Table 3 Study design of the 17 integrated care programmes for individuals with multi-morbidity

 Health Network Tennengau (HNT)

Cross-sectional and retrospective quasi-experimental; PSM

Residents of Tennengau region in Salzburg receiving integrated care services from HNT, a network of social and health service providers and voluntary organisations

Residents of similar region in Salzburg, insured by the same regional health insurance fund as the intervention group, not treated by HNT

(1) Population-level claims data of all residents of Tennengau and comparator region; n~ 37,000 per group

(2) SELFIE-questionnaire administered once to clients of HNT with multiple chronic conditions and a sample of similar individuals of the comparator region; n~ 155 per group; data from (2) are linked to claims data

 Sociomedical Centre Liebenau (SMC)

Cross-sectional and retrospective quasi-experimental; PSM

Drug users receiving services by SMC, insured at the regional health insurance fund of the state of Styria

Drug users treated by other facilities offering usual care, insured at the regional health insurance fund of the state of Styria

(1) SELFIE-questionnaire administered once in intervention and comparator group; n~ 70 in intervention group and n~ 150 in comparator group; data from (1) are linked to claims data

(2) Individual-level claims data; n~ 70 per in intervention group and n~ 150 in comparator group

 GeroS

Prospective quasi-experimental; PSM

Geriatric patients in 2 homes for elderly that provide integrated care using specific modules to monitor and evaluate health needs and functional ability

Geriatric patients in 2 different homes for elderly that have not implemented the GeroS modules

(1) SELFIE-questionnaire administered at baseline and after 6 and 12 months; n~ 200 per group

(2) Data from (1) linked to data from health insurers, GPs, and social care information systems; n~ 200 per group

 Mobile Multi-disciplinary Specialist Palliative Care Team (MMSPCT)

Prospective quasi-experimental; PSM

Palliative care patients from 3 counties that implemented the MMSPCT

Palliative care patients from 3 different counties that have not implemented the MMSPCT

(1) SELFIE-questionnaire administered at 1st home visit and after 1 and 3 months; n~ 200 per group

(2) Data from (1) linked to data from health insurers, GPs, and social care information systems; n~ 200 per group

 Casaplus

(A) Cross-sectional and retrospective quasi-experimental; difference in difference analyses

(B) Prospective before-after study

(A) People ≥55 yrs. with multiple chronic conditions and a high risk of hospitalization, insured by Viactiv BKK, receiving case management incl. a mandatory risk assessment, individual education, a 24/7 crisis service

(B) People newly enrolled in the Casaplus programme described above

(A) People ≥55 years with high hospitalization risk insured by AOK receiving usual care

(B) No comparator group

(A) Claims data of all individuals enrolled in Casaplus in the years 2013–2018; n~ 1500 in the intervention group and max. 500,000 in comparator group

(B) SELFIE-questionnaire administered at baseline and after 12 months; n~ 200 per group

 Gesundes Kinzigtal (GK)

(A) Retrospective quasi-experimental; PSM

(B) Cross-sectional

(A) Residents of the Kinzigtal region insured by LKK/AOK enrolled in GK population health management

(B) Enrollees of GK that visit GP or specialist between Sept and Dec 2017

(A) Residents of the Kinzigtal region insured by LKK/AOK not enrolled in GK

(B) Residents of Kinzigtal not enrolled in GK that visit GP or specialist between Sept and Dec 2017

(A) 2005–2016 claims data of all LKK/AOK insured enrolled in GK and ~ 20,000 LKK insured not enrolled

(B) SELFIE-questionnaire administered once in both groups; n~ 300 in intervention and n~ 2100 in comparator group

 Onko Network

(A) Prospective quasi-experimental study; multi-variate regression

(B) Comparison of cohort before and cohort after Onkonetwork; multivariate regression

(A) Target population newly admitted to the hospitals that implemented OnkoNetwork, i.e., individual path management

(B) Cohort of individuals suspected of solid tumour in year after implementing OnkoNetwork

(A) Target population newly admitted to a hospital that had not implemented OnkoNetwork

(B) Cohort of individuals suspected of solid tumour in year before implementing OnkoNetwork

(A) SELFIE questionnaire administered at first suspect of cancer, at time of the Tumour Board meeting and 6 months after start treatment; data from electronic health records; n~ 300 in each group

(B) Data from medical systems before OnkoNetwork (sept 2014-aug 2015) and after OnkoNetwork (Dec 2015-Nov 2016); n~ 3600 in year before and n~ 3600 in year after

 Palliative Care Consult Service (PCCS)

(A) Prospective quasi-experimental study; regression + propensity score weighting

(B) Retrospective quasi-experimental study; regression + propensity score weighting

(A) Cancer patients with low performance status score for whom the PCCS is newly requested

(B) Metastatic cancer patients for whom the PCCS was requested

(A) Comparable cancer patients from the same hospital for whom the PCCS is not requested (some physicians refer to the PCCS, others don’t)

(B) Comparable metastatic cancer patients from the same hospital for whom the PCCS is not requested

(A) SELFIE questionnaire administered at hospital admission, hospital discharge and 1 month after discharge; data from electronic health records; n~ 80–100 in intervention and 200–250 in comparator group

(B) Hospital administrative and claims data from Jan 2014-Dec 2016; n~ 500–600 in intervention and 1500–2000 in comparator group

 Proactive Primary Care Approach for Frail Elderly (U-PROFIT)

(A) Prospective Regression Discontinuity design

(B) Re-analysis of cluster RCT extending the follow-up

(A) Frail elderly ≥75 living at home, identified by screening with U-PRIM who participate in U-PROFIT care programme

(B) Frail elderly ≥60 in the U-PRIM or the U-PRIM+U-PROFIT group of a cluster RCT

(A) Frail elderly just below 75 from the same GP practices living at home, identified by screening with U-PRIM who do not participate in U-PROFIT

(B) Frail elderly ≥60 in control group of cluster RCT not receiving U-PRIM or U-PROFIT

(A) (1) A questionnaire (with additional items from the SELFIE questionnaire) administered at baseline and after 12 months in each group; n = 480 in intervention and 130 in comparator group

(2) Data from (1) are linked to claims data

(B) Re-analysis of cluster RCT extending the follow-up for the claims data (from 2000 to 2016 instead of 2013); n = 790 in U-PRIM only, n = 1446 in U-PRIM & U-CARE, and n = 856 in the comparator group.

 Care Chain Frail Elderly (CCFE)

Prospective quasi-experimental, PSM

Frail elderly living at home with complex care needs and loss of control, from 3 primary care groups participating in a bundled care programme for frail elderly

Similar frail elderly from same region, receiving usual care from GPs of 1 the 3 primary care groups that not implemented the programme

(1) SELFIE-questionnaire administered to elderly at baseline and after 6 and 12 months in each group; n~ 200 per group

(2) Data from (1) are linked to claims data, data from electronic medical records and GP information systems

(3) CarerQol administered to related informal caregivers at baseline and after 6 and 12 months; n~ 100 per group

 Better Together in Amsterdam North (BSiN)

Prospective quasi-experimental, PSM

Individuals with limited self-sufficiency in multiple life domains referred for participation in BSiN programme

Individuals with limited self-sufficiency identified in the ‘Amsterdam Health Monitor’

(1) A questionnaire (with additional items from the SELFIE questionnaire) administered at baseline and after 6 and 12 months in each group; n~ 70 per group

(2) Data from (1) are linked to claims data from same period

 Learning Networks

Prospective quasi-experimental, PSM

Frail elderly referred to home care services or a short-term stay in a nursing home who are newly enrolled in a programme for whole, coordinated and safe care pathways offered by 11 municipalities

A similar group of frail elderly from similar municipalities who do not offer such a care pathway programme

(1) SELFIE questionnaire at 2 fixed time periods, 6 months apart; n = 300 per group

(2) Municipality-level registry information on centrality, staffing, economics etc. over the years 2017–2018

 Medically Assisted Rehabilitation Bergen

Prospective and retrospective quasi-experimental, PSM

People with opioid addiction participating in a programme integrating health and social care services of specialists and the municipalities in Bergen

People with opioid addiction participating in a conventional care programme in Oslo, Stavanger and Trondheim

(1) SELFIE questionnaire in Bergen at 2 fixed time periods, 12 months apart

(2) Data from Status report (SERAF) over 2016 and 2017; n = 300 in intervention group and n = 300 in comparator group

(3) National registry data over 2016 and 2017; n = 300 in intervention group and n = 300 in comparator group

 Barcelona-Esquerra (AISBE)

(A) Retrospective quasi-experimental population-based evaluation, PSM

(B) Cross-sectional programme-component evaluation

(A) Residents served by the Barcelona-Esquerra healthcare provider organizations that offer integrated care services for chronic patients across healthcare tiers.

(B) Patients admitted to the hospital at home/early discharge programme offered by Hospital Clinic

(A) Residents of the entire region and residents served by other provider organisations in the same region of Barcelona-Esquerra

(B) Comparable group of patients from a comparable hospital (Hospital Sagrat Cor) that does not offer hospital at home/early discharge

(A) Data from Catalan Health Surveillance system of 540,000 residents in AISBE over the years 2011 to 2017 and a similar number in the comparator group.

(B) (1) SELFIE questionnaire administered at 1 month and 6 months post-discharge; n = 200 per group

(2) Data from (1) are linked to data from electronic medical records of hospitals and primary care providers

 Badalona Serveis Assistencials (BSA)

Prospective and retrospective quasi-experimental, PSM

Individuals living in Badalona who participate in BSA’s integrated care programme for frail elderly that includes: (i) Early Discharge support; (ii) Long-term home-based support services and (iii) Residential care

For each of the three intervention groups, a corresponding control group was selected among individuals living in Badalona but attended by providers or living in residencies not included in the BSA program

(1) For service (i): SELFIE questionnaire administered at start of service and 3 months thereafter; n = 50 per group

(2) For service (ii) and (iii): SELFIE questionnaire administered once; n = 50 per group (service ii) and n = 100 per group (service iii)

(3) Data from (1) and (2) are linked to data from electronic medical records of hospitals and primary care providers

(4) For the evaluation of the BSA’s entire integrated frail elderly care approach: registry data from the Catalan Health Surveillance System over the years 2011–2017; n = 2000 per group

 Salford Integrated Care Programme (SICP)/Salford Together

(A) Retrospective quasi-experimental population-based evaluation; difference-in-differences analyses (using matching), exploiting gradual roll-out and geographical limits, and examining differential effect by multi-morbidity status.

(B) Retrospective quasi-experimental programme-component evaluation

(A) Individuals 65+ with long-term conditions that are eligible for the following 3 services by 1 clinical commissioning group, i.e., case management services and self-management.

(B) Individuals 65+ receiving case-management, community groups, a centralised telephone hub to help with navigating

(A) Entire population of 65+ in England and populations of 65+ from other geographical regions (i.e., other clinical commissioning groups not offering a similar integrated care programme) and other time periods

(B) Salford population 65+ with similar multi-morbidity not receiving case management

(A) Routinely collected population-level English NHS data (Hospital Episode Statistics and GP Patient Survey) over the years 2011–2016; n~ 35,000 65+ in Salford and n~ 9.3 million 65+ in England as a whole

(B) Re-analysis of data from CLASSIC cohort study over the years 2014–2015 including the core outcome-concepts of SELFIE; n~ 4000 65+ in CLASSIC; n~ 35,000 65+ Salford population

 South Somerset Symphony Programme (SSSP)

(A) Retrospective quasi-experimental population-based evaluation; difference-in-differences analyses (using matching if necessary), exploiting gradual roll-out and geographical limits, and examining differential effect by multi-morbidity status.

(B) Retrospective quasi-experimental programme-components evaluation

(A) Population of the Clinical Commissioning Group that offers the SSSP including complex care hubs of GPs in the hospital and co-location of health coaches in all GP practices

(B) i) Individuals using the complex care hubs

ii) Individuals in GP practices incorporating health coaches (enhanced primary care) as this was gradually rolled out in three waves

(A) Entire population of England and other geographical regions and other time periods

(B) (i) Propensity matched persons within South Somerset not using the complex care hubs

(ii) Practices act as controls until they roll-out the intervention

(A) Routinely collected population-level English NHS data (Hospital Episode Statistics and GP Patient Survey) over the years 2011–2016; n~ 115,000 (1500 with 3 or more selected chronic conditions that the programme initially focused on) in South Somerset and n~ 54.8 million (0,5 million with 3 or more conditions) in England as a whole

(B) Routinely collected population-level English NHS data (Hospital Episode Statistics and GP Patient Survey); (i) n~ 750 in intervention group and n~ 1500 in comparator group;

(ii) 19 GP practices joining intervention in three waves