Impact of multidisciplinary navigation services on timely treatment initiation: A difference-in-differences analysis.

Abstract

163 Background: Patients with a new cancer diagnosis encounter social, structural, and clinical barriers that can delay treatment initiation. Patient navigation programs have been shown to improve cancer treatment timeliness, but few studies have utilized a multidisciplinary approach. NYU Langone Health (NYULH) implemented a multidisciplinary care navigation model integrating social workers (SWs), community health workers (CHWs), and oncology nurse navigators (ONNs) to comprehensively address barriers to care. The NYULH Brooklyn (BK) site implemented six multilingual CHWs and four ONNs in 2023, while Manhattan continued a SW-only model, allowing evaluation using a difference-in-differences (DiD) design. We aimed to describe navigation service utilization and evaluate its impact on cancer treatment within 60 days. Methods: Adults with a new cancer diagnosis who attended at least two appointments at NYULH between January 1, 2022 and January 31, 2026 were eligible for inclusion. The primary outcome was treatment initiation within 60 days of referral to the cancer center. We incorporated inverse probability of treatment weights (IPTW) to adjust for differences in age, sex, race, language preference, and insurance type between sites. DiD analyses compared changes in the proportion of patients initiating treatment within 60 days before and after navigation implementation in BK relative to Manhattan. Results: 26,430 patients were included (BK: n=4,837; Manhattan: n=21,593). In BK, the proportion of patients initiating treatment within 60 days increased from 35% to 40% after implementation. In Manhattan, the corresponding increase was from 49% to 50%. By 2025, 56% of BK patients received at least one navigation service. In unadjusted analyses, navigation implementation increased treatment initiation within 60 days by 3.5 percentage points relative to Manhattan (95% CI: 0.2, 6.8). After IPTW adjustment, the estimated effect was a 2.2 percentage-point increase in treatment initiation relative to Manhattan (95% CI: -1.3, 5.8). Conclusions: The point estimate suggests a modest increase in treatment initiation within 60 days following implementation of a multidisciplinary navigation program. However, the confidence interval was compatible with a range of effects, from a small decrease to a modest increase in treatment initiation. Further evaluation is needed to better characterize the impact of multidisciplinary navigation on timely cancer care.