Critical Questions in the Governance of Health Care Big Data
by Richard Self, University of Derby.– 26 November 2014.
Various reports claim that the use of Big Data Analytics in health care can improve efficiency by significant amounts. A 2011 report by McKinsey suggests that the USA could save some $300B by leveraging Big Data.
However, in order to achieve this, a wide range of challenges need to be overcome, as are discussed in the report from the Institute for Health Technology Transformation which evaluated the issues for the USA.Various reports claim that the use of Big Data Analytics in health care can improve efficiency by significant amounts. A 2011 report by McKinsey suggests that the USA could save some $300B by leveraging Big Data.
The Vs of Big Data can offer some insights to these challenges.
A critical problem in the Health Care industry is that there is very little standardisation of the systems and data structures. All participants tend to operate a wide range of different systems, with few standardised APIs to enable integration of the data for a single individual.
Additionally, there are many legacy systems which have little compatibility either between themselves or with more modern systems. History also teaches us that modifying stable legacy systems tends to introduce fragility into the overall operations.
Whilst the focus of the report is towards the overall national benefit of $300B, there are also questions of value at a more atomised level, such as the patient, the hospital, the pharmacist, the Big Pharmas and at many other levels. Many of these benefits can only be gained by explicit knowledge of patient identifying details, where existing levels of anonymisation cannot be guaranteed to unsure the permanent hiding of individual identity, given the power of Big Data Analytics to associate data from across the Cloud and internet to identify individuals.
Security of data is clearly of continuing concern. During 2014 alone, there are reports of come 800M sets of ID data having been extracted from a range of organisations by a range of exploits. Identity theft is becoming one of the great issues. A report today (26 November 2014) identifies that one UK University has just lost approximately 125k sets of personably identifiable data relating to applicants and students when a laptop was stolen from a car.
The involvement of so many organisations in the health care chain will, inevitably, lead to security problems.
Critical types of data in the health care chain include:-
- Personally identifiable information
- Clinical Data
- Financial Data
- Behavioural Data
The loss of any of this data will be significant if it can be attached to a named individual.
Necessary Governance Strategies
The report offers some guidance for the future developments of health care Big Data, key features of which are:-
- Implement a strong Governance framework
- Engage the Providers
- Ensure that Analytics is embedded into training
- Provide for flexibility in information delivery
- Where possible, choose in-house solutions for their ability to meet local needs
- Develop simple tools to visualise the data for clinicians
- Close the quality look