Working Group to Push Wireless Handheld Applications in Medicine

PDAA formative meeting was held November 30 at UCSD's School of Medicine to promote progress in the application of wireless handheld applications in medicine. The meeting was organized by Leslie Lenert, M.D., to inform the university medical community about current initiatives that the UCSD Department of Medicine and Scripps Clinic have taken to apply this technology. In line with its "living in the future" philosophy, Calit² plans to work with these groups to map relevant research projects to potential testbed clinical environments.

Bill Griswold, Calit² layer leader for Interfaces and Software Systems and a member of the Computer Science and Engineering department at UCSD, addressed the benefits of wireless applications for the university. In particular, he described his Active Campus project and how its principles might be adapted to the clinical world. David Wetherhold, head of the Division of Internal Medicine at Scripps Clinic, discussed his efforts to implement wireless technologies within the current clinical system. And Brandon Savage, associate director of the residency program at the UCSD Medical Center-Hillcrest, and Lenert discussed how location-aware technologies can be used in medicine.

The Active Campus project was conceived to address a number of looming issues at UCSD: The campus is expected to grow by 10,000 students over the next 10 years, class sizes will grow at least temporarily (reducing participation and intimacy), more students will be commuters, and students will interact increasingly through cell phones. All these trends indicate likely disintegration of the campus community. How can technology address this problem? Active Campus is based on the idea that location-based applications can enhance the campus community's "culture of learning." Its goal is to point students back into the physical world, not draw them further into handheld devices.

Griswold's group is developing several applications. One, called Active Class, enables students to ask questions of the instructor, which are then prioritized by the class as a whole. A second application, Active Campus Explorer, uses location, time, personal interests, and the Web to notify students of nearby interesting activities and relevant information (such as faculty office locations) as they are walking around campus. Some 400 freshmen in Computer Science and Engineering are receiving Hewlett-Packard Jornada handheld devices to experiment with these applications and develop additional ones. The entering class of Sixth College, likely to be less technologically savvy, will be equipped similarly next fall.

In a presentation entitled "Using Future Wireless Technologies with the Current Clinical System to Bridge the Gap to the Future," Wetherhold talked about the need for "evidence-based medicine" and bringing relevant information (via a handheld) to the point of care. For the medical community to embrace technology in large numbers, the devices need to be small enough, powerful enough, and, perhaps most importantly, integrated with legacy systems. Wetherhold's current solution (for a startup cost of about $1000/user) is the Compaq iPAQ Pocket PC. It can accept 802.11 PC cards, has sufficient memory capacity, can integrate with a PC network, and the medical applications software base is growing. These devices gain access to legacy systems and applications via a Citrix Windframe client. When asked about security, Wetherhold said his system uses a 128-bit wireless equivalency protocol, which is the best available but still not sufficiently robust.

Lastly, Savage discussed the importance of contextual handheld wireless computing and location-aware applications in health care. As the complexity of information for physicians is increasing, the points of access to that information become the "rate-limiting step." So any method of increasing the ability of the physician to access that information (to increase patients' understanding, increase detection of potential errors in diagnosis, etc.) ought to improve health care. Sources of information are myriad: the patient, the hospital computer system, test results, bedside charts that document nurses' care, radiology reports, etc.

Handhelds eliminate the rate limitation on accessing information; they're also low cost and take up minimal space (by comparison with computer workstations). Pitfalls, of course, include lack of sufficient screen real estate (which makes it difficult to navigate information) and lack of an easy method for data input. So Savage believes handhelds are actually better for distributing than recording information. Once the information is distributed, though, it needs to be navigated; that problem can be addressed by providing context-specific information. Physicians experience many different contexts with different pieces of information in the course of a day, including the physician's own context (the physician's specialty and personal information), patient context, time context (different tasks are typically performed at different times of the day), location context (e.g., how far is it to the Radiology Department?). Savage has identified several projects of interest:

  • Development of a prioritized patient list, in which patients are listed in order of their distance from the physician to optimize the physician's path or , if relevant, the degree of severity of their conditions.
  • Distribution of information that requires an action (e.g., hospital alerts; they should be sent to physicians when they're most likely to be able to respond).
  • "Pre-rounding" during which physicians optimize the process of gathering the necessary information before going on rounds.
  • Context tracking (as a physician navigates through a normal day, relevant information is pushed to her/him depending on what "context" s/he finds him/herself in; the context, in fact, might be controlled from the handheld).
  • Creation of "context XML" to characterize contexts and determine which contexts are applicable to a given individual (with critical values reported immediately).

The assembled group agreed to continue as a working group to exchange ideas and identify clinical "testbed environments" to implement technologies. Plans call for monthly meetings. Next topics of interest are medical applications for the consumer and adapting wireless, handheld technology to emergency scenarios.

Related Links
School of Medicine, UCSD
Interfaces and Software Systems layer