Optimization of the control of compounded capsule preparations: Feedback on the implementation of an acceptance value for the mass uniformity test
23 November 2020V.Delannoy (1), F. Maillard (1), M. Warembourg (1), I.Soulairol (1-2) 1 Department of Pharmacy, Nîmes University Hospital, Nîmes, France
2 ICGM, University of Montpellier, CNRS, ENSCM, Montpellier, France
The control of compounded capsule preparations (CCP) in hospital pharmacy units is mainly based on the mass uniformity (UM) test (2.9.5) of the European Pharmacopoeia (Pharm. Eur.). In 2017, out of the 187 CCP manufactured in our pharmacy unit, none were non-compliant (NC) according to UM. In order to improve their control, an acceptance value of the UM (VAUM), has been developed based on the acceptance value described by the Pharm. Eur. for the uniformity test (UT) of dosage units (2.9.40). The objective of this work is to assess the impact and relevance of VAUM for CCP control.
Material and method
The VAUM considers the percentage difference (E) between the average capsule mass observed and the theoretical mass calculated from the manufacturing data, and the coefficient of variation (CVUM) of the sample. Its formula has been defined as follows: VAUM = Abs(E) + (k*CVUM) (with k=2) and its acceptance limit was initially set at 11. A retrospective review of the controls of CCP and batch capsule preparations (BCP) manufactured between March 2018 and March 2020 was carried out. The NC rates of BCP based on UM test and VAUM were collected and compared to those obtained from UT test. On CCP, the NC rates of UM test and VAUM were compared.
Over the period under study, 263 BCP were manufactured. None were NC according to the UM test. According to the UT test, 24% (n=63) were NC of which 12.7% (n=8) had a VAUM greater than 11. The false positive (FP) rate (i.e VAUM NC and UT test conform) was of 1.14% (n=3). Over the same period, 694 CCP were realized. Only one was NC according to the UM test and had a VAUM of 20.7. The VAUM showed 3.7% (n=26) of CCP NC with values between 11.1 and 37 corresponding to a raw material loss between -2.5% and -32.81% and a dispersion between ± 8.7% and ± 20%.
Discussion and conclusion
On BCP, the low NC rate observed with VAUM compared to the rate obtained with the UT test can be explained by the difference of the factors considered by these 2 controls. The VAUM considers the filling uniformity and manufacturing errors such as addition or loss of raw material. However, unlike the UT, it does not evaluate the homogeneity of the mixture. Nevertheless, on CCP, the VAUM has permitted the detection of 3.7% of CCP NC against 0.15% for the UM test. The VAUM thus allows to complete the UM test which only evaluates the dispersion of the filling and does not consider the theoretical mass related to the quantities used. On the other hand, these results showed a low FP rate making it possible to use VAUM routinely without unnecessary re-manufacturing. Then, the VAUM represents an interesting specification, enhancing the control of CCP and improving their level of safety by simply using a scale.