JL513. Improving Data Quality and Knowledge Management Through the Application of Artificial Intelligence to Oncology Care Model Patients
Jacqueline T. Norrell, DNP, BS-CIS, FNP-BC, Renee Kurz, DNP, MSN, FNP-BC, Janet Gordils-Perez, DNP, ANP-BC, AOCNP®, and Doris Fonseca, BS Public Health; Rutgers Cancer Institute of New Jersey, New Brunswick, New Jersey
© 2018 Harborside™
JADPRO Live at APSHO 2017
Marriott Marquis, Houston, Texas • November 2–5, 2017
The posters for the abstracts below can be found at:
Background: The Center for Medicare Services Oncology Care Model (OCM) is defined as a quality improvement initiative focused on improving care for Medicare patients receiving chemotherapy. The six CMS requirements include: provide patient navigation; document a care plan containing the 13 components in the IOM Care Management Plan; provide 24/7 symptom management access; treat patients with therapies consistent with national guidelines; use data to drive continuous quality improvement; and use an ONC-certified EHR. Rutgers Cancer Institute as the only NCI-designated CCC in NJ, a state that ranks among the worst for overall cancer incidence, is a critical resource for NJ. In keeping with its mission, the Institute seeks to continually provide new clinical programs and participate in value-based payment models to meet the needs of cancer patients to improve quality care in a cost-efficient manner. In May of 2015, the Cancer Institute pursued participation in the OCM by submitting an application and in April of 2016, was chosen to participate.
Methods: An executive core steering committee was formed to strategize operational processes to meet all OCM requirements. The subcommittee, Quality Reporting Committee’s (QRC), overarching initial goal was to develop the tracking mechanism to assess the quality of submitted OCM data for validity, accuracy, and completeness, and make recommendations for remediation of data quality issues to facilitate reporting of OCM measures. The methodology included an evaluation of all clinical systems, workflow, and complex structured data. Statistical analysis was performed, a gap analysis completed, reports created, and a scorecard developed that identified data with steps for remedial action. The QRC and OCM team members rectified the data quality issues.
Results: Results revealed limitations of the legacy EMR including missing, undefined, or inaccessible data elements, precluding OCM reporting or analysis. By correcting data quality issues, there was an improvement in the quality of care patients received by the organization and the Advanced Practice Nurse (APN).
Conclusions: Data democratization, robotic process automation (RPM), and the use of artificial intelligence (AI) facilitated our ability to predict, rather than react, to newly discovered data quality issues. Guided by the OCM framework, the use of RPM and AI provided an efficient method of data capture and analysis for reporting OCM measures.
Recommendations: Ongoing QA practices of OCM measures using AI, and RPM, improving data governance and clinical informatics, and implementing an information-based strategy to address the quality of oncology service delivery by the organization is essential.
For access to the full length article, please sign in