Hans van den Heuvel

Patient satisfaction and health related quality of life will be analysed with a general linear model for continuous outcomes. Comparison of questionnaires will be made for each time point, with the survey at 4 weeks post delivery being the most important. Assumptions for general linear model (i.e. normality, homoscedasticity) will be checked with residual analyses. In case of heteroscedasticity, the analyses will be repeated with robust (Hubert- White) estimators for standard errors. If distributional assumptions are violated, first a log transformation of the outcome will be analysed. If this transformation does not result in a valid regression analysis, intervention effects will be evaluated with a Mann-Whitney test without any corrections. Time to delivery with account for different durations of gestation at study entry, will be evaluated with Cox regression with control of the stratification factors and parity as a predifned covariate. For the cost-effectiveness analysis, all health care resources use will be transformed into cost estimates, by multiplying number of units of health care use, i.e. number of days in hospital, number of laboratory tests and other diagnostic tests with standard unit prices as provided by the Dutch guideline for costing research in health economic evaluation studies (National Health Care Institute, Zorginstituut Nederland, 2016). For medical costs, the process of care is divided into three cost stages (antenatal stage, delivery/childbirth, postnatal stage). Cost differences between the two treatment arms will be related to effect differences (primary outcome) between the treatment arms (if any). If non-inferiority of telemonitoring is confirmed, cost differences between the two treatment arms will be analysed (cost-minimization analysis). The cost effectiveness analysis will be performed from both the healthcare perspective and the societal perspective. Study monitoring and safety Tomonitor the conduct of the trial and safeguard the interest of participants, an independent Data Safety Monitoring Board (DSMB) will be established, including a professor of biostatistics, an obstetrician and a neonatologist. A study monitor will periodically visit participating centres, assessing quality of data and auditing trial conduct. All serious adverse events, reported by either participant or local clinician, will be recorded, and reported to the accredited ethics committee and the DSMB following international GCP guidelines. Trial data will be analysed and stored in the UMC Utrecht (study sponsor). No formal interim analysis of efficacy outcome is planned. Ethics and dissemination This trial has been approved by the Medical Research Ethics Committee (MREC) of the UMC Utrecht. Trial reference number: 16-516. The MREC of the UMC Utrecht is accredited by the Central Committee on Research Involving Human Subjects (CCMO) since November 1999. Approval by the boards of management of Amsterdam University Medical Center, Diakonessenhuis Utrecht, OLVG Amsterdam, Martini Ziekenhuis Groningen and St. Antonius CHAPTER 10 172