You are using an outdated browser. Please upgrade your browser to improve your experience.

You are using an outdated browser.

To ensure optimal security, this website will soon be unavailable on this browser. Please upgrade your browser to allow continued use of ACP websites.

You are here

ACP Innovation Challenge 2019 Winners and Finalists

These finalists presented their ideas to a panel of judges who awarded prize money to implement their best and brightest ideas.

ACP awarded $20,000 to Jonathan Dau, MD, an internal medicine resident at The University of Texas Health Science Center at Houston, for “Prescriptions (Rx) for Produce: A Grassroots Medicine, Farm‐to‐Hospital Initiative.”

Miguel A. Chavez Jr., MD

Internal Medicine Resident, Texas Tech University Health Science Center El Paso

DNR, DNI, or CMO: To Code or Not to Code?

After completing my first rotation in the ICU, I more fully understood the responsibility of initiating conversations regarding goals of care and end-of-life decisions. In many cases, patients had not had the opportunity to discuss code status with their PCP or their loved ones. I was motivated by two specific situations where both patients had an Advance Directive in place but could not provide the physical documentation. In both cases, a family member was chosen to act as durable power of attorney. In one situation, the document was lost, but in both situations, someone other than the patient was able to override the patient's wishes.

An Advance Directive is a legal document and varies by state, but not many patients have it readily available or carry it with them. In this new digital age where technology is connecting the world, we should take advantage of that momentum and use it to connect to our patients. If a patient has an established Advance Directive in one state, that same patient has an equal right to have an Advance Directive when traveling to other states. Allowing a website in which patients learn about Advance Directives and agree to publish theirs online is a solution to this issue. We are all very aware of HIPAA laws and a patient's right to privacy, but a patient should also have the right to for their code status to be shared in any way they choose. My idea is an active web page that it is encrypted, but is able to be viewed by health care professionals if a patient is willing to publish their code status online via the website. Legal barriers and protection laws exist, but a patient has the right to decide if they want to publish their code status. Analogously, in many states, a driver's license contains information about whether an individual is an organ donor. My idea is to provide a platform where patients have the choice to share their code status willingly and can have a notarized copy of their Advance Directive uploaded to the website as well as to indicate whether they are willing to be an organ donor. As an Advance Directive can be changed, the website would have the potential to allow alterations once identity can be confirmed. They can also carry a card with chip technology, a QR code, or even a bracelet which provides a unique identifier that allows for patients all over the country to have their code status viewed by health care professionals.

Connecting to patients is a daily task for physicians, and for internists, this is a crucial concept present in outpatient care as well as during the admitting process. Talking to patients about end of life care and specifically, their code status, should not have to be a discussion reserved for times of crisis or elderly patients. As all of us are intimately aware, life itself can be a fragile thing, but discussing code status with our patients involves an in-depth conversation which itself brings us closer to our individual patients. This website provides the potential to connect patients to any physician in the United States.

Jonathan Dau, MD Grand Prize Winner

Internal Medicine Resident, The University of Texas Health Science Center at Houston

Prescriptions (Rx) for Produce

Lyndon B. Johnson Hospital is a safety‐net hospital serving Harris County in Houston, where 55% of the population lives below the federal poverty level and over 13,000 residents suffer daily from food insecurity. Our patients face the unique healthcare burdens associated with poverty. They live in a USDA‐designated food desert with limited healthy food options and lack the understanding and support to nurture healthy habits. One year ago, we embarked on a pioneering initiative to re‐imagine how we serve our patients. We vowed to shift the paradigm in which we practice medicine—being proactive rather than reactive—and created a community farm on our one‐acre plot of land that has become a  mecca for multidisciplinary medical care involving medical providers (attendings, residents, students), nursing staff, dieticians, physical/occupational therapist, and our patients. To date, we have reached our first milestone for the farm, establishing the foundational infrastructure for our education/outreach programs. We just had our first harvest of over 1,000 lbs. of fresh vegetables and have over 30 functional plots that the community and hospital staff helped build and plant. We have also built an area for physical rehabilitation on the farm. Now, we are developing our education/outreach programs and one such project is the Prescription (Rx) for Produce program.

This program is being developed with the goals of:

  1. Addressing nutritional‐related diseases and under nutrition in our low‐income patients.
  2. Bridging the therapeutic gap between physicians and nutritionists with our patients

We will accomplish these goals via:

  1. Hospital Based Food Pantry - Patients who would most benefit from lifestyle changes, identified from our IM Resident Clinic, are “prescribed” a bag of produce every two weeks. The food pantry will be stocked with produce grown from our Farm and the Houston Food Bank.
  2. Mobile Teaching Kitchen – To ensure compliance and engagement, in conjunction with our nutritionists/dieticians at University of Texas School of Public Health (UTSPH), we will host cooking classes for patients and families. The UTSPH has a mobile kitchen RV that we can use.
  3. Community Partner Engagement - We have collaborated with the UTSPH who have already developed a successful food pantry program with the Houston Food Bank in Pasadena, TX. They will be assisting us to ensure the success of our program.

We will be the first to create a truly comprehensive farm‐to‐hospital produce prescription program integrating both therapeutic and educational components across all healthcare providers, thus becoming a national model for the revitalization of healthcare in communities that rely on safety‐net care.

David J. Savage, MD, PhD Audience Choice Winner

Internal Medicine Resident, Cleveland Clinic

A Tool for Seamless Transfer of the Electronic Medical Record

The electronic medical record (EMR) is essential for transitions of care. However, when a patient moves between hospitals or ambulatory settings that use different EMR systems, that transition often happens with paper.

Inpatient physicians spend countless hours reviewing poorly organized, printed paper records for patients who have been cared for by other providers. It seems anachronistic that in the era of the EMR, patient records are transmitted between care facilities on paper. Oftentimes, the records brought in a printed form are incomplete. Critical imaging, H&Ps, progress notes, or consultant notes are missing. A physician must review those records and recompose the narrative of the patient's health and treatment plan to date. All of this information has to then be entered into a new EMR. Many times, tests and imaging are repeated at significant cost and risk to the patient because critical images or reports are missing. It is estimated that repeat testing costs the US healthcare system at least $25 billion annually (1).

Our proposed innovation is a secure and portable Patient Data Model (PDM). The PDM will provide a universal language and protocol for making patient data portable. Conventional EMR manufacturers will be able to use the PDM to implement Export, Secure Transfer, and Import features into their products. With patient permission, it would become seamless to transfer a complete record between two entirely different EMRs. Encrypted PDM data could be quickly transmitted across the internet. Once decrypted and imported, PDM data would be instantly useable. Valuable physician time would not be wasted in reformatting data by hand to make it fit within a new EMR.

Conventional EMRs such as Epic, Cerner, or Allscripts would be encouraged to have export/import features following this model so that data can be easily transmitted from one EMR to another when a patient changes hospitals or doctors. This third-party model eliminates the need for these vendors to open up their software code. Instead, they can make their systems work with the programming interface of the data model. Eventually, legislative action could be used to encourage implementation of data sharing tools like this one. CMS reimbursement could be reduced for institutions without a universal export/import data feature in their EMR. The US government currently pays a significant subsidy for patient care through Medicaid and Medicare, and therefore has a vested interest in incentivizing portable data. Likewise, this would improve patient safety by insuring that no critical information is lost in the patient transfer process. It would also enhance the ability of the physician to help patients by facilitating seamless transfers of care. The lead author of this proposal is a resident physician and former software developer with two years of professional experience using object-oriented programming languages. His mentor, Dr. Jamal Mahar, is an academic internist with a significant interest in safe care transitions. Both lead author and mentor have the expertise and sufficient contacts within the industries of health and software to make this project a reality.


Jesse Wang, MD/PhD Candidate Audience Choice Winner

University of Rochester School of Medicine and Dentistry

The Digital Scribe

The electronic health record (EHR) is frequently cited to contribute to increased documentation times, increased physician burnout, and decreased quality and time of patient interactions [1-4]. My innovation aims to reduce the overwhelming amounts of medical documentation. It is an intelligent speech system that would automatically convert the physician-patient conversation into an encounter note. With this technology automating documentation, physicians can spend more time with patients, more time with family, and more time taking care of themselves. My technology would allow physicians to better fulfill their calling in medicine. It would return medicine to its roots where all attention is centered on the patient, while simultaneously accomplishing the documentation work necessary to survive in a digital world.

I am in the process of developing a prototype system. The technology uses artificial intelligence and machine learning to identify different speakers in a conversation, convert speech into text, identify sentence boundaries from unmarked transcripts, and identify word parts-of-speech. Other natural language processing algorithms are used to parse medically relevant information from the conversation. They will also be used to convert words in first- and second-person tenses into third-person tenses. By implementing these different components together, the system would be able to convert a conversation into a history of present illness section of an encounter note. The system would capture information during a patient interview so that documentation will be completed by the end of the visit.

My team consists of myself, an MD/PhD candidate in the Department of Translational Biomedical Science at the University of Rochester School of Medicine and Dentistry. Local mentorship includes my PhD adviser Dr. Henry Kautz, who is Division Director for Information & Intelligent Systems at the National Science Foundation, founding director of the Goergen Institute for Data Science, and Professor of Computer Science at the University of Rochester. I am also in collaboration with Dr. James Allen, who is a John H. Dessauer Professor of Computer Science at the University of Rochester and Associate Director of the Florida Institute for Human and Machine Cognition.

My relevant previous experience includes developing a biomedical search engine for clinicians called CupQ. It uses machine learning to produce word embeddings and interpret the contextual relevance of articles with respect to the user’s search query. Over 30 million articles were processed to build a model capable of understanding a user’s search intent. The engine also evaluates title context, journal H-index, and language normalization. I used Java, Python, MongoDB and MySQL to build the backend of CupQ while I used HTML, CSS, and Javascript to build the frontend. I also networked several servers together to distribute the computational workload and provide fast search times. My experience building CupQ taught me the necessary skills to develop an intelligent speech system.


  • Friedberg, M. W. et al. Factors affecting physician professional satisfaction and their implications for patient care, health systems, and health policy. Rand Health Q. 3, 1 (2014).
  • Young, R., Burge, S., Kumar, K., Wilson, J. & Ortiz, D. A time-motion study of primary care physicians’ work in the electronic health record era. Fam. Med. 50, 91–99 (2018).
  • Meeks, D. W. et al. An analysis of electronic health record-related patient safety concerns. J. Am. Med. Inform. Assoc. 21, 1053–1059 (2014).
  • Coiera, E., Kocaballi, B., Halamka, J., & Laranjo, L. The digital scribe. Npj Digital Medicine, 1(1). (2018).