregating risk factors increases predictive accuracy. Artificial neural networks effectively combine predictive factors. Aggregating predictive factors increases the degree of linkage of the surrogate outcome to the true outcome. The resulting increase in predictive accuracy allows enrollment of people most likely to benefit from intervention. This increases the trial's efficiency, reducing the number of people required to assess a chemopreventive agent.
Height reduction by means of treatment with high doses of sex steroids in constitutionally tall stature (CTS) is a well known, though still controversial, therapy. The establishment of the effect of such therapy is dependent on the height prediction method applied. We evaluated the reliability of various prediction methods together with the subjective clinician's judgment in 143 untreated children (55 boys and 88 girls) with CTS and the effect of height-reductive therapy in 249 tall children (60 boys and 159 girls) treated with high doses of sex hormones (cases). For this purpose, we compared the predicted adult height with the attained height at a mean adult age of 25 yr and adjusted the therapeutic effect for differences in bone age (BA), chronological age (CA), and height prediction between untreated and treated children. At the time of the height prediction, controls were significantly shorter, had more advanced estimated BAs (except for the BA according to Greulich and Pyle in boys), had lower target heights, and had smaller adult height predictions compared with the CTS patients (P < 0.05). At the time of the follow-up, CTS patients were significantly taller than controls for both boys and girls (P < 0.02). In controls, a large variability was found for the errors of prediction of the various prediction methods and in relation to CA. The prediction according to Bailey and Pinneau systematically overestimated adult height in CTS children, whereas the other prediction methods (Tanner-Whitehouse prediction and index of potential height) systematically underestimated final height. The mean (SD) absolute errors of the prediction methods varied from 2.3 (1.8) to 5.3 (4.3) cm in boys and from 2.0 (1.9) to 3.7 (3.5) cm in girls. They were significantly negatively correlated with CA (r = [minus 0.27 to -0.65; P < 0.05), except for the Tanner-Whitehouse prediction in boys, indicating that height prognosis is more reliable with increasing CA. In addition, experienced clinicians gave accurate height predictions by evaluating the growth chart of the child while taking into account various clinical parameters, such as CA, BA, and pubertal stage. The effect of sex hormone therapy was assessed by means of multiple regression analysis while adjusting for differences in height prediction, CA, and BA at the start of therapy between treated and untreated children. The mean (SD) adjusted effect varied from -0.5 (2.4) to 0.3 (1.4) cm in boys and from -0.6 (2.1) to 2.4 (1.4) cm in girls. The adjusted height reduction was dependent on the BA at the time of start of sex hormone therapy and was more pronounced when treatment was started at a younger BA. In boys, the treatment effect was significantly negative at BAs exceeding 14-15 yr. After cessation of therapy, additional mean (SD) growth of 2.4 (1.2) and 2.7 (1.1) cm was observed for boys and girls, respectively. The mean (SD) BA according to Greulich and Pyle at that time was 17.1 (0.7) yr for boys and 15.2 (0.6) yr for girls. These data demonstrate that height prediction in children with CTS is inaccurate in boys, but clinically acceptable in girls. With increasing age, height prognosis became more accurate. Overall, the height-reducing effect of high doses of sex hormones in children with CTS was limited, especially in boys. However, a significant effect of treatment was observed when treatment was started at BAs less than 14-15 yr, depending on the method of BA assessment. In boys, treatment appeared to be contraindicated at BAs older than 14-15 yr, because androgen administration caused extra growth instead of growth inhibition. It is recommended that referral should take place early, preferably before puberty, for careful monitoring of growth and height prediction. Moreover, it is recommended not to discontinue therapy before complete closure of the epiphyses of the hand has occurred to avoid considerable posttreatment growth.
The predictive limits of the amino acid composition for the secondary structural content (percentage of residues in the secondary structural states helix, sheet, and coil) in proteins are assessed quantitatively. For the first time, techniques for prediction of secondary structural content are presented which rely on the amino acid composition as the only information on the query protein. In our first method, the amino acid composition of an unknown protein is represented by the best (in a least square sense) linear combination of the characteristic amino acid compositions of the three secondary structural types computed from a learning set of tertiary structures. The second technique is a generalization of the first one and takes into account also possible compositional couplings between any two sorts of amino acids. Its mathematical formulation results in an eigenvalue/eigenvector problem of the second moment matrix describing the amino acid compositional fluctuations of secondary structural types in various proteins of a learning set. Possible correlations of the principal directions of the eigenspaces with physical properties of the amino acids were also checked. For example, the first two eigenvectors of the helical eigenspace correlate with the size and hydrophobicity of the residue types respectively. As learning and test sets of tertiary structures, we utilized representative, automatically generated subsets of Protein Data Bank (PDB) consisting of non-homologous protein structures at the resolution thresholds < or = 1.8A, < or = 2.0A, < or = 2.5A, and < or = 3.0 A. We show that the consideration of compositional couplings improves prediction accuracy, albeit not dramatically. Whereas in the self-consistency test (learning with the protein to be predicted), a clear decrease of prediction accuracy with worsening resolution is observed, the jackknife test (leave the predicted protein out) yielded best results for the largest dataset (< or = 3.0A, almost no difference to the self-consistency test!), i.e., only this set, with more than 400 proteins, is sufficient for stable computation of the parameters in the prediction function of the second method. The average absolute error in predicting the fraction of helix, sheet, and coil from amino acid composition of the query protein are 13.7, 12.6, and 11.4%, respectively with r.m.s. deviations in the range of 8.6 divided by 11.8% for the 3.0 A dataset in a jackknife test. The absolute precision of the average absolute errors is in the range of 1 divided by 3% as measured for other representative subsets of the PDB. Secondary structural content prediction methods found in the literature have been clustered in accordance with their prediction accuracies. To our surprise, much more complex secondary structure prediction methods utilized for the same purpose of secondary structural content prediction achieve prediction accuracies very similar to those of the present analytic techniques, implying that all the information beyond the amino acid composition is, in fact, mainly utilized for positioning the secondary structural state in the sequence but not for determination of the overall number of residues in a secondary structural type. This result implies that higher prediction accuracies cannot be achieved relying solely on the amino acid composition of an unknown query protein as prediction input. Our prediction program SSCP has been made available as a World Wide Web and E-mail service.
In this study, we examined the effect of prediction accuracy on reaction time (RT). Subjects performed on three blocks of choice RT trials, all of which involved the mapping of four stimuli (red, green, 1, or 0) onto two response keys. The subjects were told that the four stimuli were equally probable and that their task was to respond to each stimulus onset by pressing the correct key. In one block (stimulus-prediction), the subjects predicted, prior to each trial, the precise stimulus that would appear. In a second block (category-prediction), the subjects predicted the category of the stimulus (i.e., color or digit) that would appear. In a third block (no-prediction), the subjects simply responded to each stimulus without making a prior prediction. In the stimulus-prediction block, RT was faster for correct predictions than for incorrect predictions. In addition, RT was faster on trials in which an incorrect prediction involved the correct category than on trials in which it involved the incorrect category: that is, a "half-wrong" prediction was better than an "all-wrong" prediction. In the category-prediction block, RT was faster when the stimulus category was predicted correctly than when it was not. There was little evidence of a response-facilitation contribution to the correct-prediction effect. These results permit inferences concerning the encoding and organization of information in memory.
To study the accuracy with which long-term prognosis can be predicted in patients with coronary artery disease, prognostic predictions from a data-based multivariable statistical model were compared with predictions from senior clinical cardiologists. Test samples of 100 patients each were selected from a large series of medically treated patients with significant coronary disease. Using detailed case summaries, five senior cardiologists each predicted one- and three-year survival and infarct-free survival probabilities for 100 patients. Fifty patients appeared in multiple samples for assessing interphysician variability. Cox regression models, developed using patients not in the test samples, predicted corresponding outcome probabilities for each test patient. Overall, model predictions correlated better with actual patient outcomes than did the doctors' predictions. For three-year survival, rank correlations were 0.61 (model) and 0.49 (doctors). For three-year infarct-free survival predictions, correlations with outcome were 0.48 (model) and 0.29 (doctors). Comparisons by individual doctor revealed Cox model three-year survival predictions were better than those of four of five doctors (model predictions added significant [p less than 0.05] prognostic information to the doctor's predictions, whereas the converse was not true). For infarct-free survival, the Cox model was superior to all five doctors. Where predictions were made by multiple doctors, the interphysician variability was substantial. In coronary artery disease, statistical models developed from carefully collected data can provide prognostic predictions that are more accurate than predictions of experienced clinicians made from detailed case summaries.
This study develops a "time-insensitive" predictive instrument for acute myocardial infarction mortality that would be useful both as a real-time clinical decision aid in the emergency medical setting and also for retrospective assessment and comparison of medical care based on risk-adjusted mortality predictions. This was done using prospectively-collected data on 5,773 patients with chief complaints of chest pain or other symptoms suggesting acute cardiac ischemia who came to six New England hospitals over a 2-year period. In phase one, based upon 4,099 patients, multivariate logistic regression was used to develop the predictive instrument. In phase two, its accuracy and diagnostic performance were tested on an independent sample of 1,387 patients presenting with symptoms compatible with acute cardiac ischemia. Discrimination between patients who lived and those who died was reflected by receiver-operating characteristic (ROC) curve areas of 0.85, 0.80, and 0.76, respectively, for all emergency department study subjects regardless of final diagnosis, subjects who proved to be having acute cardiac ischemia, and subjects who proved to be having acute infarction. Good calibration was shown by the fact that the predicted mortality was found to not vary significantly from actual mortality rates across deciles of predicted probabilities from 0% to 100%. In phase three, based on all 945 study subjects with acute myocardial infarction, each of the six hospitals' actual mortality rates were compared to their rates predicted by the predictive instrument. Actual hospital mortality rates ranged from 9.9% to 19.3%, with one hospital having a significantly higher rate (P = 0.005) and two hospitals both). Predicted mortality rates ranged from 13.4% to 19.4%, with one hospital having a significantly higher predicted rate (P = 0.005) and two hospitals having significantly lower predicted rates (P = 0.04 and P = 0.03). Individual hospitals' differences between actual and predicted mortality ranged from -3.4% to +3.1% (all NS). When grouped by hospital type, the actual mortality rates were 14.9%, 17.3%, and 13.0%, respectively, for urban teaching, smaller city teaching, and rural nonteaching hospitals (all NS). The predicted mortality rates were 16.5%, 17.1%, and 13.6%, respectively, with the rate for rural nonteaching hospitals being significantly lower (P = 0.009). No hospital type had significant differences between their actual and predicted mortality rates (NS). The time-insensitive predictive instrument for acute infarction mortality shows potential for risk-adjusted studies of hospitals mortality for multihospital groups, hospital-to-hospital comparisons, and within-hospital assessment.(ABSTRACT TRUNCATED AT 400 WORDS)
Three methods for predicting difficult births were tested on 4,140 yearling heifers measured before breeding and(or) at pregnancy check approximately 6 mo later. These heifers were from 115 beef herds in Alberta and British Columbia. The overall incidence of dystocia for normal presentations was 26.5%, which included 17.2% easy assists, 7.7% hard pulls, and 1.6% Caesarean sections. In Method 1, heifers were predicted as difficult (hard pull and Caesarean section) or easy (unassisted and easy pull) calvers by dividing their pelvic area (PA) by previously calculated PA to calf birth weight (PA/BWT) ratios. The ratio used depended on heifer weight and age. Method 1 predicted 63.7% to be difficult calvers. Of these only 10.4% were actually difficult calvers. The accuracy of this method was 40.0% and was not a useful on-farm method for predicting difficult births in first-calf, 2-yr-old heifers. In Method 2, 3,278 heifers measured before breeding and 1,125 heifers measured at pregnancy check were predicted as difficult or easy calvers by dividing their PA by 4.19 at prebreeding or 5.51 at pregnancy check. These values were PA/calf BWT ratios previously determined to be threshold levels. The accuracy of Method 2 was 78.5% and culling by this method would have reduced difficult birth rate by 9.6%. However, of the 738 heifers (16.8%) predicted to be difficult calvers, 86.0% actually calved easily. Heifers predicted to be easy calvers by Method 2 were heavier (P < .001), had a larger PA (P < .001), had more PA per kilogram of BW (P < .001), and had heavier (1.0 kg) calves at birth (P = .05) than heifers predicted to be difficult calvers. In Method 3, 3,269 heifers measured before breeding and 1,087 heifers measured at pregnancy check were predicted as difficult or easy calvers by dividing their PA by their BW. Heifers having ratios that were among the lowest 16% of the herd were predicted to be difficult calvers, and the rest were predicted to be easy calvers. The accuracy of Method 3 was 79.4% and culling by this method would have reduced difficult birth rate by 9.5%. However, of the 677 (15.5%) heifers predicted to be difficult calvers, 85.7% actually calved easily. Heifers predicted to be easy calvers by Method 3 were lighter (P < .001), had a larger PA (P < .001), had more PA per kilogram of BW (P < .001), and had lighter (1.2 kg) calves at birth (P = .04) than heifers predicted to be difficult calvers.(ABSTRACT TRUNCATED AT 400 WORDS)
The prediction of outcome following severe traumatic brain injury has received considerable attention in recent years. Previous prediction studies have focused on a long-term follow-up or prediction period. The reported outcome measures generally adopted a global approach (e.g. independent living) in terms of the prediction of physical function. The objective of the present study was to construct clinically useful predictive equations of motor system status, as represented by selected postural reactions (indicators of central nervous system function). Specifically, these equations would serve to predict the recovery of equilibrium and protective reactions both at 3 and 6 months post-injury, respectively. A stepwise multiple logistic regression analysis was performed, where nine predictive variables were considered using a multivariate approach. The results indicate that coma duration followed by age contribute significantly to the predictive capability of the models at both 3 and 6 months post-injury. Specifically, at 3 months, the predictive variables 'coma duration' and 'age' enabled an 84.62% correct prediction rate, whereas, at 6 months, 'coma duration' and 'age' enabled a 79.49% correct prediction rate. In addition, the exact probabilities (for given sample ages and coma durations) and associated 95% confidence intervals were calculated based on the predictive models obtained. The theoretical framework underlying these predictive models can form the basis for further studies. Furthermore, these preliminary predictive models have potential implications for early treatment planning and patient management.
OBJECTIVES: This study examines the efficacy of the predicting power for hospital mortality and functional outcome of three different scoring systems for head injury in a neurosurgical intensive care unit (NICU). DESIGN: On the day of admission, data were collected from each patient to compute the Acute Physiology, Age, and Chronic Health Evaluation (APACHE) II and III, and Glasgow Coma Scale (GCS) scores. Hospital mortality was defined as the deaths of patients before discharge from hospital. Early mortality was defined as death before the 14th day after admission. Late mortality was defined as death after the 15th day from admission. Functional outcome was evaluated by Index of Independence in Activities of Daily Living (Index of ADL). SETTING: An 8-bed NICU in a 1270-bed medical center in Taichung Veterans General Hospital. PATIENTS AND PARTICIPANTS: Two hundred non-selected patients with acute head injury were included in our study in a consecutive period of 2 years. Patients less than 14 years old were not included. INTERVENTIONS: None. MEASUREMENTS AND RESULTS: Sensitivity, specificity and correct prediction outcome were measured by the chi-square method in three scoring systems. The Youden index was also obtained. The best cut-off point in each scoring system was determined by the Youden index. The difference in Youden index was calculated by Z score. A difference was also considered if the probability value was less than 0.05. The area under Receiver Operating Characteristic (ROC) curve was computed. Then the area under ROC of each scoring system was compared by Z score. There was statistical significance if p was less than 0.05. For prediction of hospital mortality, the best cut-off points are 55 for APACHE III, 17 for APACHE II and 5 for GCS. The correct prediction outcome is 82.4% in APACHE III, 78.4% in APACHE II and 81.9% in the GCS. The Youden index has best cut-off points at 0.68 for APACHE III 0.59 for APACHE II, and 0.56 for GCS. The area under Receiver Operating Characteristic (ROC) curve is 0.90 in the APACHE III, 0.84 in the APACHE II and 0.86 in the GVS. There are no statistical differences among APACHE III and II, and GCS in terms of correct prediction outcome, Youden Index and the area under the ROC curve. Other physiological variables excluding GCS in APACHE III and II (AP III-GCS, AP II-GCS) have less statistical value in the determination of mortality for acute head injury. For the prediction of late mortality, APACHE III and II yield significantly better results in the area under the ROC curve, correct prediction and Youden index than those of GCS. Other physiological variables (AP III-GCS and AP II-GCS) play an important role in the prediction of late mortality in APACHE scores. For prediction of the functional outcome of surviving patients with acute head injury, the APACHE III yields the best results of correct prediction outcome, Youden index and the area under the ROC curve. CONCLUSION: The APACHE III and II may not replace the role of GCS in cases of acute head injury for hospital or early mortality assessment. But for prediction of the late mortality, the APACHE III and II have better accuracy than GCS. Other physiological variables excluding GCS in the APACHE system play a crucial contribution for late mortality. GCS is simple, less time-consuming and economical for patients with acute head injury for the prediction of hospital and early mortality. The APACHE III provides better prediction for severe morbidity than GCS and APACHE II. Therefore, the APACHE III provides a good assessment not only for hospital and late mortality, but also for functional outcome.
OBJECTIVE: Our aim was to determine the effect of lung resection on spirometric lung function and to evaluate the accuracy of simple calculation in predicting postoperative pulmonary function in patients undergoing lung resection. DESIGN: We reviewed preoperative and postoperative pulmonary function test results on patients who were followed in the multidisciplinary lung cancer clinic between July 1991 and March 1994 and who underwent lung resection. The predicted postoperative FEV1 and FVC were calculated based on the number of segments resected and were compared with the actual postoperative FEV1 and FVC. SETTING: This study was conducted at a university, tertiary referral hospital. PATIENTS: All patients were evaluated at a multidisciplinary lung cancer clinic and underwent lung resection by one surgeon (L.A.L.). MEASUREMENTS AND MAIN RESULTS: Sixty patients undergoing 62 pulmonary resections were reviewed. The predicted postoperative FEV1 and FVC were calculated using the following formula: predicted postoperative FEV1 (or FVC) = preoperative FEV1 (or FVC) x (1-(S x 0.0526)); where S = number of segments resected. The actual postoperative FEV1 and FVC correlated well with the predicted postoperative FEV1 and FVC for patients undergoing lobectomy (r = 0.867 and r = 0.832, respectively); however, the predicted postoperative FEV1 consistently underestimated the actual postoperative FEV1 by approximately 250 mL. For patients undergoing pneumonectomy, the actual postoperative FEV1 and FVC did not correlate as well with the predicted postoperative FEV1 and FVC (r = 0.677 and r = 0.741, respectively). Although there was considerable variability, the predicted postoperative FEV1 consistently underestimated the actual postoperative FEV1 by nearly 500 mL. Of the patients undergoing lobectomy, eight also received postoperative radiation therapy. When analyzed separately, patients receiving combined therapy lost an average of 5.47% of FEV1 per segment resected. This contrasts with a 2.84% per segment reduction in FEV1 for patients who did not receive radiation therapy. CONCLUSIONS: This simple calculation of predicted postoperative FEV1 and FVC correlates well with the actual postoperative FEV1 and FVC in patients undergoing lobectomy. The predicted postoperative FEV1 consistently underestimated the actual postoperative FEV1 by approximately 250 mL. The postoperative FEV1 and FVC for patients undergoing pneumonectomy is not accurately predicted using this equation. The predicted postoperative FEV1 for patients undergoing pneumonectomy was underestimated by an average of 500 mL and by greater than 250 mL in 12 of our 13 patients. Thus, by adding 250 mL to the above calculation of predicted postoperative FEV1, we improve our ability to we identify a minimal postoperative FEV1 for patients undergoing pneumonectomy. Finally, combined modality treatment with surgery followed by radiation therapy may result in additive lung function loss.
Patients with impaired pulmonary function are at increased risk for the development of postoperative complications. We therefore analyzed the value of preoperative lung scanning and exercise testing for the prediction of postoperative complications and of the short- as well as long-term performance in lung resection candidates at increased risk for complications. Twenty-five (mean age 63 y; 17 m) out of 84 consecutive lung resection candidates were considered at increased risk for postoperative complications due to impaired pulmonary function (FEV1 < 2L or diffusion for carbon monoxide (DLCO) < 50% predicted, or FEV1 and DLCO < or = 80% predicted combined with New York Heart Association dyspnea index > or = 2). Candidates underwent radionuclide perfusion scans and exercise testing to predict postoperative ( = ppo) values for FEV1, DLCO and maximal O2-uptake (VO2max). They all underwent thoracotomy for neoplastic lesions; 7 had pneumonectomies, 18 lobectomies. Six had postoperative complications (within 30 days), of whom three died. Three and 6 months postoperatively, pulmonary function tests and VO2max were repeated. In the 22 survivors, the observed values were then compared with the predicted values. At 3 months, there were excellent correlations (absolute/predicted values): for FEV1r = 0.78 and 0.81; for DLCOr = 0.77 and 0.74; and for VO2max r = 0.71 and 0.83. The means of FEV1 and VO2max did not differ from the predicted values, whereas the predicted DLCO was lower than the observed value (ml/min/mmHg: 15.1 vs 17.9; percent predicted: 59.6 vs 70.9) (p < 0.05). At 6 months, correlations remained very good for FEV1 (r = 0.81 and 0.84) and for DLCO (r = 0.76 and 0.74), but had decreased for VO2max to 0.56 and 0.65, respectively. All means were higher than predicted (p < 0.05) owing to recovery in the lobectomy group. Patients with postoperative complications (group B) had a lower preoperative VO2max in percent predicted (62.8 +/- 7.5% vs 84.6 +/- 19.7%) (p < 0.01) and also a lower VO2max-ppo (10.6 +/- 3.6 vs 14.8 +/- 3.5 ml/kg/min and 44.3 +/- 13.5 vs 68.0 +/- 20.7 percent predicted) (p < 0.05) than patients without complication (group A). AVO2max-ppo < 10/ml/kg/min was associated with a 100% mortality. Although FEV1-ppo and DLco-ppo were lower in group B the difference did not reach significance. We conclude that radionuclide-based calculations of postoperative VO2max are predictive of perioperative morbidity and mortality: a VO2max-ppo of < 10 ml/kg/min may indicate inoperability. Further, short-term postoperative performance is accurately predicted by FEV1-ppo and VO2max-ppo, but long-term function is underestimated after lobectomy.
The success rates reported for secondary structural class prediction with different methods are contradictory. On one side, the problem of recognizing the secondary structural class of a protein knowing only its amino acid composition appears completely solved by simply applying jury decision with an elliptically scaled distance function. Chou and coworkers repeatedly (see Crit. Rev. Biochem. Mol. Biol. 30:275-349, 1995) published prediction accuracies near 100%. On the other hand, traditional secondary structure prediction techniques achieve success rates of about 70% for the secondary structural state per residue and about 75% for structural class only with extensive input information (full sequence of the query protein, its amino acid composition and length, multiple alignments with homologous sequences). In this article, we resolve the paradox and consider (1) the question of the secondary structural class definition, (2) the role of the representativity of the test set of protein tertiary structure for the current state of the Protein Data Bank (PDB); and (3) we estimate the real impact of amino acid composition on secondary structural class. We formulate three objective criteria for a reasonable definition of secondary structural classes and show that only the criterion of Nakashima et al. (J. Biochem. 99:153-162, 1986) complies with all of them. Only this definition matches the distribution of secondary structural content in representative PDB subsets, whereas other criteria leave many proteins (up to 65% of all PDB entries) simply unassigned. We review critically specialized secondary-structural class prediction methods, especially those of Chou and coworkers, which claim almost 100% accuracy using only amino acid composition, and resolve the paradox that these prediction accuracies are better than those from secondary structure predictions from multiple alignments. We show (i) that these techniques rely on a preselection of test sets which removes irregular proteins and other proteins without any class assignment (about 35% of all PDB entries); and (ii) that even for preselected representative test sets, the success rate drops to 60% and lower for a 4-type classification (alpha, beta, alpha + beta, alpha/beta). The prediction accuracies fall to about 50% if the secondary structural class definition of Nakashima et al. is applied and only few irregular proteins are preselected and removed from automatically generated, representative subsets of the PDB. We have applied two new vector decomposition methods for secondary structural content prediction from amino acid composition alone, with and without consideration of amino acid compositional coupling in the learning set of tertiary structures respectively, to the problem of class prediction and achieve about 60% correct assignment among four classes (alpha, beta, mixed, irregular) as well as single sequence-based secondary structure prediction methods like GORIII and COMBI. Our results demonstrate that 60% correctness is the upper limit for a 4-type class prediction from amino acid composition alone for an unknown query protein and that consideration of compositional coupling does not improve the prediction success. The prediction program SSCP offering secondary structural class assignment for query compositions and sequences has been made available as a World Wide Web and E-mail service.
We have studied the use of nearest-neighbor classifiers to predict the secondary structure of proteins. The nearest-neighbor rule states that a test instance is classified according to the classifications of "nearby" training examples from a database of known structures. In the context of secondary structure prediction, the test instances are windows of n consecutive residues, and the label is the secondary structure type (alpha-helix, beta-strand, or coil) of the center position of the window. To define the neighborhood of a test instance, we employed a novel similarity metric based on the local structural environment scoring scheme of Bowie et al. In this manner, we have attempted to exploit the underlying structural similarity between segments of different proteins to aid in the prediction of secondary structure. Furthermore, in addition to using neighborhoods of fixed radius, we explored a modification of the standard nearest-neighbor algorithm that involved defining an "effective radius" for each exemplar by measuring its performance on a training set. Using these ideas, we achieved a peak prediction accuracy of 68%. Finally, we sought to improve the biological utility of secondary structure prediction by identifying the subset of the predictions that are most likely to be correct. Toward this end, we developed a nearest-neighbor estimator that produced not the traditional "one-state" prediction (alpha-helix, beta-strand, or coil) but rather a probability distribution over the three states. It should be emphasized that this scheme estimates true probability values and that the resulting numbers are not pseudo-probability scores generated by simple normalization of the raw output of the predictor. Applying the mutual information statistic, we found that these probability triplets possess 58% more information than the one-state predictions. Furthermore, the probability estimates allow one to assign an a priori confidence level to the prediction at each residue. Using this approach, we found that the top 28% of the predictions were 86% accurate and the top 43% of the predictions were 81% accurate. These results indicate that, notwithstanding the limitations on overall accuracy of secondary structure prediction, a substantial proportion of a protein can be predicted with considerable accuracy.
Cancer incidence predictions may be constructed for administrative and scientific purposes. For administrative purposes it is often important that the predictions come true. The resources planned on the basis of the predictions and allocated on the diagnostics, treatment and rehabilitation can then be optimally utilized. However, predictions that do not materialize can also be useful. The effects of intervention or early detection programmes express themselves as failures of predictions that have been made in the absence of such programmes. Predictions of cancer incidence in Finland are used as examples. The prerequisite for the predictions is a well-functioning population-based cancer registry. The predictions were constructed using time trends and differentials in cancer incidence with or without the aetiological or other risk factors. Short-term, 10-15 year predictions with no explicit use of risk factors, have proven successful with most cancers, e.g. those of the colon, rectum, pancreas and urinary organs, and lymphomas. The marked prediction failures have occurred for cancers of the lung and breast. Predictions for these cancers have been improved by taking aetiological or other risk factors explicitly into account. The cancer consequences of the preventive cardiovascular programme in North Karelia have been evaluated using predictions. The effectiveness of screening for cervical cancer at population level was predicted on the basis of estimated parameters of the natural history of the disease.(ABSTRACT TRUNCATED AT 250 WORDS)