Addressing Physical Therapy Access Gaps: Impact of Embedded Scheduling Tools on Patient Language Preference

Abstract Importance Patients with Non-English language preference (NELP) face barriers accessing outpatient rehabilitation services, with disparities at the referral-to-scheduling stage. Despite decades of Federal mandates to provide meaningful access, patients with NELP remain operationally invisible in high-volume scheduling workflows, contributing to lower completion rates and delays compared to patients identified as English-proficient (EP). Objective The objective was to evaluate whether implementing an electronic health record (EHR)-integrated language filter and visible language preference field improves referral scheduling for patients with NELP. Design This was a quasi-experimental study using interrupted time-series analysis. Setting This study took place in outpatient rehabilitation clinics at an urban academic medical center. Participants We studied 19,260 physical therapy referrals, including 985 (5.1%) patients with NELP and 18,275 (94.9%) patients with EP, from September 2024 to May 2025. Intervention We used a sortable “Preferred Language” field and toggle filter in the EHR referral workqueue enabling schedulers to identify and proactively schedule patients with NELP. Main Outcomes Our primary outcome was the proportion of referrals scheduled within 10 days of referral. Secondary outcomes included overall scheduling rates, time from referral to scheduling, scheduling to visit, and referral to visit; no-show and unscheduled referral rates; and Press Ganey patient satisfaction scores related to appointment scheduling. Results Patients with NELP demonstrated a larger increase in the likelihood of scheduling within 10 days following the intervention (adjusted RL = 2.86, 95% CI = 1.82–4.48) relative to patients with EP. Scheduling within 10 days improved from 27.4% to 31.3% pre- and post-intervention. However, baseline disparities persisted, with patients with NELP experiencing a longer median time to scheduling throughout the study period (7 days vs 6 days). Implementation fidelity varied, with 22% of schedulers regularly using the tool. Conclusions An EHR-integrated language visibility intervention improved 10-day scheduling rates among patients with NELP (adjusted RL = 2.86), yet baseline disparities persisted. Low implementation fidelity suggests that visibility intervention requires pairing with capacity enhancement and accountability structures for sustained, equitable access to rehabilitation services. Relevance Language-visibility EHR tools improve but do not eliminate scheduling disparities, pointing to the need for implementation infrastructure alongside technical solutions.