- Email: [email protected]
Shelf life of condoms
Accepted Manuscript Shelf life of condoms John Gerofi, Morten Sorensen
PII:
S0142-9418(16)30813-3
DOI:
10.1016/j.polymertesting.2016.10.031
Reference:
POTE 4811
To appear in:
Polymer Testing
Received Date: 19 August 2016 Accepted Date: 26 October 2016
Please cite this article as: J. Gerofi, M. Sorensen, Shelf life of condoms, Polymer Testing (2016), doi: 10.1016/j.polymertesting.2016.10.031. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Product Performance
Shelf Life of Condoms
RI PT
John Gerofia and Morten Sorensenb a Enersol Pty Ltd, 235 Nelson St, Annandale, NSW, 2038, Australia ([email protected])
SC
b. United Nations Population Fund, UN City, Marmorvej 51, 2100 Copenhagen, Denmark ([email protected])
Abstract
TE D
M AN U
Condom quality is, in many countries, regulated through ISO (international standard) 4074. It prescribes a maximum shelf life of 5 years and also a real time stability requirement to ensure the products are fit for use until the expiry date. This article reports on tests done on condoms well past their expiry date, as well as some which were near their expiry date and then submitted to the additional challenge of storage at 50oC for 90 days. The results show that, with two exceptions, the condoms continued to comply comfortably with the requirements of the standard. It thus appears that the 5 year maximum shelf life currently allowed for condoms should be reviewed.
Key-Words
Introduction
AC C
1.
EP
condom, shelf-life, packaging, deterioration, inflation, tensile
In a previous article1, we summarized existing data on shelf life of condoms. The methods of predicting shelf life from accelerated aging data are of limited value. It appears that ovenconditioning at 50oC is the most reliable method currently known, and has been included in the 2014 and 2015 editions of ISO 40742,3. Nonetheless, ISO 4074 requires real time aging of typical lots of each product to verify shelf life claims. Despite these assurances, the standard arbitrarily limits shelf life to 5 years. Compliance with ISO 4074 has become widely accepted as evidence that condoms are fit for use, for example under the European Medical Device Directive4. Even so, an in vitro trial to verify
ACCEPTED MANUSCRIPT
conclusively that condoms had a shelf life of, say, 7 years would be real-time, and would require setting aside samples at least a few years in advance.
RI PT
A more fundamental approach to shelf life would involve an in-vivo trial5,6 to measure slippage and breakage of condoms of various ages. One such trial was done in 19917, using one condom brand. There is now a draft standard8 for a more rigorous protocol for such trials, but it does not deal with obtaining suitable samples of condoms made to the same formulation and aged naturally for vastly differing periods. The cost of such a trial would be very high.
SC
The 2014 and 2015 editions of ISO 4074 define shelf life as the period for which condoms are required to conform to the key provisions of the standard. Compliance with the standard can thus be used as an indication that a product is fit for use, provided there is no evidence to the contrary.
M AN U
The UNFPA condom program9,10,11 is sometimes faced with expiring condoms in its stock, and it is, therefore, sometimes necessary to destroy condoms that may be fit for use. This trial was designed to generate indicative information about how condoms behaved after their expiry date, without waiting for many years for the result. Due to requirements within UNFPA, the time allowed for the study was limited, and the period for accelerated aging had to be restricted to 90 days. The trial involved collection and testing of expired condoms, to see whether they remained in compliance with ISO 4074. The samples used were the most relevant ones available, and it is recognised that most were not stored under tropical conditions.
Approach
EP
2.
TE D
The tests were conducted in the second half of 2014.
There were three parts to the study:
2. 3.
Testing of condoms that were well past their expiry date before being tested.
AC C
1.
Examination of data supplied by a manufacturer Accelerated aging, followed by testing, of products near to expiry date.
Test results were examined for compliance with the standard and, where possible, compared with the original test results, or with results generated before accelerated aging. Test methods were in accordance with ISO 4074:2014, but sample sizes were adjusted to match the availability of samples. Inferences regarding compliance were based on tables in ISO 2859112.
ACCEPTED MANUSCRIPT
Testing of expired product
3.1
Products tested
RI PT
3.
SC
The products involved are in Table 1. Brand names have been replaced with codes.
Table 1 - Products tested long after expiry date Expiry date
Dingo
09/2010
Koala
12/2010
Owl
12/2010
Shark
04/2013
Storage location
TE D
M AN U
Product Code
Sydney Sydney Sydney Germany
All products except Shark had expired about 4 years before being tested. Shark had expired one and a half years before testing.
AC C
EP
Samples available for this part of the study were limited, because of their age. In some cases, the sample had to be made up of different batches for the inflation and leaks tests. In these cases, the batches were chosen so the manufacturing dates were very close to each other, and the products were always from the same shipment and of the same design. Original pre-shipment test reports were available in 3 of the 4 cases.
ACCEPTED MANUSCRIPT
3.2
Results
The test results are shown in Table 2.
Code Name
Dingo
Koala
Age at retest
9 years 2 months
8 years 11 months
Mean
32.49
30.34
37.68
31.92
Median
33.00
31.00
38.00
32.00
SD
2.97
2.69
2.98
Range
19.0 - 39.5
21.5 - 35.0
21.0 - 45.0
No tested
315
125
315
No failed
0
0
Mean
2.06
Median
Owl
Shark
8 years 11 months
6 yrs 7 m
New
Aged
Aged
38.49
32.50
34.70
38.50
33.00
34.50
SC
Aged
2.29
3.01
2.93
1.54
26.0-38.0
24.5-47.0
17.0 - 38.0
30.0 - 38.5
125
315
125
125
0
0
0
1
0
1.92
1.97
2.07
1.98
1.97
1.95
2.05
1.95
2.00
2.05
2.00
2.00
1.95
SD
0.17
0.14
0.17
0.15
0.13
0.14
0.14
Range
1.30 - 2.45
1.30 - 2.20
0.95 - 2.35
1.70- 2.45
1.35- 2.25
1.15 - 2.25
1.65 - 2.20
No tested
315
125
315
125
315
125
125
No Failed
0
0
1
0
0
0
0
Test Date
20/01/06
22/10/14
16/02/06
22/10/14
27/02/06
22/10/14
23/10/14
Mfg Date
10/2005
10/2005
01/2006
01/2006
01/2006
01/2006
----
Exp Date
09/2010
09/2010
12/2010
12/2010
12/2010
12/2010
04/2013
No Tested
315
200
315
200
315
200
200
No Failed
0
0
0
0
0
0
0
Test Date
23/01/06
24/10/14
20/02/06
23/10/14
6/03/06
23/10/14
23/10/14
Mfg Date
10/2005
01/2006
01/2006
12/2005
01/2006
01/2006
----
12/2010
12/2010
11/2010
12/2010
12/2010
04/2013
AC C
EP
Leaks
New
M AN U
Press (kPa)
Aged
TE D
Vol (L)
New
RI PT
Table 2 - Test results for condoms tested long after expiry date
Exp Date
09/2010
All products tested in this part of the study passed the inflation and leaks tests. For leaks, the sample size was 315 where possible, and 200 otherwise. No holes were found. For inflation, 125 condoms were used for the aged condoms, because of limited availability. Of the 125 condoms tested, there were no non-compliers except in one case (one non-complier). According to ISO 4074 and ISO 2859-1, up to 5 non-compliers would be allowed. 10 non-
ACCEPTED MANUSCRIPT
compliers would be allowed for a sample of 315, and 7 for a sample of 200. Also, the means and standard deviations of volume and pressure indicate that the distributions were well clear of the pass/fail limits.
RI PT
One can conclude that condoms stored in temperate to sub-tropical climates meet the ISO standard long after their expiry dates.
Data received from a manufacturer
M AN U
4.
SC
Results for the three products for which the original test data were also available show that there is no consistent trend in the burst properties, except that the mean burst volume decreased a little.
One manufacturer, Rosella, submitted results of recent (June 2014) inflation tests on plain condoms made in each of the years 2005 to 2013. The condoms had been stored at a mean temperature of approximately 30oC.
Accelerated aging of condoms nearing expiry date
EP
5.
TE D
These results appeared inconclusive, partly because of the low sample size. On the basis of this small sample, the lots made in 2007, 2009, 2011, 2012 and 2013 met the requirements, while the others (ie 2005, 2006, 2008, 2010) no longer did.
AC C
The condoms used in this section were taken from Enersol's retained samples or supplied by manufacturers. Lion condoms were stored in ambient tropical conditions, while the other condoms for which original data are available were stored in Sydney, with a mean kinetic temperature of approximately 20oC. Condoms which were 4 to 6 years old were conditioned at 50oC for 90 days. The two Dingo products are an exception to this. They had been stored in a sub-tropical location with an annual mean temperature similar to Sydney's until September, 2014, and then sent to Enersol where they were oven-conditioned for 1 month at 50oC. According to ISO 4074, 50oC for 90 days can be used as a provisional equivalent to 3 years’ storage at 30oC (the standard temperature used for real time shelf life determination). In this case, the condoms had been stored at room temperature in Sydney (mean kinetic temperature of
ACCEPTED MANUSCRIPT
approx. 20oC) for at least 4.5 years before being exposed to the 50oC oven. The ovenconditioning time was limited due to the time available for the study. The purpose of this combined exposure was to simulate, as far as practicable, a period of storage significantly longer than 5 years, with relatively warm conditions.
RI PT
12 lots were tested. Where possible, 125 condoms were used in the inflation test and 200 were tested for leaks. In one case, the numbers were reduced to 80 for inflation and 125 for leaks. Sample sizes were smaller than what is generally used for compliance testing by WHO13, because of the limited availability of condoms. The leaks test used was the ISO 4074:2014 version of the conductivity test.
M AN U
SC
For 5 of the 12 batches, Enersol had original test data available from the time of manufacture. For another 4, there were sufficient samples available to be able to be able to compare the results before and after oven conditioning. For the remaining 3 batches, no information is available about the properties of the condoms before oven conditioning, except the presumption that they met the ISO 4074 requirements when they were new.
AC C
EP
TE D
The inflation test results are summarised below.
ACCEPTED MANUSCRIPT
Table 3 - Mean burst volume of condoms with accelerated aging after expiry date
Shark
May 2008
Frog
Aug 2009
Ant
Feb 2009
Koala
Apr 2010
Owl 1
Feb 2009
Date of test Oct 2014 Nov 2014 Sept 2009 Nov 2014 Apr 2009 Nov 2014 Jun 2010 Nov 2014
Mean vol
No tested
34.7 35.8 43.5 33.7 42.0 33.7 30.3 25.3
125 125 200 125 315 125 315 125
No of noncompliers 0 0 0 0 2 0 1 1
50oC 90 days
Nov 2014
36.0
125
0
50oC 90 days Shelf since 2009 50oC 90 days
Nov 2014 Oct 2014 Nov 2014
36.6 32.7 32.6
125 125 125
0 2 1
Nov 2014 Nov 2010 Nov 2014 Jan 2010 Dec 2014 Nov 2014 Oct 2014 Nov 2014 Oct 2014 Nov 2014
28.3 33.2 23.3 42.2 39.0 16.3
125 200 125 200 50 13
0 1 2 0 0 8
34.2
125
0
33.9
125
1
Cat
Nov 2009
Eel
Jan 2010
Lion*
Nov 2010
Bat
Nov 2009
Dingo 2
Aug 2008 Aug 2008
EP
Dingo 3
50oC 90 days Original test 50oC 90 days Original test Shelf since 2010 o 50 C 90 days Shelf since 2008 o 50 C 30 days Shelf since 2008 o 50 C 30 days
M AN U
Mar 2009
TE D
Owl 2
Sample Conditioning Shelf since 2008 50oC 90 days Original test 50oC 90 days Original test 50oC 90 days Original test 50oC 90 days
RI PT
Mfr date
SC
Product
AC C
One brand, Bat, had suffered severe degradation, and it was pointless to do the tests. The test results on the same batch when new were acceptable. Additional retained samples found later also passed when tested without the additional oven conditioning. The other 11 batches all passed the ISO 4074 requirements. The consistency of the results is convincing, despite the small sample sizes. Among the 5 products where original data are available, all suffered significant drops in mean burst volume although, except for Bat, there was no apparent increase in the rate of noncomplying condoms. After 4 years’ natural aging and 3 months at 50oC, Lion condoms had a relatively low mean volume, but still complied with the standard. The effect on volume of the oven-conditioning on the four batches of condoms for which testing was also done before oven-conditioning was relatively minor. Once again, there was no apparent increase in the rate of non-compliers.
ACCEPTED MANUSCRIPT
Table 4 - Mean burst pressure of condoms with accelerated aging after expiry date
Shark
May 2008
Frog
Aug 2009
Ant
Feb 2009
Koala
Apr 2010
Owl 1
Feb 2009
Date of test Oct 2014 Nov 2014 Sept 2009 Nov 2014 Apr 2009 Nov 2014 Jun 2010 Nov 2014
Mean press
No tested
1.95 1.72 2.16 2.00 1.97 1.99 2.51 2.16
125 125 200 125 315 125 315 125
50oC 90 days
Nov 2014
2.00
50oC 90 days Shelf since 2009 50oC 90 days
Nov 2014 Oct 2014 Nov 2014
1.93 2.08 2.10
Nov 2014 Nov 2010 Nov 2014 Jan 2010 Dec 2014 Nov 2014 Oct 2014 Nov 2014 Oct 2014 Nov 2014
0
125 125 125
0 2 0
Nov 2009
Eel
Jan 2010 Nov 2010
Bat
Nov 2009
Dingo 2
Aug 2008
Dingo 3
Aug 2008
EP
Lion*
50oC 90 days Original test 50oC 90 days Original test Shelf since 2010 o 50 C 90 days Shelf since 2008 o 50 C for 30 days Shelf since 2008 o 50 C for 30 days
1.65 2.00 2.01 2.04 1.66 0.85
125 200 125 200 50 13
0 1 0 0 0 8
1.98
125
0
1.98
125
0
M AN U
Cat
No of noncompliers 0 0 0 0 2 0 0 1
125
Mar 2009
TE D
Owl 2
Sample Conditioning Shelf since 2008 50oC 90 days Original test 50oC 90 days Original test 50oC 90 days Original test 50oC 90 days
RI PT
Mfr date
SC
Product
AC C
The mean burst pressure of most products was relatively unaffected by the oven conditioning, and by storage on the shelf.
ACCEPTED MANUSCRIPT
Table 5 - Leaks found in condoms exposed to accelerated aging after expiry date
Shark
May 2008
Frog
Aug 2009
Ant
Feb 2009
Koala
Apr 2010
Owl 1
Feb 2009
Sample Conditioning Shelf since 2008 o 50 C 90 days Original test o 50 C 90 days Original test o 50 C 90 days Original test o 50 C 90 days o
Owl 2
Mar 2009 o
Cat
Nov 2009
Eel
Jan 2010
50 C 90 days Shelf since 2009 o 50 C 90 days o
Nov 2010
Bat
Nov 2009
Dingo 2
Aug 2008
200
0
Nov 2014 Oct 2014 Nov 2014
200 200 200
0 0 0
Nov 2014 Nov 2010 Nov 2014 Jan 2010 Nov 2014 Oct 2014 Nov 2014 Oct 2014 Nov 2014
200 315 200 315 10
0 0 0 0 2
200
0
200
0
Discussion
AC C
6.
Aug 2008
200 200 315 200 315 200 315 200
No of noncompliers 0 0 0 0 0 0 3 0
EP
Dingo 3
50 C 90 days Original test o 50 C 90 days Original test o 50 C 90 days Shelf since 2008 o 50 C for 30 days Shelf since 2008 o 50 C for 30 days
TE D
Lion*
Nov 2014
No tested
M AN U
50 C 90 days
Date of test Oct 2014 Nov 2014 Sept 2009 Nov 2014 Apr 2009 Nov 2014 Jun 2010 Nov 2014
RI PT
Mfr date
SC
Product
The purpose of the studies was to expose the condoms to aging conditions that were in excess of the normal 5 year shelf life study, but to do so over a relatively short period. In order to achieve this, it was necessary to find samples that had already aged naturally for several years. The oldest ones were tested without further challenge, while those around 4 to 5 years old were ovenconditioned at 50oC. The study design was limited by what was available for testing and UNFPA’s time constraints. Among the condoms aged naturally (6 to 9 years old), all passed the tests quite comfortably.
ACCEPTED MANUSCRIPT
From results submitted by a manufacturer whose products had been stored at a mean kinetic temperature of around 30oC, some batches passed and some failed. In particular, those made before 2007 failed. Thus condoms up to 7 years old from this manufacturer can pass the tests, although some of the newer products did not.
RI PT
From tests on condoms that had had 4 to 5 years’ natural aging, plus oven conditioning at 50oC, all but one product passed all the tests at the end. Apart from Bat, none of the products showed an increase in the number of holes.
1.
SC
Because the exact equivalence between aging at elevated temperatures and that at room temperature cannot be established, one must rely on approximate conclusions. The following can be inferred: Different products behave differently in the face of the challenges of aging.
M AN U
2. One product out of 12 performed very badly after exposure to the combined challenge of several years of ambient aging, followed by 3 months in an oven at 50oC. 3. The other 11 products still complied with ISO 4074, and the rate of non-compliers was not measurably higher as a result of the natural or accelerated aging.
TE D
4. All 5 products with original test data available showed a loss of mean burst volume as a result of the combined aging protocol 5. The four products for which tests were done before and after oven conditioning showed relatively little change in mean burst volume
EP
6. Mean burst pressure falls in some products (even over the 90 days' oven conditioning), while in others, it remains relatively unchanged. It remains high enough to ensure that the products will pass the ISO requirement
AC C
7. Only one product demonstrated any increase in the rate of holes. That product also showed catastrophic decay in its inflation properties. The ability of the study to accurately reflect real time aging is limited by the sample sizes available for testing, and by the time available for the study. To obtain more definite assurance that products do indeed continue to comply with the requirements of the standard after their expiry date, it is necessary to conduct a real time study, which would last at least 6 years. If continued compliance with the standard were thought to be insufficient evidence of fitness for use, then it would probably be necessary to conduct in vivo breakage trials using suitably aged condoms - a very lengthy and expensive project.
ACCEPTED MANUSCRIPT
7.
Conclusions
RI PT
Most of the products tested, whether after natural aging well beyond the expiry date, or by four to five years’ natural aging followed by 3 months’ oven conditioning, passed the ISO 4074 test requirements comfortably. With one exception, all products were still in compliance with ISO 4074 at the end of the exposure to elevated temperature.
SC
The exception, Bat, was made in 2009, and oven conditioning caused it to fail inflation testing drastically, and to stick to the pack when opened. It was concluded that this batch of Bat products was in some way defective in manufacture, and did not necessarily represent the manufacturer's overall capabilities.
M AN U
10 of the 12 products tested appeared to withstand the combination of ambient temperature aging under moderately warm conditions for 4 to 6 years plus 3 months at 50oC without serious degradation to their properties. According to trials conducted previously14, 3 months at 50oC is approximately equivalent to 2 to 3 years at 30oC. It thus appears that these products would withstand at least 6 years' storage in appropriate warehouses, even in tropical conditions.
EP
TE D
If it is accepted that continued compliance with ISO 4074 indicates continued fitness for use, then it should be acceptable to determine the shelf life by real time aging at 30oC as required in the standard, and the five year arbitrary limit on shelf life could be gradually removed. Manufacturers should be free to extend their shelf-life (in modest steps) based on demonstrated compliance with the requirements of clause 11 of the ISO 4074 standard at the end of the time they stipulate for shelf life. Manufacturers may be able to use their retained samples of past expiry date product to do this.
AC C
If it is not accepted that continued compliance with the physical property requirements of ISO 4074 indicates continued fitness for use, then a fundamental re-think of the condom standard is required.
Funding
The experimental work in this article was funded by the United Nations Population Fund (UNFPA). UNFPA did not exercise any editorial control over the contents of the article.
ACCEPTED MANUSCRIPT
References 1
AC C
EP
TE D
M AN U
SC
RI PT
J Gerofi and M Sorensen, Re-evaluation of data and requirements on condom shelf life. Polymer Testing, 54, September 2016, pp 260–269 2 ISO 4074, Natural Rubber Latex Male Condoms – Requirements and Test Methods, International Organization for Standardization, 2014. 3 ISO 4074, Natural Rubber Latex Male Condoms – Requirements and Test Methods, International Organization for Standardization, 2015. 4 Council Directive 93/42/EEC of 14 June 1993 concerning medical devices, as amended. http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:31993L0042 5 http://www.who.int/topics/clinical_trials/en/ (accessed June 2016) 6 http://www.nhs.uk/Conditions/Clinical-trials/Pages/Healthresearch.aspx (accessed June 2016) 7 Steiner M, Foldesy R and Cole, D Study to Determine the Correlation between Condom Breakage in Human Use and Laboratory Test Results, FHI, November 1991 8 ISO/DIS 29943-1 Condoms-Guidance on Clinical Studies – Part 1: Male condoms, clinical function studies based on self-reports, International Organization for Standardization, 2014 9 http://www.unfpa.org/publications/condom-programming-hiv-prevention-0 (accessed June 2016) 10 http://www.unfpa.org/resources/male-latex-condom (accessed June 2016) 11 https://www.unfpa.org/sites/default/files/resourcepdf/20429%20UNFPA%20RFQ%20QA%20 Framework_low_V11_2.pdf (accessed June 2016) 12 ISO 2859-1, Sampling Plans for Inspection by Attributes – Part 1: Sampling plans indexed by acceptable quality limit for lot-by-lot inspection, International Organization for Standardization, 1999 13 Male Latex Condom: Specification, Prequalification and Guidelines, 2010, World Health Organization, UNFPA and FHI, 2010. ISN 978 92 4 159990 0. 14 M C Bó, J P Gerofi, L L Y Visconte and R C R Nunes, Prediction of Shelf Life of Natural Rubber Male Condoms - A Necessity, Polymer Testing, 26, 3, 2007 - pp. 306-314.
PII:
S0142-9418(16)30813-3
DOI:
10.1016/j.polymertesting.2016.10.031
Reference:
POTE 4811
To appear in:
Polymer Testing
Received Date: 19 August 2016 Accepted Date: 26 October 2016
Please cite this article as: J. Gerofi, M. Sorensen, Shelf life of condoms, Polymer Testing (2016), doi: 10.1016/j.polymertesting.2016.10.031. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Product Performance
Shelf Life of Condoms
RI PT
John Gerofia and Morten Sorensenb a Enersol Pty Ltd, 235 Nelson St, Annandale, NSW, 2038, Australia ([email protected])
SC
b. United Nations Population Fund, UN City, Marmorvej 51, 2100 Copenhagen, Denmark ([email protected])
Abstract
TE D
M AN U
Condom quality is, in many countries, regulated through ISO (international standard) 4074. It prescribes a maximum shelf life of 5 years and also a real time stability requirement to ensure the products are fit for use until the expiry date. This article reports on tests done on condoms well past their expiry date, as well as some which were near their expiry date and then submitted to the additional challenge of storage at 50oC for 90 days. The results show that, with two exceptions, the condoms continued to comply comfortably with the requirements of the standard. It thus appears that the 5 year maximum shelf life currently allowed for condoms should be reviewed.
Key-Words
Introduction
AC C
1.
EP
condom, shelf-life, packaging, deterioration, inflation, tensile
In a previous article1, we summarized existing data on shelf life of condoms. The methods of predicting shelf life from accelerated aging data are of limited value. It appears that ovenconditioning at 50oC is the most reliable method currently known, and has been included in the 2014 and 2015 editions of ISO 40742,3. Nonetheless, ISO 4074 requires real time aging of typical lots of each product to verify shelf life claims. Despite these assurances, the standard arbitrarily limits shelf life to 5 years. Compliance with ISO 4074 has become widely accepted as evidence that condoms are fit for use, for example under the European Medical Device Directive4. Even so, an in vitro trial to verify
ACCEPTED MANUSCRIPT
conclusively that condoms had a shelf life of, say, 7 years would be real-time, and would require setting aside samples at least a few years in advance.
RI PT
A more fundamental approach to shelf life would involve an in-vivo trial5,6 to measure slippage and breakage of condoms of various ages. One such trial was done in 19917, using one condom brand. There is now a draft standard8 for a more rigorous protocol for such trials, but it does not deal with obtaining suitable samples of condoms made to the same formulation and aged naturally for vastly differing periods. The cost of such a trial would be very high.
SC
The 2014 and 2015 editions of ISO 4074 define shelf life as the period for which condoms are required to conform to the key provisions of the standard. Compliance with the standard can thus be used as an indication that a product is fit for use, provided there is no evidence to the contrary.
M AN U
The UNFPA condom program9,10,11 is sometimes faced with expiring condoms in its stock, and it is, therefore, sometimes necessary to destroy condoms that may be fit for use. This trial was designed to generate indicative information about how condoms behaved after their expiry date, without waiting for many years for the result. Due to requirements within UNFPA, the time allowed for the study was limited, and the period for accelerated aging had to be restricted to 90 days. The trial involved collection and testing of expired condoms, to see whether they remained in compliance with ISO 4074. The samples used were the most relevant ones available, and it is recognised that most were not stored under tropical conditions.
Approach
EP
2.
TE D
The tests were conducted in the second half of 2014.
There were three parts to the study:
2. 3.
Testing of condoms that were well past their expiry date before being tested.
AC C
1.
Examination of data supplied by a manufacturer Accelerated aging, followed by testing, of products near to expiry date.
Test results were examined for compliance with the standard and, where possible, compared with the original test results, or with results generated before accelerated aging. Test methods were in accordance with ISO 4074:2014, but sample sizes were adjusted to match the availability of samples. Inferences regarding compliance were based on tables in ISO 2859112.
ACCEPTED MANUSCRIPT
Testing of expired product
3.1
Products tested
RI PT
3.
SC
The products involved are in Table 1. Brand names have been replaced with codes.
Table 1 - Products tested long after expiry date Expiry date
Dingo
09/2010
Koala
12/2010
Owl
12/2010
Shark
04/2013
Storage location
TE D
M AN U
Product Code
Sydney Sydney Sydney Germany
All products except Shark had expired about 4 years before being tested. Shark had expired one and a half years before testing.
AC C
EP
Samples available for this part of the study were limited, because of their age. In some cases, the sample had to be made up of different batches for the inflation and leaks tests. In these cases, the batches were chosen so the manufacturing dates were very close to each other, and the products were always from the same shipment and of the same design. Original pre-shipment test reports were available in 3 of the 4 cases.
ACCEPTED MANUSCRIPT
3.2
Results
The test results are shown in Table 2.
Code Name
Dingo
Koala
Age at retest
9 years 2 months
8 years 11 months
Mean
32.49
30.34
37.68
31.92
Median
33.00
31.00
38.00
32.00
SD
2.97
2.69
2.98
Range
19.0 - 39.5
21.5 - 35.0
21.0 - 45.0
No tested
315
125
315
No failed
0
0
Mean
2.06
Median
Owl
Shark
8 years 11 months
6 yrs 7 m
New
Aged
Aged
38.49
32.50
34.70
38.50
33.00
34.50
SC
Aged
2.29
3.01
2.93
1.54
26.0-38.0
24.5-47.0
17.0 - 38.0
30.0 - 38.5
125
315
125
125
0
0
0
1
0
1.92
1.97
2.07
1.98
1.97
1.95
2.05
1.95
2.00
2.05
2.00
2.00
1.95
SD
0.17
0.14
0.17
0.15
0.13
0.14
0.14
Range
1.30 - 2.45
1.30 - 2.20
0.95 - 2.35
1.70- 2.45
1.35- 2.25
1.15 - 2.25
1.65 - 2.20
No tested
315
125
315
125
315
125
125
No Failed
0
0
1
0
0
0
0
Test Date
20/01/06
22/10/14
16/02/06
22/10/14
27/02/06
22/10/14
23/10/14
Mfg Date
10/2005
10/2005
01/2006
01/2006
01/2006
01/2006
----
Exp Date
09/2010
09/2010
12/2010
12/2010
12/2010
12/2010
04/2013
No Tested
315
200
315
200
315
200
200
No Failed
0
0
0
0
0
0
0
Test Date
23/01/06
24/10/14
20/02/06
23/10/14
6/03/06
23/10/14
23/10/14
Mfg Date
10/2005
01/2006
01/2006
12/2005
01/2006
01/2006
----
12/2010
12/2010
11/2010
12/2010
12/2010
04/2013
AC C
EP
Leaks
New
M AN U
Press (kPa)
Aged
TE D
Vol (L)
New
RI PT
Table 2 - Test results for condoms tested long after expiry date
Exp Date
09/2010
All products tested in this part of the study passed the inflation and leaks tests. For leaks, the sample size was 315 where possible, and 200 otherwise. No holes were found. For inflation, 125 condoms were used for the aged condoms, because of limited availability. Of the 125 condoms tested, there were no non-compliers except in one case (one non-complier). According to ISO 4074 and ISO 2859-1, up to 5 non-compliers would be allowed. 10 non-
ACCEPTED MANUSCRIPT
compliers would be allowed for a sample of 315, and 7 for a sample of 200. Also, the means and standard deviations of volume and pressure indicate that the distributions were well clear of the pass/fail limits.
RI PT
One can conclude that condoms stored in temperate to sub-tropical climates meet the ISO standard long after their expiry dates.
Data received from a manufacturer
M AN U
4.
SC
Results for the three products for which the original test data were also available show that there is no consistent trend in the burst properties, except that the mean burst volume decreased a little.
One manufacturer, Rosella, submitted results of recent (June 2014) inflation tests on plain condoms made in each of the years 2005 to 2013. The condoms had been stored at a mean temperature of approximately 30oC.
Accelerated aging of condoms nearing expiry date
EP
5.
TE D
These results appeared inconclusive, partly because of the low sample size. On the basis of this small sample, the lots made in 2007, 2009, 2011, 2012 and 2013 met the requirements, while the others (ie 2005, 2006, 2008, 2010) no longer did.
AC C
The condoms used in this section were taken from Enersol's retained samples or supplied by manufacturers. Lion condoms were stored in ambient tropical conditions, while the other condoms for which original data are available were stored in Sydney, with a mean kinetic temperature of approximately 20oC. Condoms which were 4 to 6 years old were conditioned at 50oC for 90 days. The two Dingo products are an exception to this. They had been stored in a sub-tropical location with an annual mean temperature similar to Sydney's until September, 2014, and then sent to Enersol where they were oven-conditioned for 1 month at 50oC. According to ISO 4074, 50oC for 90 days can be used as a provisional equivalent to 3 years’ storage at 30oC (the standard temperature used for real time shelf life determination). In this case, the condoms had been stored at room temperature in Sydney (mean kinetic temperature of
ACCEPTED MANUSCRIPT
approx. 20oC) for at least 4.5 years before being exposed to the 50oC oven. The ovenconditioning time was limited due to the time available for the study. The purpose of this combined exposure was to simulate, as far as practicable, a period of storage significantly longer than 5 years, with relatively warm conditions.
RI PT
12 lots were tested. Where possible, 125 condoms were used in the inflation test and 200 were tested for leaks. In one case, the numbers were reduced to 80 for inflation and 125 for leaks. Sample sizes were smaller than what is generally used for compliance testing by WHO13, because of the limited availability of condoms. The leaks test used was the ISO 4074:2014 version of the conductivity test.
M AN U
SC
For 5 of the 12 batches, Enersol had original test data available from the time of manufacture. For another 4, there were sufficient samples available to be able to be able to compare the results before and after oven conditioning. For the remaining 3 batches, no information is available about the properties of the condoms before oven conditioning, except the presumption that they met the ISO 4074 requirements when they were new.
AC C
EP
TE D
The inflation test results are summarised below.
ACCEPTED MANUSCRIPT
Table 3 - Mean burst volume of condoms with accelerated aging after expiry date
Shark
May 2008
Frog
Aug 2009
Ant
Feb 2009
Koala
Apr 2010
Owl 1
Feb 2009
Date of test Oct 2014 Nov 2014 Sept 2009 Nov 2014 Apr 2009 Nov 2014 Jun 2010 Nov 2014
Mean vol
No tested
34.7 35.8 43.5 33.7 42.0 33.7 30.3 25.3
125 125 200 125 315 125 315 125
No of noncompliers 0 0 0 0 2 0 1 1
50oC 90 days
Nov 2014
36.0
125
0
50oC 90 days Shelf since 2009 50oC 90 days
Nov 2014 Oct 2014 Nov 2014
36.6 32.7 32.6
125 125 125
0 2 1
Nov 2014 Nov 2010 Nov 2014 Jan 2010 Dec 2014 Nov 2014 Oct 2014 Nov 2014 Oct 2014 Nov 2014
28.3 33.2 23.3 42.2 39.0 16.3
125 200 125 200 50 13
0 1 2 0 0 8
34.2
125
0
33.9
125
1
Cat
Nov 2009
Eel
Jan 2010
Lion*
Nov 2010
Bat
Nov 2009
Dingo 2
Aug 2008 Aug 2008
EP
Dingo 3
50oC 90 days Original test 50oC 90 days Original test Shelf since 2010 o 50 C 90 days Shelf since 2008 o 50 C 30 days Shelf since 2008 o 50 C 30 days
M AN U
Mar 2009
TE D
Owl 2
Sample Conditioning Shelf since 2008 50oC 90 days Original test 50oC 90 days Original test 50oC 90 days Original test 50oC 90 days
RI PT
Mfr date
SC
Product
AC C
One brand, Bat, had suffered severe degradation, and it was pointless to do the tests. The test results on the same batch when new were acceptable. Additional retained samples found later also passed when tested without the additional oven conditioning. The other 11 batches all passed the ISO 4074 requirements. The consistency of the results is convincing, despite the small sample sizes. Among the 5 products where original data are available, all suffered significant drops in mean burst volume although, except for Bat, there was no apparent increase in the rate of noncomplying condoms. After 4 years’ natural aging and 3 months at 50oC, Lion condoms had a relatively low mean volume, but still complied with the standard. The effect on volume of the oven-conditioning on the four batches of condoms for which testing was also done before oven-conditioning was relatively minor. Once again, there was no apparent increase in the rate of non-compliers.
ACCEPTED MANUSCRIPT
Table 4 - Mean burst pressure of condoms with accelerated aging after expiry date
Shark
May 2008
Frog
Aug 2009
Ant
Feb 2009
Koala
Apr 2010
Owl 1
Feb 2009
Date of test Oct 2014 Nov 2014 Sept 2009 Nov 2014 Apr 2009 Nov 2014 Jun 2010 Nov 2014
Mean press
No tested
1.95 1.72 2.16 2.00 1.97 1.99 2.51 2.16
125 125 200 125 315 125 315 125
50oC 90 days
Nov 2014
2.00
50oC 90 days Shelf since 2009 50oC 90 days
Nov 2014 Oct 2014 Nov 2014
1.93 2.08 2.10
Nov 2014 Nov 2010 Nov 2014 Jan 2010 Dec 2014 Nov 2014 Oct 2014 Nov 2014 Oct 2014 Nov 2014
0
125 125 125
0 2 0
Nov 2009
Eel
Jan 2010 Nov 2010
Bat
Nov 2009
Dingo 2
Aug 2008
Dingo 3
Aug 2008
EP
Lion*
50oC 90 days Original test 50oC 90 days Original test Shelf since 2010 o 50 C 90 days Shelf since 2008 o 50 C for 30 days Shelf since 2008 o 50 C for 30 days
1.65 2.00 2.01 2.04 1.66 0.85
125 200 125 200 50 13
0 1 0 0 0 8
1.98
125
0
1.98
125
0
M AN U
Cat
No of noncompliers 0 0 0 0 2 0 0 1
125
Mar 2009
TE D
Owl 2
Sample Conditioning Shelf since 2008 50oC 90 days Original test 50oC 90 days Original test 50oC 90 days Original test 50oC 90 days
RI PT
Mfr date
SC
Product
AC C
The mean burst pressure of most products was relatively unaffected by the oven conditioning, and by storage on the shelf.
ACCEPTED MANUSCRIPT
Table 5 - Leaks found in condoms exposed to accelerated aging after expiry date
Shark
May 2008
Frog
Aug 2009
Ant
Feb 2009
Koala
Apr 2010
Owl 1
Feb 2009
Sample Conditioning Shelf since 2008 o 50 C 90 days Original test o 50 C 90 days Original test o 50 C 90 days Original test o 50 C 90 days o
Owl 2
Mar 2009 o
Cat
Nov 2009
Eel
Jan 2010
50 C 90 days Shelf since 2009 o 50 C 90 days o
Nov 2010
Bat
Nov 2009
Dingo 2
Aug 2008
200
0
Nov 2014 Oct 2014 Nov 2014
200 200 200
0 0 0
Nov 2014 Nov 2010 Nov 2014 Jan 2010 Nov 2014 Oct 2014 Nov 2014 Oct 2014 Nov 2014
200 315 200 315 10
0 0 0 0 2
200
0
200
0
Discussion
AC C
6.
Aug 2008
200 200 315 200 315 200 315 200
No of noncompliers 0 0 0 0 0 0 3 0
EP
Dingo 3
50 C 90 days Original test o 50 C 90 days Original test o 50 C 90 days Shelf since 2008 o 50 C for 30 days Shelf since 2008 o 50 C for 30 days
TE D
Lion*
Nov 2014
No tested
M AN U
50 C 90 days
Date of test Oct 2014 Nov 2014 Sept 2009 Nov 2014 Apr 2009 Nov 2014 Jun 2010 Nov 2014
RI PT
Mfr date
SC
Product
The purpose of the studies was to expose the condoms to aging conditions that were in excess of the normal 5 year shelf life study, but to do so over a relatively short period. In order to achieve this, it was necessary to find samples that had already aged naturally for several years. The oldest ones were tested without further challenge, while those around 4 to 5 years old were ovenconditioned at 50oC. The study design was limited by what was available for testing and UNFPA’s time constraints. Among the condoms aged naturally (6 to 9 years old), all passed the tests quite comfortably.
ACCEPTED MANUSCRIPT
From results submitted by a manufacturer whose products had been stored at a mean kinetic temperature of around 30oC, some batches passed and some failed. In particular, those made before 2007 failed. Thus condoms up to 7 years old from this manufacturer can pass the tests, although some of the newer products did not.
RI PT
From tests on condoms that had had 4 to 5 years’ natural aging, plus oven conditioning at 50oC, all but one product passed all the tests at the end. Apart from Bat, none of the products showed an increase in the number of holes.
1.
SC
Because the exact equivalence between aging at elevated temperatures and that at room temperature cannot be established, one must rely on approximate conclusions. The following can be inferred: Different products behave differently in the face of the challenges of aging.
M AN U
2. One product out of 12 performed very badly after exposure to the combined challenge of several years of ambient aging, followed by 3 months in an oven at 50oC. 3. The other 11 products still complied with ISO 4074, and the rate of non-compliers was not measurably higher as a result of the natural or accelerated aging.
TE D
4. All 5 products with original test data available showed a loss of mean burst volume as a result of the combined aging protocol 5. The four products for which tests were done before and after oven conditioning showed relatively little change in mean burst volume
EP
6. Mean burst pressure falls in some products (even over the 90 days' oven conditioning), while in others, it remains relatively unchanged. It remains high enough to ensure that the products will pass the ISO requirement
AC C
7. Only one product demonstrated any increase in the rate of holes. That product also showed catastrophic decay in its inflation properties. The ability of the study to accurately reflect real time aging is limited by the sample sizes available for testing, and by the time available for the study. To obtain more definite assurance that products do indeed continue to comply with the requirements of the standard after their expiry date, it is necessary to conduct a real time study, which would last at least 6 years. If continued compliance with the standard were thought to be insufficient evidence of fitness for use, then it would probably be necessary to conduct in vivo breakage trials using suitably aged condoms - a very lengthy and expensive project.
ACCEPTED MANUSCRIPT
7.
Conclusions
RI PT
Most of the products tested, whether after natural aging well beyond the expiry date, or by four to five years’ natural aging followed by 3 months’ oven conditioning, passed the ISO 4074 test requirements comfortably. With one exception, all products were still in compliance with ISO 4074 at the end of the exposure to elevated temperature.
SC
The exception, Bat, was made in 2009, and oven conditioning caused it to fail inflation testing drastically, and to stick to the pack when opened. It was concluded that this batch of Bat products was in some way defective in manufacture, and did not necessarily represent the manufacturer's overall capabilities.
M AN U
10 of the 12 products tested appeared to withstand the combination of ambient temperature aging under moderately warm conditions for 4 to 6 years plus 3 months at 50oC without serious degradation to their properties. According to trials conducted previously14, 3 months at 50oC is approximately equivalent to 2 to 3 years at 30oC. It thus appears that these products would withstand at least 6 years' storage in appropriate warehouses, even in tropical conditions.
EP
TE D
If it is accepted that continued compliance with ISO 4074 indicates continued fitness for use, then it should be acceptable to determine the shelf life by real time aging at 30oC as required in the standard, and the five year arbitrary limit on shelf life could be gradually removed. Manufacturers should be free to extend their shelf-life (in modest steps) based on demonstrated compliance with the requirements of clause 11 of the ISO 4074 standard at the end of the time they stipulate for shelf life. Manufacturers may be able to use their retained samples of past expiry date product to do this.
AC C
If it is not accepted that continued compliance with the physical property requirements of ISO 4074 indicates continued fitness for use, then a fundamental re-think of the condom standard is required.
Funding
The experimental work in this article was funded by the United Nations Population Fund (UNFPA). UNFPA did not exercise any editorial control over the contents of the article.
ACCEPTED MANUSCRIPT
References 1
AC C
EP
TE D
M AN U
SC
RI PT
J Gerofi and M Sorensen, Re-evaluation of data and requirements on condom shelf life. Polymer Testing, 54, September 2016, pp 260–269 2 ISO 4074, Natural Rubber Latex Male Condoms – Requirements and Test Methods, International Organization for Standardization, 2014. 3 ISO 4074, Natural Rubber Latex Male Condoms – Requirements and Test Methods, International Organization for Standardization, 2015. 4 Council Directive 93/42/EEC of 14 June 1993 concerning medical devices, as amended. http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:31993L0042 5 http://www.who.int/topics/clinical_trials/en/ (accessed June 2016) 6 http://www.nhs.uk/Conditions/Clinical-trials/Pages/Healthresearch.aspx (accessed June 2016) 7 Steiner M, Foldesy R and Cole, D Study to Determine the Correlation between Condom Breakage in Human Use and Laboratory Test Results, FHI, November 1991 8 ISO/DIS 29943-1 Condoms-Guidance on Clinical Studies – Part 1: Male condoms, clinical function studies based on self-reports, International Organization for Standardization, 2014 9 http://www.unfpa.org/publications/condom-programming-hiv-prevention-0 (accessed June 2016) 10 http://www.unfpa.org/resources/male-latex-condom (accessed June 2016) 11 https://www.unfpa.org/sites/default/files/resourcepdf/20429%20UNFPA%20RFQ%20QA%20 Framework_low_V11_2.pdf (accessed June 2016) 12 ISO 2859-1, Sampling Plans for Inspection by Attributes – Part 1: Sampling plans indexed by acceptable quality limit for lot-by-lot inspection, International Organization for Standardization, 1999 13 Male Latex Condom: Specification, Prequalification and Guidelines, 2010, World Health Organization, UNFPA and FHI, 2010. ISN 978 92 4 159990 0. 14 M C Bó, J P Gerofi, L L Y Visconte and R C R Nunes, Prediction of Shelf Life of Natural Rubber Male Condoms - A Necessity, Polymer Testing, 26, 3, 2007 - pp. 306-314.