Development of EHR-Integrated Cancer Staging Module Identifies Barriers, Increases Discrete Capture

The implementation and subsequent optimization of a health information technology module within the electronic health record (EHR) has allowed clinicians across a diverse health care system to identify barriers to improve discrete capture of cancer staging and design solutions to address them, according to findings from a quality improvement pilot study published in JCO Oncology Practice. Study authors noted that the approach is scalable and could be applied to other initiatives in the realm of health information technology.

At baseline, a period spanning March 2018 to December 2019, 6628 patients with cancer across 11 centers had an eligible first visit encounter; 31% of patients had their cancer staging discretely captured via the EHR-integrated module and 19% were staged discretely in what investigators had determined to be a timely manner. After identifying key barriers that were thought to be driving the lack of discrete staging capture and implementing solutions, the improved module resulted in a discrete capture rate of 58% and a timely discrete capture rate of 35%, representing respective increases of 87% and 84%, among 4346 patients with cancer across 55 outpatient clinics who had a first visit encounter between the July 2021 go-live date and the September 2022 data cutoff.¹

“The documentation of cancer staging within a standardized module embedded in the EHR is a distinct form of staging documentation that has its own potential benefits and challenges compared with providing this information solely within a physician’s note,” Neelima Vidula, MD, and Jeffrey Peppercorn, MD, wrote in an editorial in JCO Oncology Practice, which accompanied the study. “If used correctly and routinely, staging modules may facilitate quality improvement, outcomes analysis, recruitment to trials, targeted supportive care screening and interventions, and accurate billing, as collection of staging information from a physician’s note can be labor intensive.”²

In 2018, a discrete data module for cancer staging was made available in the EHR of the Northwestern University health care system, however the adoption rate was poor across the 11-hospital health care system and the impact on staging data capture was minimal. Although clinicians were reminded to use the module to stage patients, staging documentation was often nonexistent or included as free text in the form of note documentation outside of the module. Without the imbedded staging documentation in the EHR, the health care system faced challenges in driving downstream programs to improve the delivery of care. Thus, investigators set out to identify barriers to utilization of the module, identify health information technology solutions to expand discrete capture of cancer staging data, and increase capture across the oncology divisions of the health care system.¹

To conduct their study, investigators used Six Sigma quality improvement methodology (define, measure, analyze, improve, control [DMAIC]) to define the problems with discrete staging capture, perform data-driven analyses to better understand these problems, prioritize key root causes of the issues, implement solutions, and sustain the improvements that were observed. Additionally, a failure modes and effects analysis (FMEA) was utilized to determine and prioritize high-risk process points that caused problems with discrete staging documentation.

The study recruited a multidisciplinary team to take on the quality improvement project of improving the staging module and implementing the staging module. The group was split into 2 pilot teams: team 1 included gastrointestinal (GI) cancer clinicians from multiple regional sites and team 2 was made up of genitourinary disease specialists and additional GI clinicians. Both teams contained mid- and advanced-level clinicians and included medical oncologists, surgical oncologists, and surgery specialists.

Pilot team 1 was tasked with developing the user requirements needed for optimized health information technology solutions, then using these to develop prototype interventions and modifications. During pilot testing, clinicians used real patient encounter workflows and real-time feedback to develop the enhancements to the module. Interventions were rolled out in 2 separate 4-week testing cycles, each of which ended with data review and qualitative user evaluation. Members of both teams were surveyed to determine which interventions would be deployed as the system standard, and the module went live across 55 outpatient clinics.

The patient population included those with a solid organ tumor diagnosis at an outpatient encounter who either had treatment administered within 30 days or had a minimum of 2 established visits within 30 days. The 11-site health care system consisted of 1 academic medical center, 1 tertiary community hospital, and 9 community hospitals.

The primary outcome of the study was discrete capture, defined as the number of patients with cancer staging—clinical or pathologic—discretely captured, irrespective of the amount of time that elapsed. Timely discrete capture, defined as discrete capture that occurred within 30 days of the first encounter or date of treatment, represented a secondary outcome.

After applying the DMAIC and FMEA processes the 3 key barriers responsible for the lack of discrete cancer stage capture were determined to be workflow inefficiency, limited clinician accountability, and incongruence between the technical design and clinical ideals. To break down each barrier, investigators decided the best path was to continue to move forward with the EHR-integrated module and optimize the process. Thirty-six user-centered design enhancements were proposed, a majority of which (70%) were deemed to be within the study authors’ control and were implemented (Table¹).

For example, to address issues with workflow, investigators created a comprehensive tool within the EHR to provide quick access to commonly reviewed laboratory results, including tumor markers, pathology reports, and imaging to reduce information seeking. To address accountability, they added reminders to alert the user when there was a deficiency in terms of staging data and if the capture was not complete within 24 hours, routing another reminder at 72 hours.

In total, study authors implemented 7 enhancements to address the accountability barrier, 11 to address technical design flaws, and 6 to address workflow barriers. They also identified additional limitations in workflow (n = 2) and technical design (n = 9) and proposed enhancements for them but were not able to implement them.

Additional findings from the study revealed that, overall, the mean time from first cancer encounter to discrete staging entry was reduced from 186 days to 22 days (88%) from baseline to the data cutoff. The academic medical center showed the greatest increase in discrete staging capture from baseline to data cutoff, at 15% and 45%, respectively (a 200% increase). “It is possible that the improvement of Walesa et al in rates of completion of staging modules was due in part to rigorous evaluation of inefficiencies, stakeholder feedback, and serial adaptation of the process to address clinician concerns and barriers in their pilot design,” the editorial authors noted.

The study authors highlighted that their research had multiple limitations, including that the customization of health information technology is often discouraged and can cause organizational burden. Additionally, they noted clinician distrust of the module, behavioral factors influencing staging capture, and the inability to evaluate patient clinical outcomes based on the data in the collected in the study.

“The potential benefits of cancer staging modules are clear, but how to best realize this potential is an ongoing challenge,” the editorial authors wrote in conclusion. “Capitalizing on artificial intelligence technology to pull the information required to complete staging modules from notes, pathology reports, and other sources, without substantial time from any clinician, but allowing for editing and updating by clinicians, may help increase efficiency. If the responsibility of staging modules falls on the clinician, it is important to incorporate the clinician’s voice and feedback to ensure this task is designed and understood to be compatible with the goals of high-quality patient care and not seen as a distraction from those efforts.”²

References

1. Walesa MB, Denny A, Patel A, et al. Enhancement and implementation of a health information technology module to improve the discrete capture of cancer staging in a diverse regional health system. JCO Oncol Pract. 2023;19(10):e957-e966. doi:10.1200/OP.23.00104
2. Vidula N, Peppercorn J. Clicking away to capture cancer Staging—the benefits and challenges of completing standardized staging modules. JCO Oncol Pract. 2023;19(10):835-838. doi:10.1200/OP.23.00500