The Promise of Data - Day Two - Customization, Collaboration and Solutions

Search

Loading...

News

Latest News

Mar 24, 2015
by Louise Hallman
The Promise of Data - Day Two - Customization, Collaboration and Solutions

How can big data help individual patients and populations at large?

Elliot Fisher of the Dartmouth Institute leads a panel in discussing the value of big data for populations at large

The amount of data available to 21st century clinicians is vast, but how best can this be analyzed and how can this analysis be best applied to both the individual patient and populations at large?

Customizable health care

Tackling issues surrounding the collection, use and application of data for individual patients were panelists Jens Deerberg-Wittram from the Executive Board of ICHOM (the International Consortium for Health Outcomes Measurement) and Sofia Ernestam from the Karolinska Institutet, a medical university in Sweden.

The advent of big data means now that doctors have more than just their own knowledge and patient data sample to draw upon; data is being collected and shared not only within countries but also across borders, too. 

It also means that patients can become more aware of their condition and treatment options, but also leads to the possibility of self-misdiagnosis and resulting anxiety.

Generating more precise big data requires the collection of “small data” directly from patients. In Sweden, for example, where there are over 100 information registries, patient trust of this data collections has been gained through clear consent for collection and ethical guidelines for usage. 

Ensuring that this data can be used globally means harmonization and standardization of data collection and analysis is needed. Even when this data is standardized, what is useful and important to a doctor may not be what concerns their patient most. 

ICHOM has a mission to “define a global standard of outcome measures that really matter to patients”; this means measuring more than just the ultimate outcome – did a patient survive after an intervention? – but also all the other outcomes, including side effects, a patient must live with. 

Using the example of prostate cancer treatment, a patient is likely to not only be concerned with whether he survives the cancer treatment (be that surgery, chemo- or radiotherapy) but also whether or not he will experience urinary or erectile dysfunction as a result of a chosen course of treatment.

Just as patients and doctors value different outcomes differently, different patients have diverse values, too. Data does not account for the different risk assessments of patients – this can only be ascertained through shared decision making with the (human) doctor and patient.

Data collaboration

The challenge no longer is from where can we get data, but which data should we use and how can they be used together?

Addressing these questions and the issues surrounding how population health and health care can benefit from the enhanced availability and management of data were representatives from large corporate- and publicly-funded data collection projects.

Big data collection and analysis can enable better prediction of and thus response to epidemics, measurement of the efficacy and safety of new drug treatments, reduction of wasteful and harmful interventions, and cost-savings in resource-strapped health services. 

In the US, this data is being derived not only from the extensive claims records (collected from over 148 million Americans from a total population of 315 million), but also clinical records, government records, consumer behavior data, employement data, demographic data and genomic data. 

For both the corporate-funded and the public-backed/academic-led data analysis projects, collaboration between multiple data collectors and analysts is vital to ensuring accurate and valuable applications of the data. However, this can be difficult given not only the differences in methodologies and purposes, but also regarding data ownership and patients’ privacy concerns. Acquisition of this data can require monetary or credit-sharing incentives, and can raise ethical questions. 

One main takeaway for many of the participants in Salzburg is that patient-centeredness, not corporate goals, should remain key.

Data-driven health care solutions

This vast amount of data collected and analyzed in health care is like a gold mine: there are pieces of significant value, but it takes work to find them. How to find and apply this knowledge more quickly was the topic for the final panel of the day.

As one panelist pointed out, there is an average of 17 years between a medical discovery and its application as a best practice, and much of the health data that is generated has a reserach half-life of only five years. 

“Yes we need to act upon data more quickly,” admitted the speaker, “but we have to do this smartly too.”

One way of being smarter about this action is to better engage the “end users” of the data: the doctors and the patients. However, four main barriers exist to transferring this knowledge into best practice: 1. clinicians and patients are unaware that such knowledge exists; 2. a lack of resources, including time, hinders them from finding out more about such knowledge; 3. once the time is taken to read any of the 2000+ scholarly medical papers published everyday, the reader lacks understanding or information of the context necessary to put this knowledge into practice; 4. there is a lack of motivation to put new methods into practice, with many individuals assuming that these changes must implemented first on a policy or institutional level rather than leading the change themselves. 

Two of the main methods in which data is being put to use by clinicians is through predictive risk models and decision-making tools.

Predictive risk models can be used, for example, in anticipating return admissions of patients. While this can help allocate resources more effectively, it is not foolproof and can generate false positives and false negatives that can lead to either unnecessary and potentially harmful interventions on one hand, or incorrect non-treatment on the other.

Decision making tools can combine an individual patient’s condition with larger data to enable a clinician to reach a decision about the best course of treatment. However, some of the preferences built into these tools are “hidden” rather than explicit and transparent, and therefore are not adjustable as they should be by the clinician together with the patients’ input. Thus these tools should enhance – not dictate – the human doctor’s decision. 

An analogy can be found in a common decision-making tool: an in-car GPS/sat nav system – don’t follow its directions off a cliff!

Download today's newsletter


To read and join in with all the discussions in Salzburg, follow the hashtag #SGShealth on TwitterFacebook and Instagram. The session The Promise of Data: Will This Bring a Revolution in Health Care? Is part of the Salzburg Global series “Health and Health Care Innovation in the 21st Century” and is being held in collaboration with theMayo ClinicArizona State UniversityThe Dartmouth Center for Health Care Delivery Science, and in association with the Karolinska Insititutet.

Related Content

The Promise of Data - Will This Bring a Revolution in Health Care?

Nov 10, 2015

Andrew Muhire - "Cholera and Other Diseases Don't Last Long in Rwanda"

Apr 23, 2015

Hamish Tomlinson and Yan Yu - How Can We Teach Students About How Big Data Is a Huge Opportunity to Improve the Health Care of Patients?

Apr 22, 2015

Birgir Jakobsson - "We Have a Unique Situation with Our System to Get Health Care That Is Really State of the Art"

Apr 22, 2015

Elliott Fisher - "How Can We Use Big Data to Improve the Value of Health Care?"

Apr 22, 2015

Amel Farrag - "I Thank Salzburg Global Seminar for Giving Me This Chance to Learn from These Expert People"

Apr 21, 2015

Jörgen Nordenström - The Purpose of Big Data in Health Care Is to Improve Value for the Patient

Apr 21, 2015

Keith Lindor - "People Want Health and We Sell Them Health Care"

Apr 20, 2015

Veronique Roger - "The Highlights Are the Desire, Passion and Commitment to Action That Came out of This Week"

Apr 14, 2015

Al Mulley - "I Have Found Salzburg to Be a Wonderful Place to Cross Contexts and Borders"

Apr 13, 2015

Sara Riggare - "At the Quantified Self I Forget I Have Parkinson's Disease"

Apr 13, 2015

The Promise of Data - Day Four - Ownership and Possession of Data

Mar 26, 2015

Salzburg Global Fellows Call for Big Data Revolutions in Health Care

Mar 31, 2015

The Promise of Data - Day Three - Generational Differences, Patient Expertise and Health Equity

Mar 25, 2015

Bertalan Mesko - Is 3D Printing the Most Disruptive Trend for the Future of Medicine?

Mar 24, 2015

The Promise of Data: Will this Bring a Revolution in Health Care?

Mar 22, 2015

The Promise of Data - Will this Bring a Revolution in Health Care?

Mar 22, 2015

The Promise of Data - Day One - Hopes and Fears

Mar 23, 2015