The status of the HIS in Ireland
Based on key informant interviews, the HIS in Ireland was, in 2020, generally assessed to be able to provide data that was, when available, of good quality, accuracy, validity and timeliness, including demographic and geographic disaggregation possibilities. However, significant gaps were identified in the completeness of the data, with it mostly being available for publicly provided acute care services and by-large missing for services provided in the private sector, notably independent hospitals, general practice and community health centres. For a similar reason, respondents reported good data coverage for the lower income groups, but much less for the middle income groups, due to the latter using both public free-of-charge and paid private provider services. Registry data were reported to be readily available and of good quality. This was explained by population- and disease-registries having long histories and tradition in Ireland, which often meant dedicated and well-established operational and research teams, high levels of attention to data quality, extensive international collaboration and sustainable funding. In line with Sláintecare reform priorities, using HIS data to assess and address regional differences in the uptake of policies around Ireland were found crucial. However, informants concluded that the current HIS is not able to fully support these regionalisation efforts due to varying levels of data completeness across regions. Additionally, inadequate focus on case-mix and risk-adjusted indicator calculations, despite data being able to support both, hindered its use for benchmarking between regions. Finally, lack of high-level agreement on key priority indicators in the health system, a national oversight body, and consistent and standardised tracking of performance results over time were identified as hindering factors in working with health data in Ireland.
Analysis of information collected through stakeholder interviews also enabled mapping data availability, across health care system’s services. Figure 3 presents the heatmap of availability of certain categories of data for various health care services. Data were described as generally readily available for acute hospital care services across different types of data. Survey-sourced data, from the annual “Healthy Ireland” survey, were also obtainable across many services. Administrative data, especially those on health workforce and financing health services were available for the majority of acute, social and primary care services.
The fitness of the HIS to support implementation of a national HSPA framework
Based on interview data and using the same categories of health data as in the previous heatmap, the five main clusters of indicators, proposed for the Irish HSPA framework, were mapped for data availability (Figure 4). The most notable finding was generally lower availability of data pertaining to the health outcomes and cross-cutting indicator clusters of the HSPA framework. Due to its focus on structures, processes and outputs, at the time of the assessment, informants generally assessed HIS data in place as inadequate in supporting the framework’s focus on outcomes (and linking outcomes to inputs). The HIS was also assessed as suboptimally prepared to provide data on adaptability, resilience and up- and down-scaling capacity of the system, including its infrastructure, services and workforce.
When discussing the HIS in Ireland in relation to its ability to populate the proposed HSPA framework, key informants pointed to and elaborated on three topics of interest: (i) data linkage and the ability to analyse and manage care pathways and integration of care, (ii) collection, reporting and the use of people-reported data and (iii) recent developments, as a direct and indirect consequence of the ongoing (at the time of conducting this work) COVID-19 pandemic.
Pathways of care were not well captured in data and the data to measure care integration were mostly unavailable. Informants mainly attributed this to suboptimal data linkage efforts which, instead of enabling to patient follow-up through the system, mostly inferred care pathways and subsequent correlations. Slow implementation of the individual health identifier (IHI) and historical use of proprietary data standards hindered data linkage capabilities. Informants also attributed the limited data linkage capabilities to the lack of a dedicated national health data coordination body and a coherent strategic approach to health data. Proxies were used, such as bed days by people with primary care sensitive conditions, but it was pointed out by informants that it is impossible to have integrated care without fully integrated data. However, attitudes to data use and sharing, among stakeholders in the system, have recently changed, with a newly found appreciation for data use for policy, by policy- and decision-makers. Primary care was exemplified as an area in which an improved flow of data between general practitioners, hospitals and HSE would lead to better coordination and integration of care. Lack of clinical data flow between most community service providers and the HSE was also emphasised as a shortcoming. Other causes and examples of data linkage issues were presented, such as unaligned data standards and technical solutions among private hospitals with different systems for patient data collection and reporting and subsequently non-interoperable information systems.
According to informants, there seemed to be increasing interest and recognised need for the collection of people-reported data in the Irish health system. This included both patient-reported outcome and experience measures (PROMs and PREMs) as well as staff- and carer-reported data. However, apart from the annual “Healthy Ireland” survey for some services and a few smaller pilot projects, these kinds of data were not structurally collected, reported or used. Collecting people-reported data was seen as challenging and hence often done in small, research-focused ways rather than consistently by health care organisations as a way of gaining and learning from patients, staff and carers. Focusing these qualitative metrics on issues that matter to people, rather than on questions such as satisfaction with the cleanliness of the hospital was another theme, often repeated by informants. So was the lack of patient experience measures in primary care. Person-centeredness was recognised as key to patients and citizens, with suggestions by informants for Ireland to join on-going international initiatives, developing this area further.
With this assessment being conducted during the COVID-19 pandemic, issues of health data management in relation to the situation were also mentioned by informants. They agreed that it was difficult to say whether changes to the data landscape, catalysed by the pandemic situation, will persist, but that it was necessary to consider the potential of these developments. For instance, a temporary, emergency version of the IHI was rolled out to follow patients with COVID-19 through the system and for vaccination purposes [10, 22]. Also, a novel, emergency data hub for researchers to access data about COVID-19 was established, where data could be linked and accessed through the Central Statistics Office’s infrastructure [23]. According to informants, the views on how data is collected, accessed, reported and used have changed due to the pandemic. The crisis emphasised both the system’s shortcomings as well as its strengths. These improvements made data for acute care of COVID-19 readily available on a very granular level across organisations. However, the quality of non-COVID-19 care data had not improved and, some informants felt, it might have even worsened. Questions of data quality, with such rapid data infrastructure developments, were raised and informants noted that, due to the urgency, less attention was put to the minimum requirements of datasets. In general, informants were hopeful that the positive developments will sustain after the pandemic.
Following pre-workshop preparatory work by stakeholders on “indicator passports” and discussions during all six workshops, a total of 266 HSPA indicators were assessed as potentially useful for the first Irish HSPA framework. Data availability and potential data sources were assessed for each. Differing levels of data availability across the five indicator clusters were found, as shown in Figure 5. Data for the proposed framework was mostly available for indicators in the structures and outputs cluster and sourced from administrative, registry and survey data. Data sourced from electronic health records (EHRs) were able to populate only 0.5% of all proposed indicators (13/266).
Data for the 266 potential indicators of the proposed HSPA framework were available from various sources and managed by different data custodians within the Irish health system. Based on the results of stakeholder consultation workshops, two organisations were responsible for most of the indicator data. One quarter of all proposed indicators (72/266, 27%) could have been sourced from datasets held by the HSE, namely data collected and managed by the National Quality Improvement Team as well as the Hospital In-Patient Enquiry, National Patient Experience Study and the Healthy Ireland survey datasets. Another 18% of indicators could have been sourced from data managed by the Central Statistics Office, and its annual Survey on Income and Living Conditions dataset. Other relevant data custodians and datasets included Health Information and Quality Authority (14/266, 5%), Irish Cancer Society (12/266, 5%), Higher Education Authority (9/266, 3%), National Vaccine Information System (6/266, 2%) and the Health Research Board (5/266, 2%). Data to populate nearly a third of proposed indicators required further clarification of custodianship and sources (79/266, 30%).
This assessment also identified several specific data source type-related challenges. The coverage and timeliness of data collected primarily for administrative purposes was generally good, especially for acute care services, but its usefulness for monitoring population-level health and individual patient-level outcomes, which are important for an HSPA framework, remained limited. The use of population and patient survey data in Ireland has been increasing in prevalence and importance. At the time of the assessment, it covered many of the services, as well as clusters and domains conceptualised in the proposed HSPA framework. Predominantly ad-hoc PROMs and PREMs collection efforts could not fully support the envisioned framework and its focus on the use of indicators based on patient-reported data. Despite recent localised rollouts of new standardised EHRs among acute care services in Ireland, this modality of capturing health data was still very limited in its linkage and re-use capabilities. Legal requirements to adopt EHRs and adhere to standards were listed as some of the possible reasons. The increasingly important role of software solution vendors was also mentioned in both interview and workshop stages of this assessment.