This qualitative feasibility study investigates a digital health app designed for patients who are frequently admitted to the psychiatric acute ward, to facilitate crisis support outside of office hours to avoid unnecessary readmissions, as well as self-harm. As this is a novel app to be implemented in an already existing service, key uncertainties regarding the intervention content and its delivery needs to be addressed before potentially conducting a full-scale trial, or, alternatively, to inform further intervention refinement. Such uncertainties are appropriate to explore in a feasibility study according to the Medical Research Council (MRC) framework for the development and evaluation of complex interventions.

There is a need to investigate whether the app is acceptable and useful to the patients using it, and to the healthcare providers who respond to it. There is also a need to know whether the healthcare providers responding to the app find the task manageable. The context in which the staff are to deliver the intervention, i.e. the in-bed ward, may be subject to different barriers, such as a staff shortage due to sick leave and heavy workload in the ward. Such barriers are essential to identify ahead of possible future trials in other contexts.

The present study is a part of the larger Revolving Door Project, a collaborative service innovation project between a university hospital and two municipalities in Norway. The revolving door project aims to create a more sustainable service to prevent unnecessary hospital readmissions, reducing the strain repeated rehospitalisations can produce and also reducing financial costs. By collecting valuable insights from key stakeholders, critical points in the healthcare services have been identified and outlined in the "emergency loop". This loop illustrates how the current service solution can worsen the revolving door problem by failing to offer timely and adequate help and ultimately risking an escalation of the patient's difficulties. To address these issues, a new healthcare service has been developed to meet the needs of the stakeholders. The service is aimed at individuals who are repeatedly readmitted and discharged from the psychiatric acute ward, who have emotionally unstable traits (not necessarily diagnosis specific), where previously attempted measures have proven insufficient and where there is an insecurity in the help apparatus as to how to help the patient.

The Muli-service The muli-app is a part of a new service, called The Muli-service. This service comprises a collaborative multidisciplinary team (Muli-team), with representatives from municipal mental health services, local psychiatric hospital, psychiatric acute ward, and police. This team coordinates admissions and collaborates with individuals to improve communication and cooperation between services and patients. As such, the Muli-service isn't a new treatment but an intervention meant to complement and enhance existing treatment and mental health provision.

The collaborative part of the service is already established. In addition, the revolving door project has developed the Muli-app, which is the focus of the present feasibility study. A Person-Based Approach has been used to guide the development the app, where stakeholders have been consulted iteratively. The app was created based on stakeholders reported needs, to provide them with an alternative means to seek help outside of office hours, which is a typical time for emergency hospital admissions. The goal is to provide timely and adequate help and thus prevent further escalation of the situation. This aligns with recommendations in the European Psychiatric Association (EPA) guidance on the quality of mental health services.

The Muli-app At the point of download, the app consists of a few predefined sets of emotion regulation exercises and contact information for emergency services. Tailored content (i.e crisis plan and crisis measures) will be written in collaboration between patient and therapist/Muli-contact. Possible crisis measures for each individual patient will be discussed and agreed upon in the Muli-team in advance, ensuring that these measures are in fact feasible. The patient will be presented with the options, and from this the patient and therapist agree upon the most fitting crisis measures to be put into the app. This is communicated back to the Muli-team.

Patients admitted into the Muli-service will be informed of the app through their Muli-contact in the collaborative team and receive an invitation to participate in this study. Study participants who consent to participation will be set up with a user profile in the Muli app and receive a text message from the developers informing them of how to access the app on their own smartphones. The Muli-app will be available to download in app-stores but won't work unless an existing profile is connected to the patient. study participants will receive training from their Muli-contact on how to use the app.

One in-patient ward at the local psychiatric hospital responds when patients activate the contact-based measures through the app outside of office hours. The ward will be provided smart-phones connected to the app-software, notifying them when a patient makes contact through the app. For every evening and night shift (and all shifts during weekends), one healthcare worker will be assigned to carry the Muli-phone and reply when a patient activates the contact-based functions.

The staff who respond at the ward all have bachelor's degrees as registered nurses or learning disability nurses, while some are specialized mental health nurses. Staff have received training in principles of Mentalization-Based Therapy by a licensed psychologist, which will be applied when study participants make contact. Staff will also receive training on the app technology and protocol by a Muli representative.

As this intervention is meant to complement existing treatment, no limitations are made regarding the type of treatment study participants receive parallel to the feasibility testing of the app.

The present feasibility study addresses several aims. To assess acceptability, perceived usefulness, implementation feasibility and fidelity, narrative data will be collected through qualitative interviews with participants. App usage data and information regarding the number of hospital admissions and emergency services during the trial phase will be collected at the end of trial, to identify possible change and enable further contextualization of the narrative data. The findings may bring forth hidden aspects of how the intervention works in practice.

Data collection methods Patient perspectives will be monitored using semi-structured interviews (N=5). These interviews will be conducted in a safe and undisturbed location chosen by the users themselves or digitally if they prefer. Each interview is expected to last approximately one hour and will be audio recorded.

Three months after the app has been tested, focus group interviews with staff (N=20) will be conducted. These interviews will take place in a suitable location at the district psychiatric center. All interviews will be transcribed into text, excluding identifying information. All audio files will be deleted after transcription.

Quantitative data on patient app usage will be collected through the Muli app, with the patient's consent given via the app. Data on the number of readmissions and/or contacts with emergency services will be gathered through interviews with patients.

Any adverse events during trial phase, will be monitored by the Muli-service team and reported to research team on a weekly basis.

Data analysis Qualitative data from in-depth interviews and focus group interviews will be analyzed using reflexive thematic analysis. Quantitative and qualitative data will be triangulated and analyzed together to gain a broader understanding of the findings.

Inclusion Criteria patients:

• older than 18 years
• reside in one of the study municipalities
• admitted into the Muli service
• have had at least three admissions into the psychiatric acute ward in the last 12 months

Inclusion criteria, healthcare workers:

- staff members at the local psychiatric hospital who respond to patient contacts made through the app

Exclusion criteria, staff:

- staff members at the local psychiatric hospital who rare not tasked with responding to patient contacts made through the app.

To protect the integrity and anonymity of study participants, IPD will not be shared due to the sensitive nature of the data being collected from a very limited sample size.