Big Data and End-of-Life Care: Promise and Peril

End-of-life care in general practice: clinic-based data collection

BACKGROUND

There are no processes that routinely assess end-of-life care in Australian general practice. This study aimed to develop a data collection process which could collect observational data on end-of-life care from Australian general practitioners (GPs) via a questionnaire and clinical data from general practice software.

METHODS

The data collection process was developed based on a modified Delphi study, then pilot tested with GPs through online surveys across three Australian states and data extraction from general practice software, and finally evaluated through participant interviews.

RESULTS

The developed data collection process consisted of three questionnaires: Basic Practice Descriptors (32 items), Clinical Data Query (32 items) and GP-completed Questionnaire (21 items). Data extraction from general practice software was performed for 97 decedents of 10 GPs and gathered data on prescriptions, investigations and referral patterns. Reports on care of 272 decedents were provided by 63 GPs. The GP-completed Questionnaire achieved a satisfactory level of validity and reliability. Our interviews with 23 participating GPs demonstrated the feasibility and acceptability of this data collection process in Australian general practice.

CONCLUSIONS

The data collection process developed and tested in this study is feasible and acceptable for Australian GPs, and comprehensively covers the major components of end-of-life care. Future studies could develop an automated data extraction tool to reduce the time and recall burden for GPs. These findings will help build a nationwide integrated information network for primary end-of-life care in Australia.

Technical Blossom in Medical Care: The Influence of Big Data Platform on Medical Innovation

Medical innovation has consistently been an essential subject and a source of support for public health research. Furthermore, improving the level of medical research and development is of great concern in this field. This paper highlights the role of big data in public medical innovation. Based on a sample of China’s listed firms in the medical industry from 2013 to 2018, this paper explores the exogenous shock effect of China’s big data medical policy. Results show that the construction of the medical big data platform effectively promotes innovation investment and the innovation patent of medical firms. In addition, the heterogeneity of this promoting effect is reflected in firm size through the overcoming of different innovation bottlenecks. The research conclusions support the positive significance of the macro-led implementation of the medical big data platform, and suggest that the positive economic externalities generated by this policy are critical to public health.

Improving palliative and end-of-life care with machine learning and routine data: a rapid review

Introduction: Improving end-of-life (EOL) care is a priority worldwide as this population experiences poor outcomes and accounts disproportionately for costs. In clinical practice, physician judgement is the core method of identifying EOL care needs but has important limitations. Machine learning (ML) is a subset of artificial intelligence advancing capacity to identify patterns and make predictions using large datasets.  ML approaches have the potential to improve clinical decision-making and policy design, but there has been no systematic assembly of current evidence. Methods: We conducted a rapid review, searching systematically seven databases from inception to December 31st, 2018: EMBASE, MEDLINE, Cochrane Library, PsycINFO, WOS, SCOPUS and ECONLIT.  We included peer-reviewed studies that used ML approaches on routine data to improve palliative and EOL care for adults.  Our specified outcomes were survival, quality of life (QoL), place of death, costs, and receipt of high-intensity treatment near end of life.  We did not search grey literature and excluded material that was not a peer-reviewed article. Results: The database search identified 426 citations. We discarded 162 duplicates and screened 264 unique title/abstracts, of which 22 were forwarded for full text review.  Three papers were included, 18 papers were excluded and one full text was sought but unobtainable.  One paper predicted six-month mortality, one paper predicted 12-month mortality and one paper cross-referenced predicted 12-month mortality with healthcare spending.  ML-informed models outperformed logistic regression in predicting mortality but poor prognosis is a weak driver of costs.  Models using only routine administrative data had limited benefit from ML methods. Conclusion: While ML can in principle help to identify those at risk of adverse outcomes and inappropriate treatment near EOL, applications to policy and practice are formative.  Future research must not only expand scope to other outcomes and longer timeframes, but also engage with individual preferences and ethical challenges.

Improving palliative and end-of-life care with machine learning and routine data: a rapid review

Introduction: Improving end-of-life (EOL) care is a priority worldwide as this population experiences poor outcomes and accounts disproportionately for costs. In clinical practice, physician judgement is the core method of identifying EOL care needs but has important limitations. Machine learning (ML) is a subset of artificial intelligence advancing capacity to identify patterns and make predictions using large datasets.  ML approaches have the potential to improve clinical decision-making and policy design, but there has been no systematic assembly of current evidence. Methods: We conducted a rapid review, searching systematically seven databases from inception to December 31st, 2018: EMBASE, MEDLINE, Cochrane Library, PsycINFO, WOS, SCOPUS and ECONLIT.  We included peer-reviewed studies that used ML approaches on routine data to improve palliative and EOL care for adults.  Our specified outcomes were survival, quality of life (QoL), place of death, costs, and receipt of high-intensity treatment near end of life.  We did not search grey literature and excluded material that was not a peer-reviewed article. Results: The database search identified 426 citations. We discarded 162 duplicates and screened 264 unique title/abstracts, of which 22 were forwarded for full text review.  Three papers were included, 18 papers were excluded and one full text was sought but unobtainable.  One paper predicted six-month mortality, one paper predicted 12-month mortality and one paper cross-referenced predicted 12-month mortality with healthcare spending.  ML-informed models outperformed logistic regression in predicting mortality but poor prognosis is a weak driver of costs.  Models using only routine administrative data had limited benefit from ML methods. Conclusion: While ML can in principle help to identify those at risk of adverse outcomes and inappropriate treatment near EOL, applications to policy and practice are formative.  Future research must not only expand scope to other outcomes and longer timeframes, but also engage with individual preferences and ethical challenges.

Secondary Analysis of Existing Datasets for Dementia and Palliative Care Research: High-Value Applications and Key Considerations

OBJECTIVE

To provide a guide to researchers selecting a dataset pertinent to the study of palliative care for people with dementia and to aid readers who seek to critically evaluate a secondary analysis study in this domain.

BACKGROUND

The impact of dementia at end-of-life is large and growing. Secondary dataset analysis can play a critical role in advancing research on palliative care for people with dementia.

METHODS

We conducted a broad search of a variety of resources to: 1. identity datasets that include information germane to dementia and palliative care research; 2. review relevant applications of secondary dataset analysis in the published literature; and 3. explore potential validity and reliability concerns.

RESULTS

We synthesize findings regarding: 1. Methodological approaches for determining the presence of dementia; 2. Inclusion and measurement of key palliative care items as they relate to people with dementia; and 3. Sampling and study design issues, including the role and implications of proxy-respondents. We describe and compare a selection of high-value existing datasets relevant to palliative care and dementia research.

DISCUSSION

While secondary analysis of existing datasets requires consideration of key limitations, it can be a powerful tool for efficiently enhancing knowledge of palliative care needs among people with dementia.