\[Intervention\] The intervention included in this study is called "Parents Round," which aims to integrate parents into the discussion of medical rounds. Parents Round is organized by healthcare staff following the protocol. The protocol details are described below.
Essential requirements
* Limited number of participants: infant, parents, other important family members if present, up to 2 doctors, and up to 4 healthcare staff including doctors
* At the bedside of the infant
* All participants, excluding the infant, sit on the chair
* Privacy consideration: ask the other parents in the same room to go out or wear headphone
* Ask parents about their infant's condition and their opinion before the healthcare staff speaks
* Mention the respiration, feeding, skin-to-skin contact, and involvement in care-taking and try to make a decision collaboratively with parents about the short- or long-term plan in each item
Recommended requirements
* Quiet environment: turn off or reduce the volume of the monitor alarm
* Effective communication to promote parents' participation and involvement: parents speaking first, empathetic attitude, not interrupting while parents speak, and not frequently using the technical term, for example.
* Encourage psychologists to participate
* Nurses complement parents' role and may lead some parts of the discussion: e.g. discharge preparation
\[Sample size\] To detect the mean difference of 0.5 in the FCC-Q average scores (standard deviation, 0.77), with power 0.8 and alfa 0.05 in the cross-over trial, at least 40 families are necessary. We estimate the possibility of dropout as 10% and, finally, the target sample size is decided to be 46 families (or 92 parents) in total.
\[Randomization\] After families consent to participate, they are randomly assigned to Group A (intervention first) or Group B (no intervention first). The randomization sequence will be created using Excel 2019 (Microsoft, Redmond, WA, USA) and will be stratified based on the gestational age at birth (preterm births \[born before 37 weeks of gestation\] or full-term birth \[born 37 weeks of gestation or later\]) with a 1:1 allocation using random block sizes of 2 and 4.
The sequence allocation was conducted by a researcher in the research group who will not be involved in family enrollment. The allocated number was concealed from the researcher enrolling and assessing participating families in sequentially numbered, opaque, sealed, and stapled envelopes. This sealing will be conducted by another research member who will not be involved in the enrollment process. Black thick paper will be used inside the envelope to prevent the allocated number from being visible from the outside. The neonatologists in the study group, except for one person working outside the unit, will be responsible for the family enrollment. One of the envelopes will be opened soon after the family's consent by the person who is responsible for the enrollment according to the number written outside the envelope. To prevent subversion of the allocation sequence, the infant's name, identification number, and date of birth will be written on the envelope before opening it and on the paper on which the allocated number is printed soon after opening it. Any blinding is not possible in this study design after allocating the families.
\[Statistics\] The differences in each continuous outcome measures (FCC-Q, Short-STAI, and Satisfaction questionnaire) between the intervention and the control will be analyzed using the linear mixed model: including intervention/control, Period I/II, an interaction effect of the intervention and the period, and the infant's days of life at each intervention/control phase as covariates and parent as a random effect.
In addition, the proportion of parents with possible clinical state or trait anxiety will be also compared using the cutoff score of \> 9.5 in state and \> 13.5 in trait anxiety 8. The generalized mixed model will be used to compare the two groups: including intervention/control, Period I/II, an interaction effect of the intervention and the period, and the infant's days of life at the assessment as covariates and parent as a random effect.
The differences in the outcome measures for healthcare staff will be analyzed using the Wilcoxon rank sum test. Additionally, the adjusted outcomes will be estimated using the linear mixed model. The following variables will be included in the model: the outcome measure, occupation, and years of experience as fixed effects, and each healthcare staff as a mixed effect.
All analyses will be conducted for both parents together and individually. Separate analyses will be conducted including and excluding those who never participate in the Family Round during the study period. The differences in each question in FCC-Q or the Satisfaction questionnaire will be also compared. The state and trait anxiety measured by Short-STAI will be analyzed separately. The proportion of each rating in the Bliss Baby Charter audit tool will be compared between the groups using the Wilcoxon rank sum test. R and RStudio will be used to analyze the data. The analyses performed are two-tailed, and P \<.05 is considered statistically significant.
\[Data management\] Running numbers will be assigned to every family after the consent. All data will be handled pseudonymously after converting the data into electronic data. Paper documents including parents' answers will be stored in a safe archive of the unit. Electronic data will be stored in a password-protected computer. All paper and electronic data will be destroyed 10 years after the final publication.
The principal investigator will be responsible for managing access to the data. The other research members will also have the right to access the digital data. The data is co-owned by the members of the research team. We do not have the plan to make the data open in the future. However, it can be shared with another personnel or research team for another scientific purpose with the permission of the principal investigator.
