Integrated Care Coordination for Managing Chronic Conditions: Views of Health Staff on the Implementation of a Program Using an Algorithm to Identify People at Higher Risk of Hospitalisation in Sydney,Australia

Abstract

Background

Integrated care interventions can improve patient outcomes and reduce the burden on acute health services, but need a strong evidence base to ensure their effectiveness. Understanding the meso and macro context in which care is delivered and determining whether patient needs are met are essential to successful implementation. Care coordination in New South Wales (NSW), Australia has evolved over time to meet the needs of an ageing population with chronic health conditions and multi-morbidity with the aim of reducing potentially preventable hospitalisations.

Objective

To examine how an integrated care coordination program was understood and implemented at state, district and clinician levels in NSW. The Integrated Care for People with Chronic Conditions (ICPCC) program was implemented statewide, however local implementation varied. Patients who were suitable for integrated care coordination were identified via a hospitalisation risk prediction algorithm and/or referrals from health professionals.

Methods

Understanding and implementation of ICPCC were assessed via interviews and a focus group with a range of health staff. Qualitative data were analysed using NVivo software and normalisation process theory.

Results

There was a strong sense of program coherence from management, clinicians and referrers. They viewed ICPCC as effective in coordinating care for patients at risk of hospitalisation and incorporating self-management at home. All health staff interviewed understood the program purpose and necessity, including the importance of achieving patient and systemic goals. Networking, linking services and program promotion were important, as was reporting on benefits. While the algorithm effectively identified previously hospitalised patients, it did not identify all suitable patients in the community with an increasing risk of requiring acute health care intervention. Referrals from health professionals familiar with patient needs and complexity were an important additional mechanism for patient selection.

Conclusions

There was a shared sense of coherence and understanding of the ICPCC program among health staff at the three levels of implementation within NSW. The program played an important role in assisting patients with a range of chronic conditions to access and benefit from integrated care coordination, while increasing their capacity to self-manage at home. Program intake via hospitalisation risk prediction algorithm plus referrals from health professionals familiar with patient needs and complexity can effectively identify those who may benefit from integrated care coordination.

Introduction

As the life expectancy of the population increases, so does the proportion of people living with multiple chronic conditions.^1,2 This has placed increased pressure on publicly funded acute health services to provide complex care and improve health outcomes. ³ With multimorbidity, there is an increased need to manage information between providers, coordinate care, and reduce care fragmentation. ⁴ Integrated Care aims to place people with complex needs at the centre of seamless care within the health system and its connections with social services, so patients can access care when and where they need it. ⁵ Person-centred care is an approach to planning, implementation, and evaluation that aims to provide mutual benefits to patients, families and health care provider partners. ⁶

The population at risk of potentially preventable hospitalisation (PPH) and emergency department (ED) presentations could be assisted through a model of integrated care management at home. ⁷ Care coordination aims to meet a range of systemic, provider and/or patient goals, ⁸ facilitating self-management, appropriate care, health outcomes and greater efficiency. ⁹ Evidence regarding care coordination effectiveness indicates it reduces care fragmentation and improves health outcomes^10,11; reduces ED presentations and length of stay ¹² ; and increases the efficiency of health care delivery. ¹³ However, critics emphasise the importance of also addressing the social needs of patients; highlight that a one-size approach does not fit all; acknowledge that a coordinated program does not fix health system fragmentation; and argue prevention should also be considered. ¹⁴ Information technology algorithms have been trialed to support clinical decision making and understand markers for intervention, ¹⁵ but the evidence for improving health outcomes for people with multiple chronic conditions is still emerging.^2,16 Assessment of care coordination effectiveness has been hampered by a lack of consistent definitions, interventions and measures. ⁸ This highlights the need for programs to be evaluated and have strong conceptual understandings to ensure effective management, design and delivery.^10,17 In integrated care, there are a plethora of complex interventions which have been scaled up or applied in different contexts. Hence, the need to understand the context in which care is provided, ¹³ including how it is defined by localised services. ¹⁰

New South Wales (NSW) has taken an iterative approach to provide integrated person-centred care to patients with chronic health conditions at risk of PPH. Encouraging patients to take a more active role in managing their own health has proven successful in models of integrated care.^18,19 From 2010, the NSW Chronic Disease Management Program provided care coordination and self-management support for patients with the five most common chronic diseases requiring support at the time. ²⁰ The program targeted patients with chronic illnesses assessed to have potentially preventable acute service use and entry into it was via health professional referral. In 2014, a review revealed that patients were satisfied, but the program did not reduce hospital admissions. ²¹ In fact, there was an increase in potentially preventable readmissions. ²⁰ At this time, the NSW Integrated Care Strategy was developed to reduce the costs of inappropriate and fragmented care across hospital and primary care services. ⁷ Aligning with these objectives, the Chronic Disease Management Program was redesigned into the Integrated Care for People for Chronic Conditions (ICPCC) program; aiming to integrate care while reducing fragmentation and duplication of services and incorporating a new algorithm to identify patients at increased risk of unplanned hospital admissions. ²²

Hospitalisation Risk Algorithm and ICPCC

In 2014, NSW Health (the publicly funded state government health service in NSW) developed complexity and vulnerability indicators of a patient’s increased risk of unplanned hospital admissions, which were incorporated into the Chronic Conditions Patient Identification Algorithm (CCPIA). This algorithm used linked routinely collected clinical and demographic data held in electronic databases to assess the suitability of patients for inclusion in integrated care programs. ⁷ The CCPIA was a simplified version of the Hospital Admission Risk Prediction tool developed in Canada based on research that stated clinical judgement and standalone checklists were not enough to predict future hospitalisation.^7,23 The NSW CCPIA assigned patients a score between 0-14 for their risk of hospitalisation in the next 15 months based on risk factors including: being ≥65 years old; having acute care hospital admissions and/or ED visits in the 6 months prior; and having chronic conditions, including chronic obstructive pulmonary disease, heart failure, liver cirrhosis, inflammatory bowel disease, diabetes and gastrointestinal obstruction. ²⁴ In 2019, CCPIA was integrated into the Patient Flow Portal (PFP) across NSW as a means of systematically identifying patients at high risk of readmission to be considered for enrolment in the program. Health care professionals (internal or external to NSW Health) could also refer patients with chronic conditions to ICPCC that could potentially benefit from care coordination, health coaching and care navigation.

From May 2019, clinicians at the South Eastern Sydney Local Health District (SESLHD) Integrated Care Unit assessed the suitability of patients, either identified via the algorithm in the PFP or referred from health professionals, for inclusion in the ICPCC program. If a patient was suitable for the program, ²⁴ they were offered enrolment if they consented. A Care Coordinator would then follow-up with the patient to assess their health goals and develop an individualised care plan.

Since ICPCC commenced, implementation has varied across the state, ²¹ indicating the need for consistency in the program components of care coordination, care navigation and/or health coaching. ⁹ As part of an integrated care evaluation framework in SESLHD in NSW, the Integrated Care Unit and researchers from the University of NSW (UNSW) collaborated to evaluate how the ICPCC program was conceptualised and implemented into practice. ²⁵ This included assessing the views of clinicians, health managers and health bureaucrats about the program and assessing the characteristics of patients enrolled in the program. The qualitative data analysis builds on previous studies assessing the views of providers and service users about interventions for reducing avoidable hospital admissions. ²⁶ The evaluation sought to: assess the understanding and implementation of the program at clinician, health district and state health department levels; and assess the characteristics of patients identified through the algorithm vs those referred to the program via other processes.

This manuscript presents the findings of the qualitative evaluation, with the aim of demonstrating how an integrated care program such as ICPCC can work for patients and health staff at the Local Health District (LHD) level.

Methods

Setting

In 2021, NSW Health provided acute health care for 8.1 million people, two thirds of whom lived in the Greater Sydney area. ²⁷ It delivered health care via 15 LHDs and 3 specialty health networks across the state; and SESLHD, as one of the six LHDs in the Sydney metropolitan region, provided care to a local population of 979 370 residents in 2021. ²⁸ Our study included NSW Ministry of Health staff involved in integrated care (referred to here as senior management) plus SESLHD staff who either worked within the Integrated Care Unit, or referred patients to integrated care programs delivered locally by that unit. Other chronic disease management services were available to SESLHD residents, some of which referred patients to the Integrated Care Unit, but most were single disease focused.

Care Coordination and Liaison with Primary Care

The ‘SESLHD Integrated Care Unit Care Coordinator Role and Responsibilities Guide’ used during ICPCC has a section on the ‘Importance of Liaising with GP/Primary Carer’, which describes the critical role of the LHD Care Coordinator in facilitating linkage to care and communication between General Practitioners (GPs) who work in primary care external to NSW Health and hospital staff/LHD Integrated Care. ²⁹ It acknowledges the central role that GPs have in the care of the patients enrolled in ICPCC and how care delivered in the community is integrated with the care coordination delivered by the LHD. The guide includes the following tasks for the LHD Integrated Care team to deliver: (i) care co-ordination between multiple health care system streams and GP/patient; (ii) engaging with patient/carer and GP for collaborative health goals and plans; (iii) collaborating with key stakeholders to support GP management plan; (iv) liaising with GPs/primary carers, hospital-based services, NGOs etc regularly to ensure up-to-date information is received and to support high quality patient care; and (v) providing a letter to the GP/primary carer when the patient is discharged from the program. ²⁹

Patient Population

The findings of quantitative patient data analysis for the first year of the ICPCC program in SESLHD after the algorithm was introduced have been published elsewhere. ³⁰ Briefly, for people enrolled in the ICPCC program in SESLHD in 2019-2020, their main chronic conditions involved the circulatory (14%), respiratory (13%), musculoskeletal (10%) and digestive (10%) systems. ³⁰ Nineteen percent of people enrolled in the program had one condition, 18% had two conditions, 17% had three conditions and the remainder had four or more conditions. Of the people enrolled in the program, 83% were identified via the algorithm, while 17% came from a health professional or other referral, including: chronic care programs 74%; allied health/community health nursing 12%; GP/other primary care 11%; and self/family 3%. While patients sourced via the algorithm and referrals were similar in some demographics (eg. marital status), there were more older patients ≥75 years (61% vs 40%) among those identified via the algorithm than those referred from health professionals. While there were equal proportions of males and females among those identified via the algorithm, there was a higher proportion of females than males among health professional referrals (62% vs 38%). Patients referred by health professionals were more culturally diverse than those identified via the algorithm; specifically, more were born outside Australia (60% vs 50%) and more spoke a language other than English (25% vs 20%), which may indicate identification of patient’s additional social care needs. ³⁰

Qualitative Analysis

Potential health staff participants were identified through a SESLHD Integrated Care committee (investigator JO was a member). Staff were invited to participate in interviews and/or a focus group via the university-based principal investigator and not through management channels. Potential participants were invited to contact the principal investigator if they were interested. Interviews and the focus group were conducted in person and online using semi-structured question guides (see Supplemental Appendix) and recorded using the Microsoft Teams software function on the computer to enable audio transcription.

We conducted 17 semi-structured individual interviews between November 2019 and August 2020 with a range of health staff, including ICPCC Care Coordinators, Integrated Care Unit managers, referrers to the program, and staff from NSW Ministry of Health (senior management). These interviewees came from nursing (n = 9), allied health (n = 2) and health management (n = 6) backgrounds. Not all the selected participants who were invited were available to be interviewed, including two GPs and some senior managers. We also conducted a focus group with six of the ICPCC program staff (including Care Coordinators and their team leader) within SESLHD who had already been interviewed individually.

Interview and focus group transcript data were analysed to identify common themes ³¹ and coded into categories according to patterns in the data. A mixture of inductive and deductive approaches was taken so that data was analysed according to the research questions but also so that new themes could emerge directly from the data.^32,33 The qualitative software package NVivo 12 assisted with data management and analysis. ³⁴ To ensure reliability and validity of the qualitative data, the coding framework of the interviews and focus groups were reviewed by project members so that variations in understandings could be discussed and refined. Being reflective of the reliability and validity of methods is an accepted approach to reduce bias in qualitative research. ³⁵

The normalisation process theory (NPT) framework guided the analysis and coding framework. ³⁶ Four constructs representing the types of work needed to implement a new intervention included its coherence, cognitive participation, collective action and reflexive monitoring ³⁷ (see Figure 1 below). These components were used to understand the factors that inhibit or promote incorporation. ³⁸ To ensure reliability and validity of the qualitative data, the coding framework was reviewed by project members so that variations in understandings could be refined and validity increased. ³⁵ The study was approved by SESLHD Human Research Ethics Committee [Ref 2019/ETH12918].

Figure 1.

Implementation of the ICPCC Program using Normalisation Process Theory and its four constructs.

Findings

A strong sense of program coherence was identified by all participants who viewed ICPCC as an effective model of coordinating and integrating care for patients at risk of PPH through self-management at home. While management staff focused on keeping patients out of hospital, program staff focused on achieving patient-centred goals, promoting patient enablement and integrating local services to reduce hospital use. Program staff worked to enhance the cognitive participation of stakeholders and referrers by promoting understanding of the distinctiveness and benefits of the ICPCC program. Various forms of collective action were used to capture the right patient cohort. Program staff felt the algorithm was a systematic way to identify patients. However, health professional referral was also supported by program implementers as an effective way to identify patients that might not be identified through the algorithm. While those directly implementing the program recognised its benefits, the recent changes involving the use of an algorithm were not yet communicated across the LHD. This process could be enhanced through more internal communication and reflexive monitoring of program messaging.

Program Coherence and Understanding

The NPT concept of ‘coherence’ is the sense making that people do about an intervention, so that practices occur in a normal way. It also involves a shared understanding of program objectives and benefits, of individual tasks, and of the program's inherent value. ³⁷ All participants had a strong understanding of the overall program purpose. At the systemic level, there was a shared understanding of the importance of integrated and coordinated care for patients with complex chronic conditions to reduce unplanned hospital admissions.

There were slightly different understandings of responsibilities depending on participants’ roles. Senior management stated the program would be implemented somewhat differently at the LHD level depending on the context and available local services. This understanding was also reflected in the views of program middle management who discussed the need for flexible delivery in relation to changing programs and resources, to reduce duplication and enhance service efficiency.

We need that level of flexibility to be able to implement the program. Looking at what other resources are already available in the District… not duplicating what’s already out there, and of course what’s out there keeps changing. (Middle Management)

ICPCC implementation staff were primarily focused on meeting patient goals, supporting enablement through coaching, linking to services and coordinatinating chronic disease management. Program management and implementation staff were united in the need for the program to meet person-centred goals through identifying and addressing barriers to care. Care Coordinators provided person-centred care, assisted with care navigation to other services and supported self-management. They also assessed the patient’s current level of motivation and readiness to engage in behaviour change.

At the clinician-patient level, staff thought the program was unique by encompassing the physical and social determinants of health and not focusing on a specific disease; establishing a collaborative goal between the patient, GP and existing supports; and Care Coordinators improving the patient’s self-management and health literacy.

It’s that really reflexive work of trying to meet them where they’re at and move them to the next step. If they’re ‘contemplative’, how do I get them to ‘action’?…Can I reduce the barriers that are in place in getting them to the ‘action’ phase? That might be it for my intervention at that point in time, then maybe the diabetes [nurse] picks it up from there… (Care Coordinator)

There are a lot of chronic care teams to deal with specific chronic illnesses, like the respiratory team or the heart failure team, which these [patients] may already be involved with. But for some people, they have a range of co-morbidities or they have some other issues or social issues that actually might make it more difficult for them to manage their chronic condition (Care Coordinator)

Our aim is always to link people back to primary care so that their GP becomes the case manager, which is the whole aim, and then they can continue to work forward… it’s that two-way information flow as well. So we can provide the GP with information, that we’ve seen, and they will provide us with what they’re already doing in terms of managing people, so that part is working well … (Care coordinator)

To the health professional referrers to the ICPCC program, the program fulfilled the need to provide a coordinated care management approach for patients with complex needs which could not be met in the scope of the disease specific care they provided. The program was able to address the underlying reasons for re-admission and provide an overall assessment of patients' complex chronic conditions, health coaching and coordination between care providers. Other program staff needed to understand the purpose of the program in order to work in partnership with ICPCC staff and to avoid duplication.

For the patients that we see that are re-presenters…it is great to have these guys on board …, so that there’s a plan in place, so that there’s ongoing care management … (Referrer)

Cognitive Participation and Program Engagement

The NPT concept of ‘cognitive participation’ refers to the work done to ‘build and sustain’ practices, engage others, and ensure sustainability. ³⁷ ICPCC staff formed strong relationships with other integrated care and chronic disease staff and services in the process of navigating and coordinating care. This process included understanding the range of other chronic disease programs, knowing eligibility criteria and maintaining flexibility to adapt to a changing integrated care policy context. Networking and promotion were important for securing future referrals, linking patients to services and highlighting the uniqueness of the program.

Participants outside the ICPCC program felt it offered a coordinated integrated care service. Other chronic disease management teams had limited capacity to provide a thorough assessment of the patient’s health goals and assess their other conditions and social situation. Care Coordinators could assess the barriers for patients in self-managing and understanding their conditions.

An [older] gentleman…has poor health literacy, multiple conditions… he is not getting it at all...He keeps forgetting appointments…he needs to understand the changes in his medications…He speaks English, but he does not speak very well…I referred him to integrative care to do a cognitive screen, with an interpreter, and to assist with diabetes education and referral...every time I see him, it is extremely time consuming, and I haven’t addressed all the issues that need addressing. (Referrer)

Other chronic disease management teams in the community and specialist services needed to have a general awareness of the ICPCC program as they may need to refer their patients there or one of their patients may have already been enrolled in the ICPCC. The Care Coordinator could provide links between different chronic disease management programs to ensure better care navigation for patients.

A lot of [the implementation] was around the networking that we started to pursue quite actively …Most of the programs that are out there are usually aligned to specific disease processes…we tried to market ourselves as ‘the linkers’. Looking at the whole picture of filling those gaps around the edges [for patients]. (Care Coordinator)

Collective Action and Implementation

Collective action is the work that enables interventions to occur, including ensuring the workforce has the right skills and that resources are appropriately allocated. ³⁷ For the successful implementation of the ICPCC program, it was imporatnt to accurately identify the patient cohort and establish clear referral pathways.

It was essential for staff to think about the types of patient that would benefit from the program. The program focused on enabling self-management of chronic conditions, linking them to services and coordinating these services so they could stay at home.

Who can benefit? It’s people who can really respond to those interventions that the Care Coordinators [provide]… [who can]…improve their health behaviors to keep them out of hospital. (Middle Management)

They’re very appropriate referrals with them [patients referred by health professionals]. We can actually make some really good inroads with them. We provide them with support around health coaching, ongoing management and linkage to broader based services. Because once again it’s a very disease specific, sort of position they will often show and when we put them onto the PFP they’re often scoring under the eight, but they are usually some of the most appropriate patients. (Care Coordinator)

SESLHD Integrated Care Unit developed triage criteria for patients identified by the algorithm, including those who were socially vulnerable and/or living alone. The algorithm developed at the Ministry level, was undergoing review to improve its efficiency and equity. Some ICPCC implementers expressed frustration about the algorithm selecting patients at a later stage of their disease trajectory who they felt were less able to benefit. Analysis of the patient referral data revealed that patients identified via the algorithm were slightly older than those referred directly to ICPCC by clinicians.

People at the upper end of the algorithm are less able to respond. They are more advanced in their disease… [but] should we be targeting people who aren’t as advanced? If one scores lower in the algorithm, because they still have more functional capacity to respond to the interventions, we actually might get essentially a bigger bang for our buck by intervening early, in that chronic disease process. (Middle Management)

Senior management staff explained that the next instalment of the algorithm would better identify the right patients earlier in their disease trajectory who would be more likely to benefit from the program. It would also consider those patients with higher needs who were socially vulnerable, isolated, culturally diverse, and at the lower end of the socio-economic gradient. It would take into consideration mechanisms at the LHD level to prioritise the needs of patients, including those referred by health professionals.

There is more of a convergence between what the algorithm would predict and what the clinician would. (Senior management)

Reflexive Monitoring

Reflexive monitoring is the appraisal undertaken by people to assess the intervention impact on themselves and others including its effectiveness; and how its procedures can be refined based on feedback. ³⁷ Appraisal of the ICPCC program was intended to be based on regular feedback, but recent program changes associated with the implemtnaton of the algorithm meant that there was not the usual level of reporting to stakeholders. There had also been changes to the program governance structures, which affected workforce promotion and communication.

The messaging is really important because that’s not just about messaging to the integrated care team, but also the broader hospital community. (Middle Management)

For some referrers, there was confusion about the communicaiton regarding regarding the program's structure and staffing model. This included the extent of resources provided to the program relative to the staffing in other programs, leading to confusion about how the ICPCC differed from other programs providing a higher intensity of care.

I don’t understand how the service is structured…The integrative care service sent me this email…I don’t quite understand why there were three people working in this service, with nine people [patients] on their books, and we have one and a half [staff] for 90 [patients](Referrer)

Another referrer to the ICPCC program felt that there could be improved communication regarding patient eligibility and background for the program.

Discussion

The project aimed to understand how the ICPCC program was conceptualised and implemented in practice. Various theoretical frameworks can be used to understand how integrated care interventions are implemented including the impact of context on effectiveness.^10,13,39,40 Understanding implementation is important for program designers and managers working with various factors that are out of their control. ¹³ The NPT framework used in this evaluation allows analysis of the different components of work needed for successful implementation of a complex intervention. Like other integrated care programs, there are challenges in defining, measuring and evaluating care coordination. ⁸ The NPT framework demonstrates the importance of understanding and refining the program purpose and work processes to ensure better evaluation and measurement. Understanding how the program is unique can complement other programs in an environment of cost efficiency and reduced duplication.

Understanding of Program Goals by Level

This evaluation examined how the ICPCC program was implemented at the clinician, LHD and senior management levels. All participants coherently conceptualised the purpose of the program in terms of both systemic approaches of reducing PPH and enhancing person-centred care and self-management at home. This was unique as international research has reported differing understandings of care coordination which focus either on systemic goals or patient perspectives. ⁸ Meeting both patient and systematic goals supports progress toward achieving integrated care quadruple policy aims of improved experiences for individuals, families, and service providers, improved health outcomes for the population and improved health system cost efficiency. ⁷

While there was participant consensus on the program's purpose, there were slight differences in implementation responsibilities depending on participants’ roles. Senior and program managers' focused on flexible understandings of program design and delivery, which is useful in times of change, as cited in other research, ³⁹ and helps to meet patient needs as well as reduce service duplication. Program managers and staff discussed assessing whether the patient already had a multi-disciplinary team assisting them before considering what additional contribution the ICPCC program could provide.

In achieving the program's purpose, program implementation staff discussed addressing patient goals through health coaching, referrals, linking services and care coordination. Peart et al have discussed how Care Coordinators first negotiate patients' goals in the context of their lives, then as a trusting relationship is established, more systematic issues of chronic disease management outside the hospital can be addressed.^41,42,43 Focusing on more person-centred goals means the health professional first works with the person’s values and beliefs, and incorporates shared decision-making processes.^41,44 Building partnerships and linking services were identified as effective strategies for integrating service delivery. Care Coordinators built strong relationships with local community and primary care health services, which have been highlighted in care coordination literature as supporting long-term commitment and integration of services.^10,39

Reviews of care coordination in practice highlight the need for strong work processes and capacity at systemic and local levels to support the program. ³⁹ Care coordination is inherently resource intensive but has benefits in consolidating care and reducing duplication. As stated by McDonald et al, ‘coordinating care better is only beneficial if other aspects of care delivery are optimized as well’ (p. 114). ³⁹ Work processes must be established at both systemic and local levels to ensure stability and enable long-term strategies. This is difficult as most integrated care programs focus on organisational level service provision. ¹⁷ The importance of program cost efficiency was also emphasised by program managers in the LHD, and it is therefore important to inform others about the benefits and outcomes of the ICPCC program.

Enhancing Patient Self-Management and Reducing Risk

The purpose of the evaluation was to assess health staff views on the understanding and implementation of ICPCC. One of the main patient-centred goals of Care Coordination and the ICPCC was to assist patients in developing the capacity and skills to self-manage aspects of their own care. This goal incorporates international understandings^19,45 and, as outlined in statewide integrated care goals,^7,25 patients should be involved in the self-management of their chronic conditions for better health outcomes. Providing health coaching improved a person’s education and understanding, and facilitated their use of this knowledge to support self-care. Such an approach enacts the process of enhancing people’s health literacy in relation to “competencies to access, understand, appraise and apply information to make health decisions in everyday life”.^46,47 The findings demonstrate that the program assisted people in managing their complex chronic conditions and overcome barriers to achieving optimal health. Health crises can reduce a person's capacity to adjust to required life changes or self manage effectively. Focusing on person-centred care and addressing equitable access acknowledges the vulnerabilities of certain population groups; and recognise that people have different capabilities and may not respond in the same way to standard approaches to improve their health as outlined in established frameworks for patient enablement and chronic care. ⁴⁸ It recognises the impact that wider social determinants of health have on their situation. Improved health literacy for patients with chronic conditions has been shown to reduce hospitalisation. ⁴⁹

Identifying the correct patient cohort is important for program efficiency. For a patient to be a right fit for the program, they need to have the capacity to self-manage. Despite moves to promote chronic disease self-management, there are no widely accepted approaches to measure ‘self-care’. ⁵⁰ Yet the costs of treating self-manageable conditions are high for the health system. ⁵¹ The need for government to implement self-care initiatives has been promoted in Australian ⁵² and international policies. ¹⁰ Implementing self-management strategies can assist patients who are at the stage of being willing to change their behaviour and who are ready to change their behaviour and capable of contributing to the management of their own conditions. ⁵³

Effective Ways to Identify Program Participants

The process of identifying the right patient for the ICPCC program determined its successful implementation. While the algorithm identified equal proportions of males and females as potentially being suitable for integrated care, patients identified via algorithm were older and less culturally diverse than those referred by health professionals. However, program managers and implementors found that direct referrals from health professionals to the program were an effective way to identify potential patients that could benefit which had not been identified via the algorithm. At the same time, the algorithm was an effective systematic way to use linked data to identify potential patients. Predictive algorithms have proved useful elsewhere in Australia and internationally in analysing routinely collected patient data and identifying those at risk of PPH as suitable for a health management program.^23,54 However, such algorithms do not identify all suitable patients especially those in the community with rising risk who have not been admitted to hospital. ⁵⁵ The need to integrate hospital and community health care data (such as data from private primary care clinics or claims data from public and/or private health insurance programs) in algorithms used to identify patients at risk of hospitalisation has been highlighted in other research.^2,56,57 These considerations could be an option for the next iteration of this program. Senior management staff stated that in the future the algorithm will better predict patients at risk of future PPH and include more of the types of patients included through health care professional referrals. It will also consider the socio-demographic risk factors that were considered at the LHD level including social isolation or being from lower socio-economic and/or culturally diverse backgrounds. Assessment is needed beyond just quantifying multiple chronic conditions to include organisational and staff engagement, and validated measures of social risk. ⁵⁸ While the algorithm is still being refined and modified, health care referrals are still integral to capturing more patients who may benefit from enrolment in the ICPCC program.

Comparison With Other Work

While authors based at the NSW Ministry of Health presented a paper on the ICPCC in the World Congress on Integrated Care in 2016, this was to announce that the previous NSW Chronic Disease Management Program was being replaced by ICPCC and described the features of the new program. ²² We are not aware of any published original research reporting quantitative and/or qualitative data analysis findings and outcomes for the ICPCC. However, there are published papers which report on the outcomes of integrated care programs in NSW other than ICPCC.^{12,20,59,60,61,62,63,64} While the quantitative analysis findings reported by Plant et al (Care Navigator trial) ⁶⁴ and Billot et al (Chronic Disease Management Program) ²⁰ were not favourable, Mallitt et al (Health One Mount Druitt) ¹² and Hartati et al ⁶¹ (Planned Care for Better Health) did report successful outcomes and the outcomes reported by Campain et al (Winter Strategy), ⁶⁰ were mixed. There are also three papers reporting on outcomes for integrated care programs at NSW Government demonstrator sites in Western Sydney Local Health District and Central Coast Local Health District.^59,62,63 Two of these studies were qualitative and provide some additional insights into some of the issues raised by our work, including: the importance of timely high-quality communication; trust in new ways of working; building a common language between providers; and having clear objectives for integrated care programs that meet the goals of multiple providers, their patients and carers.^59,62 Trankle et al, ⁵⁹ especially, have highlighted the challenges faced when trying to deliver integrated care across primary and hospital care sites when the data and information technology infrastructure is not truly integrated. Hospitals and GP clinics use different clinical software requiring third party systems for information sharing - systems that face implementation barriers and demand additional training and support.

Limitations

A limitation of our analysis is that it did not include the views of private practitioners (ie. GPs or allied health) working in primary care or patients enrolled in the program. However, we did include LHD allied health staff, and two GPs were invited to participate but were unavailable. We believe that reporting the views of government health bureaucrats plus LHD clinical and management staff on ICPCC implementation does make a meaningful contribution to the literature. While this paper relates to ICPCC implementation in one LHD in NSW, its findings nevertheless illustrate the experiences of health staff post-implementation of an algorithm-based integrated care coordination program which may have relevance in other LHDs in NSW or jurisdictions elsewhere.

Lessons Learned and Recommendations

The experience gained from ICPCC implementation and delivery informed the planning for and design of Planned Care for Better Health as the next iteration of a statewide integrated care program. The new program built upon its predecessor ICPCC by enabling LHDs to guide local program implementation and delivery, and by including a broader range of chronic conditions and considering wider social determinants of health in patient identification. ⁶¹ The hospitalisation risk algorithm used during ICPCC (ie., CCPIA) was improved upon when developing the Risk of Hospitalisation (ROH) algorithm used for selecting patients for Planned Care for Better Health. The ROH has more inputs than CCPIA in terms of chronic conditions (6 vs 20 conditions) and also includes socioeconomic status, smoking status, Indigenous status and culturally and linguistically diverse status.^61,65 While adding these data inputs should improve the ability of ROH to identify suitable patients for selection, ROH still relies on NSW Health hospital administrative datasets.

As ROH does not include GP health record data, it would still be challenging to identify those in the community who are at risk of hospitalisation and could benefit from an integrated care approach but are not selected because they have not yet been hospitalised or received hospital-based care and so do not feature in the NSW Health datasets. We recommend that a new hospitalisation risk algorithm be developed in NSW that incorporates private GP electronic health record data that is linked to NSW Health administrative datasets. ⁶⁶ This has been attempted in Victoria, though there have been some issues around implementation and usability with clinicians and researchers involved in that program still recommending that there be some mechanism for patient selection by clinicians.^55,57 This is based on the premise that clinicians’ knowledge of patients, and the complexity of their conditions and circumstances, may in some cases be better at selecting patients than a risk stratification algorithm. ⁵⁵

Going forward, we recommend a comprehensive mixed methods approach to evaluation of integrated care programs in NSW to gain the best understanding of barriers and facilitators to implementation and delivery. Working in partnership with Primary Health Networks and including the perspectives of private practitioners (GPs and allied health) working in primary care, plus those of patients and their carers/families, is vital to inform and refine our approach. Finally, linking NSW Health administrative datasets with health records from GP practices and Medicare (the federal Australian Government’s universal public health insurance scheme) would enable assessment of a wider range of health outcomes for monitoring and evaluation.

Conclusion

We used the NPT framework as an effective way to assess the implementation of the ICPCC program in practice. We demonstrated the importance of a shared and coherent understanding of the program's purpose, the impact of local context and the type of work required to achieve benefits. The ICPCC program played an important role in assisting patients with a range of chronic conditions to access and benefit from integrated care coordination, while enhancing their capacity to self-manage at home. All study participants understood the program purpose and necessity, including the importance of achieving both patient and systemic goals. Networking, linking services and promotion were important, as was reporting on benefits. Program messaging should consider the key information that it is important for both service managers and referrers to know. Program managers and implementers found that referrals to integrated care programs from health professionals who are familiar with the complex history and needs of their patients, when combined with a hospitalisation risk prediction algorithm, can effectively identify patients that may benefit from integrated care coordination. These findings may inform the design and implementation of future iterations of integrated care coordination programs in NSW and elsewhere.

Supplemental Material

Supplemental Material - Integrated Care Coordination for Managing Chronic Conditions: Views of Health Staff on the Implementation of a Program Using an Algorithm to Identify People at Higher Risk of Hospitalisation in Sydney, Australia

Supplemental Material for Integrated Care Coordination for Managing Chronic Conditions: Views of Health Staff on the Implementation of a Program Using an Algorithm to Identify People at Higher Risk of Hospitalisation in Sydney, Australia by Cathy O’Callaghan, Julie Osborne, Margo Barr, Damian P Conway, and Ben Harris-Roxas in Global Advances in Integrative Medicine and Health