Hans van den Heuvel

(maternal or fetal or neonatal), a 5-minute Apgar score below 7 and/or an arterial pH below 7,05, maternal morbidity (one or more of the following: eclampsia, HELLP syndrome, thromboembolic events), NICU admission of the new-born and caesarean section rate. The components of the composite outcome are both chosen for either (or both) the possibility to be affected by the new intervention as well as the severity as a stand-alone adverse outcome. All components will be reported separately as a secondary outcome for interpretation of study results. Secondary outcome will consist of patient satisfaction, quality of life and cost effectiveness. The satisfaction, experience and quality of life of every participating pregnant woman will be surveyed with help of the EuroQol 5D (EQ-5D), State Trait Anxiety Inventory (STAI) and Edinburgh Postnatal Depression Score (EPDS) questionnaires. 14,15,16 Surveys are sent by e-mail at study start, and 1, 3, 5 weeks after randomisation and 4 weeks after delivery. With the help of focus group discussion (see under Patient involvement), we created a questionnaire which will be filled out 4 weeks after delivery. The cost effectiveness and budget impact analyses (CEA and BIA) will be assessed from different perspectives, i.e. hospitals, health insurance companies and from the societal perspective. The budget impact analysis will follow ISPOR guidelines for budget impact analyses to calculate the differences in budgetary impact of telemonitoring and hospital admittance in high-risk pregnancies. For the CEA and the BIA, we will record duration of telemonitoring and duration of admittance (number of days), number of consultations and health care provider involved, number and length of CTG registration, number of maternal blood analyses and ultrasound assessments, emergency transport to the hospital and emergency caesarean sections. Besides this maternal use of health services, all health service use of the newborn during the follow-up period (until discharge to home) will be recorded. Sample size Before the start of the trial, we formed an expert panel, consisting of gynaecologists, and paediatricians, methodologists, and statisticians to conceive the design, content, and execution of the trial. The sample size calculation is based on the assumption that the composite of adverse perinatal outcome will be equal in the telemonitoring and the hospital admittance patient groups: a non-inferiority trial. To estimate this risk for each individual component of adverse perinatal outcome in our inclusion criteria, we made use of the results of three large Dutch randomised controlled trials for patients with PPROM, FGR and preeclampsia. 17,18,19 No data on perinatal outcome of telemonitoring in high risk pregnancy are available to use in our sample size calculation. The incidence of this composite primary outcome in the high-risk pregnancy group is assumed to be 20% in either group. The panel made a reasoned choice about the acceptable difference in adverse perinatal outcome and feasibility of the trial, since this is the first ongoing trial of telemonitoring in complicated pregnancies. As a result, the non-inferiority margin (Δ) was defined as a 10% absolute CHAPTER 10 170