Page 160 - Strategies for non-invasive managementof high-grade cervical intraepithelial neoplasia - prognostic biomarkers and immunotherapy Margot Maria Koeneman

Chapter 8
Table 2. Baseline characteristics of the respondents
Age in years (mean, range) Degree of education (absolute)
38.5, 21–70
62 35 3
39 23 35 3
27, 18–36
34 53 9 4
77 21 2
18 81 1
Characteristic
N = 100
- - -
Parity - - - -
Age in
Secondary education or intermediate vocational education Advanced vocational school or university
Unknown
Nulliparous Primiparous Multiparous Unknown
years at first birth (mean, range)
Pregnancy wish in the future - Yes
- No
- Unsure
- Unknown
Previous colposcopy - Yes
- No
- Unknown
Previous LLETZ treatment or conization - Yes
- No
- Unknown
LLETZ = Large loop excision of the transformation zone.
DCE results: Main-effect model
The results of the main-effect multinomial logit regression model are shown in Table 3. Women derived a positive utility from choosing the standard LLETZ treatment, as shown by the significant positive coefficient of 4.25. All attributes representing imiquimod treatment were considered important, and women significantly preferred a higher treatment success rate and a lower risk of premature birth, infertility, and side effects. A higher utility score for imiquimod treatment compared to LLETZ was only reached when the success rate is 80% with a low risk of subfertility and premature birth in subsequent pregnancies and low levels of side effects (utility score 4.4: V(imiquimod) = (0.08* 80% treatment success) + (-0.13* 6% premature birth) + (-0.11* 8% infertility) + (-0.03* 10% vaginal pain) + (-0.013* 10% vaginal discharge) + (-0.029* 10%flu)). In the case of a low risk of subfertility and premature birth (as is expected), but a high risk of side effects, treatment efficacy should reach 95% in order to provide a higher utility score compared to LLETZ treatment.
DCE results: Interaction model
The results of the interaction model (2) are presented in Table 3. All three interactions were statistically significant, indicating that preferences for the attributes “risk of subfertility”, “risk
158