Study protocol for a randomised controlled trial evaluating an evidence-based, stepped and coordinated care service model for mental disorders (RECOVER).

INTRODUCTION: Healthcare systems around the world are looking for solutions to the growing problem of mental disorders. RECOVER is the synonym for an evidence-based, stepped and cross-sectoral coordinated care service model for mental disorders. RECOVER implements a cross-sectoral network with managed care, comprehensive psychological, somatic and social diagnostics, crisis resolution and a general structure of four severity levels, each with assigned evidence-based therapy models (eg, assertive community treatment) and therapies (eg, psychotherapy). The study rationale is the investigation of the effectiveness and efficiency of stepped and integrated care in comparison to standard care. METHODS AND ANALYSIS: The trial is conducted in accordance to the Standard Protocol Items: Recommendations for Interventional Trials Statement. The study aims to compare the RECOVER model with treatment as usual (TAU). The following questions are examined: Does RECOVER reduce healthcare costs compared with TAU? Does RECOVER improve patient-relevant outcomes? Is RECOVER cost-effective compared with TAU? A total sample of 890 patients with mental disorders will be assessed at baseline and individually randomised into RECOVER or TAU. Follow-up assessments are conducted after 6 and 12 months. As primary outcomes, cost reduction, improvement in symptoms, daily functioning and quality of life as well as cost-effectiveness ratios will be measured. In addition, several secondary outcomes will be assessed. Primary and secondary outcomes are evaluated according to the intention-to-treat principle. Mixed linear or logistic regression models are used with the direct maximum likelihood estimation procedure which results in unbiassed estimators under the missing-at-random assumption. Costs due to healthcare utilisation and productivity losses are evaluated using difference-in-difference regressions. ETHICS AND DISSEMINATION: Ethical approval from the ethics committee of the Hamburg Medical Association has been obtained (PV5672). The results will be disseminated to service users and their families via the media, to healthcare professionals via professional training and meetings and to researchers via conferences and publications. TRIAL REGISTRATION NUMBER AND REGISTRY NAME: ClinicalTrials.gov (NCT03459664), RECOVER PROTOCOL VERSION: 19 March 2020 (V.3.0). © Author(s) (or their employer(s)) 2020. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

RCT Entities:   Population Interventions Outcomes

Implementation of an evidence-based, cross-sectoral care network for mental disorders with managed care, comprehensive diagnostic procedures and a crisis resolution for all patients in acute crises was achieved.

Fidelity and integrity of the RECOVER service model was established by 12 standard operating procedure manuals for all core components.

The study provides a unique opportunity to evaluate the impact of such a stepped care service model on healthcare costs, cost-effectiveness and on patients’ outcomes with respect to symptoms, functioning, quality of life and satisfaction with care.

The RECOVER project was initially supported by 4 and now 19 health insurance funds, which represent a high proportion (about 80%) of all insured persons.

Network management was and is a central task, because there are no established incentives in the German healthcare system that promote binding participation.

Background and rationale

About 30% of the German population are affected by a mental disorder per year,1 and about 20% of the patients experience relevant losses of their functional level.1 2 This means that approximately 15 million people in Germany are affected by a relevant mental disorder every year. The Organisation for Economic Co-operation and Development (OECD)3 4 has calculated that the share of direct and indirect costs for mental disorders in Germany in the gross domestic product was 4.8% in 2015, approximately 146 billion €.

These costs are also caused by structural problems of the German healthcare system for mental disorders.3 5 6 The OECD,3 the Advisory Councils on Healthcare5 and Macroeconomic Development,6 professional society (German Society of Psychiatry, Psychotherapy and Neurology; DGPPN,7 statutory health insurance providers (DAK-Gesundheit,8 BARMER9 as well as patient and family associations (BapK10 criticise the fragmented structures and services, the lack of trans-sectoral coordination, permeability and cooperation, inefficient use of funds due to overuse and underuse, as well as strong regional discrepancies. Additional problems remain likely access to care, inadequate standardisation in diagnostics and indications, long waiting times for psychotherapy, misdistribution in outpatient psychotherapy to the detriment of severe mental illnesses (SMI) and the lack of implementation of assertive treatment models for short-term acute treatment and for long-term treatment of patients with SMI. Furthermore, the digitalisation of the care system and the use of e-mental health has not yet been implemented.6

Like many other countries, Germany has responded to these structural deficits with a largely non-systematic increase in quantity of care. Since 2005, almost all indicators of the healthcare system have shown an increase: number of hospitals (especially clinics for psychosomatic medicine), inpatient and day-clinic treatment places, number of psychiatric outpatient departments and treated cases, number of psychotherapists in private practice, etc.11 In contrast to almost all other European countries, which have often promoted the development of the community psychiatry, even inpatient treatment places have risen (+14% in Germany vs −17% in other European countries).12 In other healthcare systems such an increase has not led to a reduced prevalence of common mental disorders.13 Accordingly, many experts and associations like the OECD recommend an improvement in quality of care and the implementation of evidence-based structures and interventions.3 13

Accordingly, the Advisory Councils on Healthcare5 and Macroeconomic Development6 as well as professional associations7 in Germany call for the ‘implementation of stepped, needs-based, person-centred, cross-sector and setting-spanning care’ and the ‘implementation of digital health including the use of e-mental health solutions in all sectors of the healthcare system’. However, when implementing such an evidence-based, stepped and coordinated care service model, including e-mental health, some evidence-based principles must be taken into account.14 15

Stepped care models exist for certain mental disorders (eg, for major depression,16–18 anxiety disorders,17 personality disorders19 20 or psychosis21 or so-called ‘service models’,14 15 22 in which evidence-based therapy models and therapies are logically linked in one evidence-based stepped care model. The inter-sectoral and trans-sectoral treatment processes are based on components of managed and coordinated care. Service models work cross-sectoral in a network of service providers that jointly cover the entire spectrum of care, including inpatient, day-care, outpatient and rehabilitative care.14

Stepped care is a system of treatment delivery and monitoring in which the most effective and resource-saving treatment is the first treatment option.14 Coordinated (or collaborative) care refers to care that is coordinated between service providers across sectors and disciplines and is also referred to as integrated care. Stepped and coordinated care has four main principles: (a) Service providers work together and coordinate across sector boundaries; (b) Interventions depend on the severity of the disease, the most effective and resource-saving treatment is always initiated first; (c) As many treatment models and therapies as possible are evidence-based and demonstrably effective (effective therapies are more efficient) and (d) an upgrading or downgrading takes place according to predefined rules (eg, disease progression).22

Severity levels are based on epidemiology with respect to the severity distribution of 20% of patients with relevant functional deficits: (a) 9% to 12% have a mild severity, (b) 4% to 6% have a moderate severity and (c) 1% to 2% have a moderate-to-severe severity.2 Most of these patients suffer from a so-called common mental disorder (CMD), that is, mental diseases with a comparatively high prevalence, but a low risk for the development of a severe mental illness (eg, unipolar depression, anxiety disorders). The remaining 1% to 2% of the 20% of patients suffer from a SMI.2 23 24 The definition of SMI comprises (a) the diagnosis of a mental disorder and (b) a functional level that is consistently and severely impaired by the disorder.22 23 The highest risk for SMI is in schizophrenia (90% will develop an SMI), followed by schizophrenia spectrum disorders (60%), bipolar I disorder and unipolar severe depression with psychotic symptoms (both 40%) and personality disorders (30%; especially the emotionally unstable personality disorder.23 Relative to 100%, 60% of all SMI are psychotic disorders.24

Regarding the integration of evidence-based therapy models and guidelines therapies into the model, the OECD Report of 20143 systematises evidence-based interventions for patients with CMD and SMI. With regard to CMD, these are psychotherapy, e-mental health and work (re)integration. For patients with SMI, these are short-term crisis resolution, early intervention services and assertive community treatment as well as psychotherapy, e-mental health and work (re)integration (eg, supported employment) and peer support.

In principle, the approach is to achieve improved care without increasing resources.14–21 To this end, various cost approaches could be integrated: (a) outpatient care before inpatient or day-clinic care, (b) stepped outpatient care, (c) care for mild (or moderate) mental disorders primarily by e-mental health instead of face-to-face psychotherapy, (d) stepped psychotherapy (eg, group or short-term psychotherapy before long-term), (e) outreach crisis resolution to prevent or shorten inpatient or day-clinic care, (f) assertive community treatment for people with SMI to prevent and shorten inpatient or day-clinic care, (g) rapid access to supported employment to reduce days with inability to work and (h) access to evidence-based care with better recovery and less consecutive costs.

The overall objective of RECOVER is to improve the care of those affected by mental disorders and their relatives on an evidence-based and sustainable basis through structured cross-sectoral cooperation between service providers and targeted additions to the care system, particularly for the treatment of severely ill patients.

Objectives

Efficiency and effectiveness of the RECOVER care model are evaluated from 2017 to the end of 2020 in a prospective, monocentric, randomised controlled trial (RCT). This article reports on the study protocol for the RECOVER RCT. The trial is conducted in accordance to the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) Statement for reporting parallel group randomised trials.25 The primary hypotheses include that RECOVER leads to cost savings compared with standard care within the 1 year treatment period, that RECOVER leads to greater benefits in terms of improving the patient's state of health and that RECOVER has a better efficiency (cost-effectiveness) than standard care.

Trial design and conceptual framework: RECOVER model

The so-called RECOVER model was developed on the basis of these structural, therapeutic and cost-saving approaches. RECOVER is the synonym for an evidence-based, stepped and coordinated care service model for mental disorders. The evaluation is funded by the Innovation Fund of the Joint Federal Committee (G-BA) from 2017 to 2020 (funding code: 01NVF16018). The G-BA is the highest decision-making body of the joint self-government of physicians, dentists, hospitals and health insurance funds in Germany.

The RECOVER model was developed by a consortium of representatives from Hamburg, Itzehoe and Germany including the Hamburg Health Authority, Hamburg patient and family associations, the Hamburg Chambers and Associations of Physicians, General Practitioners and Psychotherapists, the behavioural therapy centre ‘Behavioural Therapy Falkenried clinics GmbH’ and the work integration centre ‘ARINET GmbH’, the German expert associations of adult and child and youth psychiatry and psychotherapy (DGPPN, DGKJP), the Centre for Psychosocial Medicine of the Hospital Itzehoe and the UniversityMedical Center Hamburg-Eppendorf (UKE) as consortium leader with nine departments and institutes. The accompanying research is carried out by three independent institutes for health economics and health services research, clinical healthcare research and medical biometry and epidemiology. The application and execution of studies within the Innovation Fund is tied to the participation of health insurances. RECOVER was initially supported by four statutory health insurance providers, including BARMER, DAK-Gesundheit, AOK Rheinland/Hamburg and HEK. Since the introduction of the model in January 2018, the network is constantly growing to by now over 270 participating institutions, registered physicians, general practitioners, psychotherapists and staff. In addition, 15 further statutory health insurance providers joined the RECOVER model. In 2018, the German Society for Psychiatry, Psychotherapy and Neurology awarded the RECOVER model as the reference model for sustainable psychiatry in the future in Germany.7

RECOVER combines three approaches:14 15 First, managed and coordinated care across sectors within a sectoral partner network. Second, stepped care within four severity levels from mild mental disorders (level 1) to severe mental disorders (level 4) with associated treatment packages, always with the proviso that the most effective resource-saving interventions are used first. Third, as many interventions as possible are evidence-based, because evidence-based interventions are more efficient and thus save resources.

The RECOVER service model consists of nine innovative care components, which are described in more detail in the following section. Each care component has been documented in a standard operating procedure manual (eg, see www.recover-hamburg.de).26–28 For more details, see figure 1.

Figure 1

The RECOVER evidence-based, stepped and coordinated care model.

Improvement of managed and coordinated care

The UKE has established a Competence Centre for Integrated Mental Healthcare. This centre has the task of improving the management and coordination of all cross-sectoral forms of care. This includes, for example, the involvement of institutions and clinicians through cooperation agreements, the establishment of a sectoral care network, care management (ie, case management, allocation of therapy appointments, documentation), training and quality assurance. Access to care is improved by immediate appointments mostly within 3 to 5 days and the possibility of 24 hours crisis intervention. Information on access to care is available from all cooperation partners and can be accessed by patients and their relatives via the publicly accessible website. In addition, the centre is the operator of a new e-mental health platform (eRECOVER; see www.erecover.de), which was developed within the framework of RECOVER. An online outpatient clinic for digital therapy has been integrated.26

Improvement of diagnostics and crisis resolution

The improvement of diagnostics and crisis intervention is achieved through the implementation of a crisis resolution team (so-called AID and CARE Team; AID stands for Ambulance for Indication and Diagnostics and CARE stands for Crisis And REsolution). It is a specialised, multi-professional and interdisciplinary team of physicians, psychologists, nursing staff, social workers and recovery counsellors from adult psychiatry, child and youth psychiatry, psychosomatics, medical psychology, general practice, sexual medicine and forensic psychiatry as well as a network partner for supported employment. The tasks include standardised interdisciplinary biological, psychological and social initial diagnostics, indication and treatment planning, cross-sector outreach crisis intervention and managed care (implementation of the cross-sector treatment plan). The team works with an electronic board (called AID and CARE Board), in which all patients are discussed twice a day, especially those in crisis resolution treatment. The team assigns patients to one of four severity levels and the corresponding treatment plan is implemented by the Competence Centre in cooperation within the care network. The CARE treatment can be used whenever necessary during the entire therapy period.27

Improvement of care for people with severe mental illness

Improvement of care for people with SMI is supported by the integration of Therapeutic Assertive Community Treatment (TACT) for severe psychotic disorders including schizophrenia spectrum disorders (F2) and bipolar I disorder (F31), severe unipolar depression with psychotic features and severe borderline personality disorders (F60.3).28 These indications were chosen because these diagnoses have the highest risk for the development of SMI and account for about 80% of all patients with SMI.24 This so-called ‘Hamburg integrated care model’ has been financed since 2007 as Integrated care contract by five health insurances and was included into the RECOVER model for people in severity level 4.29–32 Currently, there are four TACT teams: two for multiple-episode patients with severe psychoses, one team for adolescents and young adults with a first-episode psychosis and one team for patients with borderline personality disorders. Each team is responsible for around 80 to 100 patients in a 1:15 to 1:25 clinician-patient ratio with 24 hours/365 days emergency interventions. The TACT teams have extensive expertise and are multidisciplinary including psychiatrists, psychologists, nurses, social workers and peers, ≥80% are psychiatrists and psychologists. The Hamburg model was examined in three major evaluations with regard to effectiveness and efficiency. These studies showed a good efficacy regarding symptoms, functional level, quality of life, remission and recovery29–32 with high efficiency.33

Integration of general practice

People with mental disorders, especially those with SMI, display a high morbidity and mortality risk.31 34 Various models have attempted to improve coordination between primary care and psychiatry with unclear success.31 One of the most recommended models is the so-called Reverse Integrated Care model, in which primary healthcare providers are co-located in the mental health setting.31 In RECOVER, this is achieved by integrating general practitioners into the AID and CARE team. They are responsible for all somatic assessments, the organisation of further examinations and therapies in the network and the establishment, management and training of a network of general practitioners.

Integration and increased flexibility of psychotherapy

Due to the long waiting times for psychotherapy of 5 months35 on average and the preference of patients with mild and moderate mental illness,36 RECOVER has developed various incentives for psychotherapists together with the Psychotherapists' Chamber in Hamburg. The objectives are to shorten waiting times, to take over patients with higher degrees of severity through joint treatment with the Crisis Resolution Team and case manager and more frequent use of stepped and flexible psychotherapy. The incentives include, for example, the waiver of the application procedure, which is now supported by all health insurances, the increase in short-term and group psychotherapies, the possibility of utilisation of the crisis resolution team at any time and treating crises together on an outpatient basis as well as the qualification of staff through certified, training courses, case conferences and quality circles. In the future, psychotherapists in private practice can also use the e-mental health platform eRECOVER.

Integration of e-mental health

Despite its great potential and meanwhile also evident benefits,37–39 e-mental health is hardly integrated into the German healthcare system, it is not part of the standard care and is currently used by less than 1% of all clinics as well as outpatient clinicians and psychotherapists. Within RECOVER, e-mental health is integrated into the stepped care model. The e-mental health platform eRECOVER (see www.erecover.de) provides digital diagnostics and therapy for all severity levels. In level 1, this is the first intervention before face-to-face psychotherapy. In other levels it accompanies other interventions. Within eRECOVER, the following variants of digital therapy can be performed: (a) digital self-help programmes, (b) guided digital therapy, (c) blended digital therapy. In the future, video individual therapy and video group therapy will be added.

Integration of supported employment

Individual placement and support is a further development of supported employment (SE). It includes training and work (re)integration with reintegration or integration on the first (paid) training or labour market with promotion of the sustainability of the intervention through job coaching.40 The basis of this intervention is that 95% of all days of incapacity to work in Germany are generated by patients with CMD and that these patients in particular do not have access to evidence-based work (re)integration. Accordingly, a partnership was initiated with a provider of SE (ARINET GmbH) which offers the following interventions: (a) systematic screening and examination of occupational perspectives, (b) for patients with incapacity to work, measures such as job coaching or clarification assistance for early employability with initial counselling, in-company training, training on the job and support on the job at the workplace, (c) advice and support for taking a vocational rehabilitation measure. Supported employment offers counselling for people who are unable to work, clarification of prerequisites or integration and placement in the existing labour market. The know-how is passed on to network partners and gradually a cooperation network with employers is established.

Integration of culture-sensitive and language-sensitive care for migrants and refugees

The integration of cultural aspects in the therapy of mental disorder is becoming increasingly important. A number of measures have been implemented to improve integration: Within the AID and CARE team, specially trained employees work who in turn instruct other employees and provide further education in regard to cross-cultural competencies. In addition, culturally sensitive diagnostics has been implemented. A manual has been developed to ensure quality standards for culturally sensitive care.

Participation of peers and relatives and implementation of peer support

The aim is to improve the empowerment and participation of patients and their families in the organisation, treatment and research. This is achieved by representing patient and family associations on the RECOVER advisory board and by peer support in all major clinical units (crisis resolution team and assertive community treatment teams). In addition, the goals of the project and the accompanying research were coordinated with a special committee of patients and relatives.

Improvement of evidence-based treatment is achieved by assigning evidence-based treatment models and therapies to the four severity levels (levels 1 to 4) as shown in the nine innovative care components. Regardless of severity, all patients will have access to managed and coordinated care, diagnostics and crisis resolution, social work, supported employment and peer support. Depending on the degree of severity, patients in levels 1 to 4 have access to the following treatment packages:

Level 1: mild severity (mostly CMD): counselling, active waiting, self-help, guided digital therapy, social work, supported employment and peer support.

Level 2: moderate severity (mostly CMD): coordinated standard care with stepped individual and/or group psychotherapy (≤12 hour), digital therapy, social work, supported employment and peer support.

Level 3: moderate-to-severe severity (mostly CMD): coordinated standard care plus case management with stepped individual and/or group psychotherapy (>12 hour to long-term), digital therapy, social work, supported employment and peer support.

Level 4: severe mental illness (1% to 2%, SMI): therapeutic assertive community treatment including 24 hours crisis resolution and individual and/or group psychotherapy, digital therapy, social work, supported employment and peer support.

Methods and analysis

Study design

The RECOVER study is a prospective, monocentric, RCT conducted in the catchment area of the University Medical Center Hamburg-Eppendorf, Germany.

Changes of trial design

In addition to the 4 statutory health insurance funds, another 15 statutory health insurance funds have joined the model, which has not resulted in any changes of the study design.

Study setting

The study takes place within the catchment area of the University Medical Center Hamburg-Eppendorf, Hamburg, Germany, from January 2018 to December 2020. The Department of Psychiatry and Psychotherapy covers an area of approximately 330 000 inhabitants. The area comprises about 20 psychiatric institutions and about 100 registered specialists in psychiatry and psychotherapy, general practice and psychosomatics and about 200 registered psychological psychotherapists. The RCT is conducted by three independent institutions: the Department of Medical Psychology (UKE), the Department of Health Economics (UKE) and the Department of Biostatistics (UKE). As part of the study implementation, the research institutions also have the task of data monitoring.

The recruitment of the participating patients took place via a systematic, daily screening in the psychiatric regular care of the UKE. In addition, all partners involved have received a screening form to refer patients to the UKE. We have also made a screening form available on the homepage that interested parties could use to contact us directly. This form is adapted to the following inclusion and exclusion criteria.

Inclusion criteria

Eligible participants are people at the age of ≥16 years, insured with one of the 19 health insurances involved and living in the catchment area of the UKE (8 km radius), when they suffer from at least one relevant mental disorder according to the International Statistical Classification of Diseases and Related Health Problems - 10th Revision, German Modification (ICD-10):41 schizophrenic spectrum disorders (ICD-10: F20, F22, F23, F25), bipolar disorder (ICD-10: F31), major depression (ICD-10: F32, F33), anxiety disorder (ICD-10: F40, F41), obsessive-compulsive disorder (ICD-10: F42), post-traumatic stress disorder (ICD-10: F43.1), adjustment disorder (ICD-10: F43.2), somatoform disorders (ICD-10: F45), eating disorder (ICD-10: F50), personality disorder (ICD-10: F60, F61) or attention deficit hyperactivity disorder (ICD-10: F90).

Exclusion criteria

Subjects are excluded from the study when fulfilling the criteria for organic mental disorders (ICD-10: F00-09); with a main diagnosis of addiction disorders (ICD-10: F10-19) (comorbid addiction disorders do not lead to exclusion), with severe or moderate mental retardation (pre-diagnosed ICD-10: F72/F73), with insufficient knowledge of German, with uncorrectable impairment of vision and/or hearing.

Interventions

On inclusion in the study, all participants receive a detailed psychological assessment. This consists of standardised questionnaires regarding the main diagnosis, comorbid disorders, social problem areas, functional level and quality of life. On the basis of combined criteria consisting of severity of the disease and functional level, the classification into the four severity levels is carried out (see table 1). Subsequently, the randomisation and thus the allocation to the intervention and control group takes place. With regard to the comparison of structures and interventions between intervention and control groups, table 2 compares all essential care components (see table 2).

Table 1

Classification into four severity levels

Measurement	Severity levels
	Level 1(mild)	Level 2(medium)	Level 3(medium-to-severe)	Level 4(severe)
Main disorder according to DSM-V	296.x, 300.x, 307.x, 309.x, 314.0x	296.x, 300.x, 301.x, 307.x, 309.x	296.x, 300.x, 301.22, 307.x, 309.8, 301.x	295.x, 296.4/5 (incl. psychosis), 296.34, 297.1, 298.x, 301.x
Main disorder according to ICD-10	F32, F40, F41, F43.2, F45, F90	F32, F40, F41, F42, F43.1, F43.2, F45, F50, F90	F20, F22, F23, F25, F31, F33, F41, F42, F43.1, F45, F50, F60, F61	F20, F22, F23, F25, F31, F32.3, F33.3, F60
Global Assessment of Functioning (GAF)	GAF score 61–100: no or mild symptoms in the last 4 weeks	GAF score 51–60: moderate symptoms in the last 4 weeks	GAF score 31–50: serious symptoms or impairments in the last 4 weeks	GAF score ≤50 for the last 6 months: serious or major impairments
Clinical Global Impressions -Severity scale (CGI-S)	CGI 1–3	CGI 3–4	CGI 4–6	CGI 5–7

DSM-V, Diagnostic and Statistical Manual of Mental Disorders – Fifth Edition; ICD-10, International Statistical Classification of Diseases and Related Health Problems - 10th Revision; incl., including.

Table 2

Key characteristics of RECOVER intervention and TAU control groups

Dimensions	RECOVER group	TAU group
(1) Access to care	Outpatient appointment within 3–7 days, crisis resolution 24 hours/day	Often long waiting time for outpatient appointments, with respect to psychotherapy 4–6 months Emergency department 24 hours/day
(2) Standardised assessment at service entry	Standardised psychological, somatic and social assessment Multi-professional and interdisciplinary review	Assessment often not standardised, often focus solely on psychological issues No multi-professional and interdisciplinary review
(3) Indication and treatment planning	Multi-professional and interdisciplinary indication and treatment planning	Mostly no multi-professional and interdisciplinary indication and treatment planning in outpatient care
(4) Managed and coordinated care	Organisation of the therapy plan in the network and coordination of therapy	Managed and coordinated care not part of standard care
(5) Crisis resolution (CR) for people with all mental disorders	Multi-professional and interdisciplinary crisis resolution team with 24 hours crisis resolution Coordinated inpatient and day-clinic care	Inpatient care Day-clinic care
(6) Assertive community treatment (ACT) for people with severe mental illness	Multi-professional ACT teams including psychotherapy and 24 hours crisis resolution	ACT not part of standard care ≤5% of patients with SMI receive psychotherapy
(7) Access to primary care	Integrated access to primary care physicians in the network	Access to primary care physicians with waiting time Not integrated into other mental healthcare
(8) Access to psychotherapy	Access to stepped psychotherapy within the network with short waiting time	Access to short-term or long-term psychotherapy with long waiting time
(9) E-mental health	Digital self-help, guided or blended digital therapy	Not part of standard care Dependent on health insurance access via special supply contracts Not integrated into other mental healthcare
(10) Supported employment	Access to supported employment workers	Not part of standard are
(11) Culture and language-sensitive care	Access to specialists within the crisis resolution team Systematic involvement of interpreters	Not part of standard outpatient care Systematic involvement of interpreters in inpatient care available
(12) Peer support	Peer Support workers in CR and ACT teams	Not part of standard outpatient care

TAU, treatment as usual.

RECOVER treatment (intervention group)

Patients assigned to the RECOVER model are first admitted to the centre and registered in the electronic patient file. The patient is then automatically assigned to a case manager of the AID and CARE team. On the basis of the preceding standardised diagnostics, the case manager carries out a re-evaluation of the psychosocial diagnostics, transfers the data to the AID board and calls in a social worker in case of social problems. As a standard, each patient receives somatic diagnostics from the general practitioner. Subsequently, the patient is discussed in the twice-daily multi-professional and interdisciplinary meetings and a treatment plan is drawn up. In acute crises, the patient is admitted to the CARE treatment. In contrast to standard care, the treatment plan is organised in the network for the patient. The case manager always remains the patient's primary contact person, even when referrals are made to the network. If another acute crisis occurs, the patient can be treated again with CARE at any time.

Treatment-as-usual (control group)

The control group receives standard care that is provided in the sector of the University Medical Center Hamburg-Eppendorf. This includes the use of full and day-clinic inpatient treatment within the Clinic for Psychiatry and Psychotherapy of the UKE and the Clinic for Child and Adolescent Psychiatry of the UKE, the treatment in the Psychiatric Outpatient Clinic of the UKE, the use of therapy from specialists in general practice and psychiatry as well as psychological psychotherapists and treatment at institutions of assisted living and rehabilitation of mental illnesses.

Twelve months treatment in RECOVER compared with TAU result in a reduction of (a) costs of healthcare as covered by the statutory health insurance (SHI), (b) costs of care as covered by other payers, (c) costs due to productivity losses (indirect costs). RECOVER is cost-saving compared with TAU.

Twelve months treatment in RECOVER compared with TAU is associated with (a) improved disease remission and response, (b) reduced symptoms and illness severity, (c) improved functioning and (d) improved health-related quality of life. These measures will be linearly transformed and added up to one measure ‘psycho-functional level’.

Twelve months treatment in RECOVER compared with TAU lead to a gain in quality-adjusted life years (QALYs) across all patients, with simultaneously unchanged or reduced direct (SHI perspective) or direct and indirect (societal perspective) costs.

Outcomes and hypotheses

Secondary outcomes

Twelve months treatment in RECOVER compared with TAU leads to the following secondary outcomes: (1) a reduction of inpatient and day-care admissions and inpatient and day-care days, (2) a reduction of days with inability to work, (3) a lower service disengagement rate, (4) a reduction of waiting time until start of psychotherapy aid by SHI, (5) a higher percentage of patients with SMI receiving group and individual psychotherapy, (6) a higher use of digital therapy and (7) a higher use of peer support.

Changes to trial outcomes after trial commenced

None.

Sample size

The sample size is based on a power calculation to detect a statistically significant difference between the intervention and control group of a small-to-medium effect size (Cohen's f of 0.175) after 12 months (t12). Two hundred and thirty-four study participants in total (117 in each group) are required given a statistical power of at least 80% (with a type I error rate of 5% in a two-sided test and 10% explained variance by the baseline value). The sample size is increased to 384 to include interactions of the interventions with a small-to-medium effect size (Cohen's f of 0.175). Diagnostics is performed centrally. After the individual randomisation of each study participant, the therapy in RECOVER and in TAU takes place in approximately 50 clusters with approximately 21 participants each. To take this cluster effect into account, an intra-class correlation=0.05 is assumed. So we obtain a design effect of 2.0 for the primary, continuous outcome, which leads to a total number of 890 patients that have to be included (dropout rate of about 30% is included).

Assignment of interventions

Single blinded study (outcomes assessor): Individuals who evaluate the outcomes of interest will remain blinded regarding a participant's condition (RECOVER/TAU) over the course of the study. The individual stratified randomisation (ie, four severity levels) will be conducted after baseline assessment and communicated to the participant by a person that is not the outcome assessor. We have used the procedure ‘ralloc’ (Stata/SE 14) with variable block sizes within each stratum. During follow-up assessments after 6 and 12 month the outcome assessor will remain blinded regarding a participant's condition.

Data collection, management and analysis

Data will be collected before intervention (t0) after 6 (t6) and 12 months (t12) (see figure 2 for the Consolidated Standards of Reporting Trials flow diagram). The following instruments are used: Diagnostic and Statistical Manual of Mental disorders - Fifth Edition (DSM-V), structured clinical interview (SCID I and II),42 psychiatric use of care services (FIMPsy questionnaire),43 general use of health services (FIMA questionnaire),44 disease remission or responses (HEALTH-49)45; ClinicalGlobal Impressions.46 Moreover, the use of additional, study specific health services like the use of digital therapy is assessed in a questionnaire. The severity of the symptoms is also measured for the different diagnostic groups (diagnosis-specific). Further questionnaires measure everyday functioning level (observer rated: GlobalAssessment of Functioning,47 health-related quality of life (EQ-5D-5L,48 SF-12,49 ReQOL,50 and QALYs (based on EQ-5D-5L index).48 Various risk parameters and comorbid diseases are recorded across all diagnoses. A sample of relatives will be interviewed (questionnaire and interview). The collected health data are enriched with secondary data obtained from external data owner (eg, inpatient performance data, outpatient medical performance data, etc). After 12 months (at t12), this data is requested from the health insurance companies for the past 36 months. A monetary valuation is then performed according to standardised monetary valuation rates.51 52 For more details, see table 3.

Figure 2

Consolidated Standards of Reporting Trials flow diagram of the RECOVER study.

Table 3

Measurement used for measuring primary and secondary outcomes

Outcome measure	Measurement	Details of the measurement	Completed by
Primary outcomes
Direct costs	FIMA,43 FIMPsy44	Assessment of psychiatric (FIMPsy) and general (FIMA) use of healthcare services and monetary evaluation using standardised unit costs40 41	Interviewer
Indirect costs	RECOVER questionnaire	Assessment of indirect costs as productivity loss due to days off work/sick leave or early retirement	Interviewer
Disease remission and response	Health-49,45 CGI46	Rating of general aspects of psychosocial health (Health-49) and severity of patient’s illness (CGI-S)	Study participant/interviewer
Symptoms and illness severity	Diagnosis-specific questionnaires	Rating of the severity of symptoms using several diagnosis-specific questionnaires	Interviewer
Functioning level	GAF47	Rating of everyday functioning level	Interviewer
Health-related quality of life	EQ-5D-5L,48 SF-12,49 ReQoL50	Rating of health-related quality of life and calculation of QALYs using the results of the EQ-5D-5L	Study participant
Secondary outcomes
Inpatient and day-care admissions, inpatient day-care days	Clinic documentation, FIMA,44 FIMPsy45	Assessment of psychiatric (FIMPsy) and general (FIMA) use of healthcare services	Clinician/interviewer
Days with inability to work	RECOVER questionnaire	Assessment of days off work/on sick leave	Interviewer
Service disengagement rate	Clinical documentation	Patient interrupts contact with the treatment facility and cannot be reengaged again	Clinician
Waiting time until start of psychotherapy	RECOVER questionnaire	Assessment of active search for outpatient psychotherapeutic treatment after 6 months (t6) and 12 months (t12)	Study participant
Group and individual psychotherapy for patients with SMI	Clinic documentation, RECOVER questionnaire	Assessment of service use of specific interventions after 6 months (t6) and 12 months (t12)	Clinician/study participant
Use of digital therapy	RECOVER questionnaire	Assessment of service use of specific interventions after 6 months (t6) and 12 months (t12)	Study participant
Use of peer-support	FIMPsy43 (t0), RECOVER questionnaire	Assessment of service use of specific interventions after 6 months (t6) and 12 months (t12)	Study participant/interviewer

CGI-S, Clinical Global Impressions - Severity scale; GAF, Global Assessment of Functioning; QALYs, quality-adjusted life years; SMI, severe mental illnesses.

The primary and secondary outcomes are evaluated according to the intention-to-treat principles. For the primary outcome psycho-functional level, the change from baseline to 12 months follow-up will be analysed by calculating a linear mixed model with group (RECOVER, TAU), severity level and interaction between group and severity level as fixed effects, cluster as random effect and the baseline value of psycho-functional level as a covariate. Disease remission and response to treatment are analysed using mixed logistic regression. Remission is assumed if a predefined cut-off value is achieved at follow-up (eg, Patient Health Questionnaire-9 ≤5). Response to treatment is assumed if a patient has improved by 50% of the initial symptoms. Changes in disease symptoms, everyday functioning level and health-related quality of life are analysed using mixed linear regression models. For the evaluation of the primary outcome direct and indirect costs during the 12-month follow-up, multiple difference-in-difference regressions are used. We assume that the intervention is cost saving, if the negative difference in costs is statistically significant (p<0.05). Results are interpreted as cost neutral, if the difference in costs is negative and not statistically significant (intervention is less expensive) or if the difference in costs is positive with a p value >0.5 (intervention is more expensive). All models are used with the direct maximum likelihood estimation procedure which results in unbiassed estimators under the missing-at-random assumption. Adjusted means and ORs, respectively, with their 95% CIs and p values will be reported. The two-sided type I error will be set at 0.05. Interim analyses are not planned. Cost-effectiveness will be analysed by incremental cost-effectiveness ratios (ICER). Cost-effectiveness acceptability curves based on net benefit regressions will be used to evaluate uncertainty of the ICER. A detailed statistical analysis plan will be prepared and finalised before the code is broken. Results will be reported according to the SPIRIT guidelines. Only the analysis of primary outcomes will be considered in a confirmatory manner. For sensitivity analyses, missing values are replaced by multiple imputations and per-protocol analyses are performed.

Secondary outcomes are examined with multivariate methods (logistic regression, difference-in-difference regression). The secondary outcomes will be evaluated according to the scale level with mixed linear or logistic regression models. For the analysis of inhibiting and promoting factors for the overall treatment model and an effective and efficient implementation of RECOVER in clinical routine, qualitative methods are used. The results of the study will be evaluated using descriptive and inferential statistical analyses of sociodemographic and diagnostic data. The predictive value of different factors will be tested by logistic regression.

Patient and public involvement

Within RECOVER, patients, relatives and the public were systematically involved: (1) Peer support is a separate intervention module, which provides the systematic integration of trained patients into the provision of care, for example, in the crisis resolution team and in the assertive community treatment teams; (2) The entire care model and research project RECOVER was planned and carried out in coordination with the patient and relatives organisation ‘EmPeeRie - Empower Peers to Research’ regarding content and study questions; (3) RECOVER was led by a steering committee; patient and family member organisations from Hamburg are represented in this committee; (4) The public was informed via a separate project website. Here, all materials developed are also available for download.

Discussion

Health systems around the world are looking for efficient solutions to the growing problem of mental healthcare and its funding. In Germany, the Advisory Councils on Healthcare and Macroeconomic Development as well as professional associations call for the introduction of stepped, integrated and coordinated care. RECOVER is the synonym or such an evidence-based, stepped and cross-sectoral coordinated care service model for all main common and severe mental disorders. RECOVER implements a cross-sectoral care network with managed care, a comprehensive psychological, somatic and social initial diagnosis, crisis resolution for all patients in acute crises and four severity levels (from mild-to-severe mental illness), each with assigned evidence-based therapy models and therapies. The RECOVER study will be able to answer important questions regarding costs, efficiency and effectiveness of the model. In addition, it evaluates the transfer of the model to another region in Germany.

The RECOVER model could have the following limitations: (1) It is possible that not enough partners from the outpatient sector participate in the model with regard to network formation; (2) It is possible that patients at level 3 in particular already are too impaired for placement in outpatient psychotherapeutic care; (3) With regard to the sustainability of RECOVER, there is a need to introduce treatment models into standard care that are currently internationally evidence-based but are not yet part of mainstream care in Germany.

Successful confirmation of the efficiency and effectiveness of the RECOVER model can make theoretical, clinical and societal contributions. First, the findings will generate new knowledge about stepped care service models, effective integrated therapy models and therapies as well as efficient care processes. Specifically, the integration of e-mental health will help to increase acceptance and use of digital diagnostics and therapy. Second, the proof of effectiveness and efficiency creates all the prerequisites to transfer the model into standard care. How this can be achieved is already the subject of intensive cooperation between the developers of the RECOVER model and the participating health insurance funds. Third, the proof of effectiveness and efficiency, together with the accompanying research and experience with the transfer of the model as well as the 12 quality assurance manuals create optimal prerequisites for the further transfer of the whole model or essential components into other German regions.

Ethics and dissemination

This study has obtained ethics approval from the ethics committee of the Hamburg Medical Association (PV5672).

The written consent of all participants will be obtained and they will receive a detailed explanation of the study objectives, the voluntary nature of their participation, their right to withdraw their participation and the risks and benefits of the study.

RECOVER is a care model that should not cause any physical or psychological harm to participants. In the event of an unforeseen problem or if the participant experiences inconvenience or anxiety while filling out the questionnaire or answering the questions in the interview, the researcher will report this to the head of data collection. The researchers will help the participants to get additional support from experts. Participants can also choose not to answer the questions or stop the interview. Participants are asked to sign two copies of the informed consent form, one to be given to the participants and the other to be returned to the principal investigator of this study for recording purposes. The consent forms will be kept separate from the data. All data collected, without personal names, will be stored in the locked cabinet of the principal investigator (PI), while all digital or electronic records will be password-protected and kept in the PI computer for 5 years. Only the PI and the local co-chairs of this research project have access to the original experimental data set for research purposes only.

The current RCT will improve our understanding of the impact of RECOVER on the results of service users, especially as far as they are concerned:

Demonstrate the benefits and unintended consequences of recovery-oriented, strength-based services for people with mental illness.

Highlight the key therapeutic ingredients of RECOVER and how they affect outcomes.

Review how you can best use RECOVER in Germany.

Post-trial care of the study participants is ensured by the possibility of further treatment in the standard care setting.

Our dissemination policy aims at several important target groups. To share our knowledge with service users and their families, PI and the team will work with the local community and media. Healthcare professionals will benefit from the study's contribution to staff training and expert interviews. We will share our findings with researchers at home and abroad through conference presentations and publications in peer-reviewed journals. Our results are also disseminated through seminars organised by the PI Department and RECOVER websites.