Comprehensive geriatric assessments in integrated care programs for older people living at home: A scoping review

Abstract In many integrated care programs, a comprehensive geriatric assessment (CGA) is conducted to identify older people's problems and care needs. Different ways for conducting a CGA are in place. However, it is still unclear which CGA instruments and procedures for conducting them are used in integrated care programs, and what distinguishes them from each other. Furthermore, it is yet unknown how and to what extent CGAs, as a component of integrated care programs, actually reflect the main principles of integrated care, being comprehensiveness, multidisciplinarity and person‐centredness. Therefore, the objectives of this study were to: (a) describe and compare different CGA instruments and procedures conducted within integrated care programs for older people living at home, and (b) describe how the principles of integrated care were applied in these CGAs. A scoping review of the scientific literature on CGAs in the context of integrated care was conducted for the period 2006–2018. Data were extracted on main characteristics of the identified CGA instruments and procedures, and on how principles of integrated care were applied in these CGAs. Twenty‐seven integrated care programs were included in this study, of which most were implemented in the Netherlands and the United States. Twenty‐one different CGAs were identified, of which the EASYcare instrument, RAI‐HC/RAI‐CHA and GRACE tool were used in multiple programs. The majority of CGAs seemed to reflect comprehensiveness, multidisciplinarity and person‐centredness, although the way and extent to which principles of integrated care were incorporated differed between the CGAs. This study highlights the high variability of CGA instruments and procedures used in integrated care programs. This overview of available CGAs and their characteristics may promote (inter‐)national exchange of CGAs, which could enable researchers and professionals in choosing from the wide range of existing CGAs, thereby preventing them from unnecessarily reinventing the wheel.


| INTRODUC TI ON
Older people want to stay independent and live in their homes and communities until old age (Gillsjö, Schwartz-Barcott, & von Post, 2011;Wiles, Leibing, Guberman, Reeve, & Allen, 2011). As the prevalence of multiple chronic conditions and disability increases with age, older people who live at home may suffer from limitations in the physical, cognitive, psychological, social and/or environmental domains of life (Hoogendijk et al., 2014;Lette et al., 2017). These limitations may challenge older people's social participation and independent living, and result in complex health and social care needs. There is growing evidence that integrated health and social care is a promising approach to address such care needs (Boult et al., 2009;De Bruin et al., 2012;Gress et al., 2009;Hopman et al., 2016;Mattke, Seid, & Ma, 2007;Wagner et al., 2005). Based on the Chronic Care Model and related models, we define integrated care as those initiatives that proactively seek to structure and coordinate care for older people in their own home environments, centred around their needs (Barr et al., 2003;Boult et al., 2009;De Bruin et al., 2012;Epping-Jordan, Pruitt, Bengoa, & Wagner, 2004;Raleigh et al., 2014;Wagner et al., 2005).
Integrated care programs for older people living at home vary in types and numbers of intervention components. Most programs comprise frailty screening, multidisciplinary consultation meetings, case management, individualised care plans and follow-up contacts to monitor the status of older people and the implementation of personalised care plans (Boult & Wieland, 2010;Eklund, Wilhelmson, Gustafsson, Landahl, & Dahlin-Ivanoff, 2013;Hoogendijk, 2016;Looman, Fabbricotti, de Kuyper, & Huijsman, 2016). Another prevalent component within these programs to foster integration of care is the comprehensive geriatric assessment (CGA), also referred to as needs assessment, multidimensional assessment or geriatric assessment (Boult & Wieland, 2010;Eklund & Wilhelmson, 2009;Eklund et al., 2013;Hoogendijk, 2016;Kodner & Spreeuwenberg, 2002;Looman et al., 2016;Looman, Huijsman, & Fabbricotti, 2018;Oliver, Foot, & Humphries, 2014;Pilotto et al., 2017). The CGA can be defined as a multidimensional and interdisciplinary diagnostic process to identify older person's capacities, problems and needs, of which outcomes can serve as input for the development of a coordinated and shared plan for care and support and long-term follow-up (Rubenstein, Stuck, Siu, & Wieland, 1991). Through providing a uniform process for assessing needs, a CGA may promote shared understanding of older people's needs, common professional language and agreed-upon practices and standards, which are considered essential ingredients in integrated care (Kodner & Spreeuwenberg, 2002). A CGA helps health and social care professionals to understand problems and care needs older people experience, so that services professionals provide are matched to the needs and preferences of the people they serve (Hoogendijk et al., 2014;. Also for older people themselves, it is important to understand and acknowledge their (unmet) care needs, as this is expected to increase their involvement in and control over their care and support. This in turn may support them to keep living in their own home environments (Chen & Thompson, 2010;. Furthermore, a CGA enables proactive and early identification and management of care needs. Such a (preventive) assessment coupled with a plan for care and followup may have significant benefits for older people's well-being and independence including a reduction of hospital admissions, moves into long-term care, falls and long-term mortality (Beswick et al., 2008;Melis, Adang, et al., 2008;Stuck, Siu, Wieland, Adams, & Rubenstein, 1993).
In the wide range of integrated care programs for older people living at home, different CGA instruments and procedures (i.e. standardised processes that were adhered to when undertaking a CGA) for conducting a CGA are in place. Although some studies reflect the variety of CGAs in integrated care (Hoogendijk, 2016;Looman et al., 2018), a complete overview of CGAs instruments and procedures used in the context of integrated care for older people living at home is not yet available. Such an overview would provide insight into the characteristics of the various available CGA instruments and procedures for conducting them in integrated care programs, and allow for comparison between CGAs.
Additionally, for CGAs as an intervention component of an integrated care program, we hypothesise that the principles of integrated care are reflected in this specific component. However, it is yet unknown how and to what extent CGAs actually adhere to these principles. Such insights may help researchers and professionals with selecting a CGA from the range of existing instruments and procedures, and increase their awareness of differences between CGAs and their applicability in different integrated care contexts.
We therefore conducted a review of the literature in order to: (a) What is known about this topic • Integrated care programs for older people living at home often include a comprehensive geriatric assessment (CGA) to increase the understanding of an older person's care needs and preferences.
• A complete overview of CGAs instruments and procedures used in integrated care is lacking. Also knowledge about how principles of integrated care are being applied in CGAs is limited.

What this paper adds
• This study demonstrates that there is little overlap in CGA instruments and procedures used in integrated care programs, and shows their similarities and differences.
• This overview supports exchange of knowledge on CGAs available in integrated care between researchers and professionals, which could enable them applying existing CGAs in their own contexts. describe and compare different CGA instruments that are being used in integrated care programs for older people living at home, and procedures for conducting them, and (b) describe how the principles of integrated care are being applied in these CGAs.

| ME THODS
In order to explore, map and synthesise information on CGAs in the context of integrated care, a scoping review was conducted. The methodological steps outlined in the Arksey and O'Malley framework for conducting scoping reviews were followed to undertake such a review (Arksey & O'Malley, 2005). This process included five steps: (i) identifying the research question, (ii) identifying relevant studies, (iii) selecting appropriate studies, (iv) charting the data and (v) collating, summarising and reporting the results (Arksey & O'Malley, 2005).

| Identifying research questions
The research questions were as follows: In addition to the search in the electronic databases, relevant papers were identified through reference tracking.

| Selecting appropriate studies
Two reviewers (AS and PvG) independently reviewed the papers yielded by the search for their relevance by screening their titles and abstracts. When considered relevant by both reviewers, the fulltext paper was retrieved and reviewed for its relevance again. Any disagreement between the reviewers was resolved by consulting a third reviewer (SdB) to reach consensus.
In this study, three key components were selected, which were considered to represent the main principles of integrated care (De Bruin et al., 2012;Hopman et al., 2016;Kodner & Spreeuwenberg, 2002): (a) comprehensiveness: focus on problems and (care) needs in different domains of life (e.g. physical, cognitive, psychological, social and environmental), (b) multidisciplinarity: involvement of health and social care professionals from multiple disciplines and sectors and (c) person-centredness: active involvement of older people and their informal carers in decision-making and planning their care process, and putting their capacities, needs and preferences at the centre.
For the present study, papers were eligible if they met the following inclusion criteria: 1. The integrated care program itself as described in the paper complied with our main principles of integrated care: (a) comprehensiveness, (b) multidisciplinarity and (c) person-centredness; 2. The integrated care program aimed to address frailty or complex/ multiple health and social care needs in multiple domains of life which may benefit from an integrated care approach; 3. The target population of the integrated care program included people aged 65 years and older living at home. Living at home means living in their own homes or in some form of assisted living. Also papers that focus on people enrolled in an integrated care program that was started during hospitalisation and was continued after people's discharge in their home environment were included; 4. The integrated care program contained a CGA and consisted of multiple components or interventions (e.g. CGA is followed by designing and executing an individualised care plan); 5. The paper concerned a study protocol of the integrated care program or an intervention study evaluating the impact of a program.
These papers were expected to provide most extensive information about the CGA or provided a reference to a paper in which such information could be found. In case the CGA was described elsewhere, the paper which was referred to was also included in this study. A description included the CGA instrument and/or procedures for conducting the CGA. Procedures were the standardised processes that were adhered to when undertaking a CGA including established guidelines and experiences from practice.
Duplicate studies, papers not written in English or not retrievable and non-scientific papers (e.g. an editorial) were excluded.
For this study, the methodological rigour of the included studies was not evaluated (Arksey & O'Malley, 2005). This was in line with the principles of scoping reviews, as the focus of our study was to map existing literature on a specific topic (in this case CGAs in integrated care) rather than to evaluate the impact of a particular intervention.

| Collating, summarising and reporting the results
The findings of this review are presented using a narrative style as well as using tables to systematically present all specific details about the CGA instruments and procedures, and the incorporation of the principles of integrated care in the CGAs. The results section will first describe the output of the literature search. Then, the characteristics of selected integrated care programs will be described to provide insight into the context in which CGAs were conducted. Then, the results will report characteristics of the CGA instruments and procedures for conducting them, their differences and similarities and the extent to which they incorporate principles of integrated care.

| Study retrieval
A total of 972 articles were identified in the original literature search ( Figure 1). On the basis of their titles and abstracts, 128 publications were selected for full-text screening, of which 39 were included in this study. Thirteen papers were added after reference tracking to provide a more thorough description of the integrated care program and the CGA when the papers resulting from the literature search did not provide sufficient information. This resulted in a total of 52 scientific papers describing 27 integrated care programs and 21 different CGA instruments. For one integrated care program, we present both the original and the updated program (Kono et al., 2009;Kono, Izumi, Yoshiyuki, Kanaya, & Rubenstein, 2016). As the programs were rather different, they were included as two separate programs.

| Characteristics of integrated care programs
Many included integrated care programs were implemented in the Netherlands (n = 12). The integrated care programs were also implemented in the United States (n = 5), Canada (n = 4), Japan (n = 2), France (n = 1), Germany (n = 1), New Zealand (n = 1) and Sweden (n = 1) ( Table 2) The settings in which the integrated care programs were implemented varied from the primary care setting to home care organisations, hospitals and specialised centres (e.g. Veterans Affairs medical centres, community-based integrated care centres). Eighteen programs were implemented in primary care practices. In about half of the programs, all implemented in the primary care setting, the CGA process relied on a core team consisting of a general practitioner (GP; or other type of physician) and a nurse. The GP (or physician) was responsible for the care and treatment older people receive.
The nurse was often designated as case manager and responsible for coordination and continuity of care across different settings and providers of care. In many programs, the core team was supported by a multidisciplinary team of social workers, elderly welfare consultants, physical therapists, occupational therapists, dietitians, pharmacists, (specialist) nurses, psychologists, geriatricians, nursing home physicians, psychiatrists and other medical specialists.
The composition of a multidisciplinary team differed, ranging from a team of exclusively medical professionals to a team of an equal number of health and social care professionals. In most programs, the core and/or multidisciplinary team also consulted or referred to other professionals or community organisations providing services beyond the (professional) expertise of the core and/or multidisciplinary team, for instance medical specialists, social workers or meal delivery organisations.

| Instruments
Many different CGA instruments were used in the included integrated programs. In the 27 integrated care programs included in this study, 21 different CGA instruments were identified. All details with regard to the CGAs can be found in Table S1. Most programs developed or selected a unique CGA instrument, except for the programs using the RAI-HC/RAI-CHA, EASYcare instrument and GRACE tool.
These instruments (i.e. EASYcare instrument, RAI-HC/RAI-CHA and GRACE tool) were used in different included programs (Table S1 -column 'CGA instrument/tool'). The EASYcare instrument was used in five programs, all in the Netherlands. The RAI-HC/RAI-CHA was used in four programs, in the Netherlands, Canada and France. The F I G U R E 1 Flow diagram of literature screening process. *Reasons for exclusion other than the above, for example, duplicates, papers were not written in English, papers could not be retrieved or identified documents were non-scientific papers TA B L E 2 (Continued) GRACE tool was used in two US programs. The EASYcare instrument was either used as a standalone CGA instrument or as part Information provided about the exact content of the instrument varied between CGAs. For five CGA instruments, the content of the CGA was available in (the appendix of) the original study or in a study cited in the original study Buurman, Parlevliet, van Deelen, de Haan, & de Rooij, 2010;Fleischer et al., 2008;Ruikes et al., 2012Ruikes et al., , 2016Suijker et al., 2012Suijker et al., , 2016. For the other CGAs, the original study did not provide or only partially provided the exact content of the CGA nor referred to other papers describing this information.

| Procedures
Integrated care programs established procedures for conducting a CGA in order to enhance consistency in application. For a number of CGAs, papers explicitly mentioned that a structured format was followed for conducting the CGA, which means that instruments and questions were defined prior to the assessment and were preferably administered to each older person that was assessed (Table   S1-column "Procedures"). Some CGAs encompassed a basic assessment that was conducted for each older person, which could be supplemented with additional assessments when considered necessary and desirable by the involved staff, older person and informal carer.
Three CGA instruments were tested for their validity and reliability (De Stampa et al., 2014;Ruikes et al., 2012Ruikes et al., , 2016Van Hout et al., 2010;Vedel et al., 2009). For four CGAs, it was explicitly indicated that the instrument was tested for its feasibility (Daniels et al., 2011;Metzelthin, van Rossum, de Witte, Hendriks, & Kempen, 2010;Metzelthin et al., 2013;Ruikes et al., 2012Ruikes et al., , 2016Suijker et al., 2012Suijker et al., , 2016. For nine CGAs, the average duration of conducting the assessment was described, which ranged from 30 min to 3 hr. The location where a CGA was conducted differed between the integrated care programs. Most CGAs were conducted during a visit at the older people's home. In few programs, CGAs were conducted during hospital admission, via a telephone call, or during a visit at a clinic or centre. Information on whether or how (frequently) a CGA was used to monitor changes in functioning and care needs of older individuals was limited in the papers. For two programs, papers explicitly mentioned that the nurse involved used a digital instrument to conduct the CGA Muntinga et al., 2012;Van Hout et al., 2010).
Particularly professionals from Dutch integrated care programs attended a training program or (refresher) course on conducting a CGA and related topics, including care planning, goal setting, stimulating self-management, motivational interviewing, case management, interdisciplinary collaboration and the use of a computer for the assessments. Training was usually provided to the nurse responsible for conducting the CGA and acting as case manager in the program, although in several programs also other professionals, such as the GP or entire multidisciplinary team, attended a training program.

| Comprehensiveness: life domains addressed in CGA
As described in  Kono et al., 2009;Van Hout et al., 2010), finances (Fleischer et al., 2008;Suijker et al., 2012Suijker et al., , 2016, and providing informal care . In addition to addressing different life domains, some CGAs included a section in which concerns, priorities and goals of the older person and the informal carer were explicitly assessed. The different domains that were addressed during the assessment were described for all CGAs except six. Some papers described the domains and its subdomains in detail (e.g. psychological domain covers cognition, delirium, depression, anxiety and dependency), whereas most papers only mentioned the elements covered by the CGA without further distinguishing of domains and subdomains.
Furthermore, domains were defined in different ways in CGAs.

| Multidisciplinarity: professionals involved in CGA
CGAs differed in the extent to which they seemed to work in a multidisciplinary way during the inventory of the older person's situation and the development of the care plan. Depending on the integrated care program, the roles of professionals from different organisations and disciplines in conducting the CGA can be roughly divided into three different ways (Table S1-column "Multidisciplinarity"): (a) a nurse (practitioner) conducted the CGA, in some cases in collaboration with or under supervision of a GP, higher qualified nurse or social worker. In some of these programs, additional assessments by the nurse, GP or professionals from other disciplines were conducted when considered necessary, (b) all members of the multidisciplinary team conducted a part of the CGA which was related to their own professional expertise or (c) an assessor designated for performing the assessment conducted the CGA (but was not further involved in developing the care plan or providing care and support for the older person).
In line with the objective of a CGA to serve as input for developing a care plan, in each initiative, an individual care plan or recommendations for care and support were developed. They included identified problems, care needs, goals and actions to be taken, and were based on the outcomes of the assessment. The steps that were taken to develop a care plan differed between integrated care programs: (a) a nurse or GP formulated the care plan, (b) a nurse presented a draft care plan to the GP or multidisciplinary team to discuss outcomes of the CGA and the initial care plan before finalising it, (c) all core or multidisciplinary team members met to discuss the outcomes of the CGA and together formulated a care plan and decided on further action to be taken, or (d) professionals who were involved in the development of a care plan were not explicitly mentioned in the paper.
To sum up, healthcare professionals (GP, nurses, allied healthcare professionals) were always involved in the assessment of care needs and preferences and the development of a personal care plan. Social care professionals (social services, community organisations) were less frequently involved in these stages.

| Person-centredness: involvement of older people and their informal carers in CGA
Across the programs, older people and their informal carers were involved in different ways in conducting the CGA and developing the care plan (Table S1-column "Person-centredness"). CGAs generally aimed to take needs and preferences of the older person into account in the processes of conducting a CGA and developing the care plan, although the manner of working in a person-centred way varied between CGAs: (a) in the majority of the programs, older people were encouraged to share their concerns, priorities and goals they would like to achieve. As such, professionals were supposed to develop care plans or recommendations in a way consistent with older people's needs and preferences, and (b) in a few programs, professionals developed care plans or recommendations based on the outcomes of the CGA without explicitly involving older people and their informal carers in conducting the CGA and developing the care plan.
After formulating the individually tailored care plan, in almost all programs, the professionals discussed the specifics of the care plan with older people and collaborated with them to review and modify it. In two of these programs, the care plan was only implemented after final approval by the older person (Looman, Fabbricotti, & Huijsman, 2014;Looman et al., 2016;Spoorenberg et al., 2013Spoorenberg et al., , 2018Uittenbroek et al., 2016). In seven programs, older people were responsible for implementing activities and recommendations that were established in the care plan (Boult et al., 2013;Bouman, Van Rossum, Ambergen, Kempen, & Knipschild, 2008;Boyd et al., 2008;Faul et al., 2009;Fleischer et al., 2008;Ploeg et al., 2010;Rogerson et al., 2006;Suijker et al., 2012Suijker et al., , 2016. To encourage active involvement of the older person and informal carer in the decision-making and planning of their care process, some papers explicitly indicated "strategies" to stimulate professionals to do so during conducting a CGA: building trust with older people and their informal carers (for instance by making an introduction visit), allowing older people time to talk and using motivational interviewing techniques (for instance summarising and validating given answers). Also developing a user-friendly care plan could be part of this. In addition, some programs aimed to empower older people and their informal carers by providing self-management support including coaching, advice, education and health promotion materials.
Altogether, the majority of CGAs seemed to address comprehensiveness, multidisciplinarity and person-centredness, although not all CGAs reflected all three principles of integrated care to the same extent. Some CGAs reflected one of the principles less extensively than other CGAs. The CGAs conducted in the integrated care programs Embrace (Spoorenberg et al., 2013(Spoorenberg et al., , 2018Uittenbroek et al., 2016) and IMPACT (Tracy et al., 2013) seemed to incorporate the three principles most extensively, and as such showed how (extensively) the integrated care principles may be reflected in CGAs.
These CGAs aimed to cover a broad range of health and social care needs and problems in different domains of life, involve professionals from multiple disciplines and sectors-including social workersduring the CGA and the development of the care plan, and actively include the older person and informal carers in the decision-making and planning their care process.

| D ISCUSS I ON
The aim of this study was to describe and compare CGA instruments that were used in integrated care programs for older people living at home, and procedures for conducting them. There is a vast amount of literature describing CGAs conducted in integrated care programs.
Some of these studies address the heterogeneity of CGAs used in integrated care (Hoogendijk, 2016;Looman et al., 2018). However, a systematic overview of CGA instruments and procedures and of their differences and similarities was lacking. This scoping review addressed these gaps (Foreman, Thomas, & Gardner, 2004;Hermans et al., 2014), and is unique in focusing on older people living at home in the context of integrated care. In addition, it was unknown how and to what extent CGAs, being a crucial intervention component of many integrated care programs (Boult & Wieland, 2010;Hoogendijk, 2016;Kodner & Spreeuwenberg, 2002), reflected the principles of integrated care. We therefore also aimed to describe how principles of integrated care were applied in existing CGAs.
As shown in this study, many different CGA instruments were used to identify older people's problems, care needs and preferences. Most programs developed or selected a unique CGA instrument, except for the programs using the RAI-HC/RAI-CHA, EASYcare instrument and GRACE tool, as they were used in different programs included in this study. The reason for the variability of CGA instruments and procedures for conducting them in integrated care programs is largely unknown; arguments for developing new CGAs were often not presented. One possible explanation may be related to the limited knowledge about how older people's problems and care needs should best be assessed using a CGA. Existing CGA instruments and procedures may not satisfy requirements set by other researchers and professionals of the integrated care programs. As such, these researchers and professionals may develop alternative instruments which they, for instance, perceive to be more suitable for their specific context, or which explicitly involve the target population in the development of the instrument

| Principles of integrated care
Our results show that all CGAs in this study may be considered comprehensive, which means that, to a greater or lesser extent, the importance of addressing older people's needs across different domains of life seems to be recognised (Lette et al., 2017). Care and support tended to primarily focus on physical health problems and needs (Kuluski, Ho, Hans, & Nelson, 2017;Lette, Baan, Van Den Berg, & De Bruin, 2015). However, gradually a broader concept of health including for instance also psychological, social and environmental life domains is being adopted in policy making, research, education and practice. The CGAs identified in this study are reflecting this broader health concept as well.
As for the principle of multidisciplinarity, multidisciplinary teams and the CGA concept of multidisciplinary assessment and management may have different potential benefits for the older person (Johansson, Eklund, & Gosman-Hedström, 2010 With regard to the principle of person-centredness, there is broad consensus that older people should be encouraged to participate actively in the process of care and decision-making (Barry & Edgman-Levitan, 2012). In most programs, older people and their informal carers were actively involved in conducting the CGA and developing the care plan. However, preferences for the degree and the way of participation in decision-making may differ between older people (Belcher, Fried, Agostini, & Tinetti, 2006;Levinson, Kao, Kuby, & Thisted, 2005). Therefore, to actually work in a person-centred way, older people should be offered the opportunity to be engaged in the process of their care, in which their preferences regarding the role they wish to play in their care are paramount (Levinson et al., 2005).

| Methodological considerations
In this study, comprehensiveness, multidisciplinarity and person-centredness were selected to represent the main principles of integrated care. We realise, however, that integrated care includes more than these three components. Still, these selected components are considered common aspects of integrated care and are therefore most likely to be explicitly addressed in papers describing integrated care programs (Hopman et al., 2016;Kodner & Spreeuwenberg, 2002). The authors felt that the application of stricter criteria would mean that highly relevant papers would not appear in our selection.
We performed a broad literature search to collect as many relevant papers as possible. To perform this study, we conducted a scoping review instead of a systematic review. Compared to systematic reviews which address more precise questions, scoping reviews also require transparent and rigorous methods in their conduct and are considered specifically helpful for scoping a body of literature on a certain topic (Munn et al., 2018). Since this study aims to map existing literature on CGAs

| Recommendations for research and practice
This study provides an overview of available CGA instruments and procedures for conducting them, and insight into the extent to which they reflect principles of integrated care. Unfortunately, reasoning for the choice to select to develop a specific instrument mostly remains unclear. Still the overview in this study is useful in supporting the exchange of knowledge about CGAs between researchers and professionals, and helps them select a CGA appropriate for their context from the wide range of possible instruments and procedures. As such, this overview will enable researchers and professionals to make use of available CGAs where possible, in order to prevent them from unnecessarily reinventing the wheel. However, considering our methodological considerations, the current overview of available CGAs is not yet complete.
More thorough descriptions of CGAs, as well as experiences with using them, are recommended to complete the overview of available CGAs in integrated care.
Furthermore, knowledge about which qualifications a CGA should preferably meet is still limited. While this study showed the extent to which CGAs reflect principles of integrated care, there are also other key features which could determine the choice for a CGA.
These include, for example, the feasibility, validity and reliability of the CGA, whether users of CGAs have open access to the instrument and procedures or how easy it is to share CGA outcomes digitally between professionals. Also the extent to which CGAs is appropriate for monitoring changes in older people's functioning or living situation in relation to their care needs could be an important aspect.
Further research is recommended to gain more insight into the value of such additional qualifications of CGAs in the context of integrated care for older people, and whether the existing range of CGAs meet these qualifications. These insights could then guide the adaption of existing instruments or, if necessary, the design of new ones.
Additionally, little is known about the experiences and views of older people and their informal carers with specific CGA instruments and procedures. These are, however, important when assessing existing CGAs . Therefore, we would also recommend to obtain more insight into what is important for older people with regard to the content and procedures of a CGA, taking into account the wide variation in preferences of individuals.
Questions that could benefit from more research include, for example, whether older people consider a CGA relevant, and if so, what important themes are that should be addressed during a CGA (Kodner & Kyriacou, 2000;Lette et al., 2017), how long a CGA may last, who should be targeted (Stijnen et al., 2014) and which professional would be most suitable to build rapport between the older person and professionals Van Kempen et al., 2012)?

| CON CLUS ION
This study shows that many different CGA instruments and procedures for conducting them are in place, each with a different perspective. CGAs vary in the way and the extent to which they reflect principles of integrated care. Most CGAs seemed to address comprehensiveness, multidisciplinarity and person-centredness.
However, in some cases, healthcare professionals were predominant or older people and their informal carers were not explicitly involved in the process of conducting the CGA and developing the care plan. This overview of characteristics of available CGAs supports knowledge exchange between researchers and professionals, although more thorough descriptions of CGAs and experiences with using them are necessary to further complete the overview.
Sharing knowledge on CGAs available in integrated care could enable researchers and professionals apply existing CGAs in their own contexts.

ACK N OWLED G EM ENTS
This paper was published on behalf of the SUSTAIN consortium. The authors thank Wim ten Have for his help with the development of the literature search strategy and search in the different databases.
They also thank Lidwien Lemmens for her valuable comments on earlier versions of the manuscript.

CO N FLI C T O F I NTE R E S T
No conflicts of interest have been declared.

AUTH O R CO NTR I B UTI O N S
All authors contributed to the study concept and design. AS and PvG carried out study selection. SdB was consulted in case of any disagreement during study selection. AS and PvG extracted relevant data from the studies included and analysed the data. AS