Whew! I just finished reviewing 29 web-based and mobile applications designed to improve care transitions and reduce preventable hospital readmissions. This is the first of several blog pieces I will do to sum up my experience as a judge for the Ensuring Safe Transitions from Hospital to Home Challenge. It would have been fun to do this like Top Chef or Dancing With The Stars, booting one of the contestants off each week, with escalating drama. But this contest required that all of the apps be examined side-by-side in one evaluation round.
The contest is sponsored by the Office of the National Coordinator for Health Information Technology (ONC), in collaboration with the Partnership for Patients. Software developers were challenged to improve care transitions by creating an intuitive and easy-to-use application to empower patients and caregivers that fits into existing ways that providers communicate. The winner gets $25,000 plus tons of publicity for their app.
In looking at so many technical solutions as a group, I noticed some meta-issues about requirements, design, and business models that were not directly included in any of the challenge documentation. Some of these issues may be important to future design projects, so I will document them while they are fresh in my mind. I will look at some of the specific entries in more detail, but I will hold off on that until the winning team is notified on Wednesday, November 30, 2011. My comments will cover general architectural issues and design issues, and will not disclose any of the details that may be subject to embargo. In each post I will try to cover one or two of the thematic issues I want to cover.
First off, here are the judging criteria:
- Integrates design and usability concepts to drive patient adoption and engagement: Is the app user-friendly? Intuitive and easy to grasp? Interactive? Does the interface increase the app's efficacy? (weighted 25%)
- Demonstrates creative and innovative uses of mobile technology: How new or innovative is this solution? Are there similar apps or products already in the marketplace that address this health problem in a similar way? Does the technical implementation of the solution meet the technical standards as laid out in challenge? (weighted 15%)
- Demonstrate potential to improve health status for individuals and/or the community: How likely is it for patients and their families/caregivers to use the app to manage their health? Does the app promise to reduce hospital readmissions? Is the solution capable of changing the way patients and providers communicate and interact? Will it increase informed decision-making? (weighted 25%)
- Effectively integrates inpatient and outpatient data and provides structured support for self-care: How well does the application access available data sources? How well does the application provide support for discharge? (weighted 25%)
- Leverages NwHIN standards: Are NwHIN standards in content, vocabulary and transport included? (weighted 10%)
Who are the at-risk patients?
The first big issue I noticed with the submissions was that not all of the entries had a clear idea about who the high-risk patients are. Since this is the main focus of attention for this challenge it was surprising that so few of the entries were clear on the concept. Many of the apps seemed designed for yuppies who need surgery on runner's knees or problems with tennis elbow. I could see bright, self-empowered overachievers using the apps on their tablet computers while enjoying a soy latte at the nearest coffeeshop. Unfortunately for those developers, this is not where the problem lies.
While all patients deserve good discharge planning and followup services, a key to reducing preventable re-hospitalizations is to understand risk stratification. Only a few of the apps seemed to have a clue that risk stratification is the key to allocation of institutional support resources, which are expensive and limited in supply. Can we identify earlier on those patients who are most at risk for re-hospitalization and ensure they get the treatment they need? This is a really important issue for healthcare systems. If you are good at statistics and want to win a $3,000,000 prize, consider entering the Heritage Provider Network Health Prize competition. That contest is trying to come up with an algorithm to identify patients who will be admitted to a hospital within the next year, using historical claims data.
We can make some general observations about groups that are most at-risk for rehospitalization. Two major groups jump out from the rest. First off, elderly folks with multiple chronic conditions, requiring lots of different medications, who are having problems with mobility and self-care, particularly if they live alone. Failing vision may make reading and even paying bills difficult, let alone reading discharge instructions. Second, anyone with a serious, chronic mental health problem such as major depression or delusions, or serious addictions may be having problems handling day-to-day challenges of life. If you can barely keep it together on a good day, imagine how tough it is when you get sick. The third category includes anyone who has already shown a pattern of recycling back into the hospital system. This might include anyone who has been in the hospital twice in six months or in the ER a few times in a the past year. Unless these incidents were clearly due to special situations or accidents, there may be a high risk of continuing that pattern. [Learn more about tough customers for care transitions: Tough Customers - Who need care transitions most]
Who is the User of the app?
Many of the apps assumed that the patient would be the User. That is consistent with a key idea behind the challenge, which was to use technology to empower patients and caregivers. But in addition to the patient, the trick to making an impact lies in the interface between the patient, the patient's support ecosystem, and various healthcare providers. Many of the high-risk patients have serious, chronic, and complex conditions with limited support networks. Many have low levels of health literacy and tech literacy. Often, the key to helping them stay out of the hospital will come by activating or building support services around them that can identify problems before they get out of hand. Support is likely to come from one of three sources, so I looked at whether or not the apps seemed aware of these potential user groups, in addition to the patient:
- The hospital discharge planners and case managers, plus other direct care providers, can provide monitoring, data collection, and followup services to keep track of what is going on. Based on monitoring data they may activate supportive services to reduce the need for an expensive rehospitalization. These services cost money, so they must be allocated to those who need them the most. The theory is that spending money on lower-cost interventions reduces overall costs by keep someone out of the hospital. This goes back to the issue of risk stratification. Did the app have some sort of dashboard for use by care providers, or integration with existing hospital and medical record systems? Was there some way the app could faciliate decisions on allocation of scarce intervention services by care managers?
- The extended circle of care hopefully will include family and friends who can be caregivers. Did the app help form and support these circles of care? Several of the apps were driven primarily by this model, and tried to create a sort of Facebook for sick folks. This is not a new idea, and can be an important component of a solution, but by itself it is not enough to make a real dent in the institutional reality of care delivery. Only a couple of the apps ran with the idea that the most likely user would be the daughter, son, or caring friend, and explicitly showed how having a tablet in hand when talking with a healthcare provider could be a game-changer for a determined caregiver.
- Community-based agencies such as senior daycare centers, meal programs, and other supportive agencies can play a key role in monitoring changes in health for vulnerable groups. Was there some way these resources could be part of the support mix for a client? Not one of the apps showed me how a Meals on Wheels volunteer who had been authorized to do so could carry a tablet around like a FedEx delivery person, helping to spot possible problems based on regular service interactions. When you put the meal in the fridge, do you notice that nothing has been eaten for days? None of the apps took me to the lobby of the Senior Center where mom spends afternoons.
What is the business model?
Depending on your views about the two issues of who the patient is and who the User of the app is likely to be, two major categories of business model emerge:
- If you think the customer is a healthcare system, you will design something that feeds off the hospital Electronic Health Record (EHR) and has strong integration with healthcare data exchange standards. You hope the institutional customer will pay big bucks for a system that they can install on their own data center floor, or use via your cloud service. Cost justification will need to show provable net reductions of operating expense (a cost minimization model).
- If you think the customer is the patient, you will design something that has a low price, or even a free app that is supported by advertising or branding by a healthcare provider. You may charge a fee to healthcare systems that want to utilize a provider interface, but the design paradigm will largely be driven by consumer needs, following a Personal Health Record (PHR) model.
