This study was conducted to compare the dissolution variability of USP Prednisone Lot P RS Tablets to that of four commercially
available products (13). Two of the products dissolved so quickly that no meaningful comparison was possible. Reference 13
and the present article thus present data only for ranitidine hydrochloride and metformin tablets compared to the USP Prednisone
Lot P Tablets.
Study 1 was a single-laboratory study. Each of the three products was tested by two six-tablet experiments on each of three
assemblies configured as Apparatus 2 (paddles), each by two analysts, for a total of 12 experiments. Following log transformation,
the variance for each set of six was calculated and pooled over the two experiments. The pooled variance was then converted
to the intra-experiment %CV as described above. These intra-experiment %CVs include contributions from, at least, the tablets
tested, variability among positions in each assembly, the assay, and the position of the tablet when dropped into the vessel.
As conducted, position stands for the physical position as well as the particular combination of shaft, paddle, and vessel
used at that position. The "combined" results were obtained by a random-effects ANOVA. The following were included as random
effects: assembly, position nested within assembly, analyst, and experiment nested within analyst and assembly. The data were
analyzed by SAS Proc Mixed [SAS Inc., Cary, NC]. The residual %CVs thus obtained include contributions from the same sources
as for the assembly- and analyst-specific %CVs except for position within the assembly.
This study was conducted to evaluate the reproducibility of PVT results following the relocation of equipment between different
laboratory sites (16). Dissolution experiments with USP Lot P Prednisone RS Tablets were conducted at two sites (Site 1 and
Site 2). Equipment from two manufacturers was used for the experiments. Three dissolution experiments (six tablets per experiment)
were repeated by the same two analysts at each of the sites. Experiments were performed at Site 1. The testers were then moved
to Site 2, a thorough mechanical calibration was performed, and PVT experiments were repeated. Finally, the equipment was
moved back to Site 1, the mechanical calibration was repeated, and PVT experiments were repeated. Vessels, paddles, and shafts
were serialized and retained the same position in the instrument throughout the experiments. The reported %CVs were pooled
across the three experiments as described for Study 1. Contributions to these intra-experiment %CVs are also as described
for Study 1.
This study was conducted to evaluate the contributions of analyst and assembly to dissolution results using both Apparatus
1 and Apparatus 2 (2). The study focused on intermediate precision within a single laboratory. Five assemblies were used,
each by at least four of six analysts. For each analyst–assembly combination and for each of Apparatus 1 and Apparatus 2,
there were six experiments, and each experiment was performed using six to eight tablets, depending on the equipment.
The data were first analyzed by a random-effects ANOVA separately for each assembly. Random effects were position, analyst,
and experiment nested within analyst. The %CVs reported are from the residual variance from the ANOVAs. These intra-experiment
%CVs include contributions as for the Study 1 ANOVA results, namely from, at least, the tablets tested, the assay, and position
of the tablet when dropped into the vessel.
This study was conducted to evaluate the relationship between vessel dimensions and dissolution results (12). This was a
single-laboratory study, and all work was done by one analyst on one assembly configured as Apparatus 2. Three sets of six
vessels were evaluated, and each set was obtained from a different manufacturer. For each set, six dissolution experiments
were conducted, and each consisted of six tablets of USP Lot P Prednisone RS Tablets. Following log transform, the variance
for each set of six was calculated and pooled over the six experiments for a given vessel manufacturer. The pooled variance
was then converted to the intra-experiment %CV as described above. As with Study 1, these intra-experiment %CVs included contributions
from, at least, the tablets tested, variability among positions in each assembly, the assay, and the position of the tablet
when dropped into the vessel.