Amirhosein Gholami: Anatomy of a readmissions-focused mobile app


Four Binghamton University graduate students--Amirhosein Gholami, Qi Jia, Lu He, and Runpei Xu--have developed an award-winning digital health app that promises to slash hospital readmission rates. Their Post Discharge Treatment and Readmission Predictor app uses analytics to predict hospital readmissions and allows patients to more easily communicate with providers, which they, say could potentially save the healthcare industry billions of dollars.

FierceMobileHealthcare reached out to Gholami (pictured), the team leader, for deeper insight on the app and the challenges facing such innovations.

FierceMobileHealthcare: How did the idea to create such an app come about?

Amirhosein Gholami: Although readmissions research, by itself, was a valuable topic, it was not enough for building an app. Briefly, the readmission prediction is made based on some technical parameters that are meaningless for general mobile app end-users. We decided to add more features which have more benefits for the patients. As a result, the "Post Discharge Treatment" has been added to this app as a valuable feature for all end-users (providers, patients, and insurance agencies).

FMH: What was the biggest challenge in developing the tool?

Gholami: Creating software for the healthcare industry always comes with challenges. There are numbers of standards that should be considered in such an app development. One is security. Based on HIPAA, the app and communications should be secured. For this purpose, we should keep all information and prediction calculations in a secured server. To reach this goal, we developed the app based on the multi-tier architecture, which enables us to keep the end users completely separated from patient data.

Since a big portion of the end users are senior patients, another challenge was designing a user interface in a way that all users can access the features without confusion. The first version of the app was made based on a sliding menu; the second version was made based on a tiled menu. A group of senior SSIE undergraduate students have conducted usability tests on both designs and proved that the tile-based UI is the best design for new users and seniors.

A third challenge was choosing the mobile app platform. Since Android-based phones and tablets are usually cheaper and have more users in the market than iOs devices, we decided to implement the prototype app in Android.

FMH: There are several benefits for providers and patients in using this app. Which one does the research team view as most valuable or most critical?

Gholami: The readmission prediction research mainly focuses on reducing the re-hospitalization expenses. Although patients will benefit from this app by saving their time and money to get some services from hospitals, the primary benefit will be for providers and insurance agencies.

FMH: In moving ahead, what future functions are being considered?

Gholami: Right now, we only cover basic vital information provided manually by patients. One of the future works will be importing that data automatically from wearable devices and fitness equipment. Sending and analyzing other vital information, such as ECG, is also part of our future development.

FMH: What do you see as the biggest challenge facing development of digital health tools?

Gholami: Like other healthcare related projects, this project cannot be successful without provider support. It won't be useful if you have the best app but no one wants to use it.

In 2005, I was leading a small team of developers working on a real-time ECG mobile app. The app has been designed under Microsoft Windows Mobile 5, to deliver a real-time ECG from CCU and ICU into a doctor's mobile phone. So the doctor could check on his or her patients remotely and send instructions to a nurse based on their situation.

We couldn't make it commercial, however, because the healthcare industry was unfamiliar with the new technologies at that time, and it was hard for them to trust those kind of communications. As a result, they didn't support us, and unfortunately, the project was stopped at its early stages. If we could convince all end users to use these kind of technologies in their daily lives, then we can say that the project is successful.

Editor's Note: This interview has been edited for clarity and length.