Health IT, Hospitals, Startups

Two ways companies overcame health IT data challenges with better visualization

Data seems to spill out of so many parts of healthcare, but in many cases […]

Data seems to spill out of so many parts of healthcare, but in many cases it is tough to use it to great effect because the data is in a format that doesn’t easily lend itself to highlighting problems or figuring out solutions. Some of the companies that make up DreamIt Health’s third accelerator class took different approaches to surmounting this challenge and they were on display at its demo day this week.

The move to link reimbursement with patient satisfaction scores has meant hospitals are paying more attention to the patient experience in their hospitals. But the downside of these scores, as NarrativeDx sees it, is that simply ranking various aspects of their care varying levels of good or bad positive or positive or negative only tells part of the story. A poor score on nurses’ listening skills offered no context behind it. So the health IT company developed a way to include that information with natural language processing. The effect, as one case study suggests, is of painting a picture between two dots.

The health IT company was one of nine companies that took part in DreamIt Health’s demo day this week.

At one hospital, patients delivered poor grades to nurses because they didn’t seem to do anything to help them feel more comfortable, despite the fact that they are freezing. Why? The hospital was low on blankets. So when patients asked nurses for them, and they couldn’t deliver, the nurses got the blame.

The company’s analytics tool uses natural language processing as part of hospital surveys and also to dig through social media, such as reviews on Yelp, to identify clues. The idea is that hospitals can use the feedback to solve relatively simple problems or at least get a better grasp of what is motivating patient responses.

 

 

Haystack, a company that wants to do a better job of identifying patient data breaches inside hospitals, comes out of Children’s Hospital of Philadelphia as part of a special program with DreamIt. The primary factor that makes it so complicated to get to the bottom of these IT whodunnits is the paper trail tends to come in the form of a spreadsheet where words and numbers are organized like they’re staging a military coup against the paper. Adrian Talapan, CEO and co-founder, explained that it wanted to establish a baseline of who would be accessing each patients data on a normal basis to make it easier to distinguish them from outliers — those people who wouldn’t usually access  a given patient’s record.

It doesn’t necessarily mean that each incident is a breach, but it creates the ability to check and respond to these incidents easier and faster. The breaches could be caused by simple curiosity — wanting to know how a neighbor’s surgery went or how they’re lab results came out  came out or something more nefarious.  It has made a convincing case to some hospitals — it is poised to start pilots at four in the coming months.

As Talapan noted in his pitch, “Patient records are much more valuable than credit cards yet cards are much better protected.” But anyone who has been paying attention to the big box data breaches lately knows that even those card protections are limited.

 

Shares0
Shares0