Enablers and constraints influencing implementation of a novel, multi-site community-based frailty programme: perspectives of leaders and implementers

Abstract

Background

Evaluations of integrated care models for frailty anchored by comprehensive geriatric assessment have reported mixed effects on patient outcomes. Various reasons have been proposed, including how implementation varies widely across different contexts. This paper aims to identify the key enablers and constraints that influenced the implementation of a novel community-based frailty programme– the Geriatric Services Hub– in the rapidly ageing nation-state of Singapore through the perspectives of programme leaders and implementers from five sites.

Methods

Seventy-four programme leaders and implementers were recruited for in-depth interviews and focus group discussions. The Framework Method was used to summarise and compare data across different sites and perspectives.

Results

The findings on enablers and constraints were organised into four domains: System, partnership, programme and patient factors. Systemic factors include differing subsidies across public and private healthcare providers, systemic constraints of primary-care partners, a fragmented system of care, and the COVID-19 pandemic. Partnership factors include how capability-building for primary-care and community partners was welcomed by stakeholders. At the programme level, ensuring physical, financing, and timely access to services by frail older adults was prioritised. However, as noted under patient factors, some GSH patients faced complex medical, social, and financial issues that may affect service utilisation.

Conclusions

Systemic factors had a major influence on the implementation of GSH and we posit that they hold particular relevance for integrated care models in complex healthcare systems. Whole-system changes above and beyond implementation efforts of individual programmes may be needed for successful and sustainable implementation of integrated care.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12877-025-06164-5.

Background

The world is ageing rapidly, with the Asia–Pacific region at the forefront of this trend [1]. Frailty is a state of increased vulnerability due to age-associated decline in function and reserve [2, 3]. Globally, the average prevalence is 10.7% [4]. In Singapore, a small island-state set to become a “super-aged” society in 2026, prevalence is between 2.5% to 6.2% of older community dwellers. Another 32% to 46% are pre-frail [5–8].

Frailty can potentially be delayed and even reversed [9] with appropriate interventions. Early identification and intervention among older adults may support their independent living in the community [10] and delay their use of secondary and tertiary care [11].

Integrated care for frail older adults

The increased vulnerability of frail older adults exposes them to sudden health status changes when faced with minor stressors [12]. This may result in frequent care transitions between different providers and care settings [13]. Integrated care models that coordinate clinical care across different providers and levels may thus cater to their multi-faceted and changing needs [14]. Many frailty care models are anchored by a comprehensive geriatric assessment (CGA), which involves inputs from a multidisciplinary team on different aspects of well-being, including medical, psychological, social and functional needs. It considers medical diagnoses, functional impairments, environmental and social issues of older adults through the use of validated geriatric scales and tests, and ultimately produces an integrated and coordinated care plan [15–18]. In integrated care, CGA is combined with other components, such as a single entry point, comprehensive assessment and individualised care plans (ICPs), case management and coordination across the continuum of care, supported by integrated electronic information systems, and clear policies and processes on eligibility and patient care [19, 20].

Reviews on community-based integrated care programmes with CGA had reported mixed or uncertain effects [21–23]. Possible reasons included different variations of intervention components [23] and the diverse healthcare settings and target populations of such programmes [18, 21]. Furthermore, the implementation of CGA varied widely across different cultural and healthcare contexts [24].

Factors that affect implementation in different contexts should be studied more carefully. A systemic integrative review on CGAs in community settings [21] found the lack of partnership alignment and feedback, poor acceptance of preventive work, and challenges in operationalising and optimising CGAs to be obstacles in implementing CGAs. The facilitators were perceptions that CGAs were performed by highly skilled staff and resulted in holistic assessment and education, and improvements in care coordination and convenience. A scoping review on barriers and facilitators for integrated care for older and frail populations categorised them into macro (external context), meso (organisation), and micro factors (provider/patient) [25]. They ranged from health system instability (macro) to shared values and understanding among staff and teams (micro).

However, most of these studies were conducted in Europe and North America, with just a few from the Asia–Pacific region. Two studies from Australia and New Zealand was found that CGA, particularly home-based CGA whereby the older adults could be observed in their living environments, were viewed positively by healthcare professionals as enhancing their assessment of their patients’ health status and needs. However, there may initially be some uncertainty about their roles and duplication of work across different services [26, 27]. As health care beliefs, utilisation and care delivery systems differ across regions, studies from the Asia–Pacific provides another perspective and may generate insights directly applicable to this rapidly ageing region.

Healthcare in Singapore

In Singapore, the public sector dominates the delivery of acute care, while the private sector is the main provider of primary care. Intermediate and long-term care are mostly provided by voluntary welfare organisations [28]. Singapore’s rapidly ageing population and the larger chronic disease burden is placing considerable pressure on the hospitals. Major healthcare reforms are now underway to shift focus from acute care in the hospitals to preventive care in the communities. Subsidies for doctor consultations, diagnostics and investigations and medications are available for health services provided by the public sector if the patient meets the means-test criteria.

The Geriatric Services Hub

As part of this concerted shift, the Ministry of Health (MOH) Singapore funded a novel frailty programme for older adults in the community called the Geriatric Services Hub (GSH). It was piloted by five acute hospitals in collaboration with primary care and community care providers across Singapore from 2019 to 2023. The GSH is a multi-disciplinary care model that identifies and manages frailty among those aged 65 years old and above in the community [29]. Older adults categorised as 4 to 7 (very mild to severe frailty) on the Clinical Frailty Scale (CFS) [30] were eligible to enrol into the 1-year programme.

To briefly describe the overall implementation of GSH, the five hospitals partnered primary care providers and community care organisations (CCOs) to first identify frailty among older adults based on CFS and geriatric syndromes in different settings, including an emergency department, polyclinics (which are one-stop primary care providers that provide a wide range of subsidised services, from management of acute and chronic conditions to allied health and pharmacy services), general practitioner (GP) clinics and day care centres. Those found to score 4 and above in the CFS were then further assessed by CGA conducted by the hospital teams (henceforth known as GSH core teams) and/or their community partners. Multidisciplinary meetings (MDMs) involving geriatricians, geriatric nurses and allied health professionals were held to discuss the findings and to formulate an individualised care plan (ICP) for each patient. If needed, patients would be referred to frailty-related services such as physiotherapy and dietician services. All sites have at least one designated personnel to coordinate care, such as helping patients make appointments with the relevant care providers, reminding them of their appointments and following up with them on a regular basis. In addition, the GSH core teams built capability of primary care and CCO staff to identify, care and manage frail older adults through different forms of training.

From April 2019 to December 2022, a total of 1,985 patients were enrolled in the GSH. The model and the community-based partners involved are described in detail elsewhere [29] and is summarised in Fig. 1.

Fig. 1.

Summary of the Geriatric Services Hub model

MOH laid out broad objectives for GSH, including providing early identification of frailty and CGAs, followed by care-planning and coordinated care through working with primary care and community partners; increasing older adults’ access to essential services and seamless care transitions; providing core services such as nursing support and therapy services; and building capability of primary care staff in frailty identification and management.

From these, each hospital conceptualised and implemented different configurations based on their existing networks, contexts and resources. Some hospitals adopted more than one configuration (summarised in Table 1). The table listed the number of sites that adopted each configuration for most of the pilot-study period. By the end of the pilot, there were a few changes as the GSH teams had adapted or planned to adapt their programmes in response to insights gained from the ongoing implementation and evaluation.

Table 1.

Different configurations adopted by five acute hospitals in collaboration with primary care providers and community care organisations

	GSH sited in polyclinics (2 sites)	GSH sited in general practitioner (GP) clinics (2 sites)	GSH sited in community care organisations (CCOs) (2 sites)	GSH led by a mobile geriatrician who move among partners (1 site)	GSH sited in community nursing posts (1 site)
Description	GSH team builds capability for polyclinic partners/supports polyclinic partners in frailty assessment and management and facilitates multi-disciplinary meetings (MDMs)	GSH team builds capability for GP partners/supports GP partners in frailty assessment and management and facilitates MDMs	GSH team builds capability for CCO partners to lead frailty assessment and management and facilitates MDMs	GSH team leads frailty assessment and management, facilitates MDMs and.develops ICPs	Geriatric-trained community nurses lead frailty assessment and management, facilitate MDMs and develop ICPs, with inputs from geriatrician
Core functions
Identify frailty & conduct CGA	Core team builds capability of doctors & nurses to conduct CGA and develop ICP	Core team builds capability of GPs to conduct CGA and develop ICP	Core team builds capability of on-site doctors and nurses, or the core team may conduct CGA and develop ICP	Deployed to support clinics and centres that are not well resourced with clinical personnel; Core team conducts CGA and develops ICP	Geriatrics-trained nurses lead CGA, facilitate MDMs and develop ICPs, with inputs from geriatrician
Frailty management	Polyclinics follow up with patients or polyclinics refer them to specialist outpatient clinics	GPs follow up with patients or refer them to specialist outpatient clinics	On-site doctors follow up with patients or may refer to polyclinic for treatment or onward referral to specialist outpatient clinics	GSH team refers patients to polyclinic for treatment or onward referral to specialist outpatient clinics	Community nurses refer patients to polyclinic for treatment or onward referral to specialist outpatient clinics
Care coordination and case management	Core team Community nurses	GP nurses Community nurses Community partners	Core team	Core team	Core team

A mixed-methods evaluation of GSH was carried out to study the development and implementation of GSH, the patient-reported health outcomes, and a healthcare utilisation and cost analysis using a propensity-scored matched comparator group [31]. The evaluation protocol is published elsewhere [29]. As part of the larger evaluation, this paper aims to identify the key enablers and constraints that influenced the implementation of GSH through the perspectives of GSH programme leaders and implementers who piloted different configurations of integrated care in Singapore.

Methods

Study design

We used a qualitative design grounded in pragmatic epistemology and an interpretive approach to understand people’s diverse interpretations of social phenomena through the lens of researchers who also participate in these social contexts [32]. The aim is to generate useful, applicable and practice-informing knowledge.

The study adopted the use of in-depth interviews (IDIs) and focus group discussions (FGDs). Combining different qualitative methods could potentially enhance the analysis of a phenomenon and widen its conceptualisation [33, 34]. Reasons for using multiple methods can be broadly categorised into: (1) pragmatic reasons, (2) to compare and contrast participants’ perspectives (parallel use), and (3) for data completeness and/or confirmation (integrated use) [35]. For this study, parallel use was the main reason. Different methods were used with different groups of participants to explore their specific perspectives and experiences. Interview was chosen for programme leads for its flexibility and the use of open-ended and semi-structured questions [32], which allow for elaboration and reflection [36]. In addition, interviewing these key informants at the start of the study helped elicit more detailed information that helped advance the evaluation team’s understanding of GSH before we approached the implementers. FGDs were then conducted with implementers who worked together to deliver the programme, to elicit their attitudes and thoughts through group communication and interactions [37]. FGDs were also conducted with small groups of programme leads and key personnel as an opportunity for joint reflection about their implementation journey thus far, and the way forward.

During initial implementation, we conducted IDIs with programme leaders to gain insight into their conceptualisation, planning, and implementation of the programmes. Thereafter, we conducted FGDs with the GSH core teams and their community partners to understand their implementation experiences. Towards the end of the pilot, we conducted another round of FGDs with programme leaders and key personnel to explore the sustainability and scalability of each programme. This study received ethics approval from the National Healthcare Group Domain Specific Review Board (DSRB Ref: 2019/00925).

Study participants and sampling method

We adopted purposive sampling to elicit a wide range of perspectives from those involved in designing and implementing GSH across all five sites, both the GSH core teams and their community-based partners, and members of the different healthcare professions. As this was an evaluation study, the sample size was determined by whether participants from different health professions and who played key and different roles were included. Emails were sent to programme leaders and GSH core team members to invite them to participate in the IDIs and FGDs. We asked the GSH core teams to help identify community partners whom they worked with to implement GSH, and the study team then sent email invitations to them for the FGDs. We outline the eligibility criteria and characteristics of participants in Table 2. Eleven of the seventy-four participants (14.9%; primarily the programme leaders) took part in two rounds of data collection.

Table 2.

Demographic characteristics of IDI and FGD participants

Initial Implementation				End of Pilot
IDIs with programme, clinical and administrative leaders (GSH leaders) (n = 14)		FGDs with implementers (n = 54; 11 FGDs)		FGDs with programme, clinical and administrative leaders (n = 17; 5 FGDs)
Eligibility criteria: • Involved in development, setting-up, management of the programme; • Involved in the initiative/policy-related decision on starting of the programme		Eligibility criteria: • GSH core team: Have been working/providing services within the programme for at least 6 months; Part of the GSH-funded team • GSH community partners: Ever heard of the programme; have referred or received referrals from the programme		Eligibility criteria: • Involved in development, setting-up, management of the programme; • Involved in the initiative/policy-related decision on starting of the programme
Age, years, mean (min– max)	47.9 (38–55)	Age, years, mean (min– max)	36.5 (24–64)	Age, years, mean (min– max)	46.3 (29–57)
Gender		Gender		Gender
Female (%)	9 (64.3)	Female (%)	44 (81.5)	Female (%)	9 (52.9)
Male	5 (35.7)	Male	10 (18.5)	Male	8 (47.1)
Profession		Profession		Profession
Doctor	9 (64.3)	Nurse	18 (33.3)	Doctor	12 (70.6)
Nurse	1 (7.1)	Doctor	12 (22.2)	Nurse	1 (5.9)
Medical social worker	1 (7.1)	GSH programme administrator a	6 (11.1)	Administrator	4 (23.5)
Administrator	3 (21.4)	Centre manager b	4 (7.4)
		Physiotherapist/therapy aide	4 (7.4)
		Medical social worker	3 (5.6)
		Occupational therapist	3 (5.6)
		Care coordinator c	3 (5.6)
		Dietician	1 (1.9)
Years of practice		Years of practice		Years of practice
< 20 years	5 (35.7)	< 20 years	47 (87.0)	< 20 years	7 (41.2)
20–29 years	6 (42.9)	20–29 years	5 (9.3)	20–29 years	6 (35.3)
> 30 years	3 (21.4)	> 30 years	2 (3.7)	> 30 years	4 (23.5)

^a GSH programme administrators oversee coordination, operations and administrative work in the GSH core team

^b Centre managers are staff from community partners who oversee the operations and management of individual centres or a group of centres

^c Care coordinators in the GSH core team work with different service providers to help patients access required services

Data collection

The topic guides for the IDIs and FGDs were developed based on the integrated care for frail older adults framework [38]. This analytical framework evaluates integrated care models across 15 factors. Twelve were identified to be relevant to the implementation of GSH and are described in Table 3.

Table 3.

Factors based on the Integrated Care for Frail Older Adults Framework

Factors	Brief description
Patient screening	Identification of frail older persons in the community
Multidisciplinary assessment	Performing a comprehensive, multidimensional patient evaluation and developing a care plan to meet identified needs
Comprehensive service package	Putting together a broad range of health and social care services to meet identified needs
Network relationships	Nature of working arrangements among and between institutions and providers, including information sharing
Care management	Planning of care, coordination of care and follow-up across time, place and discipline
Continuity of coverage and care	Provider’s ability to help patients access the broad range of health and social care services across different settings and providers
Seamless/Ease of transition	Patient’s ability to access the broad range of health and social care services and to navigate between different settings and providers
Teamwork	Roles and responsibilities of GSH core team members; ongoing communication and collaboration among the multidisciplinary group of providers
Patient-centred care	The extent to which clinicians and patients work together to make decisions and select tests, treatments and care plans based on evidence that balances risks and intended outcomes with patient preferences and values
Strategic planning	Stakeholder involvement in joint planning and community needs assessment
Funding mechanism	Structure of funding for health and social care
System outcomes	Overall responsibility for the intended outcomes

For the initial implementation, IDIs were conducted with programme leaders. FGDs were conducted with GSH core teams and partner providers. The IDIs were conducted by an interviewer (W.S.T or H.T.R.T), with a notetaker (P.H.J.T) present. The FGDs were facilitated by a main facilitator (W.S.T, H.T.R.T or P.H.J.T) and a co-facilitator (P.H.J.T or H.N.H.L). An observer was present for a few of the FGDs to take notes (H.N.H.L, Z.L.N or G.S.).

The FGDs with programme leaders and core team members conducted at the end of the pilot were facilitated by a main facilitator (W.S.T, H.T.R.T, H.N.H.L or P.H.J.T) and a co-facilitator (W.S.T, P.H.J.T or H.N.H.L). One observer was present for the FGDs (H.N.H.L or P.H.J.T).

Due to COVID-19, all IDIs and FGDs were conducted via the ZOOM video-conferencing platform, except for one interview that was conducted face to face. All topic guides are included as Appendices 1–3. Data collection approaches across all phases are summarised in Table 4.

Table 4.

Summary of data collection approaches across all phases

Data Collection Phase	Participants	Objectives	Data Collection Method	Data Collection Period	Data Collection Tools
Initial implementation	14 key programme leaders	Explore the perspectives of key programme leaders on the factors that influenced the development and implementation of the GSH	In-depth interviews (IDI)	June to October 2020	Interview guide adapted from Kodner and Kyriacou (2000)
	54 GSH core team and partner providers	Understand the implementation experiences of the GSH core teams and partners, as well as the key enablers and constraints of implementation	Focus group discussions (FGD)	December 2020 to May 2021	Interview guide adapted from Kodner and Kyriacou (2000)
End of pilot	17 key programme leaders and core team members	Explore the perspectives of key programme leaders on sustainability and scalability	Focus group discussions (FGD)	May to July 2022	Interview guide adapted from Kodner and Kyriacou (2000), with questions on sustainability and scalability

Data analysis

We used the Framework Method to generate important categories and themes on factors that have influenced the development and implementation of GSH [39, 40]. It is a form of thematic analysis that facilitates the summarisation and comparison of data across different cases through using a matrix [39].

The inductive approach was used for coding of data. Preliminary broad codes derived from inductive coding and discussion of transcripts by four coders (W.S.T, H.T.R.T, H.N.H.L and P.H.J.T) resulted in a code book for the first round of data collection (IDIs with GSH leaders). The rest of the transcripts were subsequently coded by two coders (H.N.H.L and P.H.J.T) who had regular discussions to adjudicate differences and refine the code book. Additional codes and categories from subsequent data collection rounds were added as they emerged. The team kept an audit trail to enhance dependability of the analysis.

Two levels of analysis were conducted by the evaluation team. At the first level, a matrix was created to compare codes and categories from transcripts across the five sites. Key themes were generated through reviewing the matrix and making connections within and between the codes and categories, while being guided by the research objectives. The team further deliberated on the inter-relationships among the themes. Analysis of the implementation of GSH at each site was shared with two senior geriatricians from an acute hospital that was not involved in GSH. They served as steering committee members for this project and gave inputs based on their clinical expertise and knowledge of the local healthcare context. In addition, the study team met the GSH core teams periodically to obtain informant feedback. At the second level, findings from all phases, sites and participants were combined and synthesised to generate the common enablers and constraints observed across all sites, as reported in this paper. All analyses were performed using QSR NVivo software Release 1.0 (QSR International Pty Ltd., 2020).

Researcher team and reflexivity

Data collection, coding and analysis were conducted by researchers from diverse backgrounds such as social sciences, education, public health, economics and healthcare research. All were trained in conducting qualitative research and possessed knowledge of the healthcare sector. The heterogeneity in their training and work experiences contributed to a diversity in perspectives. Analysis was done as a team to leverage on the diverse viewpoints and to reduce the risk of bias. They had no prior relationships with the participants. In this evaluation, the programme leads and administrators helped in the recruitment of participants and provided feedback to the final manuscript, but the study team had no dependent relationships with them. Moreover, feedback from the steering committee and all GSH core teams helped counter personal, team or site-specific bias and potential gaps of understanding in the interpretation of results.

Results

From the analysis, we derived 4 broad domains that describe the enablers and constraints of implementing GSH: (1) System; (2) Partnerships; (3) Programme; and (4) Patients. These are summarised in Fig. 2 and described in this section.

Fig. 2.

Summary of broad domains and themes on the enablers and constraints of implementing GSH

Domain 1: System

This domain is on Singapore’s healthcare landscape, including the prevalent care philosophy, practices, and policies. Healthcare financing in Singapore is still geared towards supporting disease-centred care in hospital settings. Moreover, a differentiated subsidy framework across public and private healthcare providers used in Singapore results in differing out-of-pocket payments for GSH patients. GSH patients may have concerns about their ability to pay if private providers are involved in their care.

“Some patients (are) still refusing to go or don’t turn up et cetera that kind of thing. I think it’s comfort level. Patients are very familiar with (the public) hospital, very familiar with polyclinic…. Maybe patients are also thinking if I go there, find out something, then do I have to pay more cost again? You know what I mean? Because they are told that this is a private GP clinic.” (IDI 04, GSH Administrator, Hospital B).

Conducting CGAs in primary care settings was challenging due to systemic constraints of primary care partners, such as lean resources and high workload. General practitioner (GP) clinics were often not resourced to manage geriatric syndromes such as dementia or memory impairment. As such, GSH may not be compatible with their business model in the long term. For polyclinics, conducting the CGA in busy settings and without protected time was taxing for clinicians. CGAs often revealed unmet needs which were challenging to address during the average 10–15 min consultations. In addition, some diagnostics or investigations were not available in primary care settings.

“We’ve got so many people we can tap on right (in the hospital)…. But in the GP practice it’s just them. So they’re really, really lean, and we (in the hospital) have a lot of resources. So how lean they are outside is tricky, and the fact that the whole system doesn’t think in a function-focused way is a big, big barrier.” (IDI 04, GSH Administrator, Hospital B).

Operating a multi-site hub of different service providers in a fragmented system of care was a trying experience. Care coordination, or what one programme leader termed as “concierge work”, was “fragmented”. A large part of implementation had been about forging clinical and operational pathways that did not exist.

“I think GSH helps with setting the links, setting the framework. There’s more that can still be done (….) But I think it’s a good foundation to start with for patients who probably will be in the system running around different, different places and not being able to get the appropriate care. Now we have somewhere that’s a bit more centralised and (results in) quick access for them.” (FGD1 10, Primary Care Doctor, Hospital B).

The existing infrastructure and conditions in the healthcare system were less than ideal for implementing integrated care. GSH teams often encountered system-level issues. For instance, there is currently no common information-sharing platform for patient data across the public-to-private sectors as well as between medical and social care providers. A hospital geriatrician described information-sharing as ‘piecemeal” due to the lack of a common IT platform for communication between care providers.

“So it’s a bit piecemeal at times because we do not have direct access to what’s going on and we’re very dependent on people who are on the ground, who knows what’s happening, to inform us. Otherwise, we’re totally out of touch to be very honest. Yeah, but the good thing is when we do the MDM, because I’ve been going down and I have seen some of (their) patients, so when they tell me about the patients I'm actually quite familiar.” (FGD1 43, Geriatrician, Hospital D).

Despite an ageing population in Singapore, there seems a lack of understanding about frailty and its reversibility among healthcare professionals, administrators, older adults and the general public. Frailty identification and management are time- and resource-intensive and may not be prioritised in healthcare. This was even more so during the COVID-19 pandemic, when the impact on GSH was consequential. Programmes were suspended at various points due to increased infection rates. Some GSH implementers were redeployed to national or hospital-related COVID efforts. Capability-building efforts for partners were often suspended or delayed.

“And you’re right to say that sometimes we get deployed out, for example the more recent one was the deployment for mobile vaccination teams…. So on certain days I have to (be) deploy(ed) out of clinic. But I guess this is something that is beyond our control, if there’s a need we do have to move, and just stand this for a little while.” (FGD1 53, Community Care Doctor, Hospital D).

Domain 2: Partnerships

This domain looks at the importance of building strong, complementary relationships with partners with shared values, culture and vision and also standardising of work processes. Due to the continuum of GSH that spanned from the hospital to community, strong alignment of priorities among different levels of stakeholders was crucial. The levels were described by two participants as national, cluster (known locally as the three Regional Health Systems [RHS]), organisation (such as individual polyclinics), medical specialities (such as family medicine) and the individual clinicians. This entailed stakeholder discussions at different levels that were complex but necessary for different parties to align their priorities.

“Because this is a collaborative approach at the (polyclinic group) level…, And then this goes down to the individual polyclinic heads, and how they actually envision this. So there are many layers of collaboration. So many layers of discussion that comes in. To be honest it will take a minimum–at least 5 years or even longer to actually move this forward.” (FGD2 01, Geriatrician, Hospital C).

GSH core teams that built new working relationships from scratch with partners had to invest considerable time and effort, compared to those that tapped into community networks already established by their parent hospital or RHS. The former had to repeat their engagement efforts using a tailored approach to address the unique concerns of each partner.

“…so we have to provide to their [primary care partner] needs and then we have to actually engage them…. The reason why I mention all these things take time, you can’t just expect just start and then deliver in six months and one year. You have to grow with it [the GSH programme] because we are looking at a sustainable solution to this ballooning problem of frailty.” (IDI 11, Geriatrician, Hospital C).

Most primary care and community partners discerned the value of GSH and benefits conferred to their patients or clients: specifically, being holistically assessed using CGA, having direct access to geriatricians and support from a multidisciplinary GSH team in managing complex needs and navigating the healthcare system and community resources.

“Often the patients that see us live in the neighbourhood. Along the way, things crop up—their cognition deteriorates, their function for some reason deteriorates. In the usual GP clinic setting, being just the medical person, I'm not able to offer the rest of the allied health perspectives and assessments that they need. And putting them into GSH allows us to have a better overview of how they are doing, not just medically but also scoring them on their cognition, scoring them on their mood and things like that.” (FGD1 50, Primary Care Doctor, Hospital D).

However, due to their stretched resources and multiple priorities, a few polyclinics were reluctant to join during the pilot phase.

Capability-building for primary care providers and community partners was welcomed. The efforts ranged from dyadic talks and seminars to on-the-job training whereby community healthcare professionals were paired with a geriatrician, GSH nurse or therapist to conduct CGAs. During multidisciplinary meetings (MDMs), geriatricians and nurses coached and advised partners on formulating individual care plans. The geriatricians found the on-the-job training to be fruitful as partners were often keen to apply what they learned. Likewise, their partners perceived the training to have enhanced their clinical practice.

“Actually, it improved my way of assessing the patients. I used to not ask any problem with urination. If the patient said yes, then I will just tick yes. But now, I actually would ask further. ‘When did this happen? Why is this happening?’ I will try to find what is the reason causing it. So actually, it takes quite long to do the CGA.” (FGD1 09, Primary Care Nurse, Hospital B).

Domain 3: Programme

This domain examines programme-specific elements of GSH that affected its implementation on the ground. Attempts were made to keep GSH services affordable while still operating within the national financing framework. The GSH teams worked with MOH to provide direct access to selected and subsidised diagnostics and investigations at their parent hospitals (e.g., computed tomography scans and bone mineral density test). However, direct access to subsidised geriatric medicine (GRM) specialist clinics in hospitals was inconsistent across GSH programmes. Some GSH patients had to be referred to a polyclinic or a Community Health Assist Scheme (CHAS)-affiliated GP or community provider first for another referral to the GRM clinics. All Singaporeans receive subsidies for medical and/or dental care at participating GP and dental clinics under CHAS. The importance of timely access to specialists was emphasised by a participant.

“So, you see, to get appointment to see a (hospital) geriatrician right, three months [which is standard average waiting time] you know. And within these three months we have to hold the resident? (…) If let’s say we identify any patient with cognitive issue, if we realise it’s probably more of mental health so we want to seek appointment, while waiting for that appointment right we got to hold them [in the meantime] you know.” (IDI 02, Nurse, Hospital A).

As patients were cared for by multiple partners, care coordination and continuity were an ongoing concern. All programmes had designated personnel to coordinate care. They made appointments with care providers for patients, reminded them, and followed up with them regularly. A related function is case management of selected GSH patients with complex needs and multimorbidity. As case management was not designated a key GSH component, it was often contingent upon the community partners or the regional hospital to provide this service. It was unclear whether this is sustainable if GSH were to be scaled up.

“…we have not yet explored having a social worker per se. We have been working closely with community nurses who does all the care coordination and management. Of course, they have their resources to link up with their various community partners out there if there is a need.” (FGD2 11, GSH Administrator, Hospital B).

On the tracking of patients, a few polyclinic doctors commented on a lack of updates on patients referred to other care providers. It was hard for them to gauge if their care plans had been implemented as planned and led to improved outcomes. The main reasons are likely a shortage of personnel, as well as a lack of data-sharing mechanisms among partners.

“For example, if we refer them to day rehab, we wish if somebody was there (to see) that this patient is following up regularly. Or if there is some transport issues which somebody can arrange and follow up with that. That would be nice because the reason why it’s very difficult to follow up in the past is as I said, our nurses, our CC (care co-ordinators) are meant to be in polyclinic rather than the follow-up in community. So we might not have that bandwidth to stretch that far. So that is one of the issues and one of the unmet needs, for the follow-up in the community.” (FGD1 26, Primary Care Doctor, Hospital E).

Inadequate resources, both in time and dedicated personnel, were a major implementation constraint for GSH. Many GSH implementers juggled multiple commitments. Apart from running GSH, the core teams were busy building partnerships and establishing new workflows and referral pathways. Moreover, they often lent support to resource-poor partners in conducting CGAs, care coordination or case management. In considering sustainability, some implementers commented that more personnel for specific roles would be required if GSH were to be scaled up.

“I guess the issues or foreseeable problems would be more of sustainability, when the numbers ramp up. Because not all the GP (clinics) there are case managers or social workers at the sites. So as much as we want to also upskill them to do case management for [GSH], it may be a bit difficult for this to happen because you just don’t have the manpower or this role, or this person suitable to do this.” (FGD1 42, Medical Social Worker, Hospital D).

As GSH was a novel programme piloted in the community, ensuring the physical, financing and timely access to services by frail older adults was a priority. Having multiple physical sites in the community and home visits conducted by community nurses expanded the catchment and enhanced access for those with mobility challenges or who were unknown to the hospitals.

“And she (the CareHub nurse) will do a more detailed report…so much detailed information compared to my assessment because my assessment is quite brief, based on asking. So for her, one, she’s actually very comprehensive, she will do assessment head to toe –, the home environment, patient compliance of medications, for example insulin technique, so she will be there to see whether how’s patient administration. Is it correct technique or not? How’s the caregiver coping? Are they coping well in the care?…. So when (the nurse) came in she really improved the whole project tremendously, I would say, with her input. So she will periodically update the doctors what is her finding and she will give suggestion to the doctors and whether patient need to go back to polyclinic earlier for review or not. So it definitely benefit(ted) the programme a lot.” (FGD1 32, Nurse, Hospital E).

One programme is led by geriatric-trained community nurses operating out of community-nursing posts or doing home visitations. Another two had created close linkages to community nursing.

The broad objectives laid out by the ministry for GSH allowed room for experimentation as the different teams operationalised frailty care in community settings using different configurations for frailty identification and management. For example, one programme leader’s view that primary care doctors should not be turned into geriatricians led to the development of a geriatrician-in-the-community model, which aligned with the “one-care-team” philosophy of care espoused by their parent hospital.

“(…) Meaning the patient stays with the same team, gets looked after by the same team but in fact post-discharge gets followed up by the same care team, it’s one-care-team approach. And this is actually very vital for managing frail elders, because you need a continuity especially when you take care of frail elderly. In the hospital and all the way through in the community. This is how the GSH will nicely fit in place.” (IDI 11, Geriatrician, Hospital C).

The sharing of ground feedback and interim evaluation findings among the policymakers, evaluation team and the GSH teams aided programme adjustments as well. MOH initially wanted to target older adults with CFS scores of 5 to 7. However, the evaluation team found there was a consensus among GSH leaders that CFS 4 (very mildly frail) patients would benefit from GSH.

“I'm aware that what we have is basically CFS 5 to 7, that was the target group. But we’re also seeing cases that are I think the pre-frail group…. Well, they need that screening done, they need the assessments done so that it’s a means of keeping them good, healthy and well. We should be targeting groups that are a bit– not at the level where they’re already frail, you know. We should be working a bit earlier on them.” (IDI 05, Geriatrician, Hospital B).

Discussions between MOH, the leaders and the evaluation team led to the inclusion of CFS 4 patients from October 2020 onwards. During the FGDs with implementers, we found they generally agreed with the move upstream to recruit CFS 4 patients.

Domain 4: Patients

This domain describes the characteristics, needs, perceptions and preferences of patients that influenced GSH operations, enrolment and service-utilisation rates. The implementers noted that patients’ lack of understanding of frailty and the need for intervention may lead to little motivation to use GSH services. Furthermore, as GSH is a community programme, some did not see the urgency to use its services.

“Because they (patient) are being seen at the community level, they may not see their condition as crucial as if they are being seen in a hospital setting…. When you go hospital it means like wow, very jialat (Chinese dialect for “negative”), very serious already. For the follow-ups to be at the community level they may think okay, maybe not so serious so I can then miss my appointments.” (FGD1 16, GSH Administrator, Hospital B).

One participant noted some older adults seem to lack the awareness or readiness to commit to rehabilitation.

“I think maybe not just GSH but just thinking broadly (about) all my patients. Somehow, I have this impression that a lot of the elderly patients would prefer a quick fix – when they see the amount of commitment that is actually needed to improve muscle strength which is not something that happens overnight or with a pill, a lot of them just want ‘Oh give me the medication that will help me and then it will solve this problem.’.” (FGD1 50, Primary Care Doctor, Hospital D).

Some GSH patients faced complex medical, social and financial issues that may affect service utilisation. Medical issues included cognitive impairment or dementia, which reduced their ability to communicate and adhere to recommendations. Social challenges included living alone with no caregiver to manage their care.

“…we have some clients, they don’t have any NOK [next of kin]. So they don’t have any family members or kin, they are staying all alone. And then like getting appointment for their forgetfulness or to let’s say rule out dementia, they do not agree.” (FGD1 12, Nurse, Hospital A).

Not only that, when such problems were uncovered as part of the CGAs, clinicians worried that they were often “too complicated to be solved in one sitting”.

“And most of the geriatric problems we do understand (they’re) not a one-time issue, it’s going to be for a long time. We don’t know how long we can sustain things like that. I mean, what I think about it is, okay, this is a very good thing to start with but is it sustainable down the line? Is somebody going to support (this)? The reason being that polyclinic is very crowded. To get a CGA for one person is like (a) whole afternoon.” (FGD1 26, Primary Care Doctor, Hospital E).

Patients who faced financial hardship may also be less motivated to follow up. Concerns about out-of-pocket-payment may deter them from using GSH services. Medication costs were a concern for those receiving care in GP clinics. Initiatives to address these included subsidies for medications and fee waiver for patients residing in rental public flats. Charges for GSH consultations were subsidised in varying degrees at all sites, but usually only for the first few sessions. Physiotherapy and consultations with private GPs in the community were perceived as expensive in the long term, resulting in patient dropout.

“Then the second challenge is probably for repeated attendances. That means not many people might want to follow up long-term at (a private family clinic) for repeated attendances at the moment because of the OOP [out-of-pocket] factor. Because the OOP might increase on the fourth or fifth visit.” (FGD1 17, GSH Administrator, Hospital B).

Suggestions from implementers on working with constraints of patients include building rapport and trust with them through asking them about their goals for improving their mobility and increasing the number of home visits for more meaningful engagement. Or it could even be to “speak the same language” and “sayang” (Malay for “to care about”) them, lest they feel that they were being “forced” to seek medical attention.

Discussion

Through the GSH leaders and implementers, we sought to understand the contexts and factors that influenced their conceptualising and implementing of GSH. In this paper, we focused on commonalities in their experiences of implementing the different configurations. From there, we distilled the high-level enablers and constraints of implementation. We will first discuss the findings that are aligned with existing literature, and then highlight new findings that emerged from this evaluation.

The design and running of a programme often leverage on the practical wisdom of practitioners, who can help identify the conditions or factors for successful implementation [41]. Such insider knowledge can help provide a richer account of how interventions are implemented as “events in systems” [42], rather than as an isolated activity [43]. Our findings support the view that complex interventions exist within and require the support of wider systems, processes and practices [43].

Conceptualising and operationalising a novel programme that is linked to national or organisational strategic priorities is imperative to garnering support for its implementation [44, 45]. GSH was conceived and implemented at a time when Singapore healthcare is undergoing a major paradigmatic shift to focus on preventive care in the community. With the introduction of Healthier SG, a preventive-care strategy whereby residents work with their family doctors to formulate a care plan that includes check-ups and lifestyle changes [46], primary care providers may receive additional support and resources for care coordination and case management. The government is also committed to enhancing IT infrastructure to improve data submission and sharing.

System constraints mostly stemmed from the larger structures and processes. During the pilot phase, the GSH programme leaders and teams drew upon their knowledge of the healthcare systems and their networks to plug existing gaps. But to achieve sustainability and scalability, systemic and organisational support structures need to be in place. The potential of such models to spread or scale up to other jurisdictions often depends on macro-level factors [47], such as the financing, structure and governance of the wider health system and the funding mechanisms. Whilst challenging to address at the programme level, such broad prevailing conditions must be taken into consideration when designing or spreading an intervention.

Successful integration often also hinges on provider-level factors, such as having a multi-disciplinary team with shared values and in regular communication [25]. Our study participants stressed the importance of shared values and engagement with their partners. For partnership enablers, most partners understood the benefits of participating in GSH, including the increased clinical competency of their staff. Most partnership constraints arise because of inadequate resources and personnel, and differing priorities. Communication among different providers was noted in literature as a major factor for implementing integrated care [25]. This was also mentioned in our study, but sometimes as a stumbling block due to the lack of robust data-sharing mechanism among different partners.

A large part of implementation success could be attributed to programme leaders’ ability to tap into existing resources to address operational gaps and to implement their approaches in different settings. The broad objectives laid out by MOH injected a degree of flexibility into how different configurations may be employed. No single model or approach has been shown to be the best for integrated care [48, 49]. Instead, as demonstrated by GSH, broad guidelines or policies that allow for autonomy and adaption of practices and procedures over time to suit the operational environment may be more appropriate [25, 48, 50].

As documented in existing literature, stretched resources and personnel for implementation were a key concern [25]. The ability to provide case management and follow-up with GSH patients was inconsistent across sites. While the evidence is uncertain on its effects on patients’ health, service outcomes and costs [51, 52], case management is associated with increased adherence to treatment guidelines and patient satisfaction [53]. It has also been identified as a key factor in facilitating communication and coordination among multiple providers, implementation of care plans, and improving access for patients [54].

Existing literature noted that patients are not always aware of the benefits of preventive care [55] or they hold views that frailty is unmodifiable [56] and is a part of normal ageing [57]. Similarly, in this study we found through this study (and interviews with 30 GSH patients, which were reported elsewhere) (Geriatric Services Hub Evaluation Team. Geriatric Services Hubs Evaluation Report - Final Report. Singapore: Geriatric Education & Research Institute; 2023. Unpublished) that older adults’ motivation to use GSH services or adhere to recommendations may be affected by their lack of understanding of frailty. Furthermore, the complexity of their health and living conditions had implications on service utilisation and implementation. An explicit patient-centred care (PCC) approach has been shown to support shared decision-making between healthcare professionals and patients, and to improve health outcomes [58, 59]. As part of PCC, clinicians consider patients as a whole person with needs, values and preferences [60]. At present, some healthcare professionals in GSH try to embody PCC through certain practices, such as discussing goal-setting with patients. While the constraints faced by patients are undeniably multifaceted, explicating PCC as a guiding principle and embedding it more systematically into the delivery of GSH services through funding and training may be one way to begin to address them.

Overall, our findings from this study are largely in line with the international literature, indicating that the realities of implementing integrated frailty care in Singapore often reflect that of other parts in the world. However, this study also offered a few new insights. While all factors at the macro, meso and micro levels that may affect implementation should be examined [47], we noted that systemic factors had a major influence on implementation of GSH. We posit that they hold particular relevance for integrated care models where care delivery spans across different organisations and actors in a complex system. This can be seen in the way systemic factors influenced factors in other domains. For example, the differentiated financing framework in Singapore resulted in concerns among some GSH patients about out-of-pocket payments. A fragmented care landscape with a lack of a common information technology (IT) platform also made data sharing among different providers more challenging, thus affecting care continuity.

The evaluation provided a snapshot of how acute hospital teams engaged with different levels of stakeholders to build multi-provider hubs. The intensive efforts in engagement and training that were invested in early implementation or even pre-implementation were highlighted by the GSH core teams as time-consuming but necessary to align priorities and care standards. Enhancing the awareness of frailty and its potential reversibility with suitable interventions not just among the older adults and the public, but also healthcare and social-service stakeholders, is also recommended.

Another emergent finding was how GSH teams leveraged their knowledge and experiences to overcome gaps in service delivery and work around certain systemic issues. Arguably, such programme-level adaptiveness may not scale in the long term. The key learning from GSH may be that an ecosystem with structures and policies that support integrated care is needed for successful implementation, over and above implementation efforts specific to the programme. Jurisdictions and organisations around the world committed to providing integrated care as a strategy to maintain the independence and health of their older population would need to find the political will to make whole-system changes, such as adopting a healthcare funding model that incentivises preventive care and having a well-functioning and secure data-sharing system across different service providers [61]. Free from having to navigate systemic constraints, healthcare professionals could then concentrate their efforts on providing high-quality, patient-focused care for older adults, thereby gaining a better chance of improving their outcomes.

Strengths and limitations

This study has a few strengths. It is one of the few to document the implementation of integrated care for frailty outside of North America and Europe, in so providing a needed perspective into community-based integrated care in the Asia Pacific. It combined the complementary perspectives of those who designed and led GSH and those who delivered care, allowing for a richer and wider account of the implementation experience. Moreover, experiences of multiple sites were compared to identify common enablers and constraints. As these factors prevailed across the different configurations and were similar to those found in international literature, we propose that they may be emblematic of integrated frailty care.

However, this study has its limitations. Due to the COVID-19 restrictions, the evaluation team was not able to carry out non-participant observations at the GSH sites as originally planned. Non-participant observation would have allowed researchers and evaluators to observe what participants might had been unable or unwilling to share due to various reasons [62]. This may in turn have led to a richer and more detailed description and understanding of the implementation process.

Despite this, triangulation was used to increase trustworthiness of the results [63]. We recruited a wide range of participants from different stakeholder groups. For implementers, we elicited their responses in FGDs where individual perspectives could be substantiated by others working within the same GSH. For the analysis, we adopted a team approach to minimise researcher bias in the interpretation process [64]. In addition, the interpretation of the results was presented to selected respondents from all five GSH sites for member-checking.

This paper focused on enablers and constraints at a higher level of abstraction that are common to the different configurations. The reason was to identify factors pervasive across integrated frailty care. Future studies could consider linking enablers and constraints more closely to the implementation process of different configurations, to further our understanding of how different models or configurations are operationalized and implemented in specific contexts.

Conclusion

Integrated care for frailty is a form of complex intervention with no standard model or formula for success. However, how it is being implemented in different contexts deserve a closer look. In the study of the implementation experiences across multiple GSH sites, we found commonalities in what helped and what hindered the teams as they employed different configurations of frailty care in the community. Besides building the evidence base on enablers and constraints, future research may wish to focus on the potential strategies to overcome the key constraints and leverage the enablers, and their efficacy in real-world settings. More attention to implementation of integrated frailty care is needed to improve patient care and outcomes in the Asia–Pacific and elsewhere.

Abbreviations

CCOs

Community care organisations

CGA

Comprehensive geriatric assessment

CHAS

Community Health Assist Scheme

FGDs

Focus group discussions

GP

General practitioner

GRM

Geriatric medicine

GSH

Geriatric Services Hub

ICPs

Individualised care plans

IT

Information technology

IDIs

In-depth interviews

MDMs

Multi-disciplinary meetings

MOH

Ministry of Health

PCC

Patient-centred care

RHSs

Regional Health Systems

Authors’ contributions

TPHJ designed the research, collected and analysed the data, and wrote and revised the article. THTR and HNHL designed the research, collected and analysed the data, and revised the article. LWS and DYY conceived the study, obtained the funding and revised the article. TWS conceived the study, obtained the funding, collected and analysed the data, and revised the article. All authors gave final approval for the manuscript to be published.

Authors’ information

Not applicable.

Funding

This work was supported by the Ministry of Health, Singapore [HSDP Project Number 19X01]. The funder has played no role in the study design or preparation of the manuscript.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Ethics approval was sought and obtained for this study. This study received ethics approval from the National Healthcare Group Domain Specific Review Board (DSRB Ref: 2019/00925). The study was performed in accordance with the Declaration of Helsinki. Informed consent to participate was sought and obtained for all participants in this study.

Consent for publication

Not applicable.

Competing interests

Some individuals in the Geriatric Services Hub Programme Group took part in the in-depth interviews and focus group discussions as study participants. However, they did not participate in the analysis of the data and interpretation of results, which were undertaken by the evaluation team. There were no dependent relationships between these individuals in the Geriatric Services Hub Programme Group and the evaluation team.