Pooling, of data

The thermodynamic stability of coordination compounds is relatively easy to determine, and provides us with a valuable pool of data from which we may assess the importance of ligand-field and other effects upon the overall properties of transition-metal compounds. The bulk of this chapter will be concerned with the thermodynamic stability of transition-metal compounds, but we will briefly consider kinetic factors at the close. 

PK of Cetuximab Following Pooling of Data Across All Studies 357... 

To facilitate the pooling of data across a project for safety updates and integrated safety and efficacy reporting... 

Data may be organized by type of data, such as adverse events, laboratory data, demography, physical exam, etc. Since many of these categories of data exist across clinical trials, standard file structures can be designed and implemented. This standardization allows for the reuse of validated software as well as facilitates the pooling of data across studies for use in project safety summaries and other data reporting across studies. 

In addition, creating a data set on which PMKD is to be performed is not a trivial task. When data are combined across clinical trials, attention must be paid to the variables in the data sets being combined and data completeness (or incompleteness). Data access tools must be available for the pooling of data stored in different data warehouses and platforms. 

As the reported values of the thermodynamic parameters indicate, the largest pool of data is based on the Huggins-Flory relation. This is only to be expected since the theory and related to it concept of the binary interaction parameters, either B or was introduced to polymer science more than a half century ago, in 1941 to be precise. Even their well-recognized complexity of functional dependence, viz. B = B(T, P, ( ), MW, MWD, molecular stmcmre, stresses,...), does not discourage efforts to continue using this approach against all odds. 

The population PK/PD analysis, rather than individual PK/PD analysis, allows for rehable data analysis following sparse sampling, correctly accounts for sources of variabihty, allows for pooling of data across individuals and studies, and provides for an integrated model for PK/PD data (even across species). However, data analysis should be driven by the question of interest, rather than the method of analysis. It is not always necessary to use complicated methods of analysis. Sometimes a simpler individual PK/PD model approach can be used if the PK/PD data collected are relatively rich in a quite homogeneous population (especially in Phases 1 and Ha settings). 

Some applications of the kinetic data on clustering reactions are presented in subsequent sections. One application should be pointed out here, namely that to the development of kinetic theory. The existing theory of third-order association reactions and collisional activation and deactivation was developed on the basis of experiments with neutral species. However, the systems involving neutrals for which reliable data are available are quite limited. The state of the theory is also not quite satisfactory. The large variety of systems observable in ion-molecule studies and particularly the wide pressure and temperature variation which is experimentally feasible should provide a new pool of data and thus stimulate further development of the theory. 

There are important statistical considerations in meta-analyses for the evaluation of safety. In particular, statistical methods must be valid in the presence of sparse data. As discussed previously, safety outcomes may be infrequent. Some trials may not have any relevant events. We refer to these trials as zero-event trials. The statistical methods should provide estimates with good bias properties and with valid standard errors and confidence intervals in the presence of low event coimts and zero-event trials. In any meta-analysis, the overall estimator and associated standard errors and confidence intervals should be stratified by the trials. Simple pooling of data across the trials can result in misleading results because of Simspon s paradox. With stratification, the randomize comparisons within trials are maintained. 

The empirical Bayes method allows the pooling of data across a range of events to estimate an overall rate. Individual occurrence rates are then calculated as deviations from this overall rate. Empirical Bayes uses the pooled data to estimate the prior distribution parameters. The benefits of using historical data when estimating rare event probabilities are discussed by (Gelman et al. 1998). 

Based on the nature of the interatomic interactions, the effects of the medium on strength can be classified as belonging to several different types. We can illustrate this using the pool of data collected by Savenko et al. on the strength of graphite in contact with different liquid metals [43-46], Namely, one can outline the following characteristic cases ... 

---