We propose to enable a scalable clinical decision support (CDS) platform that helps clinicians and patients select cancer screening strategies that are best suited to each individual. This kind of CDS is important because increased evidence supports personalizing cancer screening decisions according to each individual's unique cancer risks. While a highly desired goal, individualizing screening at a population scale requires the implementation of patient-specific risk assessments for several types of cancer. This is quite challenging in today's overwhelmed primary care environment. Our proposed CDS platform addresses this challenge by (i) automating the risk stratification process at the population level based on EHR data and patient reported data; (ii) prioritizing patients for case review by genetic counselors; and (iii) automatically communicating risk and screening recommendations with primary care providers and their patients. We will integrate the CDS platform with the Epic EHR and test it at the University of Utah Health Care community clinics and the Huntsman Cancer Institute. We will assess the generalizability of the CDS platform with a different EHR (Cerner) at a different institution (Intermountain Healthcare). To maximize the dissemination potential for the proposed cancer risk screening platform, we will extend two well-established open source CDS platforms: OpenCDS and OpenInfobutton. These platforms are reference implementations of international EHR standards that are required for EHR certification in the US. We will also obtain software certification from the Open Source EHR Alliance, share the CDS platform and our experiences with other awardees in the ITCR Program, present the CDS platform at national and international cancer and informatics conferences; and engage with relevant stakeholders through a technical expert panel. Any healthcare organization with a certified EHR will be able to use the proposed CDS platform. Therefore, if successful, this proposal can have a significant impact on disseminating individualized cancer screening practices according to the best available evidence.
The proposed research addresses the need for individualizing cancer screening according to the risk of an individual. We propose to enable software that scans electronic health records of entire patient populations to automatically detect those at a high risk to develop specific types of cancer according to cancer guidelines. Genetic counselors will review patients identified by the software, and discuss optimal cancer screening strategies with the patient and their primary care providers.
| Bian, Jiantao; Abdelrahman, Samir; Shi, Jianlin et al. (2018) Automatic Identification of Recent High Impact Clinical Articles in PubMed to Support Clinical Decision Making Using Time-agnostic Features. J Biomed Inform : |
| Del Fiol, Guilherme; Michelson, Matthew; Iorio, Alfonso et al. (2018) A Deep Learning Method to Automatically Identify Reports of Scientifically Rigorous Clinical Research from the Biomedical Literature: Comparative Analytic Study. J Med Internet Res 20:e10281 |
| Bian, Jiantao; Weir, Charlene; Unni, Prasad et al. (2018) Interactive Visual Displays for Interpreting the Results of Clinical Trials: Formative Evaluation With Case Vignettes. J Med Internet Res 20:e10507 |
