Particulate contamination of Australian ampoules.

Theoretical predictions and computer simulations indicate that it may be impossible to achieve values of the coefficient of variation usually found in large volume parenterals (LVP) when the intrinsic particulate contamination in small volume parenterals (SVP) is investigated. Snap-opened ampoules from Australian manufacturers containing 5, 10 or 20 ml Water for Injections or Sodium Chloride Injection had a high level of particulate contamination, although within the USP XXI SVP limits. Heat-opened ampoules had much lower levels of contamination which were generally below the official LVP limits. Counts ml-1 were typically less than 10 and less than 1 for 5 and 20 micron particles, respectively. Coefficients of variation of the 5 micron data from an ampoule in any batch examined, typically ranged over 30-70%. Statistical analysis of the 5 micron data indicates significant differences between batches. Occasional ampoules had higher 5 micron counts than others in the same batch. At no time were the particulate contamination levels considered to be clinically important.