Machine Learning Times
Machine Learning Times
EXCLUSIVE HIGHLIGHTS
Some Thoughts on Analytics in a Post COVID-19 Environment
 In these most difficult times, the use of analytics...
Re-examining Model Evaluation: The CRISP Approach
 The performance of prediction models can be judged using...
An Agile Approach to Data Science Product Development
 Introduction With the huge growth in machine learning over...
Wise Practitioner – Predictive Analytics Interview Series: Haig Nalbantian at Mercer – BIZ
 By: Eric Siegel, Founder, Predictive Analytics World for Business...
SHARE THIS:

7 years ago
Making Better Predictions about Patient Health

 

 
In the future, doctors will be better able to help us ward off illnesses by analyzing information buried in medical records. That effort will be greatly enhanced by the ability to access electronic medical records in so-called health information exchanges. Those exchanges could give researchers access to large pools of anonymized health data, so they can more accurately spot trends.

One impediment is that records today are often siloed in individual, medical systems. A group of five electronic medical record providers, Tuesday, said they would cooperate to make their systems better work together. This could be one of many steps that could make it easier to conduct analytics that could help doctors more accurately predict which patients might be at risk.

“The potential for data mining to improve the way we deliver health care is enormous, and I predict that the impact on human health could equal that of vaccines,” writes Kaiser Permanente CIO Phil Fasano in his recently published book, “Transforming Health Care.”

Kaiser Permanente already uses information gleaned from its patient databases to conduct research that it hopes will improve patient care. For example, an analysis of a Kaiser database with 1.4 million members led to the discovery that the pain reliever Vioxx was dangerous and ultimately the drug was taken off the market, according to Mr. Fasano. That discovery required a large dataset and the right tools to find lifesaving information. Another Kaiser study found that as few as three to four alcoholic drinks per week increased breast cancer recurrence by 30%.

“The advent of electronic medical records and health information exchanges means that we now have the power to collect enormous amounts of medical data and make it instantly accessible across large geographical ranges, with profound implications,” writes Mr. Fasano.

Information gleaned from health information exchanges would be able to pull together records from the various doctors a patient sees, giving researchers a clearer picture of patient health than a single medical record from one physician, says Scott Lundstrom at IDC Health Insights.

Health care organizations are currently investing in software that will help them predict outcomes. “Analytics is the number one new investment for payers and providers over the last two years,” said Mr. Lundstrom.

Using analytics, NorthShore University HealthSystem was able to find that the norm temperature for post-operative patients is 100.3 degrees Fahrenheit. Typically, a temperature above 98.5 degrees Fahrenheit would be cause for concern. With this information, NorthShore was able to cut post-operative antibiotic use for patients, writes Mr. Fasano.

In the future, Mr. Fasano predicts that medical professionals will be better able to spot emerging risks, discover distant early warning signs of disease in order to stop patients from becoming ill.
 
By: Rachael King, Reporter
Originally published at The Wall Street Journal

Leave a Reply

Pin It on Pinterest

Share This