Page 159 - Strategies for non-invasive managementof high-grade cervical intraepithelial neoplasia - prognostic biomarkers and immunotherapy Margot Maria Koeneman
P. 159

Patient preferences in treatment of high-grade CIN
LLETZ was based on the existing local information brochure, the description of imiquimod treatment was based on the treatment protocol according to the RCT by Grimm et al.[5] All questionnaires were completed before the planned colposcopy.
Data analysis
The respondents’ choices were analyzed with a multinomial logit regression model in Nlogit version 5. The following model was estimated:
V(imiquimod) = β0 + (β1* treatment success)
+ (β2*premature birth) + *β3* infertility)
+ (β4* pain) + (β5*discharge) + (β6*flu) V(LLETZ treatment) = β7 (fixed constant )
where V represents the relative utility score patients derive from a treatment with imiquimod or
LLETZ treatment; β0 is a constant that, if positive, reflects a preference for the label imiquimod; β1
to β6 are the coefficients reflecting the impact of a particular attribute level on the utility score;
β7 is the coefficient (i.e., alternative specific constant) for the LLETZ treatment with fixed levels.
The sign of a coefficient shows whether an attribute has a positive or negative impact on utility.
A relative utility score can be calculated by filling in the regression equation with estimated coefficients from the multinomial logit model and levels of the attributes. A higher utility score 8 means that patients prefer one treatment over another. To examine whether preferences for the
attributes “risk of subfertility”, “risk of premature birth in subsequent pregnancy,” and the fixed alternative LLETZ-treatment are different depending on a desire for a future pregnancy, we included these interaction terms in the model. In addition, a subgroup analysis was performed based on women with or without a future pregnancy wish.
Ethics approval
The study was evaluated and approved of by the Medical Ethics Committee of the Maastricht University Medical Center (MUMC/METC 13-4-116, 6 January 2014). All study procedures were conducted according to the Declaration of Helsinki, 7th revision, 2013.
Results
Respondents and general results
Respondents were recruited in five Dutch hospitals between March 2014 and December 2015. A total of 177 women were consecutively approached for participation in the study, of which 100 gave informed consent and completed the questionnaire (response rate 56.5%). Baseline characteristics of the respondents can be found in Table 2. Three questionnaires had missing data (six questions in total). Of the 100 women, 82 evaluated the questions as clear or very clear. The internal consistency of the respondents was good, as reflected by a kappa value of 0.79 (p-value < 0.01).
 157















































































   157   158   159   160   161