J Pain Res. 2026 May 16;19:594918. doi: 10.2147/JPR.S594918. eCollection 2026.
ABSTRACT
PURPOSE: Accurate identification of neuropathic pain (NeP) remains challenging in routine clinical practice. While PainDETECT is widely used, its sensitivity is limited by its focus on somatic symptoms. This study aimed to develop a high-sensitivity screening tool by integrating PainDETECT with the Brief Scale for Psychiatric Problems in Orthopaedic Patients (BS-POP) using machine learning.
PATIENTS AND METHODS: Neuropathic pain was diagnosed based on comprehensive clinical evaluation including medical history, neurological examination, and imaging findings when available. We analyzed clinical data from 1083 consecutive patients with pain. The study involved two phases: evaluation of conventional tools via statistical modeling and construction of a random forest-based classification model.
RESULTS: The proposed system achieved an overall accuracy of 75.6%. For NeP, the sensitivity was 70.3% and specificity was 86.0%, representing higher sensitivity compared with the conventional PainDETECT cutoff method (17.6%).
CONCLUSION: Integrating psychosocial factors via BS-POP and utilizing machine learning significantly enhances NeP screening performance. This system may support earlier and more appropriate pain management in clinical practice.
PMID:42170614 | PMC:PMC13189191 | DOI:10.2147/JPR.S594918
