Setting and Study Population
Between August 15th, 2016 and September 30th, 2019, we conducted a randomized prospective open-label study of women aged 18 to 42 years who were interested in using ENG-IMPLANT and were willing to have 90 days of pre-treatment with DSG. The study was conducted in the family planning clinic of the Geneva University Hospitals and women’s policlinic of the University Hospital Basel. The exclusion criteria were pregnancy, lactation, vaginal bleeding of unknown origin, wish to become pregnant, weight > 80 kg, history of venous thromboembolism, hypertension, diabetes or other metabolic diseases, coagulation disorders, severe hepatic disorder, history of gynaecological cancer, known hypersensitivity to study drugs and current treatment with enzyme inducing drugs. The study was approved by the ethics committees of both hospitals (CCER 16–972). All patients signed an informed consent prior to recruitment.
Study procedure
Once consent was obtained, women were randomized either to the ENG-IMPLANT group or the DSG + ENG-IMPLANT group. In the ENG-IMPLANT group, the implant was inserted immediately. In the DSG + ENG-IMPLANT group, patients were given a 3 months’ supply of DSG to be started immediately after which insertion of ENG-IMPLANT was proposed. In both groups, patients were instructed to complete a bleeding calendar and a satisfaction questionnaire. A 3-month visit was pre-programmed for all participants, during which the bleeding calendar and the questionnaire were collected. Participants in the DSG + ENG-IMPLANT group who completed the 3 months on DSG pill had their ENG-IMPLANT inserted during this visit. All patients were seen or called over the phone after 12 months post ENG-IMPLANT insertion in both groups. Women who had their implant removed between 3 to 12 months after the insertion completed the satisfaction questionnaire at the time of removal.
Measurement of outcomes and other variables of interest
We collected socio-demographic characteristics. Data collectors recorded the date of implant insertion or the date of DSG initiation. The main outcome variable for this study was method discontinuation. Secondary outcomes were the side effects, tolerance and satisfaction 3 months after inclusion and 12 months after ENG-IMPLANT insertion.
We used a satisfaction questionnaire and a bleeding calendar to assess bleeding patterns throughout the first 3 months of use in both groups. Overall, irregular vaginal bleeding was defined as any deviation from their habitual menstrual patterns of bleeding.
The satisfaction questionnaire evaluated side effects on skin, mood, sex-drive, abdomino-pelvic or breast pain, headaches, changes to vaginal discharge, and irregular vaginal bleeding according to a graded 5 points Likert score, from never experiencing (1 point) to all the time experiencing (5 points). Two separate questions evaluated tolerance and satisfaction and were graded on a scale from 0 to 10 (0 not tolerating or unsatisfied with the method and 10 being totally satisfied and perfectly tolerating). Weight gain as cause for removal was noted when relevant.
Randomization and statistical analysis: Statistical analysis was performed using the Stata program version 13 (StataCorp LP: College Station, TX, USA); the significance level for all tests was p < 0.05. To detect a mean difference of 0.5 standard deviation on the numerical tolerance scale, with a power of 0.85 and type 1 error rate of 0.05, we needed 2 groups of 74 participants. The randomization plan was so as to have 33% more participants in the DSG + ENG-IMPLANT group, with a 1:3 ratio, because of 35% drop out in the ENG-IMPLANT group. This would have provided a power of 0.8 to detect a difference in proportions with implant removal of 0.35 versus 0.15 to compensate for possible loss to follow-up. We planned to recruit 80 participants in the ENG-IMPLANT only group, and 120 in the DSG + ENG-IMPLANT group. Randomization was conducted on www.randomization.com using randomly permuted blocs of 14, 21 and 28 patients. The participants' study allocation was included in opaque, sealed envelopes, prepared by the Clinical Research Platform (PGO) in the University Hospital of Geneva.
For continuous variables, means and standard deviation (SD) were calculated; for categorical data, proportions were calculated. We used the chi-square test and Fisher´s exact test when appropriate, as well as the student’s t-test and Mann–Whitney test. Descriptive statistics were used to analyse the baseline characteristics of the study population.