Patient-controlled Admissions in Inpatient Mental Health Services

BACKGROUND

Patient-controlled admissions are short self-referred inpatient admissions in mental health services without approval by clinicians (Heskestad \& Tytlandsvik 2008). The intention is to reduce a high use of inpatient care. Patients are signing a contract for a specified period with such stays limited to a maximum number of days per stay and with a minimum number of weeks between stays.

A systematic review on patient-controlled hospital admissions identified six small studies which had all been done in Norway and including patients with schizophrenia, affective disorders, anxiety disorder, personality disorder and substance abuse (Strand et al 2015). The review showed a tendency towards a reduction in involuntary inpatient care and use of inpatient days, but it could not conclude on effect due to small sample sizes and methodological limitations. A later and larger study in Denmark showed a reduction in involuntary admissions and inpatient days compared to historical controls, but not compared to another matched group with treatment as usual (Thomsen et al 2018). The model is line with research on the positive influence of shared decision-making, where the patient has a major influence on decisions on choice of treatment (Wills \& Holmes-Rovner 2006) and research on the use of low-threshold crisis units as an alternative to stay at psychiatric hospital departments (Lloyd-Evans et al 2009). But more studies are needed to explore the usefulness of patient-controlled admissions for different patient groups.

CONTEXT

The current study is being done in inpatient wards in four community mental health centers of Akershus University Hospital, Norway, after a smaller pilot study that indicated potential effect for patients with affective illness, but less for patients with psychosis (Støvind et al 2012). Based on the pilot project and earlier studies, it was estimated that two patient-controlled beds in each center would be enough to meet the needs for patient-controlled hospital admissions for approximately 120 patients within the target group. Each of the 19 health trust in Norway has 2-5 community mental health centers, and 40 % of the beds in mental health services in Norway are in community mental health centers.

AIMS

The aims of the study are to describe the use and experiences of patient-controlled admissions, compare the use of inpatient admissions and inpatient days during the 24 months before and after baseline, and to identify subgroups who may benefit from the model.

The study protocol in Norwegian was approved by the Regional Committee on Medical and Health Research Ethics in Norway South East 29 April 2011 (reg.no. 2011/790).

RESEARCH QUESTIONS

(RQ1) Has the model and use of patient-controlled admissions been implemented according to the plan? (RQ2) Are the number of admissions, involuntary admissions or days of stay reduced for the 24 months contract period compared with the previous 24 months? (RQ3) Are there any differences among subgroups in changes in admissions or total inpatient days? (RQ4) How has the patients' contact with staff, participation in activities and experiences of the stay been for patient-controlled stays? (RQ5) How has the patients' and relatives' experience been with the model of patient-controlled admissions for 24 months? (RQ6) Do the results indicate which patients seem to benefit from patient-controlled admissions?

DESIGN

The study is a pre-post prospective intervention study where the use of inpatient admissions in the contract period is compared to a similar period before baseline so that the patients are their own controls.

RECRUITMENT AND SAMPLE

The study has aimed to recruit 120 patients. The eligible patients have a severe mental illness, high use of inpatient mental health care the last two years and are expected to benefit from patient-controlled admissions. Only patients giving written consent are included. The inclusion period was planned for 12 months (April 2011 to March 2012) but was extended with 9 months (to December 2012) due to slower recruitment than expected. The patients are followed for 24 months from baseline.

INSTRUMENTS AND VARIABLES

Data on patients at baseline (start of period with contract about patient-controlled admissions, and inclusion in the study) includes socio-demographics, current use of health services, ICD-10 diagnoses, ratings on type and severity of current psychiatric problems using the Health of the Nation Outcome Scales (HoNOS) (Wing et al 1998) and use of alcohol or illegal drugs using Alcohol and Drug Use Scales (Mueser et al 1995).

Data on each patient-controlled admission and the patient's experience of this includes date and hour for self-referral, date and hour for admission, date and hour for discharge, reasons for self-referral, contacts with staff and participation in activities during the stay, and satisfaction with the stay.

Data on patients' experience with the model with patient-controlled admissions includes questions about feasibility of using patient-controlled admissions, satisfaction with the contract and the patient-controlled stays, how they think the model may have changed their use on inpatient care, and how the model may be improved. Data on relatives' experiences with the model includes the same question as to the patient but adapted to the relatives' situation.

Data on all inpatient admissions/stays in the mental health services for each participant at Akershus University Hospital for the 24 months contract period and for the 24 months period before baseline is extracted from the electronic patient records. For each stay data is extracted on department/section, admission date, discharge date, whether admission was voluntary/involuntary, and date for changing involuntary stay to voluntary stay.

DATA COLLECTION

Data on the patient's situation and mental health at baseline is collected by mental health nurses trained in completing the assessment form in collaboration with the clinician of the patient. Data on each patient-controlled admission is collected by the ward staff during the stay and in an interview with the patient at the end of the stay. Data on the patient's and a relative's experience with the model is collected by the ward staff in a brief interview or telephone interview at the end of the 24 months contract period.

Data on all inpatient admissions/stays in the mental health services for each participant is extracted later from the electronic patient journal system at Akershus University Hospital for the 24 months contract period and for the 24 months period before baseline.

DATA ANALYSIS

Descriptive analyses will be done to describe the patient sample and their situation at baseline, the use and characteristics of the patient-controlled admission/stays, and the patients' and relatives' experiences with the patient-controlled admission model.

Testing of significant differences in use of admissions and inpatient days between the 24 months contract period and the previous 24 months period with treatment as usual will be done with t-test and chi-square. Multiple regression analyses will be used to study association between baseline characteristics and the use of patient-controlled admissions. If subgroups with different patterns of use of patient-controlled admission/stays are identified, analyses may be done to identify characteristics of these groups, depending on the size of sample and subgroups.

The inclusion period was 2011-2012. Data collection was done 2011-2014. Data extraction from the patient records was done 2015-2016. Quality control and organization of data was done 2017-2018. Data analysis will start in August 2018.