Blog

Data linkage is a method of bringing information from different data sources together about the same person or entity to create a richer dataset.  How data are linked depends on the type of data within the datasets to be linked.

In this Q&A, Maria Kelly, an IT specialist working part-time with the National Cancer Registry in Cork and PhD student at the University of Limerick, discusses why utilizing health data is important for research purposes, as well as the challenges she faced when conducting her own research. Maria’s thesis examines the use of administrative health and social care data for palliative and end-of-life care (PEoLC) research in Ireland, and was partially funded by the University of Limerick.

Why is it important to harness health data for research purposes in Ireland? 

It is recognised in Ireland, and elsewhere, that health data are a valuable asset that can improve health system performance, health care quality and drive scientific discovery that improves medical treatments and save lives (Sláintecare Report).  The type and availability of health data is a product of the health system from which it derives. We cannot know the value, strengths and limitations of Irish health and social care data for research purposes until we try to use it.

Why did you carry out this study and what was the aim? 

This study draws from our experience of using administrative health data for PEoLC research in cancer patients.  It extrapolates the knowledge gained from that work to evaluate the potential and challenges of other national non-cancer health and social care data collections for PEoLC research. Details of our earlier studies can be found here and here.

How many datasets did you identify? 

We identified nine datasets with potential for PEoLC research. These included population based death certificate data, hospital episode (HIPE) data, pharmacy claims (PCRS) data, one national survey, four disease registries, (cancer, cystic fibrosis, motor neurone and interstitial lung disease) and a national renal transplant registry.

How can these health data sources be used for palliative and end-of-life care research? 

80% of deaths in Ireland between 2007-2011 were from conditions recognized as having associated palliative care needs. These included deaths from cancer, neuro-degenerative diseases and dementia. Palliative care research is difficult, there are ethical concerns for vulnerable patients who are often seriously ill so that identifying and recruiting subjects is difficult. There are also methodological concerns such as loss to follow up. Some of these issues can be addressed using administrative health and social care data.   

The general data protection regulations (GDPR) introduced in 2018 together with the health research regulations (HRR) provides a legal and ethical framework to determine who can access patient data and for what purposes. The datasets we identified, particularly the disease registry datasets and patients transplant databases concern individuals with life-limiting illness. With appropriate permissions, these datasets on their own or linked datasets to others, (e.g. HIPE, PCRS, death certificate data), could be used to explore the care these patients receive. Using administrative health data in this way, we might identify gaps in palliative care service provision and how services could be improved.  

Are there any potential challenges? 

In Ireland’s mixed public private health system, data outside the public system are governed by individual institutions so identifying who owns the data and the mechanism for accessing it can require an individual approach to each institution. This is further complicated where an institution is financed from multiple sources, for example the Health Service Executive (HSE), privately and through fundraising. The fragmented health information infrastructure means patients cannot be easily tracked from hospital to community based care. There is a paucity of community and social care data which is particularly relevant for PEoLC research since a considerable amount of palliative care is delivered in the community. There is little standardisation across institutions collecting data, so that even if the data were available for sharing, linking data from different sources would require considerable expert knowledge and oversight.  

Unlike countries with universal healthcare systems where population-level insurance claims databases have been used to identify receipt of palliative care via specific palliative care claim codes, we cannot reliably identify who has received palliative care in Ireland. Currently HIPE is the only population level administrative data source in Ireland where palliative care is recorded.

What changes can help support the use of available administrative health data for research on palliative and end-of-life care? 

Along with the GDPR and HRR regulations several recent developments will support the use of administrative health and social care data for research, including PEoLC research. These include the HSE and Department of Health’s eHealth strategy. Currently unique health identifiers for patients and service providers are being implemented. Development of electronic health care records, (EHR) is also ongoing.  These developments will improve data privacy and data sharing. The eHealth Ireland Open Data Portal simplifies finding and accessing data from across the Irish Health Sector and the Sláintecare 10 year reform plan promises continued support for the eHealth strategy. Integration of postcodes to the national data collections will allow investigation of regional disparities in access to palliative care services.

Palliative care is a very broad discipline so that identifying the delivery of palliative care from administrative health and social care data is challenging.  The introduction of EHR’s should facilitate the use of advance care plans where a patients’ preferences for care are recorded as part of their EHR.  Documented advance care plans will allow actual care received to be measured against the patients’ stated preferences. In this way PEoLC can be more fully examined.

What does open research mean to you, and why did you decide to publish your Research Article on HRB Open Research? 

Open research includes free access to published articles, access to the peer review process and where possible making raw data openly available to allow replication and verification of study results. The results from this study, using Irish national health and social care data collections for PEoLC research, are most relevant to an Irish audience.

The HRB’s efficient and easy to use publication platform was an obvious choice to reach that audience. Peer review brings a fresh perspective and expert knowledge to the subject matter and, I believe, improved our published article. Open peer review demonstrates the process.

Read the Research Article to learn more about the study and how Maria and colleagues hope it will facilitate increased use of  administrative health and social care data given the continued roll out of the healthcare reforms currently underway.