ORIGINAL ARTICLE Acute pain Factors predictive of post-operative pain and opioid requirement in multimodal analgesia following knee replacement J. Thomazeau 1,2,† , A. Rouquette 3,4,† , V. Martinez 1 , C. Rabuel 5 , N. Prince 6 , J.L. Laplanche 6 , R. Nizard 5 , J.F. Bergmann 2,6 , S. Perrot 1,7,† , C. Lloret-Linares 2,6,† 1 INSERM 987, Physiopathologie et pharmacologie clinique de la douleur, Ambroise Par e, Paris, France 2 Assistance Publique-H^ opitaux de Paris, Therapeutic Research Unit, Department of Internal Medicine, H^ opital Lariboisi ere, Paris, France 3 Assistance Publique-H^ opitaux de Paris, Biostatistics and Epidemiology Department, H^ opital H^ otel Dieu, Paris, France 4 INSERM U1178, Mental Health and Public Health, Paris-Sud and Paris Descartes Universities, Paris, France 5 Assistance Publique-H^ opitaux de Paris, Service de chirurgie orthop edique et traumatologique, H^ opital Lariboisi ere, Paris, France 6 INSERM U1144, Variabilit e de r eponse aux psychotropes, Universit es Paris Descartes and Paris Diderot, France 7 Centre d’Evaluation et de Traitement de la Douleur, H^ opital H^ otel Dieu, Paris, France Correspondence C elia Lloret-Linares E-mail: celialloret@yahoo.fr Funding sources None. Conflicts of interest None declared. †These authors contributed equally to the work. Accepted for publication 10 September 2015 doi:10.1002/ejp.808 Abstract Background: Despite the development of multimodal analgesia for postoperative pain management, opioids are still required for effective pain relief after knee arthroplasty. We aimed to identify the determinants of post-operative pain intensity and post-operative opioid requirement in this context. Methods: In this observational prospective study, we recorded patient characteristics, pre-operative pain intensity, anxiety and depression levels, sensitivity and pain thresholds in response to an electrical stimulus, and mu-opioid receptor (OPRM1) and catechol-O- methyltransferase (COMT) single-nucleotide polymorphisms. Multivariate linear regression models were used to identify predictors of post-operative pain at rest and opioid requirement. Results: We included 109 patients. Pre-operative pain at rest (p = 0.047), anxiety level (p = 0.001) and neuropathic pain symptoms (p = 0.030) were independently and positively associated with mean post-operative pain intensity adjusted for mean post-operative morphine equivalent dose (MED). Mean post-operative pain intensity at rest was lower (p = 0.006) in patients receiving celecoxib and pregabalin in the post-operative period, with all other variables constant. Mean post- operative MED over 5 days was low, but highly variable (78.2  32.1 mg, from 9.9 to 170 mg). Following adjustment for mean post-operative pain intensity, it was independently negatively correlated with age (p = 0.004), and positively correlated with associated paracetamol treatment (p = 0.031). No genetic effect was detected in our sample. Conclusions: Our findings suggest that clinicians could use the pre- operative pain profile, in terms of anxiety levels, neuropathic pain symptoms, and chronic pre-operative pain intensity, to improve the efficacy of pain management after knee surgery. 822 Eur J Pain 20 (2016) 822--832 © 2015 European Pain Federation - EFIC â