Analysis of patient-physician portal secure message content and the associated cost savings in oncology practice

1503 Background: The consistent utilization of secure messaging (“SM”) since the onset of the COVID-19 pandemic has provided cancer patients with unprecedented access to their medical providers, resulting in reduced out-of-pocket costs for patients at the expense of increased workload for oncologists. Herein we analyze oncology SM content and report estimated cost savings from reduced traditional clinic visits. Methods: This population-based retrospective cohort study examined the content of patient-initiated SM threads exchanged through the patient portal website or app over a one-year period (6/1/21 to 5/31/22) at 22 Kaiser Permanente Northern California Oncology practices. A trained physician researcher reviewed a random sample of 500 SM threads and categorized them by message content type, acuity, and appropriate level of service. Standard costs were applied using the average of 2021 and 2022 Center for Medicare Services physician fee schedule and US patient average specialty copayment. Results: During the study period, 49,230 patients initiated 398,835 unique SM threads to 238 oncology providers. The average number of replies to patients’ initial messages was 2.6, with a mode of 1 (maximum 36). Of the SM threads reviewed, only 28.0% required oncologist expertise. Based on thread content, the remaining 72.0% may have been better managed by a nurse (35.4%), medical assistant (29.2%), primary care doctor (6.2%), or a different subspecialty provider (1.2%). Emergency care was required in 2.8% of the total threads reviewed. Significant medical care was provided to patients in 26.0% of the total SM threads which would traditionally require a billable visit (i.e., in-person, telephone, or video). We estimate that these clinical actions may represent 4,299,281.77 of savings in medical care (based on traditional 10-19-minute visits), and 4,562,672.40 in potentially avoided out-of-pocket co-pay costs for patients. Conclusions: Increased utilization of remote medical care and specifically SM has recently generated additional workload for oncologists that may have been more appropriately managed by alternative providers. The magnitude of unreimbursed medical care provided via SM and the use of SM by cancer patients for emergent medical situations creates an urgent need for new models to assist with practice management. An alternative architecture for triaging, managing, and billing SM that utilizes artificial intelligence, machine learning algorithms, or medical assistants, nurses, and other providers could help manage the burden experienced by oncologists.