You are here

Guidelines for sterility testing of therapeutic goods

13 September 2006

Book pagination

1. Introduction

100. Relevant section of the BP/Ph Eur: 'Guidelines for Using the Test for Sterility'

Regulatory aspects

101. The TGA Guidelines on Sterility Testing of Therapeutic Goodsprovide guidance for sterility testing of sterile therapeutic drugs and devices supplied in Australia for human use. They are intended for use by manufacturers and the Official Analysts of the Therapeutic Goods Administration (TGA) Laboratories, and as guidance for referee testing when results are in dispute. Some sections therefore include information relating to referee testing.

102. It should be noted that these Guidelines are not mandatory for industry. The official (legal) requirements for sterility tests in Australia are currently those specified in the most recently gazetted British Pharmacopoeia(BP), which are  aligned with  the European Pharmacopoeia(Ph Eur). Therapeutic Goods Order Number 11 Standard for Sterile Therapeutic Goodswas revoked on the date of adoption of the BP 1998. The BP, Ph Eur and US Pharmacopoeia(USP) sterility test methods became harmonised with the publication of the BP 2004, Ph Eur 5.1 and USP 28 editions and as such all are considered to be acceptable by the TGA.

103. These Guidelines are based on the superseded document Australian Code of Good Manufacturing Practice for Therapeutic Goods - Medicinal Products - Appendix C Guidelines on Tests for Sterility,1990 (Appendix C). They also incorporate the BP/Ph Eur requirements with additional elements from the Pharmaceutical Inspection Convention/Pharmaceutical Inspection Co-operation Scheme (PIC/S) Recommendations on Sterility Testing,the USP and the combined experience of the TGA Laboratories, manufacturers and contract testing laboratories.

104. Sampling schedules and specific guidance for testing of medical devices are provided because these are not included in the BP/Ph Eur. Sampling schedules are based on those of the USP 28 effective January 2005.

105. Where the Guidelines describe particular procedural steps and media, it should be understood that other procedures and other media may be equally satisfactory, provided that their use can be validated. However, when the results of a sterility test performed by the TGA Official Analyst are challenged, the combined procedures of the BP/Ph Eur test and this document must be followed.

106. Where reference is made to the 'competent authority' throughout this document, the TGA fulfils this role in Australia.


107. Success in detecting microbial contamination of goods is dependent inter aliaupon statistical considerations. Generally, the odds are against detecting a low level of contamination, ie, when relatively few organisms contaminate a small proportion of items. Detection of contamination with absolute certainty would necessitate the examination of all items of the batch using media capable of supporting the growth of all possible contaminants.

108. In the BP/Ph Eur sterility test prior to 1998, when contamination was found in a primary test, a repeat test was permitted as a check against the possibility that the contamination was introduced by the operator and was not a contamination of the material tested. Repeat testing reduces the efficiency of testing because the probability of accepting a contaminated batch is thereby increased.

109. The probability of including contaminated items in two successive samples taken from a given batch is the product of the probabilities of a contaminated item being included in each of the single samples. For example, with a sample size of 10 and a contamination rate of 5% the probability of including a contaminated item is 0.4 (see Table 1).

110. If the test is repeated with another sample size of 10 units the probability of including a contaminated item is again 0.4. However, the probability of both tests being positive is 0.4 x 0.4 = 0.16 which is lower than the probability of the first test, and hence in 84 of 100 such tests the batch will be accepted as sterile.

Table 1: Probabilities of detecting a contaminated lot in a single test
Percentage of items contaminated 0.1 1.0 2.0 5.0
Sample size 10 0.01 0.09 0.18 0.40
Sample size 20 0.02 0.18 0.33 0.65
Sample size 50 0.05 0.39 0.64 0.92
Sample size 100 0.09 0.63 0.87 0.99

111. A repeat test is now permitted only if it can be clearly demonstrated that the test was invalid for causes unrelated to the product being examined (see clauses 492-496).

112. Sterility cannot be guaranteed by the quality control laboratory. It must be built into the product during processing. Experience in many countries over the years has confirmed that greater reliance must be placed on appropriate techniques and procedures throughout the manufacture of the product (including in-process sterility testing at various stages) rather than simply depending on sterility tests made on a number of samples of the final batch as the sole criterion of sterility.

113. Permission to delete the sterility test from batch release specifications may be granted by the competent authority where manufacturing standards are of a high order and the product is terminally sterilised in its final container.

114. In all other cases, sterility testing is necessary to detect contamination arising from technical malfunction, human error or mix-up between sterilised and non-sterilised goods. It must be performed as part of batch release specifications for product that is not terminally sterilised in the final container. It is also the only analytical method available for finished product testing by the competent authority.

115. It is important to realise that background contamination as detected by negative controls can obscure low levels of product contamination and for this reason every effort should be made to ensure that background contamination is kept as low as possible. Statistics compiled by the TGA Laboratories show that skilled operators working under the prescribed conditions can achieve a level as low as one contamination in five thousand control inoculations (0.02%).

Book pagination