INVESTIGATIONS OF TABLETS PREPARED FROM PELLETS PRODUCED BY EXTRUSIONAND SPHERONISATION .2. MODELING THE PROPERTIES OF THE TABLETS PRODUCED USING REGRESSION-ANALYSIS

Citation
Jf. Pinto et al., INVESTIGATIONS OF TABLETS PREPARED FROM PELLETS PRODUCED BY EXTRUSIONAND SPHERONISATION .2. MODELING THE PROPERTIES OF THE TABLETS PRODUCED USING REGRESSION-ANALYSIS, International journal of pharmaceutics, 152(1), 1997, pp. 7-16
Citations number
15
Categorie Soggetti
Pharmacology & Pharmacy
ISSN journal
03785173
Volume
152
Issue
1
Year of publication
1997
Pages
7 - 16
Database
ISI
SICI code
0378-5173(1997)152:1<7:IOTPFP>2.0.ZU;2-P
Abstract
Tablets prepared from three different types of pellets were produced e ither in a mechanical press or in a tabletting machine. The experiment s were carried out according to a centre of gravity design. The proper ties of the tablets produced were predicted from the models based on l inear equations with the variables selected by canonical analysis and principal component analysis. The results showed that the variables to be used in the equations of the models for the two machines had to be different, proving the differences in the compaction mechanisms invol ved. For the Manesty machine, five variables reflected the properties of the tablets produced, whereas for the Instron machine seven variabl es had to be considered. From the regression it was possible to select the dependent variables (properties of the tablets) that provide a be tter reflection of the independent variables (formulation and processi ng factors). Properties such as the value of 'R' (ratio between the lo wer to the upper compression pressures), the density of the tablets, t he tensile strength of the tablets, the disintegration time of the tab lets, and the mean dissolution time of the model drug (indomethacin) s eemed to be well defined by the model. On the other hand, the porosity of the tablets, the force required to crush the tablets, and the fria bility were not well reflected by the model. Predictions of the change s in the dependent variables when the independent variables were incre ased by 20% are also presented in the form of star diagrams. (C) 1997 Elsevier Science B.V.