Skip to main content

AMIA 2019 Student Design Challenge

Reinventing Clinical Decision Support

AMIA is pleased to announce the 7th Annual Student Design Challenge (SDC). In this challenge, we invite teams of graduate students from different scientific disciplines and of various backgrounds to propose creative solutions to a specified problem related to healthcare. We seek novel solutions that incorporate cutting edge computational and interactive technologies and take advantage of the considerable advances in such research areas as biomedical informatics, human-computer interaction, computer science, information visualization, pervasive and ubiquitous computing, among many others.

New for 2019! This year, we have a new format for the student design challenge that will focus on providing student teams with training and mentoring as they work on articulating their proposed solutions.

Proposal submission: We invite teams of graduate students and trainees in biomedical informatics and related fields to prepare a brief (1-2 pages) proposal for innovative solutions that take a new approach to providing clinicians and patients with decision support in regard to health. The proposal should clearly specify the selected specific problem, outline gaps in existing decision support systems targeting this problem, and describe the authors’ vision for their proposed solution. The five most innovative proposals will be invited to participate in a virtual summer training program. During this program, the teams will be provided with training on topics related to the design of interactive systems in health, including but not limited to methods for understanding human practices and requirements gathering, theories and frameworks related to decision support, design methods, evaluation methods, and others. The training will include virtual lectures, and practical exercises. Further, each team will have the opportunity to work with mentors who will provide teams with feedback on their emerging solutions. At the end of this period, the teams will be expected to fully articulate their proposed design concepts, develop an illustrative prototype of their solution, and design an approach to evaluating their prototype in user studies.

Final submission: At the end of the training program, the teams will submit their extended abstract of their final design solutions (more details will be provided to the selected teams about the structure of the extended abstract). A panel of distinguished members of the AMIA community will review the proposed solutions and rate them based on a number of criteria, including their originality and transformative potential. All five teams will be asked to attend the AMIA 2019 Annual Symposium and present their solutions during the AMIA poster session. The top three teams as selected by the SDC panel will be invited to participate in a formal presentation at AMIA.

Timeline

  • June 15: Submission deadline for proposals
  • June 30: Invitations to top five teams sent
  • July 15 - August 15: Summer training program
  • August 30: Final Submissions
  • September 15: Notifications to top three teams selected for oral presentations
  • November 19: Presentation at AMIA 2019 Annual Symposium

Call for Submissions

Helping clinicians and patients to make more informed decisions in regards to medical treatment and health management is at the heart of biomedical informatics. Clinical decision support systems (CDSS) strive to use available data, knowledge, evidence, and guidelines to help clinicians arrive at diagnostic and treatment choices with higher accuracy and efficiency and provide consistently high quality of care (1,2). Patient-centric decisions support systems (PCDSS) strive to help patients make more informed choices in regards to both treatment and self-management (3–8). However, despite decades of research in these areas, the adoption of such systems remains low, thus diminishing their potential positive impact on health and medicine (9). Moreover, many previous attempts to introduce decision support in clinical practice have led to unanticipated consequence. This includes alert fatigue, one of the most common and well-studies side-effects of clinical decision support (10).

In recent years, new advances in machine learning, natural language processing, computational modeling, and data science opened unprecedented opportunities for the design of intelligent systems that can understand the world around them and arrive at accurate inferences and decisions. These new technical abilities can have a profound impact on clinical decision-making by augmenting clinicians’ ability to understand each patient’s unique pathophysiology, and arrive at informed treatment choices, thus helping to realize a vision of augmenting human intelligence with computational capabilities (9). However, realizing this vision will require not only innovative computational solutions and data science methods, but also innovative ways to deliver decision support to both clinicians and patients. Previous research suggested poor interface design, poor fit with clinical work practices, and disconnect between CDSS design and existing knowledge on clinical reasoning and decision-making among chief reasons for their low impact on clinical practice (11). While alerts have become the most common way of delivery in CDSS, there exist many mechanisms for augmenting human intelligence, from interactive visualizations that help to identify important patterns, to intelligent simulations that help to explore different what-if scenarios, to intelligent conversational agents that provide assistance on demand. Yet few of these innovative approaches have been explored in the context of decision support in medicine and health.

In this challenge, we call on undergraduate and graduate students and trainees in biomedical informatics and related fields to envision new ways of providing clinicians and patients with intelligent decision support that augments their reasoning and decision making and aligns with existing work practices. Such system may take the form of intelligent agents, predictive analytics, intelligent simulations, and many others.

Importantly, the focus of the challenge will be on ways to augment and complement human intelligence with new ways of engagement with clinical decision support.

References

  1. Blumenthal D, Tavenner M. The “Meaningful Use” Regulation for Electronic Health Records. New England Journal of Medicine. 2010;363(6):501–4.
  2. Musen MA, Middleton B, Greenes RA. Clinical Decision-Support Systems. In: Shortliffe EH, Cimino JJ, editors. Biomedical Informatics [Internet]. Springer London; 2014 [cited 2014 May 23]. p. 643–74. Available from: http://link.springer.com/chapter/10.1007/978-1-4471-4474-8_22
  3. Wilkinson MJ, Nathan AG, Huang ES. Personalized Decision Support in Type 2 Diabetes Mellitus: Current Evidence and Future Directions. Curr Diab Rep. 2013 Apr 1;13(2):205–12.
  4. Quinn CC, Clough SS, Minor JM, Lender D, Okafor MC, Gruber-Baldini A. WellDocTM Mobile Diabetes Management Randomized Controlled Trial: Change in Clinical and Behavioral Outcomes and Patient and Physician Satisfaction. Diabetes Technology & Therapeutics. 2008 May 12;10(3):160–8.
  5. Glasgow RE, Kurz D, King D, Dickman JM, Faber AJ, Halterman E, et al. Twelve-month outcomes of an Internet-based diabetes self-management support program. Patient Education and Counseling. 2012 Apr;87(1):81–92.
  6. Christian JG, Bessesen DH, Byers TE, Christian KK, Goldstein MG, Bock BC. CLinic-based support to help overweight patients with type 2 diabetes increase physical activity and lose weight. Arch Intern Med. 2008 Jan 28;168(2):141–6.
  7. Liang X, Wang Q, Yang X, Cao J, Chen J, Mo X, et al. Effect of mobile phone intervention for diabetes on glycaemic control: a meta-analysis. Diabet Med. 2011 Apr;28(4):455–63.
  8. Costa BM, Fitzgerald KJ, Jones KM, Am TD. Effectiveness of IT-based diabetes management interventions: a review of the literature. BMC Family Practice. 2009 Nov 17;10(1):72.
  9. Middleton B, Sittig DF, Wright A. Clinical Decision Support: a 25 Year Retrospective and a 25 Year Vision. Yearb Med Inform. 2016 May 20;(Suppl 1):S103–16.
  10. Backman R, Bayliss S, Moore D, Litchfield I. Clinical reminder alert fatigue in healthcare: a systematic literature review protocol using qualitative evidence. Syst Rev [Internet]. 2017 Dec 13 [cited 2019 May 7];6. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5729261/
  11. Sittig DF, Wright A, Osheroff JA, Middleton B, Teich JM, Ash JS, et al. Grand challenges in clinical decision support. J Biomed Inform. 2008 Apr;41(2):387–92.

To qualify for participation, teams should include only students in degree-pursuing graduate programs (including clinicians in training, such as residents and fellows, as well as post-doctoral fellows pursing MA or MS degrees) or Graduate Certificates. Undergraduate students are welcome to participate in design teams, provided that they are supervised by graduate students. Given the nature of the creative process, we suggest that teams include no more than four or five individuals. Because of our focus on fostering multidisciplinary teams, the SDC will not accept submissions from single individuals. No faculty advising is required for participation; in fact, we encourage teams to work independently and with minimal faculty supervision.

Each team will be asked to identify a specific challenge related to the proposed theme. We recommend that teams select a specific context of use and target audience, for example “Using interactive simulations to examine potential impact of treatment on blood pressure in patients with hypertension” or “Using conversational agents to facilitate shared decision making among providers and patients in regards to genetic testing in cancer”. In both of these scenarios the focus of the solution should be on new ways to engage patients and providers with intelligent decision support that goes beyond traditional alerts.

To be considered for the inclusion in the challenge, the teams must begin by developing a one-page proposal that specifies the problem and context, describes gaps in existing solutions, and proposes an innovative solution for delivery of decision support. The proposals will be judged on their innovation and transformative potential, as well as on their feasibility. We expect that the best proposals will incorporate already established computational capabilities (predictive models, inference engines, etc.), and focus on delivery mechanisms. The five teams selected to develop their proposals will be expected to articulate their solution in a way sufficient to demonstrate their functionality. This could include interactive prototypes or mockups. Fully functional prototypes that integrate with computational analytical engines are encouraged, but not required for the submission.

Competition Process

Each team will submit an abstract (1-2 pages) describing the specific challenge and context for their solution, their vision for the solution, and provide evidence of its feasibility (existing computational analytical engines either developed by the teams, or in published literature).

The submission process will be done through ScholarOne (more details on the submission process are to follow). The submissions will be evaluated through a peer-review process by the SDC steering committee.


The five best proposals will be invited to participate in the virtual summer training program delivered online (no travel required). The program will include a set of interactive lectures on topics relevant to the design of interactive systems in health and clinical decision support and will be delivered by leading researchers in biomedical informatics and HCI. Further, the teams will have a chance to discuss their emerging solutions and receive feedback from mentors assigned to each team. Most of the training program is expected to occur in July-August 2019.

At the end of the training program, the teams will be asked to submit an extended abstract (5 page maximum) describing their solution. These abstracts will be submitted via email directly to the SDC chairs. A panel of judges will review submitted abstract and will select 3 finalists for the final phase of the challenge.

All five teams participating in the training program will asked to present their solutions during a poster session at the AMIA 2019 Annual Symposium. At least one member from each of the five teams will be expected to attend the conference to present a poster illustrating their solution, discuss their solution, and the design process with conference attendees. AMIA will waive the registration fee for one presenter from each of the five teams, with the expectation that the presenter holds a student membership with either AMIA or ACM.

Three teams will also be notified prior to the symposium that their proposal has been selected as a finalist for the AMIA SDC Award. They will be asked to give a presentation about their solution during the AMIA Student Design Challenge session. The three finalists will give an oral presentation and, where appropriate, demonstration of their design to the panel of SDC Judges and AMIA attendees. The judges will rank the solutions and presentations to identify the winner and subsequent second and third place teams. The winners will be announced during the last day of AMIA Annual Symposium and acknowledged during the AMIA Closing Plenary.

Proposal Preparation

The participants will prepare an abstract (1-2 pages) written in the AMIA format that must include:

  • Definition of the selected problem grounded in deep understanding of an identified health problem;
  • Discussion of gaps in existing decision support solutions addressing the same problem;
  • Description of the novelty/originality of the proposed solution;
  • Establishing the feasibility of the proposed solution (evidence of existing computational capabilities, or knowledge needed to implement the proposed solution).

The completed abstract should be submitted using ScholarOne by 11:59 p.m. EDT on June 15, 2019. If you do not already have a ScholarOne account, you will need to create one. AMIA member log-in will not provide access to ScholarOne.

Proposal Review Criteria

The proposals will be reviewed using the following criteria:

  • Justification of the selected problem (does the proposal provide sufficient justification for the selected problem and its applicability for computerized decision support)
  • Fit to the problem (how likely is the proposed solution to address the selected problem?)
  • Innovation (how novel and original is the solution?)
  • Feasibility of the solution (evidence of computational analytical capability, or knowledge resources necessary to implement the solution)

Awards

The SDC awards ceremony will take place during the last day of AMIA Annual Symposium.

Timeline

  • June 15: Submission deadline for proposals
  • June 30: Invitations to top five teams sent
  • July 15 - August 15: Summer training program
  • August 30: Final Submissions
  • September 15: Notifications to top three teams selected for oral presentations
  • November 19: Presentation at AMIA 2019 Annual Symposium