Prediction diagnostics tables

Table 5.12 smmarizes the prediction diagnostic tools discussed in this section. The first r Iumn in the table lists tlie name of each tool and the second column discusses what is expected with well-behaved and problematic data. 

Model and Parameter Sta stics (Model Diagnostic) Table 5-13 displays the variables selected for a model constructed to predict caustic. The table lists summary statistics for the regression model as weU as information about the estimated regression coefficients. Six variables in addition to an intercept are found to be significant at the 95% confidence level. 

One of the most powerful advantages of multivariate calibration is that the predictions can be validated. That is, it is possible to evaluate the reliability of the concentration estimates. Four prediction diagnostic tools are discussed below and a summarv is found at the end of the section in Table 5.4. 

The four prediction diagnostic tools found in Table 5.4 in Section 5-2.1.1 are applied to the prediction data. These tools are used to assess the reliability of the prediction results. 

This discussion follows the prediction diagnostics found above in Table 5.12, Section 5.3.1.1. 

Since Congress gave it authority to regulate in vitro diagnostic tests in 1976, the FDA has included assessment of clinical validity based on data made available by the test s developer (Holtzman, 2000). In view of the paucity of published data on predictive values of pharmacogenetic tests, and the likelihood that many of them will be low (Table 9.2), data on clinical validity should be required before pharmacogenetic tests are used clinically.12... 

Critical for predictivity in a recent comprehensive study was the number and choice of parameters measured [4]. Early, sublethal effects on cell proliferation, cell morphology and mitochondria occurred consistently and ubiquitously with toxicity and when used collectively were most diagnostic. It is noteworthy that the toxicity of many drugs is attributable to various mitochondrial targets, including oxidative phosphorylation, fatty acid oxidation, Krebs cycling, membrane transport, permeability transition pore, proliferation and oxidative stress (Table 14.4). 

Root Mea Square Error of Prediction (RMSEP) (Model Diagnostic) The RMSEP values for all four components are numerically summarized in Table 5.6. They are large owing to the bias in the predictions. Several reasons for this bias can be proposed, including an inaccurate reference method, transcription errcKS, poor experimental procedures, changes in densiw and/or pathlength, l t scatter in the instrument or sample, chemical interactions,... 

The estimated pure spectra are accepted and the ICLS model is validated using the diagnostics in Table 5.1, Section 5.2.1.1. The validation data are obtained by performing prediction on the Hdation samples using the pure spectra estimated from the calibration samples. 

The results for calibration and prediction of MCB are discussed in detail and the results for the other three components are summarized at the end of this section. The diagnostics in Table 5.18 in Section 5.3 2.1 are used to validate the MCB model. 

COPD is a chronic, slowly progressive disorder characterized by airways obstruction (FEVi < 80% predicted FEViA C ratio < 70%) which does not change markedly over several months. The airways obstruction is largely fixed but may be partially reversible by bronchodilator therapy. Unlike asthma, airflow limitation in COPD can never be returned to normal values. The diagnosis of COPD is usually suggested by symptoms. A firm diagnosis can only be made by objective measurement of airways obstruction with spirometric tests, which may be enhanced by radio diagnostic techniques (Table 4). 

---