Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Articles
www.thelancet.com/lancetgh Vol 1 September 2013 e
146
Performance and safety of the second-generation female
condom (FC2) versus the Woman’s, the VA worn-of-women,
and the Cupid female condoms: a randomised controlled
non-inferiority crossover trial
Mags E Beksinska, Gilda Piaggio, Jennifer A Smit, Junqing Wu, Yufeng Zhang, Jacqueline Pienaar, Ross Greener, Ying Zhou, Carol Joanis
Summary
Background New designs of female condom have been developed to reduce costs and improve acceptability. To secure
regulatory approvals, clinical studies are needed to verify performance. We aimed to assess the functional performance
and safety of three new condom types—the Woman’s Condom, the VA worn-of-women (wow) Condom Feminine,
and the Cupid female condom—against the existing second-generation female condom (FC2).
Methods We did a randomised controlled, non-inferiority, four-period crossover trial at three sites in Shanghai, China,
and one site in Durban, South Africa, between May 1, 2011, and Jan 31, 2012. Participants aged 18–45 years who were
sexually active, monogamous, not pregnant, and not sex workers, were eligible for inclusion if they were literate, had no
known allergies to the study products; used a reliable, non-barrier method of contraception, and had no visible or reported
sexually transmitted infections. We used a computer-generated randomisation sequence with a Williams square design
of size four to assign patients (1:1:1:1) to the FC2 control device, or the Woman’s, VA wow, or Cupid condoms, with
12 potential allocations. Randomisation was stratifi ed by site. Participants were not masked to condom type, but allocation
was concealed from study investigators. The primary non-inferiority endpoints were total clinical failure and total female
condom failure, with a non-inferiority margin of 3%. Women were asked to use fi ve of each condom type and were
interviewed after use of each type. We also assessed safety data for each type. We did both per-protocol and intention-to-
treat analyses. We calculated frequencies and percentages for each failure event and estimated diff erences in performance
with a generalised estimating equation model. This study is registered, number DOH-27-0113-4271.
Findings 616 women were assessed for eligibility, of whom 600 were randomly assigned to condom-type order (30, 120,
and 150 women in the three sites in China, and 300 women in the site in South Africa). 572 women completed follow-
up, with at least one condom of each type. Total female condom failure was 3·43% for FC2, 3·85% for the Woman’s
Condom (diff erence 0·42%, 90% CI –1·42 to 2·26), 3·02% for VA wow (–0·42%, –1·86 to 1·32), and 4·52% for
Cupid (1·09%, –0·60 to 2·78); total clinical failure was 2·88%, 3·05% (0·17%, –1·48 to 1·81), 2·49% (–0·25%,
–1·75 to 1·26), and 3·87% (0·99%, –0·55 to 2·52), respectively. Only two (<1%) participants, in South Africa, reported
serious adverse events, unrelated to use of the study products.
Interpretation Non-inferiority was shown for all condom failure events for the three new devices versus the FC2,
within the predefi ned margin.
Funding Universal Access to Female Condoms (UAFC).
Introduction
Evidence is scarce about the eff ect of choice of
contraceptive method on increased uptake of methods;
however, a systematic review1 in 2006 supported the
theory that increased contraceptive choice for women is
associated with increased uptake and better health
outcomes (eg, lower pregnancy rates and fewer sexually
transmitted infections [STIs]). Furthermore, women
continue use of their chosen contraceptives to a greater
degree than do those who are denied a choice.2 The female
condom is a barrier method that could increase choice for
women worldwide. Although distribution of female
condoms continues to increase worldwide, doubling
from 25 million to 50 million units between 2007 and
2010, it is substantially lower than that for male condoms,
and accounts for only 0·19% of global condom
procurement.3 In 2011, the female condom was identifi ed
by the Reproductive Health Supplies Coalition as one of
several underused reproductive health technologies that
could expand choice in reproductive health and family
planning programmes, add value to the method mix, and
respond to the needs of various clients.4
Several new female condoms in the fi nal stages of
development, or that have recently become available, aim
to reduce unit cost or improve acceptability.5 The fi rst-
generation female condom (FC1), made by the Female
Health Company, was approved by the US Food and Drug
Administration (US FDA) in 1993 (now discontinued).
Classifi ed as class 3 medical devices by the US FDA, the
regulatory process for female condoms is more complex
Lancet Glob Health 2013;
1: e146–52
See Comment page e119
Copyright © Beksinska et al.
Maternal, Adolescent and Child
Health (MatCH), Department
of Obstetrics and Gynaecology,
Faculty of Health Sciences,
University of the
Witwatersrand, Johannesburg,
South Africa
(M E Beksinska PhD,
J A Smit PhD, J Pienaar MSc,
R Greener MSocSc); Faculty of
Epidemiology and Population
Health (M A Beksinska) and
Department of Medical
Statistics (G Piaggio PhD),
London School of Hygiene and
Tropical Medicine, London, UK;
School of Pharmacy and
Pharmacology, Faculty of
Health Sciences, University of
KwaZulu-Natal, Durban, South
Africa (J A Smit); Department of
Epidemiology and Social
Science on Reproductive
Health, Shanghai Institute of
Planned Parenthood Research,
Fudan University, Shanghai,
China (J Wu PhD, Y Zhang MSc,
Y Zhou MSc); and Joanis
Consulting, Kennesaw, GA. USA
(C Joanis BSc)
Correspondence to:
Dr Mags Beksinska, Maternal,
Adolescent and Child Health
(MatCH), Department of
Obstetrics and Gynaecology,
Faculty of Health Sciences,
University of the Witwatersrand,
Overport, Durban 4091,
South Africa
mbeksinska@match.org.za
Open access under CC BY-NC-ND license.
Articles
e147
www.thelancet.com/lancetgh Vol 1 September 2013
than that for male condoms. This diff erence was
compounded by the absence, until 2011, of an
international standard to verify the quality of new devices.
To secure regulatory approvals, including WHO–UNFPA
prequalifi cation, manufacturers need to do clinical
studies to verify the performance of new designs of
female condoms. In 2009, the US FDA approved the
second-generation female condom (FC2) on the basis of
results of a non-inferiority study6 in which the
investigators compared the functional performance of
FC2 with that of FC1 with respect to condom failure
events. Studies of the functional performance of condoms
typically collect detailed data for small numbers of
condom uses (fi ve to ten uses) in a short period of time
(4–6 weeks). Conversely, studies of condom eff ectiveness
for prevention of pregnancy should be done over at least a
6 month period and are far more costly to do than are
functional performance studies. Studies of contraceptive
eff ectiveness were not needed for US FDA approval for
FC2 and no studies of contraceptive eff ectiveness have
been published for FC2. If new female condoms are non-
inferior to FC2 in function, choices will increase and
individual needs can be considered for women wanting to
prevent pregnancy and STIs and HIV. We assessed the
functional performance and safety of three new female
condom designs: the Woman’s Condom, the VA worn-of-
women (wow) Condom Feminine, and the Cupid female
condom versus the FC2 device.
Methods
Study design and participants
We undertook this four-period, randomised, non-
inferiority, crossover, clinical trial between May 1, 2011,
and Jan 31, 2012, at three centres in Shanghai, China (the
Shanghai Institute of Planned Parenthood Research
Hospital and two affi liated family planning districts—
XuJiahui and Xuhui), and one centre in Durban, South
Africa (an urban reproductive health clinic). Participants
aged 18–45 years who were sexually active, monogamous,
and not practising sex workers, were eligible for inclusion
if they had no known allergies to the study products
(latex, synthetic latex, polyurethane); used a reliable, non-
barrier method of contraception; and had no STIs (as
established by pelvic examination and use of a syndromic
diagnostic approach). Participants could be novice or
experienced users of condoms, but had to be literate
because take-home condom logs were used to obtain
data. We excluded pregnant women (established by urine
Figure 1: Trial profi le
The numbers of women withdrawing after period one were two for the sequence CWVF, three for FVWC, none for VCFW, and three for WFCV; after period two were three, two, one, and two,
respectively; and for period three were three, two, one, and two, respectively. C=Cupid. W=Woman’s Condom. V=VA worn-of-women. F=second-generation female condom (FC2).
75 CWVF 75 FVWC 75 VCFW 75 WFCV 50 CVFW 50 VWFC 50 WFVC 50 FCWV 25 VFWC 25 CWFV 25 FCVW 25 WVCF
66 included
in main
analysis
66 included
in main
analysis
70 included
in main
analysis
70 included
in main
analysis
50 included
in main
analysis
50 included
in main
analysis
50 included
in main
analysis
50 included
in main
analysis
25 included
in main
analysis
25 included
in main
analysis
25 included
in main
analysis
25 included
in main
analysis
1 used no
condom
1 lost to
follow-up
8 did not
use all
condom
types
3 lost to
follow-
up
5 for non-
safety
reasons
66 used all
condom
types
2 used no
condom
2 for non-
safety
reasons
7 did not
use all
condom
types
1 lost to
follow-
up
6 for non-
safety
reasons
66 used all
condom
types
5 did not
use all
condom
types
3 lost to
follow-
up
2 for non-
safety
reasons
70 used all
condom
types
5 did not
use all
condom
types
1 lost to
follow-
up
4 for non-
safety
reasons
70 used all
condom
types
50 used all
condom
types
50 used all
condom
types
50 used all
condom
types
50 used all
condom
types
25 used all
condom
types
25 used all
condom
types
25 used all
condom
types
25 used all
condom
types
600 randomised
16 excluded
16 did not meet inclusion criteria
616 women assessed for eligibility
Articles
www.thelancet.com/lancetgh Vol 1 September 2013 e
148
pregnancy test). Participants from the South African site
were similar in profi le and recruited from the same site
as that in the study used to establish the comparative
performance of the reference condom (FC2) to the
predicate device, FC1.6 Participants gave written informed
consent before screening and enrolment.
The study was approved by the institutional review
boards of the University of Witwatersrand Human
Research Ethics Committee in South Africa and the
Ethical Committee of Shanghai Institute of Planned
Parenthood Research in China. The study was also
approved by the National Population and Family Planning
Commission (NPFPC) in China and by the provincial,
district, and local departments of health in South Africa.
Male partners were informed of the study through use of
a fact sheet, which explained the purpose of the study and
provided information about the study products and their
role in the research. Our study methods are based on the
recommendations of WHO–UNFPA.7
Randomisation and masking
We used a computer-generated randomisation sequence
(SAS version 9.38) to assign patients to one of 12 se-
quences of condom use (fi gure 1). Randomisation was
stratifi ed by site. We used a Williams design, which
consisted of the random construction of a Williams
square (fi rst step) and the random allocation of women
to the treatment sequences, independently at each site
(second step).9 These sequences were balanced for
period and treatment, and each treatment had a
diff erent preceding treatment. To ensure such balance,
the Williams design uses block sizes of four in a Latin
square.
To devise a simple and eff ective method of concealment,
we designed a scratch card per participant, containing
the allocated treatment sequence, whereby each code for
a given condom type was concealed beneath a separate
foil square. For each card, research staff would remove
(scratch) the foil corresponding to the visit number
printed above the square, thereby revealing the next visit
allocation in the sequence. This novel adaption of an
existing technology had not been previously reported for
concealment of the allocation sequence in a randomised
crossover trial. The development and use of the cards for
this trial has been reported elsewhere.10 Participants were
not masked to condom type because the designs are all
quite distinct and required product-specifi c training.
Procedures
Our primary objective was to compare the functional
performance of the four types of female condom within
the selected study populations (fi gure 2). The FC2 was
the control device, because the predicate device (FC1)
ceased production in 2009. Each condom product was
shipped by the manufacturer to Family Health
International (now FHI 360) for quality assurance testing
to ensure that products were of the quality specifi ed by
The second-generation female condom (FC2)
In 2009, the FC2 replaced the polyurethane first-generation female
condom (FC1; trade names include Reality and Femidom), which had
been available since 1993. FC2 is similar in specification and
appearance to FC1, but is made of synthetic nitrile rubber latex. FC2
has a flexible inner ring to insert the device and keep it in place during
use. A ring at the open end lies flat across the genital area. FC2 was
regarded as acceptable for bulk procurement by all UN agencies in 2007.
In 2009, FC2 was approved by the US Food and Drug Administration.
The Woman’s Condom
The Woman’s Condom, designed and developed by the Program for
Appropriate Technology in Health (PATH), is made of polyurethane.
The condom sheath is tucked into an insertion capsule that dissolves
after insertion. The four foam dots on the body of the condom ensure
internal stability. The condom is not pre-lubricated and is supplied with
a water-based lubricant. Licensed to the Dahua Medical Apparatus
Company (Shanghai, China), the Woman’s Condom received the
CE Mark of the European Union in 2010, and the approval of the
Shanghai Food and Drug Administration in 2011.
The VA wow Condom Feminine
The VA wow condom (Medtech Products, India) is made of natural
rubber latex and encases a medical-grade sponge at the closed end.
The sponge is used for insertion and the outer anchoring structure has
a triangular-shaped frame. The condom is lubricated with silicone oil.
The device carries the CE Mark of the EU, and has approvals from the
India Drug Control Authority and the Ministry of Health in Brazil.
The Cupid female condom
The Cupid condom (Cupid, India), manufactured in India, is available in
some European and African countries. Made of natural rubber latex, the
condom has an octagonal outer frame and is inserted with a medical-
grade sponge, which also holds the condom in place during use. The
device is pre-lubricated with silicone oil and comes in natural latex and
pink colours. The Cupid condom is the only scented condom of all those
assessed. The condom holds the CE Mark of the EU, is prequalified by
WHO–UN Population Fund and registered by the India Drug
Control Authority.
A
B
C
D
Figure 2: Description of study products
Panel 1: Defi nitions of failure modes
• Total clinical failure: the number of female condoms that clinically break or slip, or that are
associated with misdirection, invagination, or any additional failure modes identifi ed in
the risk assessment, which result in reduction of the protective function of the condom
• Total condom failure: a female condom for which a non-clinical breakage, clinical
breakage, or slippage occurs, or that is associated with misdirection, invagination,
or any additional failure modes identifi ed in the risk assessment
• Clinical breakage: breakage of the condom during sexual intercourse or during withdrawal
of the female condom from the vagina (potential adverse clinical consequences)
• Non-clinical breakage: breakage noticed before intercourse or occurring after
withdrawal of the condom from the vagina (no potential adverse clinical consequences)
• Total breakage: the number of all condom breakages at any time before, during,
or after sexual intercourse; includes both clinical and non-clinical breakages
• Slippage: when a female condom slips completely out of the vagina during sexual
intercourse
• Misdirection: vaginal penetration whereby the penis is inserted between the condom
and the vaginal wall
• Invagination: when the external retention feature of the female condom is partly or
fully pushed into the vagina during sexual intercourse
Articles
e149
www.thelancet.com/lancetgh Vol 1 September 2013
the manufacturer and met International Organization
for Standardization (ISO 25841-2011) requirements.7
The primary endpoints were self-reported total
clinical failure and total female condom failure.
Additionally, we assessed the component failure events
of clinical breakage, non-clinical breakage, total
breakage, slippage, misdirection, and invagination.11
Panel 1 shows the defi nitions of each failure mode
analysed. We also collected data for safety and
acceptability. Failure events are recognised by WHO
and other regulatory agencies.
Our secondary objectives were to assess the safety and
acceptability of each condom type and to compare
acceptability endpoints between the four types, with
superiority hypotheses. We measured and assessed
safety of each device according to number, severity,
relatedness, and duration of adverse events. We collected
standard acceptability measures and these data will be
reported elsewhere.
In this study, each woman was asked to use fi ve of each
of the four condom types and to complete a condom log
at home after each condom use. After women completed
use of each condom type, they returned to the clinic to be
interviewed about their experiences. Condom logs were
used to obtain data for condom function and safety.
We used interviewer-assisted questionnaires to gather
acceptability and preference data.
Statistical analysis
We calculated the power to show non-inferiority
obtained for diff erent sample sizes, starting with a
minimum of 200 couples completing the study as
recommended by ISO 25841-2011 for functionality
studies of acute failure events with female condoms
based on self-reports.12 We assumed a total failure rate
of 4% for FC2 as reported from previous research,6 and
a correlation between uses of 0·15, as reported for male
condoms.13 With 3% as a clinically determined margin
of non-inferiority and a signifi cance level of 5% for the
hypothesis of non-inferiority, 500 women (250 per
country) completing the study would provide 98%
power in each country. We expected a non-completion
rate of 15% (ie, 85% would provide relevant follow-up
data for at least one condom of each type); therefore, we
increased the sample size to 600 women (300 per
country). In the power calculation, we adjusted for
various comparisons with the Bonferroni criterion.14
The hypothesis for the primary endpoints of total
clinical failure and total female condom failure and their
component failure events, was that each of the three new
condoms was non-inferior to FC2 for the rate of events
within a margin of 3·0%.
The main analysis for primary and secondary endpoints
was according to the assigned condom use sequence in the
subset of participants who provided relevant follow-up data
for at least one condom of each type (per-protocol analysis).
We did an additional analysis with women with complete
or incomplete condom series (one to four condom types,
one to fi ve uses each; data not shown). This additional
analysis was by intention to treat, but with exclusion of
three participants for whom no data about condom
function were available. All analyses were stratifi ed by
country.
To show non-inferiority, the upper limit of the two-sided
90% CI for the diff erence in the occurrence of events (new
condom minus FC2) needed to be below 3·0%. If non-
inferiority was shown, we tested superiority at a 5% level
of signifi cance using a two-sided 95% CI.
We analysed primary endpoints with a generalised
estimating equation approach assuming a binomial
China
(n=300)
South Africa
(n=272)
Total
(N=572)
Age (years)
Mean (SD) 36·3 (5·68) 27·9 (6·29) 32·3 (7·30)
Min–max 21–45 18–44 18–45
Ethnic group
Asian (Chinese) 299 (>99%) 0 299 (52%)
Black 0 271 ( >99%) 271 (47%)
Coloured 0 1 (<1%) 1 (<1%)
White 1 (<1%) 0 1 (<1%)
Completed years of school
0–5 4 (1%) 2 (<1%) 6 (1%)
6–10 79 (26%) 24 (9%) 103 (18%)
11–15 160 (53%) 244 (90%) 404 (71%)
16–20 57 (19%) 2 (<1) 59 (10%)
Primary occupation
Health or medical 39 (13%) 5 (2%) 44 (8%)
None or unemployed 34 (11%) 134 (49%) 168 (29%)
Offi ce, public service, or government 104 (35%) 4 (1%) 108 (19%)
Sales 33 (11%) 12 (4%) 45 (8%)
Self employed or other 51 (17%) 21 (8%) 72 (13%)
Student 50 (18%) 50 (9%)
Teacher or lecturer 6 (2%) 3 (1%) 9 (2%)
Technical 13 (4%) 1 (<1) 14 (2%)
Unskilled labour 20 (7%) 42 (15%) 62 (11%)
Number of living children
0 20 (7%) 50 (18%) 70 (12%)
1 266 (89%) 117 (43%) 383 (67%)
2 14 (5%) 70 (26%) 84 (15%)
≥3 35 (13%) 35 (6%)
Civil status
Married or living together 298 (99%) 62 (23%) 360 (63%)
Not married and not living together 2 (<1%) 210 (77%) 212 (37%)
Duration of relationship with current spouse or partner
<1 year 4 (1%) 17 (6%) 21 (4%)
1–5 years 61 (20%) 156 (57%) 217 (38%)
6–10 years 57 (19%) 58 (21%) 115 (20%)
>10 years 178 (59%) 41 (15%) 219 (38%)
Data are number of individuals (%) unless otherwise indicated.
Table 1: Baseline sociodemographic characteristics of the main analysis population by country and overall
Articles
www.thelancet.com/lancetgh Vol 1 September 2013 e
150
distribution, and with an identity link function, including
type of condom in the model at the couple-use level and
defi ning couples as clusters, to take into account the
crossover nature of the design. We used an
interchangeable error structure within couples and
adjusted results by multiplicity with Scheff é’s method.13
This trial is registered, number DOH-27-0113-4271.
Role of the funding source
The sponsor of the study had no role in study design, data
collection, data analysis, data interpretation, or writ ing of
the report. MEB, GP, JP, RG, and CJ had full access to the
primary data. All other authors (JAS, JW, YZha, and YZho)
had access to the data collected in their own country. All
authors had fi nal responsibility for the publication.
Results
Figure 1 shows the trial profi le. 600 women were enrolled
in the study (300 per country). 572 (95%) women
completed the study with each woman using at least one
condom of each of the four types. These women comprise
the main analysis population. Only three women, in
South Africa, did not use any condom.
Chinese participants were older than participants from
South Africa (table 1). Women in both countries were
well educated, with 72% in China and 90% in South
Africa having more than 10 years of schooling (table 1).
Most Chinese participants were married or living with
their partners, whereas most South African participants
were not married or residing with their partners (table 1).
188 (63%) of 300 Chinese participants and 272 (100%) of
South African participants had used male condoms
previously, whereas 39 (14%) of South African women
had previous experience with female condoms compared
with no participants in China.
11 350 female condoms were used in this study by
women in the main analysis population (2838 Cupid,
2850 FC2, 2827 VA wow. and 2835 Women’s Condoms).
Table 2 shows the mean failure rate and failure diff erence
of Cupid, VA wow, and Woman’s Condom compared
with FC2 for both countries combined. Non-inferiority
was shown, within the non-inferiority margin, for all
failure modes for the three new devices versus FC2 (data
not shown) and for the two countries combined (table 2
and fi gure 3). The additional analysis done with women
with complete or incomplete condom series provided
similar results (data not shown).
We noted no evidence of superiority for any of the new
condoms compared with FC2 in any of the failure modes
(table 2).
Overall, the occurrence of adverse events and medical
problems for the enrolled population was low (data not
shown). Only two (<1%) participants in South Africa
reported serious adverse events, unrelated to use of the
study products. 27 (5%) adverse events were reported in
the 597 women using at least one condom at least once.
Adverse events occurred in fewer than 1% (range
0·14–0·38%) of participants for each condom type.
Adverse event reports were vaginal burning (fi ve with VA
wow in China); vaginal itching (one with Cupid, one with
FC2, and one with Women’s Condom in South Africa);
Mean failure
rate (%)
Failure diff erence (%) p value
Diff erence
with FC2
90% CI 95% CI
Clinical breakage*
Cupid 0·10 NE NE NE ··
VA wow 0·08 NE NE NE ··
WC 0·00 NE NE NE ··
FC2 0·25 NE NE NE ··
Non-clinical breakage
Cupid 0·66 0·04 –0·61 to 0·69 –0·69 to 0·77 0·9991
VA wow 0·53 –0·09 –0·52 to 0·33 –0·57 to 0·38 0·9586
WC 0·74 0·12 –0·41 to 0·65 –0·47 to 0·71 0·9557
FC2 0·62 (Ref) ·· ·· ··
Total breakage
Cupid 0·77 –0·05 –0·76 to 0·66 –0·84 to 0·74 0·9984
VA wow 0·67 –0·14 –0·67 to 0·38 –0·73 to 0·44 0·9226
WC 0·84 0·02 –0·57 to 0·60 –0·63 to 0·67 0·9999
FC2 0·82 (Ref) ·· ·· ··
Invagination
Cupid 1·21 0·03 –0·85 to 0·91 –0·95 to 1·01 0·9999
VA wow 0·99 –0·20 –1·17 to 0·77 –1·28 to 0·89 0·9684
WC 0·47 –0·72 –1·46 to 0·03 –1·55 to 0·12 0·1247
FC2 1·18 (Ref) ·· ·· ··
Misdirection
Cupid 1·22 0·62 –0·12 to 1·37 –0·21 to 1·46 0·2264
VA wow 1·19 0·58 –0·21 to 1·37 –0·30 to 1·47 0·3309
WC 1·13 0·53 –0·31 to 1·37 –0·41 to 1·47 0·4759
FC2 0·60 (Ref) ·· ·· ··
Slippage
Cupid 1·48 0·50 –0·41 to 1·42 –0·52 to 1·53 0·5979
VA wow 0·43 –0·55 –1·17 to 0·07 –1·25 to 0·14 0·1760
WC 1·28 0·30 –0·64 to 1·24 –0·75 to 1·35 0·8880
FC2 0·98 (Ref) ·· ·· ··
Total clinical failure
Cupid 3·87 0·99 –0·55 to 2·52 –0·73 to 2·70 0·4614
VA wow 2·49 –0·25 –1·75 to 1·26 –1·92 to 1·43 0·9823
WC 3·05 0·17 –1·48 to 1·81 –1·67 to 2·01 0·9957
FC2 2·88 (Ref) ·· ·· ··
Total female condom failure
Cupid 4·52 1·09 –0·60 to 2·78 –0·80 to 2·98 0·4552
VA wow 3·02 –0·42 –1·86 to 1·32 –2·05 to 1·51 0·9809
WC 3·85 0·42 –1·42 to 2·26 –1·64 to 2·47 0·9553
FC2 3·43 (Ref) ·· ·· ··
NE=not estimable. Cupid=Cupid female condom. VA wow=VA worn-of-women Condom Feminine. WC=Woman’s
Condom. FC2=second-generation female condom. 2-sided 90% CIs for non-inferiority hypothesis at α=5% and 2-sided
95% CIs for superiority hypothesis, with p va lues, adjustment for multiplicity by Scheff é’s method. *Estimated as mean
proportions of 75 women each because generalised estimasting equation model did not converge.
Table 2: Mean failure rate and failure diff erence of Cupid, VA wow, and Woman’s Condom in relation to
FC2, complete condom series population, for both countries
Articles
e151
www.thelancet.com/lancetgh Vol 1 September 2013
vaginal pain or swelling (two with Cupid, two with FC2,
one with VA wow in South Africa); penile itching, burning,
rash, or swelling (one with FC2, two with VA wow, and
one with Women’s Condom in South Africa); general rash
(one with Women’s Condom in China); and unknown,
unclear, or unrelated rash (three with Cupid, two with
FC2, three with VA wow, and one with Women’s Condom
in South Africa). All events resolved without sequelae.
Discussion
Although individual failure rates vary by country and
condom type, the three new devices assessed in this
study were non-inferior to the reference condom (FC2)
within the non-inferiority margin for all the functional
parameters. The rates of total clinical failure, total female
condom failure, and component failure reported were
expected and are consistent with fi ndings from previous,
similarly undertaken studies.6,15–17 Furthermore, the
failure results reported in this study were from two
culturally diverse study populations. That the results are
similar between South Africa and China attests further to
the validity of the fi ndings. These data show that
clinically, the new female condoms assessed function in
a similar way to the FC2. Data from this study have been
used to fi nalise the dossier submitted to WHO–UNFPA
for the Cupid condom and resulted in its prequalifi cation
for public sector procurement in June, 2012.18 The
manufacturer of the Woman’s Condom (Dahua, China)
will use these data to fi nalise their dossier and
submissions to WHO–UNFPA and US FDA. The VA
wow condom is under review by WHO–UNFPA;
however, manufacture of the product is suspended,
possibly related to change of ownership of the factory.
On the basis of fi ndings from this trial, these new
female condoms could be important players in
contraception and STI prevention. The availability of new
types of female condoms will provide donors and users
with more options and could reduce the unit price of
devices.4 Although prediction of how far the price of
female condoms will reduce in the future is not possible,
the Reproductive Health Supplies Coalition has
published information comparing diff erent unit costs to
the donor funded public sector, indicating that the Cupid
female condom, which was prequalifi ed in 2012, will be
sold at a lower price per unit than FC2.4 However, studies
of contraceptive eff ectiveness and those showing
eff ectiveness or reductions in STI incidence with these
new female condoms are still needed.
Contraception is one strategy that can have an eff ect on
and improve maternal health by reducing unplanned
pregnancy. In 2009, estimates showed that if all women
who wanted to avoid pregnancy used modern
contraceptives, the number of unintended pregnancies
in developing countries would fall from 75 million to
22 million annually.19 Reasons for unmet need for
contraception go beyond simple access to methods, and
are attributable to a combination of factors including
personal and cultural issues, religious restrictions,
concern about side-eff ects, health concerns, and scarcity
of knowledge. For this reason, the need to have a range of
available contraceptive methods is crucial (panel 2).
A limitation of our study was that masking of the
participants and research staff to the products was not
possible because the products all diff er in appearance.
We used allocation concealment to ensure that this
limitation was minimised. This study was based on only
self-reported measures of condom use. Reporting of
condom use by participants can be inaccurate compared
with more reliable methods such as exposure to prostate-
specifi c antigen, which provides a more objective
Figure 3: Estimates of the risk diff erence between failure of new female condoms and failure of FC2, for the
primary outcomes of TFCF and TCF, showing the results for non-inferiority
TFCF=total female condom failure. VA wow=VA worn-of-women. TCF=total clinical failure. FC2=second-generation
female condom.
0·42% (–1·42 to 2·26)
–0·42% (–1·86 to 1·32)
1·09% (–0·60 to 2·78)
0·17% (–1·48 to 1·81)
–0·25% (–1·75 to 1·26)
0·99% (–0·55 to 2·52)
TFCF–Woman’s Condom
TFCF–VA wow
TFCF–Cupid
TCF–Woman’s Condom
TCF–VA wow
TCF–Cupid
New condom better FC2 better
Risk difference (90% CI)
–4·00 –3·00 –2·00 –1·00 01·00 2·00 3·00 4·00
Panel 2: Research in context
Systematic review
We searched PubMed between Jan, 2000, and Oct, 2012, for
articles published up to November, 2012, with the search
terms “female condom functional performance”. No
randomised trials, using any of the three new types of female
condom—the Woman’s Condom, the VA worn-of-women
(wow) Condom Feminine, and Cupid—were identifi ed, aside
from one randomised, non-inferiority trial6 used to establish
the functional performance of the second-generation female
condom, FC2, compared with the predicate device (FC1).
Interpretation
This study is the fi rst randomised trial of the functionality of
female condoms in China. The new female condoms were
non-inferior to the reference condom (FC2) within a margin
of 3% failure for all the functional parameters. This is the fi rst
trial that has provided important function data for these
devices and has been used to compile evidence for
WHO–UNFPA prequalifi cation. Because of this trial, the Cupid
condom has already been approved by WHO–UNFPA and is
available for public sector procurement. Manufacturers of the
other devices are using these data in their ongoing
applications to regulatory authorities. Access to various types
of female condoms could improve choice for women in need
of a contraceptive method or of dual protection against
pregnancy and infection.
Articles
www.thelancet.com/lancetgh Vol 1 September 2013 e
152
measure of condom failure.20 However, techniques that
use prostate-specifi c antigen are more expensive and
study procedures more complex.
Contributors
CJ and MEB conceived the study and designed it with GP (statistician);
CJ, MEB, JAS, JW, YZha, RG, and JP managed the undertaking of the
trial. GP was responsible for all statistical analyses, and GP, MEB, CJ,
and JAS interpreted the data. MEB, CJ, GP, JAS drafted the article with
input editing from RG, JP, JW, and YZha. All authors read and approved
the fi nal version for submission.
Confl icts of interest
We declare that we have no confl icts of interest. Since completion of the
study CJ has done paid consultancy work for Cupid, manufacturer of the
Cupid female condom used in this study. In this capacity, she is assisting
with the development of a dossier that will be submitted to the US Food
and Drug Administration for requested approval of the Cupid device. All
data analysis and statistical interpretation and the fi nal statistical report
for this study were fi nalised before CJ accepted the consultancy.
Acknowledgments
This trial was funded by the Universal Access to Female Condoms
(UAFC) Joint Programme. We thank all research nurses, interviewers,
and data entry staff in both participating sites who collected and entered
the data, in particular Claudia Ngoloyi, Gloria Rasi, Li Yuyan, and
Zhaorui Ying; José de Sousa Ramos (JRamos Consultoria, São Paulo,
Brazil) for his valuable contribution to setting up the database and
programming the tables; and all clients who gave their time to
participate in the trial.
References
1 Gray AL, Smit JA, Manzini N, Beksinska M. Systematic review of
contraceptive medicines. “Does choice make a diff erence?” October,
2006. http://archives.who.int/eml/expcom/expcom15/applications/
sections/ContraChoiceReview.pdf (accessed July 25, 2013).
2 Jain AK. Fertility reduction and the quality of family planning
services. Stud Fam Plann 1989; 20: 1–16.
3 UN Population Fund. HIV prevention gains momentum. 2011.
http://www.unfpa.org/public/op/edit/home/publications/pid/7668
(accessed Aug 1, 2013).
4 Reproductive Health Supplies Coalition. New/underused RH
Technologies caucus. 2013. http://www.rhsupplies.org/working-
groups/caucus-on-newunderused-rh-technologies.html (accessed
Nov 29, 2012).
5 Beksinska M, Smit J, Joanis C, Potter WD. New female condoms in
the pipeline. Reprod Health Matters 2012; 20: 188–96.
6 Beksinska M, Smit J, Mabude Z, et al. Performance of the Reality
polyurethane female condom and a synthetic latex prototype:
a randomized cross-over trial among South African women.
Contraception 2006; 73: 386–93.
7 ISO. ISO 25841:2011. Female condoms—requirements and test
methods. http://www.iso.org/catalogue_detail.
htm?csnumber=43282 (accessed July 6, 2012).
8 SAS Institute. SAS/STAT Software, version 9.1. http://www.sas.
com/.
9 Wang BS, Wang XJ and Gong LK. The construction of a Williams
Design and randomization in cross-over clinical trials using SAS.
February, 2009. http://www.jstatsoft.org/v29/c01 (accessed Aug 1,
2013).
10 Beksinska M, Joanis C, Smit J, Pienaar J, Piaggio G. Using scratch
card technology for random allocation concealment. Clin Trials
2012; 10: 125–30.
11 Beksinska M, Joanis C, Manning J, et al. Standardized defi nitions of
failure modes for female condoms. Contraception 2007; 75: 251–55.
12 WHO, UN Population Fund, and FHI360. Female condom
specifi cations: generic specifi cation, prequalifi cation and guidelines
for procurement. 2012. http://www.unfpa.org/webdav/site/global/
shared/procurement/08_QA/Female%20condom%20Generic%20
Specifi cation%20EN.pdf (accessed Nov 30, 2012).
13 Taylor D. Issues in the design, analysis and interpretation of
condom functionality studies. Contraception 2009; 80: 237–44.
14 Westfall PH, Tobias RD, Rom D, Wolfi nger RD, Hochberg Y.
Multiple comparisons and multiple tests using SAS. Cary, NC,
USA: SAS Institute, 1999.
15 Schwartz JL, Barnhart K, Creinin MD, et al. Comparative crossover
study of the PATH Woman’s Condom and the FC Female Condom.
Contraception 2008; 78: 465–73.
16 Smita J, Neelam J, Rochelle DY, et al. Comparative acceptability
study of the Reality female condom and version 4 of modifi ed
Reddy female condom in India. Contraception 2005; 72: 366–71.
17 Joanis C, Beksinska M, Hart C, Tweedy K, Linda J, Smit J. Three
new female condoms: which do South African women prefer?
Contraception 2011; 83: 248–54.
18 UN Population Fund. Female condoms. 2013. http://www.unfpa.
org/public/home/procurement/pid/10867 (accessed Nov 30, 2013).
19 Singh S, Darroch JE, Ashford LS, Vlassof M. Adding it up: the costs
and benefi ts of investing in family planning and maternal and
newborn health. New York: Guttmacher Institute and UNFPA,
2009.
20 Mauck CK, Weaver MA, Schwartz JL, Walsh T, Joanis C. Critical
next steps for female condom research-report from a workshop.
Contraception 2009; 70: 339–44.