Projects, designs, and writings on health IT

2014-09-02

Guide to Hacking the Healthcare System

9:31 PM Posted by David Do, MD No comments
By DAVID DO, MD AND CRAIG MONSEN, MD

Originally appeared in The Health Care Blog on 2014-07-30

We are residents and a software developers. Before starting residency, we spent time as software developers in the startup community. We were witness to tremendous enthusiasm directed at solving problems and engaging people in their health. The number of startups trying to disrupt healthcare using data and technology has grown dramatically and every day established healthcare companies appear eager to feed this frenzy through App and Design Competitions.

When we started residency, the restrictiveness of data and reliance on decades-old technology was grossly apparent. Culturally, hospitals are an environment of budgets and deadlines that are better suited towards maintaining the status quo than promoting the creative process. Hospital IT departments harbor a deep cover-your-rear-end mentality and are incentivized for two things: first, keep systems running, and second, prevent security breaches. Perhaps rightly so–privacy and security need to be prioritized–but this environment has delayed them from facing the inevitable challenges of effectively using their own data and investing in new tools, including ones that could improve the triple aim of greater quality care with greater patient satisfaction and lower cost.

In the future, as hospitals and health systems become more accountable for the long term outcome of patients, we are optimistic that they will innovate as much out of cost-cutting necessity as for providing a better product to patients. We have little evidence that established players can power innovation solely on their own engines and expect many of the solutions will come from problem-solvers outside medicine. Doctors and patients will choose from an arsenal of apps to interact with the health information in EMRs. These healthcare apps come in three major categories: education, workflow, and decision support.

Patient Education Apps
Patients are the most underutilized resource in health care. After a hospital visit, discharge summaries and medication lists that we send patients home are entirely disconnected from the tremendous amount of educational health material available to them for free on the Internet. Many patients are discharged from a hospital with more than ten medications, and rarely will they begin the instruction necessary to fully understand the indications and dosing instructions for each. Imagine sending a patient home with a smart medication list that reminds them when to take each medication, along with a description of why its beneficial. Discharges are just one opportunity where patient education can be improved and there are just as many efforts to improve education during outpatient visits and even before patients see the doctor in the first place.
The promise: Better patient education results in better patient engagement, compliance, and less ineffective care
Existing apps and services: iTriage (symptom-checking), Aircare (post-discharge continuing education), Emmi Solutions (procedure informed consent)
Data sets available for developers: Visible Human Project API, Pillbox API, Medline Plus, to name a few
Data sets not yet available: Individualized patient history and medication lists
Prior app Competitions: Health Design Challenge (http://healthdesignchallenge.com/), RWJF Aligning Forces for Quality App Challenge, Cigna Innovation App Challenge.

Workflow Apps
Hospitals pay tens of thousands of dollars for data-management applications that do a fraction of what Google Docs, for example, can do. There are many areas in the hospital for which niche software can make a huge impact, many of which are basic data-management or communication applications that allow collaboration and are HIPAA-compliant. Another area is in data visualization; current off-the-shelf EMRs organize data vertically (by the way they were collected, e.g. labs, radiology, nursing flowsheets, vital signs, etc). To make accurate medical decisions, doctors need to view data horizontally (e.g. to make a decision about fluid status, providers need to see elements from all over the categories above), which is challenging in the current systems. EMRs have been engaged to solve the health care enterprise concerns: security, billing, and data persistence, often at the cost of clinician efficiency. The backlash is already apparent and there will be a growing opportunity to use technology to make care delivery more efficient.
The promise: With better workflow tools and data visualization, physicians spend less time with technology and more time with patients.
Existing apps and services: Cureatr (provider-provider text messaging), Aidin (one-click discharge planning), Box (HIPAA-compliant cloud storage)
Data sets available for developers: N/A
Data sets not yet available: specific to the project
App Competitions: none

Decision Support Apps
Hospitals and health systems collect tremendous amounts of data from patients including vital signs, laboratory results, and imaging. The sheer volume of data prevent providers from integrating everything to make optimal decisions. Decision support often takes the form of using data from many patients to predict the needs of an individual patient. For example, when a patient develops sepsis or requires intubation, retrospective analysis often reveals clues of impending deterioration. Predictive analytics could help detect these conditions earlier and warn providers. Similarly, apps synthesizing hospital-wide microbiology data in real-time could help providers choose effective antibiotics earlier.
The bottom line: Using technology to synthesize large amounts of data can lead to better patient outcomes and safety
Existing apps and services: there are numerous publications on medication dosage-checking, Lockheed Martin is piloting sepsis-detection at several institutions, Explorys does population management
Data sets available for developers: Medicare claims data, HCUP
Data sets we need access to: ICU data sets (exist, but are generally for closed research)
App Competitions: Heritage Health Prize

David Do, MD is a resident at the Hospital of the University of Pennsylvania. Craig Monsen, MD is a resident at the Brigham and Womens Hospital. They are agile software developers and founding members of Symcat.com.

0 comments:

Post a Comment