Best Paper Selection

Appendix: Content Summaries of Selected Best Papers for the 2024 IMIA Yearbook Section Public Health and Epidemiology Informatics

Jiang LY, Liu XC, Nejatian NP, Nasir-Moin M, Wang D, Abidin A, Eaton K, Riina HA, Laufer I, Punjabi P, Miceli M, Kim NC, Orillac C, Schnurman Z, Livia C, Weiss H, Kurland D, Neifert S, Dastagirzada Y, Kondziolka D, Cheung ATM, Yang G, Cao M, Flores M, Costa AB, Aphinyanaphongs Y, Cho K, Oermann EK.

Health system-scale language models are all-purpose prediction engines.

Nature. 2023 Jul;619(7969):357-362.

doi: 10.1038/s41586-023-06160-y.

This study intends to improve clinical predictive models by leveraging recent advances in natural language processing. The authors develop, train, validate and deploy a new health system-scale Large Language Model (LLM) designed and validated for clinical use. The new clinical predictive engine uses unstructured clinical notes from the entire electronic health records of a large healthcare system to train a LLM for medical language and make predictions across a wide range of clinical and operational predictive tasks. It is then deployed to assess its impact on real-world clinical care, evaluating its ability, versus existing tools, to predict three typical clinical outcomes (in-patient mortality prediction, comorbidity index prediction and readmission prediction) and two operational tasks (insurance claim denial prediction and inpatient length of stay prediction). The authors demonstrate that their new clinical predictive engine improves existing tools by 5 to 15%. The study is unique in many aspects. At the time of publication, this was a very original and ground-breaking work. Research and application of LLMs has been boosted thereafter but this paper was certainly one of the pioneers in the field.

Javed A, Kim DS, Hershman SG, Shcherbina A, Johnson A, Tolas A, O'Sullivan JW, McConnell MV, Lazzeroni L, King AC, Christle JW, Oppezzo M, Mattsson CM, Harrington RA, Wheeler MT, Ashley EA.

Personalized digital behaviour interventions increase short-term physical activity: a randomized control crossover trial substudy of the MyHeart Counts Cardiovascular Health Study.

Eur Heart J Digit Health. 2023 Aug 9;4(5):411-419.

doi: 10.1093/ehjdh/ztad047.

Increasing engagement in physical activity is a tenet of prevention of many chronic diseases, yet studies on primary prevention have been relatively seldom in the digital public health literature. This large-scale trial (including a few thousand individuals in a free-living cohort) investigated how a digital health intervention through a smartphone app can encourage engagement in short-term physical activity. The 7-day intervention consisted of: (i) daily personalized e-coaching based on the individual's baseline activity patterns, (ii) daily prompts to complete 10 000 steps, (iii) hourly prompts to stand following inactivity, and (iv) daily instructions to read guidelines from the American Heart Association website. The primary outcome was change in mean daily step count from baseline. The novel finding is that digital interventions tailored to an individual are effective in increasing short-term physical activity, suggesting that persons are more likely to react positively and increase their physical activity when digital prompts are personalized. Beyond the large scale, the originality of this study pertains to its design which has mastered the variabilities (phones, sensors) with the use of the same devices and app (iPhone, watch, My Heart Count). The study demonstrates the power of digital tools to implement public health interventions toward improving lifestyle for prevention of chronic diseases with a precision approach.

Publication History

Article published online:
08 April 2025