This study evaluates possible effects of two different NICU designs; by comparing traditional and Family Centered Care in terms of infant nutrition, health and growth, and coping by family.

Two Norwegian neonatal intensive care units (NICU) participate in this study; Haukeland University Hospital (HUH) and Vestre Viken Hospital Trust (VVHT).

The NICU at HUH was built around 1980 as a traditional unit. It is crowded, has no single rooms or access to rooming in with the parents. The NICU at VVHT opened in April 2012 and it is unique in its kind both in a Norwegian and European context. A shift from traditional NICU to a FCC unit involves both physical structural changes and involves the entire interdisciplinary team around the individual families. The main difference is firstly the involvement of parents and their right to be present together with infant and secondly the establishment of parents as the responsible caregivers for the infant, even when the infant is hospitalized. This comprises so-called "minimal handling", the number of medical procedures are reduced to the very strictest necessities, and restructuring the common nursing interventions. The infant is given Kangaroo Mother Care (KMC) skin-to-skin rather than being in an incubator.

The first question raised and discussed is whether the FCC - with parents present in Single Room Care 24 hours and performing KMC - will affect the preterm infant's growth and attainment of breastfeeding. The second question raised and discussed is whether the surroundings in FCC-unit - with parents present - will affect parents' reports on parental health in relation to psychological stress, anxiety and depression, also in a longitudinal perspective. The purpose of the study is to evaluate possible effects of these two different NICU designs; the traditional open bay unit with limited space and no possibilities for parents to continuously stay with their infant, and the novel model physically and culturally optimized for parent-infant interaction 24 hours a day 7 days a week from birth to discharge. Although an intriguing novel asset to neonatal intensive care, the effects of FCC on the child's and family's somatic and mental health, on NICU staff and cost of care, have not been scientifically scrutinized. There is so far lack of valid knowledge about effect of FCC in a single room unit with parents present to the extent we can offer in Nordic countries.

A randomized controlled study (RCT) in a single unit would be ethically and methodologically impossible to conduct, e.g. since it will be impossible to avoid "contamination" between groups, and a Cochrane review argued that comparison between different units would be the most valid method. This recommended approach will be followed in the present study, and the challenge is to minimize factors that may act as confounders. To minimize confounding the following details on design will be adhered to:

• Recruitment of comparable infant - parent dyads
• Strict definitions of independent and outcome variables, e.g.nutrition
• Prospective registration: Parental presence and duration of KMC, infant nutrition, growth, investigations and occurrence of morbidity will be prospectively and similarly registered during the hospital stay under supervision by a designated research nurse.

All data on growth will be transferred to z-scores based on standardized growth charts for Norwegian infants for pre and post term growth. Relevant statistical methods for comparisons will be applied and statistical methods will be chosen in cooperation with a biostatistician.

The main outcome is weight at 34 weeks and 40 weeks post menstrual age (PMA), basically calculated as z-score increments between 34 and 40 weeks. Judged from a pilot study it is actually sufficient with only ten children in each group to detect a difference in weight of 100 gram 34 Gestational age (GA) weeks with a power of 80% at a significance level of less than 0.05. However, due to the other variables and analyzes (among other mixed model) it is appropriate with about 100 infants, 50 in each group. An important secondary outcome is length of stay in terms of PMA at discharge. From data at HUH a difference in PMA at discharge of 5 days (which is considered clinically relevant) will require 42 infants in each leg (calculated from a standard deviation of 13 days, power of 80% and p < 0.05). From these data one year recruitment will be sufficient in that, in our experience, more than 90% will agree to participate. We will consider expanding inclusion to 1 ½ years (75 infants in each branch) to allow for power to assess effects of other explanatory variables.

Parents who agree to participate are registered in a database containing name, personal identification number, a code number which corresponds to other files, information relating to the date the questionnaires are sent or received, and whether or not reminders have been sent. No other information is stored in this database. The data will be stored at the respective Research servers at VVHT and HUH according to the requirements of the hospitals and the Regional Committees for Medical and Health Research Ethics (de-identified study file and a file with a key to the study file). Transfer of data from HUH to VVHF will be performed as encrypted files according to requirements. All participants will be asked to sign an informed consent form at the enrolment in the study.

Inclusion Criteria:

• All eligible infants (gestational age 28-32 weeks) at VVHT and HUH.

Exclusion Criteria:

• If infant suffers from congenital malformations or chronic disease, i.e intraventricular hemorrhage grade III or IV, undergo surgical treatment for Necrotizing enterocolitis or if the birth weight is less than 800 grams.
• Furthermore, we will exclude families from participation if the mother or father suffers from serious mental illness
• If they don't understand Norwegian (oral and written)
• If the mother was taking illicit drugs or was treated with methadone or Subutex during pregnancy or in cases of severe social challenges in the family, e.i. infant being submitted to Child Protective Services immediately after birth.

All infants (gestational age 28-32 weeks) and their parents hopitalized at VVHT and HUH.The recruited parent-infant dyads will thus be population based, i.e. representing the county of Buskerud and Hordaland. Patients will be recruited consecutively at admission to the NICU and included after written informed consent by both parents.