Shared Care and Usual Health Care for Mental and Comorbid Health Problems

Overview

The aim is to study the impact on patients and health care by an adaption of the Hamilton Family Health Team model of shared care between general practitioners (GPs), mental health services and other primary or specialized services. The study is done at six GP centers with 30 GPs in three boroughs in Oslo, Norway. At each GP center with shared care one clinical psychologist from a CMHC works two and a half day a week and a psychiatrist 1-2 hours a week. They are available for the GPs for discussions, and they give assessment and brief treatment as requested by the GPs. Other primary health and social care and other specialized mental health or substance addiction services are consulted according to needs of the patients. The study is a cluster randomized controlled study of GP patients and their health care in GP centers with shared care (experimental group) compared with usual health care in other GP centers (control group). In each of three boroughs one GP center is randomized to shared care for 18 months while another center continues with usual health care. Register data on patients and services are extracted for 12 months (T0) before implementation of shared care and for 12 months (T1) with shared care after 6 months implementation. The register data on patients and health care are extracted from the GP patient records, mental health and substance abuse services, and NAV (social/welfare services). The cohorts at T0 and T2 include all patients 16-65 years old seen by the GPs. Patient-reported mental health, functional impairment due to health problems, overall quality of life and satisfaction with health services are studied for a sub-sample of the register cohort at both T0 and T1. A qualitative sub-study of experiences with collaboration, the shared care model, implementation of the model, facilitators and barriers is done at the end of the 18 months period of shared care. The informants include patients, GPs and coworkers, and managers and professionals involved with shared care. The study protocol was approved by the regional and national committees on medical and health research, as well as by the data protection officer at health trust. Progress and status are described in Detailed description. Data analysis starts in September 2018.

BACKGROUND There have been various forms of collaboration in Norway between community mental health centres (CMHCs), GPs and primary health and social care. But with a lack of research on such models, there is little knowledge on whether they contribute to better health care, outcome and satisfaction, e.g. for the large group of patients with moderately severe mental and comorbid health problems mostly seen by GPs. There is a great need for clearly defined collaboration models and for research-based knowledge of their feasibility and impact in the context of the Norwegian health services. Reviews of studies on collaboration show inconsistent results, including insufficient evidence to demonstrate significant benefits from shared care (Smith 2009), modest reductions in primary care consultations and mental health referrals (Harkness 2009), partly improved depression management (Butler 2011), and research primarily on simpler models of collaboration between two professionals (Hviding et al. 2008, Craven et al 2006). The Family Health Teams model developed in Hamilton, Canada, is a promising model of shared care (Kates 2011a, 2011b). It is built around GP practices. Each team include one or more GPs, nurses, mental health counsellors, a visiting consultant psychiatrist, and some other part-time staff (pharmacist, nutritionist, physiotherapist, occupational therapist). The model has increased access to care for persons with mental health and addiction problems, reduced waiting time, reduced referrals to outpatient mental health services, increased patient satisfaction, improved patient health, improved communication and co-ordination, and increased GPs' confidence and skills in treating mental health problems. CONTEXT The study is done in three Oslo boroughs (Alna, Grorud, Stovner) with a total population of 108 000. They are served by a total of 85 GPs in GP centers, by the primary health and social care in the boroughs, and by a community mental health center (CMHC), child and adolescent mental health services (CAMHS) and other departments of the Division Mental Health Services at Akershus University Hospital. INTERVENTION The intervention is an adapted version of shared care with close collaboration by services and professional groups, mainly localized in three GP centers. At each GP center one clinical psychologist specialist from the community mental health center (CMHC) works two and a half day a week, and a psychiatrist from the CMHC work at each GP center 1-2 hours a week. The psychologist and the psychiatrist are available for the GPs for consultation and discussions, and they give assessment and brief treatment to patients as requested by the GPs. Other primary health and social care and other specialized mental health or substance addiction services are involved or consulted according to needs of the patients. The decisions on the adaption of the model were taken by the collaborating services themselves in October-December 2015 after a joint visit to Hamilton by 14 persons and with input from the research group. This secured that the services had ownership and commitment to the model and contributed to make it feasible and useful. The shared care in the three experimental GP centers was established during the winter of 2016 and were operational for 18 months from the April/May 2016 to October/November 2017. The implementation was supported by discussions in monthly meetings at each GP center for the professionals collaborating on shared care there, and in monthly meetings at the CMHC between the psychologists and psychiatrists in shared care, managers at the CMHC and the principal investigator of the research study. AIMS The aim is to study the impact of a Norwegian adaption of the Hamilton Family Health Team model of shared care between general practitioners (GPs), mental health services and other primary or specialised services. RESEARCH QUESTIONS 1. What are the characteristics of patients seen in the usual GP practices (ICPC-2 codes, mental health problems, comorbidity of mental and somatic health problems)? 2. How is the total usual health care for the GP patients from GPs, primary health and social care, and mental health and substance addiction services? 3. How is the patient-reported mental health, functioning and satisfaction with health services and collaboration in among patients seen in usual GP practices? 4. Is shared care associated with changes in access to mental health care, referrals to specialized mental health care, use of health services and distribution of patients across services? 5. What is the patient course and outcome (waiting time, sick leave, mental health, functioning, duration of treatment) with shared care compared to usual health care? 6. Is there a more integrated care for comorbidity (concurrent mental/somatic/substance addiction problems) in shared care than in usual health care? 7. How are the GP patients' experiences with shared care compared to usual health care? 8. What do health professionals in shared care experience as main advantages and disadvantages of shared care? 9. What do health professionals in shared care identify as facilitating factors or barriers for implementing shared care? METHODS CLUSTER RANDOMIZED CONTROLLED REGISTER STUDY ON IMPACT OF SHARED CARE The study is a cluster randomized controlled study of GP patients and their health care in shared care (experimental group) compared with usual health care (control group). In each of three boroughs one GP center is randomized to have shared care for 18 months while another GP center continues with usual health care. Register data on patients and health care are extracted for 12 months (T0) before the implementation of shared care and for 12 months (T1) with shared care (after 6 months with implementation of the shared care model). Retrospective register data on patients and use of services for the two 12 months periods (T0 and T1) are extracted from the GP patient records, Akershus University Hospital (mental health and substance addiction services for adults and youth), and NAV (social/welfare services). Patient cohorts at T0 and T1 include all patients 16-65 years old seen by the 30 GPs in 12 months. Based on earlier studies each cohort was expected to be 20-25 000 patients. In 2014 the 30 GPs referred 840 patients to outpatient mental health clinics and 190 to inpatient mental health wards. Based on this, significant changes in referrals to outpatient clinics (main primary outcome) are expected to be shown by comparing periods of 12 months. In the data analyses of the impact of shared care the experimental sites are compared with control sites at T1, as well as with all sites at T0, including a difference-in-difference analysis (research questions 4-6). Data analysis on usual health care at T0 will include descriptive analyses and analyses to identify and compare different patient subgroups or patterns of use of health care in this unique combined data set from the several services for a large group of GP patients (research questions 1-2). SUB-STUDY ON PATIENT-REPORTED HEALTH AND SATISFACTION Patient-reported mental health, functional impairment due to health problems, overall quality of life and satisfaction with health services are studied for a sub-sample of the register cohort at both T0 and T1. GP patient 16-65 years old are recruited during two weeks at each GP center with some assistance from research staff. Included patients give written consent and complete a brief questionnaire in Norwegian or English. Inclusion criteria and secondary outcome measures in the questionnaire are shown in the tables. The sample is a sub-sample of the register cohorts at T0 and T1. To identify a difference of 6 points on the mental health measure CORE-10 between T0 and T1 with 5% two-tailed significance and 90 % power 70 patients per arm is needed in a cluster randomized trial with ICC=0.03 and 12 GPs in each arm (Eldridge 2012, Schultz 2010, Moher 2010). Aiming to recruit 600-750 patients (20-25 per GP) will secure a more than large enough sub-sample for the main secondary outcome measure (CORE-10). Data analyses of patient-reported information will answer research question 3 and parts of research questions 4 and 5. QUALITATIVE SUB-STUDY ON EXPERIENCES WITH SHARED CARE A qualitative sub-study of experiences with collaboration in general, the shared care model, implementation of the model, facilitators and barriers is done at the end of the 18 months period of shared care. Focus groups are conducted with the GPs in each of the GP centers (experimental and control centers), and individual qualitative interviews are done with approximately 15 patients with experience of shared care, nurses and other co-workers at the GP centers, professionals involved with shared care in the primary health and social services, managers at the CMHC and the psychologists and psychiatrist from CMHC working in the shared care. This qualitative study is done in stead of a planned survey with a questionnaire to health personnel, as the number of potential informants would be too small for statistical analyses compared to what was expected when planning the study. Analyses of the qualitative data will answer the research questions 7-9. ETHICAL APPROVAL The study protocol was approved by the Regional Committee on Medical and Health Research Ethics Health Region South East 08 May 2014 (reg.no. 2014/435), by the National Committee on Medical and Health Research Ethics in Norway 10 November 2014 (reg.no. 2014/160) and by the Data Protection Officer at Akershus University Hospital (reg.no. 13/138). The National Committee on Medical and Health Research Ethics approved that the study fulfilled the necessary legal criteria to extract structured register data from the services without giving the patients information or asking for written consent. The approved protocol of 2014 is attached to this registration. PROGRESS AND STATUS OF THE STUDY Data extraction of the register cohorts was done from the GP electronic patient records in 2015 and 2017, from the mental health services at Akershus University Hospital in 2017-2018 and from the NAV (social/welfare services) in 2017-2018. Data collection from the two clinical cohorts at the GP centers was done in 2015 and 2017. Quality control and organization of data was done 2017-2018. Data analysis starts in September 2018.