Community health worker contacts for high utilizer vs. other sickle cell patients in an adult sickle cell disease medical home

Abstract Background Medical center-based Community Health Workers, also known as Patient Navigators (PNs), often increase patient engagement by giving high-touch community care, in an effort to substitute for high-cost emergency and hospital care. Adult sickle cell disease (SCD) PN contacts may therefore focus on high-utilizer patients to meet program needs. We were interested in how much more time SCD-specific PNs spent among high-utilizer SCD patients versus other SCD patients, and whether they spent their time differently on the two groups. Methods We analyzed all standardized case contact forms recorded by PNs working in an adult SCD medical home from 2018 to 2024. Contacts were categorized as inpatient, ambulatory visit, phone (live or voicemail), visual telemedicine, text messaging, community visit, home visit, or email. PN-initiated contacts were distinguished from patient-initiated contacts. We annually classified patients as high-utilizers (> $200,000 annual charges) or non-high-utilizers. We compared high-utilizer contacts (HUCs) vs non-high-utilizer contacts (NHUCs). Measures described or compared were number of contacts, reason for the contact, category of contact, initiator of the contact, and contact duration, in minutes (annual sums were compared). Wilcoxon rank-sum tests and descriptive analyses were used. Results There were 27,188 PN-initiated and 2,867 patient-initiated contacts during the analysis period. Ambulatory visit contacts dominated, followed by hospital visits. Community visits took the longest by far, nearly an hour each. Appointment-related issues were the most frequent reason for contacts (45.2%). Emotional support (18.1%) and medication-related concerns (16.9%) were the second and third most frequent reason for contact. HUCs used more annual PN time compared to NHUCs (Wilcoxon test in 2021 p=0.05, in 2022 p=0.002, in 2023 p<0.0001, and in 2024 p=0.02). Patient- initiated contact visits took significantly more of PNs' time compared to PN-initiated contact visits (Wilcoxon test Z=29.34, p<0.0001). Conclusion Between 2018 and 2024, PN-initiated contacts dominated all PN contacts in one adult SCD medical home. Contacts occurred primarily during ambulatory clinic visits, and centered around appointments, emotional support, and medication concerns. In most years analyzed, HUCs used more PN contact time than NHUCs. And though patient-initiated contacts were a small minority of all PN contacts, they took significantly more time on average than PN-initiated contacts, potentially because they were generated by more complex needs, or were comprised of less structured communication. Our findings suggest the efficiency of high-touch engagement, initiated by PNs, for high-utilizing patients with SCD. They emphasize the high resource intensity of patient-driven, unplanned vs. staff-driven, planned patient communication. Future SCD studies should test the ability of high-touch, planned PN communication to prevent more resource-intense contacts, including ED visits and hospitalizations.