Clinical Trial Details
— Status: Not yet recruiting
Administrative data
| NCT number |
NCT02280681 |
| Other study ID # |
S57341 |
| Secondary ID |
|
| Status |
Not yet recruiting |
| Phase |
N/A
|
| First received |
October 29, 2014 |
| Last updated |
October 29, 2014 |
| Start date |
October 2014 |
| Est. completion date |
October 2015 |
Study information
| Verified date |
October 2014 |
| Source |
Universitaire Ziekenhuizen Leuven |
| Contact |
n/a |
| Is FDA regulated |
No |
| Health authority |
Belgium: Institutional Review Board |
| Study type |
Observational [Patient Registry]
|
Clinical Trial Summary
To gain insight by means of a discrete choice experiment questionnaire into characteristics
of medically assisted reproduction (MAR) with artificial gametes (AGs) valued by couples
confronted with ovarian failure and their physicians and the trade-offs they make between
these treatment characteristics.
Description:
Introduction 3-17% of heterosexual couples trying to conceive will be faced with fertility
problems (Boivin et al, 2007). Of these couples 50% will be confronted with a female factor,
which includes amongst others ovarian failure (WHO, 2000). Ovarian failure can occur
pre-maturely (i.e. before the age of 40 years old, in 1% of women; Coulam, 1982; Coulam et
al., 1986) or at a physiologically natural age of after 40 years but before reaching the
limit of 44 years old at which modern women that still want children can rely on reimbursed
medical assisted reproduction (MAR) in European countries, such as Belgium and the
Netherlands.
Although some couples with ovarian failure will have attempted MAR without success, they
will be left with IVF with donor ooyctes or embryo reception as only treatment option. These
treatment options will not result in genetic parenthood for one or both of the partners.
Researchers are currently developing new forms of MAR that do not require patients to have
functional gametes but rely on artificial gametes (AGs) developed by means of different
techniques (Hendriks et al, submitted).
Reproductive medicine experts are starting to reflect on the need for minimal thresholds of
effectiveness and safety to introduce these and other new forms of MAR into clinical
practice (Harper et al, 2012; Schatten, 2002; Winston & Hardy, 2002). Research, however,
showed that specific subgroups of infertile patients value other treatment characteristics
besides safety and effectiveness, such as, for example, naturalness (Hendriks et al, 2014;
De Groot et al, submitted; Lankreijer et al, in preparation). The perspectives of couples
with ovarian failure on MAR with AGs has yet to be explored. Clinicians' perspectives on MAR
with AGs - valuable because of their expertise and their direct involvement in possible
future decision-making - has not been studied. A previous study showed that clinicians
differ from patients in their evaluation of the importance of characteristics of fertility
clinics (van Empel et al, 2011) and would be interesting to compare the perspective of
patients and clinicians on MAR with AGs
Objective The aim is to gain insight into characteristics of MAR with AGs valued by couples
confronted with ovarian failure and their physicians and the trade-offs they make between
these treatment characteristics.
Methodology The treatment characteristics valued by patients are identified by literature
review. Treatment characteristics valued by clinicians will be identified by preparatory
qualitative research. The (relative) importance of these treatment characteristics will be
examined with a discrete choice experiments analysed with multinomial logistic regression.
The study population will be heterosexual couples confronted with infertility due to ovarian
failure between the age of 18 and 44 years old treated in the K.U.Leuven and clinicians who
indicated an interest in reproductive medicine in their membership of the Belgian Society
for Reproductive Medicine (BSRM).
Both the preparatory qualitative study and discrete choice experiment questionnaire will be
preceded by giving written information supported by figures on MAR with AGs (i.e. referred
to in the rest of the protocol as 'preceding information on MAR with AGs') which will be
pilot tested. This preceding information on MAR with AGs will define MAR with AG and will be
exemplified by explaining the possible techniques to create AG and conceive a child
(appendix 3).
The preparatory qualitative study Physicians treating heterosexual couples confronted with
infertility due to ovarian failure will be selected from the professional network of the
authors, resulting in a convenience sample. Clinicians will be interviewed at their clinic.
All interviews will be conducted by a female interviewer (SH) who is familiar with all
aspects of the treatment options but not involved in patient care. The interview will start
with the preceding information on MAR with AGs'. The in-depth interviews will take 30-60 min
and will be guided by a semi-structured interview guide and a topic list developed based on
the literature on the biology of AG (Hendriks et al, submitted), on the consequences of the
clinical application of AGs (Hendriks et al, submitted), and previous qualitative and
quantitative research on the importance of treatment characteristics for patients (Hendriks
et al., 2013; Lankrijker et al, in preperation; Hendriks et al, in preperation).
Interviews will be recorded, transcribed verbatim and analyzed with content analysis by two
researchers independently.
As appropriate for qualitative research, the number of interviews per group (i.e. couples
and physicians) will be deļ¬ned by the achievement of data saturation per group (i.e. when a
sense of closure is achieved because no new themes answering the research question emerged
in three successive interviews; Francis et al., 2010; Pope et al., 2000; Tong et al., 2007).
Part 2: Discrete choice experiment (DCE) A discrete choice experiment questionnaire (DCE-Q)
will be designed based on the results of the preparatory qualitative study, incorporating
all identified valued treatment characteristics. A DCE aims to identify attributes (in this
case treatment characteristics) significantly affecting choice and trade-offs of all
attributes with one attribute generally considered the most important (in this case, either
pregnancy rate or safety (Dancet et al., 2013), depending on the insights gathered by the
qualitative study.
A discrete choice experiment (DCE) is the best available method to investigate the relative
importance of characteristics of multi-dimensional concepts (e.g. quality of care,
characteristics of fertility treatments) and allows the calculation of respondents
trade-offs between characteristics (Ryan and Farrar, 2000; Bryan and Parry, 2001; Sculpher
et al., 2004; Telser and Zweifel, 2007; Albada and Triemstra, 2009). In a DCE, respondents
have to choose repeatedly between choice sets, each described by characteristics with
varying levels.
In this DCE-Q respondents will be asked to choose between choice sets developed to represent
the hypothetical but realistic fertility treatments based on AGs. The fertility treatments
will be described by a number attributes that can each vary across a number of levels (e.g.
pregnancy rate varying between 20% and 35%). The attributes will be the valued treatment
characteristics identified by the preparatory qualitative study in couples and physicians.
The levels over which the attributes can vary in the description of the hypothetical
fertility treatments will reflect a range realistic for current MAR and future MAR with AGs
(Hendriks et al, submitted).
A mathematical model (i.e. a fractional factorial design, which is both orthogonal and
balanced, created according to published principles; Street and Burgess, 2007) will define
how many of all combinations of attributes and levels will be questioned. To realize 'choice
sets' (i.e. define to which hypothetical fertility treatment to compare each selected
combination) with maximum differentiation between the compared hypothetical fertility
treatments, a fold-over technique will be applied to result in 'mirror' scenarios. We will
make sure that the efficiency of the final design, calculated according to Street and
Burgess (2010b), will ensure 100% informative value.
We will make sure that maximal 17 choice sets (Sculpher et al., 2004; Christofides et al.,
2006; Bech et al., 2010) are presented in each questionnaire, if necessary choice sets will
be divided over different questionnaires.
Four DCE-Q questionnaire addenda will be developed in addition to the preceding information
on MAR with AGs. First, to clarify the DCE method, the DCE-Q will be preceded by detailed
instructions. Couples and physicians will be explained that from each choice set they need
to select the hypothetical fertility treatment which they prefer. An example choice set,
where respondents have to select their preferred airline ticket for a holiday will be
exemplifying. Second, a form to collect demographic characteristics and professional
characteristics or medical characteristics, for physicians or couples, respectively will be
developed. Third, to test the face-validity of the modelled treatment characteristics,
respondents will be asked for any additional treatment characteristics on which they would
have liked to be informed to choose between fertility treatments. Fourth and only applicable
to couples, not to physicians, couples' actual past choice behaviour will be examined by
asking them which fertility treatments they have opted for, and for which reasons they
decided for or against the offered treatment options.
The DCE-Q and its addenda will be distributed by post. With regards to couples, all couples
of which the woman was diagnosed with ovarian failure between October 2009 and October 2014
at the K.U.Leuven will be addressed. With regards to physicians, all physicians who
indicated an interest in reproductive medicine in their membership of the Belgian Society
for Reproductive Medicine (BSRM) will be addressed. Couples will be requested to fill out
the questionnaire addenda together and to each fill out the DCE-Q separately. Physicians
will be requested to fill out the questionnaire individually. The questionnaire will be
coded, so that respondents who do not reply either by filling out the questionnaire or the
statement of non-participation, can be send a maximum of two reminders. Prior to filling out
the questionnaire respondents will be informed on the purpose of the study and provided with
contact information. This way participation results in giving informed consent.
Data of incoming questionnaires will be entered in the Statistical Package for the Social
Sciences (SPSS; version 21.0 for Mac, Chicago, IL, USA). For analysis of the DCE-Q the data
will be exported to the STATA Software (version 11.1.). Data from couples and physicians
will be analysed separately. P-values of <0.05 will be considered statistically significant.
The data gathered from the DCE-Q addenda will be described with descriptive analyses (e.g.
demographic characteristics) or content analysis (e.g. question on missed treatment
characteristics).
The data gathered with the DCE-Q will be analyzed with multinomial logistic regression. For
each treatment characteristic, analyses will examine whether it contributes significantly to
respondents preference. Trade-offs that respondents are willing to make between pregnancy
rate and any other attribute (i.e. the willingness to trade-off pregnancy rate; WTpreg) will
be calculated as pregnancy rate is considered the 'gold standard' to benchmark the other
fertility treatment characteristics (van Empel et al, 2011). Confidence intervals (95%) for
the WTpreg were computed using a non-parametric bootstrap approach (2000 replications).
To investigate heterogeneity in preferences between subgroups of couples or physicians, we
will included interaction terms in the model (i.e. confounders' tests). Furthermore, the
results of the analysis of the data of the couples will be compared with the results of the
analysis of the data of the physicians.
