Generalized Sensory Sensitivity for Prediction of Post-Surgical Analgesic Outcomes: An Observational Cohort Study of Total Hip Arthroplasty and Hysterectomy

Authors:
Preprint
Publisher website
Google Scholar
Abstract
Background: A substantial minority (∼20%) of patients fail to achieve meaningful pain reduction following surgery intended to relieve pain. Risk is elevated in patients with nociplastic pain features, but available self-report measures were not designed for pre-surgical screening. We aimed to develop a brief, data-driven screener for poor analgesic response to surgery. Methods: Participants were recruited from tertiary orthopedic and chronic pelvic pain clinics. Total hip arthroplasty participants had Kellgren–Lawrence grades III–IV with hip pain ≥1 year; hysterectomy participants had chronic pelvic pain ≥6 months. The primary outcome was a 50% reduction in worst pain at six months. Items were selected via elastic net regression with k-fold cross-validation from 68 candidates. Results: Of 428 participants (81% female; mean age 51), 35% failed to achieve a 50% pain reduction. The resulting 11-item screener — the GenerAlized sensory sensitivity for sUrGical rEsponsiveness (GAUGE) — comprises pain across seven body regions and four symptom items measuring interoception (nausea, numbness/tingling) and exteroception (sensitivity to sound, sensitivity to odors). GAUGE outperformed the Central Sensitization Inventory, Fibromyalgia Survey Criteria, and PainDETECT for predicting surgical non-response (RR 1.535, 95% CI 1.342–1.55; AUC 0.738; sensitivity 0.741, specificity 0.635) and for predicting Patient Global Impression of Change. In an independent validation cohort of 54 total knee arthroplasty patients, GAUGE outperformed the Fibromyalgia Survey Criteria in predicting pain severity at six-months. Conclusions: GAUGE is a data-driven, theoretically grounded screener for poor analgesic response to surgery, with potential utility for pre-surgical counseling and clinical trial enrichment.
Funding Information
  • NIH (P50AR070600)
  • NIH (R01HD088712)

This publication has 34 references indexed in Scilit: