Speaking with my colleague Ki Mae Heussner at GigaOM’s Structure:Data conference on Wednesday, Aetna’s head of innovation, Michael Palmer, opened up about the Innovation Labs’ initiatives, the challenges they face and the neat opportunities ahead.

The Innovation Labs has focused on a few conditions since starting last year, including cancer, heart disease and metabolic syndrome. There are five telling signs to metabolic syndrome, including a large waist circumference and high blood pressure. Metabolic syndrome is a sort of gateway for diabetes, cardiovascular issues and other conditions, so Aetna wants to prevent patients from getting metabolic syndrome in the first place. It has started using data on 18 million of its customers’ employees to tell doctors which of the five signs of metabolic syndrome their patients are likely to get in the next year, Palmer said.

Aetna appears to want to pull in more kinds of data for patients and physicians to examine to provide better care. Asked if Aetna will start working in genomic data, Palmer said the Aetna Innovation Labs are running a pilot with some companies to allow patients to get their own genetic information and receive genetic counseling. “That will drive a wellness program driven around genomics,” he said.

And social information from patients could provide much-needed feedback to show how effective a medication is in real time, for that patient and for others. “Some companies are doing that,” Palmer said, referring to incorporating social streams into health care. “The challenge is, a lot of HIPAA laws prevent the ability to connect those two in a way that would be ultimately useful,” he said.

Aside from regulatory compliance, patients simply might not want to share as much data as insurers and doctors want them to, as it could get out to the wider public. Keeping health data out of employers’ hands, for example, is playing out right now, as Ki Mae mentioned, with media outlets reporting on Wednesday that CVS employees provide health information or pay up. But Palmer said Aetna has been working with large patient data sets for many years, though, and is cognizant of complying with HIPAA laws.

It’s ironic, but Aetna — and probably others in the health care community — finds a challenge in getting people to work on their own health. Companies want to see that, so their employees will stay healthier and work more, Palmer said.

Toward that end, Aetna is looking to work with other companies, mainly small startups, on what Palmer called “medication adherence” — texting, calling or otherwise contacting patients about taking their medications and doing other things to improve their health. It’s not easy, though, because, as Palmer said, what works for some patients doesn’t necessarily work for others.

Check out the rest of our Structure:Data 2013 coverage here, and a video embed of the session follows below:

A transcription of the video follows on the next page

Go to page 2 (of 2) on GigaOM .

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Connected world: the consumer technology revolution
• A near-term outlook for big data
• Sector RoadMap: Social customer service in 2013

Michael Palmer, formerly the Global Managing Partner of Accenture’s Medical Technology business, heads the effort and shared some thoughts on how the program works and what opportunities lie ahead in the world of big data and medicine. Aetna has three primary goals with its data strategy; patient safety, patient care and what he calls patient engagement, Palmer says. Patient safety involves screening for ineffective treatment plans or harmful suggestions in certain patient populations and could develop into something as real-time and useful as watching for medications that might react badly together.

On the patient care front the goal is getting doctors the best information. Maybe it’s data that shares the optimal cancer treatment for a specific type of cancer. Because there is genetic component to some cancers as well as drugs that can better treat specific types, doctors can use data analysis to find more effective treatments, although Palmer estimates that using custom genetic information in treatments is still three to five years away and would require doctors to get new training.

Big data helps in big decisions

Michael Palmer of Aetna

And finally, patient engagement encompasses the tricky ways to get people to change their behavior and live healthier lives. And here’s where Palmer’s work has borne the most fruit by making his big data small enough that it can change the life of an individual. Take for example, metabolic syndrome, the catch-all name for a group of maladies associated with heart attacks, strokes or diabetes. Symptoms can range from having a big waist, elevated blood pressure, low HDL and high glucose among others.

But symptoms are often difficult to treat because a patient hearing a diagnosis of metabolic syndrome will likely be told to eat right and exercise and perhaps be given a cholesterol medication. Aetna tried a program to help improve patient outcomes for those with metabolic syndrome but only 102 patients out of thousands completed the program and saw any improvement. That’s not lowering costs and it’s not improving many lives.

So Palmer’s first big task was to use Aetna’s data to take on metabolic syndrome — with more impressive results. The program worked with an independent lab that looks at patient results on a series of metabolic syndrome-detecting tests. Then Aetna crunched a patient’s data against all the data Aetna has on 36,000 other patients from the same employer. The pilot project involved two annual screenings held at the beginning of consecutive years and scanned through 600,000 lab results and 18 million claims events in a one-year period.

The result is a highly personalized treatment plan that assesses patient risk factors and focuses on treating one or two things that will have the most impact (statistically speaking) on improving their health. So for one patient a doctor might say, “Here’s a statin and lose five pounds around your middle and you will be 50 percent less likely to have a heart attack in the next decade,” while for another the doctor might say, “You are at a 20 percent risk of developing high glucose so you should focus on lowering your triglycerides.”

Personalized interventions yield results

So the data is used to personalize both the risk to the patient of suffering some catastrophic health event as well as the treatment designed to most effectively prevent that event for their statistical profile. That’s pretty powerful to hear as a patient and the Aetna data so far backs that up — based on the pilot project, 90 percent of patients who didn’t have a previous visit with their doctor would benefit from a screening, and 60 percent would benefit from improving their adherence to their medicine regimen.

“We’re now taking [that data] and making it actionable,” Palmer said. “Telling doctors to focus on two intervention programs per factor and looking at the right interventions. The customer medicine is already here.”

And the data involved wasn’t quite as mammoth as one might expect. Aetna worked with about 1.3 terabytes of data in a proprietary system for this project. Of course, that only covered one disease and 36,000 people, so anything pertaining to all Aetna customers could get unwieldy quickly.

Now that Aetna has seen its efforts bear fruit, Palmer said he plans research into natural language processing so he can incorporate notes written by doctors into Aetna’s data stores. As for new efforts to bring personalized medicine to Aetna’s customers, Palmer is evaluating different diseases or symptoms that cost Aetna a lot of money. That’s one reason the Aetna second data-related program is tied to cancer care — which is the most expensive disease from Aetna’s point of view.

Dedicating data to cancer care

“We spend $5 billion on behalf of our insured customers on cancer care,” Palmer said. “And a relatively small percentage of oncologists use evidence-based, well-proven interventions based on patient information.” Palmer’s second pilot has doctors use a program called Eviti to track their patient and the interventions.

The program will pull in data from the patient’s primary care physician and oncologist as well as cancer staging information and even genetic markets if applicable. Then it recommends a treatment plan. When the doctor suggests a treatment plan that is backed by the evidence provided by Eviti, Aetna pre-authorizes payment.

Both pilots show how data can change and improve medicine, for both patients and insurers. What’s less certain is how this data will continue to benefit a patient if he or she switches insurers, or how the data will benefit the uninsured. From a big picture-point of view most medical data is locked up in patient records, private insurers or in clinical trials which can be expensive to mine.

Sharing that data and providing the same standards or treatment codes will help take the benefits experienced by Aetna’s patients to the wider U.S. population. Of course, As Palmer explained in the interview, in many ways Aetna views these efforts as a way to improve customer experience and thus, retention, which means it might not want to share, even if the laws allowed it. As is usually the case with healthcare, a mish-mash of incentives, an imbalance of power and a lack of standards could make mining medical data an impossible dream.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• AWS Storage Gateway jolts cloud-storage ecosystem
• Dissecting the data: 5 issues for our digital future
• Connected world: the consumer technology revolution

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Who and what to watch in the new era of the living room
• 12 tech leaders’ resolutions for 2012
• The future of mobile: a segment analysis by GigaOM Pro

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Connected world: the consumer technology revolution
• Mobile 2012 and beyond
• CES 2012: a recap and analysis