The size distribution of particles contaminating parenteral solutions

Abstract

Some problems in the formulation of a standard for the minimum acceptable level of particulate contamination in large volume parenteral solutions are discussed. The proposed Australian Biological Standards Laboratory standard, in which an electrical method is used to detect particulate matter, makes some assumptions about the nature of the particle-size distribution. These are critically examined from the published results. Experimentally, it is confirmed that the relationship between particle diameter D and number N is dN/d(D)=CD^–M where C and M are constants. A membrane filtration method for the determination of particle numbers is shown to have serious limitations for the type of solution at present available, as there are few particles present that are larger than the lowest practical limit of detection (10 to 15 µm).

Examination of forty-five containers of sodium chloride injection B.P. from seven different batches with a modified Coulter Counter shows that, generally, the material in each container can be regarded as being unique to that container. The assumption that the size distributions are parallel within a batch is shown to be unsound. The main problem in setting an adequate standard is that of accurately defining the count of some threshold level in a solution that is clean and, therefore, relatively free of particles. The statistics of the particle-size distribution are examined and a simplified procedure suggested to define the count at a threshold and the slope of the particle-size distribution. There is no direct relationship between these two parameters, but materials that would have been regarded as unsatisfactory by the standard proposed earlier could be eliminated by a relationship of the form [graphic omitted] where N_1·0 is the estimated number of particles of diameter 1·0 µm and M is the slope of the log-log distribution.