Towards universal palliative care in Ireland: Scoping decision-makers’ needs for an Atlas of Variation data mapping tool

Research overview

Ireland published its first national policy on adult palliative care in 2001, effectively instituting a universally accessible system of publicly funded palliative care¹. The country’s palliative care system is thus something of an outlier within the broader healthcare system as Ireland is the only country in Europe without a universal healthcare system, though major initiatives are ongoing to this end under the country’s 10-year programme of health reforms, Sláintecare². Despite a policy of universal palliative care, there is evidence of persistent inequalities in service coverage and access^3,4. Significant obstacles to evidence gathering in this area include gaps in current administrative palliative care data collection and interoperability, and in the country’s weak health data infrastructure more broadly^5–7. To facilitate improvements in palliative care service planning and delivery in future, Ireland’s updated policy calls for development of an atlas of variation for palliative care⁸.

Atlas of variation methodology uses small area geographic analysis to identify warranted and unwarranted variation in specific areas of healthcare provision, access, and health outcomes⁹. Pioneered more than half a century ago by developers of the Dartmouth Atlas Project in the US¹⁰, atlases of variation have since been employed by health systems around the world as a powerful data visualisation tool to guide health service planning and delivery aimed at improving health equity. While atlases of variation have been developed to identify variations in a broad range of healthcare services and outcomes, few atlases of variation exist that focus on palliative and end of life (EoL) care even though access to palliative care is recognised by international health bodies as an essential component of universal healthcare¹¹.

This paper presents findings from a qualitative study exploring key stakeholders’ views on development of an atlas of variation for palliative care for Ireland¹². Interviews with stakeholders followed a review of the literature on existing atlases of palliative care developed in other jurisdictions. The research identified critical enablers and barriers to development of an atlas of variation, offering important lessons for other countries considering application of atlas of variation and alternative health atlas methodologies. Ultimately, the research found that while atlases of variation are undoubtedly a valuable tool, they are a challenge to realise in countries where the digitalisation and harmonisation of health data is nascent, as in Ireland, and an alternative hybrid approach that utilises both administrative, quantitative data sets and qualitative data may be more appropriate and indeed desirable given stakeholders’ data needs and priorities. The study was funded by Ireland’s Health Service Executive (HSE) through a collaborative partnership award, with the Lead Knowledge User based in the HSE’s National Health Intelligence Unit.

Definition of palliative care

Palliative care is defined as “an approach that improves the quality of life of patients and their families facing the problems associated with life-threatening illness, through the prevention and relief of suffering by means of early identification and impeccable assessment and treatment of pain and other problems, physical, psychosocial and spiritual”¹¹. Access to palliative care is considered an essential component of universal healthcare by the World Health Organization¹¹. As such, ensuring access to palliative care is integral to the achievement of UN Sustainable Development Goal (SDG) addressing Universal Health Coverage (SDG 3.8). In 2014, a World Health Assembly resolution (Resolution WHA67.19) aimed at strengthening palliative care as a component of comprehensive health and social care delivered across the lifecourse declared equitable access to palliative care an ethical imperative, stating, “it is the ethical duty of health care professionals to alleviate pain and suffering, whether physical, psychosocial or spiritual, irrespective of whether the disease or condition can be cured”¹³.

Palliative care in Ireland

Palliative care has a long history in Ireland, originating in the establishment of hospices by religious orders as places for care of the dying in the late 19th century¹⁴. Voluntary organisations continue to provide most of the specialist palliative care, though their services are increasingly integrated into mainstream health service provision across the acute and community healthcare sectors and are funded primarily by the public sector. In 1995, Ireland became only the second country in Europe to recognise palliative medicine as a distinct medical specialty, and in 2001, became one of the first in the world to publish a national policy on palliative care¹. As mentioned earlier, an update to the 2001 policy, focused on addressing inequalities in access and capacity issues in the sector, was published in 2024⁸.

Since implementation of the 2001 policy, significant advancements have been achieved in terms of the development and delivery of palliative care services. Nonetheless, gaps in service provision persist, leading to inequities in access^3,4, with specialist services historically concentrated in the most populous cities of Dublin, Cork and Limerick. These urban areas and the regions surrounding them have more advanced infrastructure and provision, while less populated areas continue to have poorer infrastructure, with residents of certain areas lacking a specialist palliative care unit in the vicinity^4,8.

Palliative care services in Ireland are organised into three levels, structured by level of healthcare providers’ specialisation: 1) Level one – Palliative care approach: all healthcare professionals should be aware of, and appropriately apply, palliative care principles; 2) Level two – Generalist palliative care: healthcare professionals not engaged full time in palliative care but have some relevant training; 3) Level three – Specialist palliative care: healthcare professionals whose core activity is providing palliative care (see Figure 1). Specialist palliative care units, referred to as hospices, act as hubs for specialist palliative care delivery in their respective geographic regions. A fully developed specialist palliative care unit comprises in-patient unit beds, day hospice, out-patient and bereavement services. It serves as the base for the community palliative care team and for palliative care education and research activities for the region. The specialist palliative care unit coordinates with acute specialist palliative care services provided in hospitals in the same region through the consultants in palliative medicine who are contracted to provide services in both specialist palliative care units and hospital settings. According to the national policy, all three levels of palliative care should be accessible to all individuals with palliative care needs, with access determined by the level of care need, irrespective of a person’s ability to pay or place of residence⁸.

Figure 1. Palliative care expertise and specialisation levels in Ireland’s model of integrated palliative care.

Source: Adapted from Department of Health⁸, Figure 2, p.14.

Broad consensus on data priorities

There was consensus among interview participants that, broadly speaking, mapping data on existing capacity within the system across the HSE’s six Health Regions should be a core element of an atlas. More specifically, data on specialist palliative care service capacity (provided by specialist palliative care in-patient units, community specialist palliative care teams, and specialist palliative care teams in acute hospital settings); and generalist palliative care (provided across primary, acute, and long-term care settings). One interviewee emphasised the importance of clearly delineating the different care settings where specialist and general palliative/EoL care is provided across the country, information that is not necessarily available even to those working within the system:

Asked how, in the absence of comprehensive data on existing capacity within the system, decisions are taken regarding where additional investment and resources are needed, another interviewee stated that decisions are still in part being made on the basis of local lobbying, rather than being wholly driven by evidence grounded in data.

Beyond providing information on the ‘the state of play’ in terms of service capacity and service availability, several key informants imagined an atlas having what one interviewee referred to as discrete ‘buckets of data’, consisting of: population demographics, population needs data for palliative care, which could be juxtaposed with service capacity/availability, system performance, and care recipient outcomes. The same interviewee highlighted the vital link between understanding population demographics and need, and effective long-term care planning:

While the focus of the atlas was acknowledged to be on mapping variation in specialist palliative care, several interviewees stressed the importance of ensuring that an atlas capture data on both specialist and general palliative care provision; and on the specialist as well as non-specialist palliative care workforce. Interviewees stressed that failing to capture non-specialist palliative care provided by nursing home care professionals, general practitioners (GPs), and public health nurses (PHNs) results in a significant underestimation of palliative care capacity, utilisation, and ultimately access. It is also a missed opportunity to integrate non-specialist providers into the palliative/EoL care eco-system and acknowledge the important palliative care they provide as part of their wider care work. One interviewee noted that nursing home care professionals and other general palliative care providers often do not recognise that they are in fact providing palliative care because they are “just not able to label it”, to record and report it [15, Service provider, NHI]. At the same time, interviewees noted the difficulty in meaningfully measuring, quantifying non-specialist provision of general palliative care.

Interviewees expressed that in an ideal world, an Irish atlas would include data that capture the transition between different types and levels of health care services - crucially, between community-based services and acute hospital services. A significant barrier to achieving this is the absence of population level data linked across care settings. An interviewee working at regions level said:

Other categories of data prioritised by key informants interviewed include indicators of health system performance that capture the patient experience of care, such as timely access to palliative care across different care settings; data on geographic variations and variations across care settings in patients living with chronic and progressive fatal illnesses who are seen by a specialist palliative care provider; data on palliative care provided by private hospitals, community, and residential care providers as, at present, little to no information is available concerning patients with palliative/EoL care who seek care outside Ireland’s public healthcare system; and inclusion of palliative/EoL care for children as one aspect or domain, while recognising that within the current model of care, services for children and related data collection is distinct from specialist palliative care for adults.

Some key informants expressed uncertainty about the type of data that an atlas of palliative/EoL care for Ireland should include. Others held strong views on the categories or domains of data an atlas should include, with a few interviewees providing specific indicators they would want captured. While there was substantial overlap among the key informants in terms of the high-level categories of data to include, presented in the section above, additional consensus-building activities would need to be undertaken in order to agree the aim(s) of the atlas, and following from that, the domains and the formulation of indicators.

Integration and harmonisation

Interviewees mostly agreed that while important gaps exist, substantial data are being collected on palliative care. The challenge lies in the integration of existing data sets, and in pulling specific variables from different data sets, which are often under the ownership of different governmental agencies.

GDPR and other data sharing legislation can add significantly to timelines and resources. Following on from the GDPR, in 2019 the government introduced the Data Sharing and Governance Act, which brings in extra stipulations for government bodies sharing sensitive personal data. And for government bodies that do not explicitly identify health research (conducting surveys) as part of their mandate or remit, a further requirement is that a public consultation lasting six weeks be organised to ensure that the public is informed of the purpose of the proposed data sharing. The public consultations are organised centrally through the Office of the Government Chief Information Officer (OGCIO).

While the personal health data of every living individual is covered by GDPR, it does not cover the health data of deceased individuals. Processing the data of deceased individuals falls under the Data Sharing and Governance Act. While not necessarily a barrier, in developing an atlas, consideration should be given to the different types of data being processed and shared across governmental and non-governmental institutions, what the regulatory requirements for data sharing and protection are at each stage, and how these will impact on timelines, if at all.

According to the majority of key informants, a significant barrier is the lack of investment historically in the country’s healthcare IT systems, resulting in the siloed, un-linked nature of the many IT systems operating within the healthcare sector, and the continued use of paper-based records in some care settings. One interviewee spoke of the challenge in accessing WTE data for the palliative/EoL care workforce for monitoring and planning purposes, sharing that they had to chase down accurate figures over the phone for one service area in particular because they were not being inputted into the system. Another interviewee pointed out the difficulty in linking data across care settings in which palliative/EoL care is provided, stating:

The lack of implementation or use of a unique patient identifier—known as the Individual Health Identifier (IHI) number¹ in Ireland—was seen by more than one interviewee as a major barrier to integrating data.

Data quality was another key issue raised related to the integration/harmonisation of data for the purposes of an atlas. In the reporting of Hospital In-Patient Enquiry (HIPE) data, where two new data quality fields for each indicator were introduced as part of the 2024 HIPE Data Dictionary, some of the newer indicators added, of great relevance to palliative/EoL care (e.g. when a specialist palliative care team attended a patient during their episode of care), are in the development phase and the data reported are likely to be incomplete and of unknown quality. Data quality issues were also reported related to data on place and cause of death.

A critical point raised by several interviewees is that the geographic unit(s) of analysis used in an atlas must be aligned with the health system’s sub-national structures. In Ireland, these are in first order the six newly formed Health Regions, and within each of these, Integrated Health Areas (IHAs), Community Health Networks (CHNs), and Community Specialist Teams (CSTs). One interviewee, a service provider and expert in service model/delivery was of the view that the IHAs, which are intended to bring together the different service types and care settings within the HSE Health Regions, could prove to be the most useful level to disaggregate data to map variation:

Multiple interviewees highlighted that the fragmented nature of palliative care data is indicative of the fragmentation of the palliative care and wider health and social care system, where until recently, acute hospital groups were not necessarily aligned geographically or administratively with outpatient care services. The new HSE Health Regions are working to address this issue, but some interviewees maintained that many of the details are still being worked out at the sub-regional level. For example, a key informant working within one of the six regions said that work to identify exactly who falls within the region’s catchment area is still underway:

Relatedly, the omission of Eircodes or postal codes as part of all health data collection generally and palliative care/EoL care data specifically was also raised by a key informant with data infrastructure expertise as a current gap. They noted that, “we’re trying to encourage it because if we have the Eircode, then it’s a single point and we can then find the data at whatever level” [03, Data infrastructure expert, DoH].

Interviewees were hopeful that ongoing developments related to the introduction of the Health Information Bill and the national shared patient record would bring about better integrated healthcare data, including palliative care data in future. For the time being, however, there was consensus that the collection and sharing of patient level data across settings and services is limited and would pose an obstacle to the development of an atlas of variation for palliative and EoL care.

Resources and capacity

A key consideration in developing an atlas of variation for palliative/EoL care for Ireland is identifying the level of resourcing and capacity required, and the extent to which the latter are available within the system. All interviewees were of the view that the resources could be found, but that the aims and design of the atlas would need to be clearly defined before the level of resourcing necessary could be determined and that resource constraints could impact on the ambitiousness of the exercise. A strong emphasis was placed on the need to clearly articulate the potential value and utility of the atlas early on, i.e. its potential to improve the efficiency of resource allocation for palliative and EoL care.

Interviewees noted that a ‘data culture’—a culture of using data to improve service delivery and quality and recognising its value for same—has not penetrated the healthcare system in Ireland and that this can result in reluctance on the part of care providers to comply with new data requests, particularly if the data compares services across the different provider organisations and makes them identifiable in the public domain. One key informant noted that it had been their experience that service providers became concerned about how their services were being represented after data they had submitted had been analysed and (re-)shared with them:

It was therefore suggested that resources should be invested in raising awareness among and training providers in any new data reporting requirements or data collection instrument associated with an atlas.

Some interviewees indicated that developing, interpreting, and maintaining an atlas would require considerable human resources over an extended period, in particular if the atlas were to involve the deployment of a new data collection instrument. While potentially resource intensive, most interviewees were of the view that the necessary skills could be found from both within and external to the public sector.

Governance

To a certain extent, the governance of the atlas relates to the issue of resourcing and capacity discussed above and was raised by several interviewees in that context. Governance here refers to responsibility and accountability for the atlas. While there was no consensus among the key informants interviewed as to which institution or institutions should have governance oversight of an atlas, some felt strongly that whichever institution(s) ends up responsible should already have—or legally be able to gain—access to and be able to process and integrate relevant data sets.

Just as if not even more important for some interviewees was the perceived need for an atlas’ governing body to have the capacity to interpret and make appropriate use of the information provided by the atlas, to ensure that it is both actionable and dynamic:

Interviewees suggested both the National Clinical Programme for Palliative Care and the National Health Intelligence Unit within the HSE as possible governing bodies, respectively. A few interviewees were of the view that an atlas of palliative/EoL care could simply be an extension of or incorporated into the existing Health Atlas currently maintained by the Health Intelligence Unit. Others felt that while the Health Atlas could be a vital resource, an atlas of palliative/EoL care should be a separate, distinct tool, with an emphasis on being public facing and on visually mapping variation across a range of domains and indicators.

While a public, government institution such as the HSE was most frequently suggested as the governing body for an atlas, key informants inside and outside the public sector stressed the importance not only of competence but also of a degree of independence for the group or body ultimately responsible:

Stakeholder engagement

Dovetailing with the last point under the previous thematic area, several interviewees expressed the view that engaging relevant stakeholders in the development and operation of an atlas is about more than just getting input and being able to say that broad participation was achieved. Instead, most interviewees agreed that engaging stakeholders can serve the dual purpose of integration work, bringing providers, health care system operators, policymakers, researchers together around a common cause:

One interviewee took the idea of stakeholder engagement with the atlas a step further, highlighting its value in bringing people together beyond its value as a technical tool or platform and suggesting the creation of a ‘community of practice’ around it centred on the ‘people element’ [07, Data infrastructure expert, HSE].

Other interviewees raised certain challenges or considerations in taking a participatory approach. One key informant, working in a service provider capacity, said that some providers may need to be convinced of the strategic value of collecting and analysing data on the services they provide and that any requests for participation should be framed as an enabler, not simply as a reporting requirement, and that the emotional aspect of palliative/EoL care providers’ work must be taken into account in terms of messaging.

Another key informant cautioned that bringing all stakeholders to the table, so to speak, can lead to long lists of wishes and preferences in terms of data to include and ‘mission creep’. It was suggested that if a participatory approach was desirable, that stakeholders’ expectations were carefully managed and stakeholder engagement activities carefully planned and facilitated by skilled moderators.