Shah's Score After a Decade: Is It Still a Reliable Tool for Psychiatric Screening in the Emergency Department?

BACKGROUND

In 2024, 6 million patients with mental health issues visited the Emergency Department (ED). No consensus exists on diagnostic tests for medical clearance and limited scoring systems are available for triage psychiatric patients. Shah's score screens patients with behavior-health-related complaints to decide who can be safely transferred to a psychiatric unit without ED evaluation.

OBJECTIVES

To evaluate Shah's score accuracy and potential improvements, aiming to balance care for this underserved population with appropriate ED resource utilization.

METHODS

Retrospective cohort study (May 1, 2020 to May 30, 2021) at an urban academic Level I trauma center. We reviewed the electronic medical records (EMRs) of adult ED patients with psychiatric chief complaints in a large urban teaching hospital. The exclusion criteria included pregnancy, incarceration, or incomplete records. EMRs were reviewed for acute medical diagnoses and management, reported as the mean (SD), and analyzed using Pearson's χ2 test.

RESULTS

Of the 1613 patients screened, 507 met the inclusion criteria. The mean age was 39.1 years (±14); 66.7% were male and 49.7% were African American. Suicidality was the most frequent complaint (49.9%); 55.0% presented voluntarily, and 89.9% had laboratory tests performed. Dispositions included discharge from ED (61.9%), psychiatric unit admission (35.5%), and medical admission (2.6%). The Shah score cleared 76.73% of patients. Eight patients were cleared by the score required admission for acute kidney injury, uncontrolled hypertension, thyroid cancer, anticoagulation, and inability to perform daily activities (ADLs). The Shah score demonstrated a negative predictive value (NPV) of 97.94% (95.99-99.12%) for identifying psychiatric patients who did not require medical workup.

CONCLUSION

The Shah score is a promising tool with the potential to enhance clinical safety and effectiveness.