Analysis of preoperative measures that predict interference with sleep recovery after surgery

Previously, we reported a prospective observational study of the determinants of delayed opioid cessation after 5 distinct surgical procedures.¹ We conducted an exploratory analysis on this same data set to discover determinants of time to sleep recovery after surgery. This was defined as a decrease of reported pain interference with sleep on the Brief Pain Inventory to pre-operative values (or lower) for 5 consecutive days after surgery. All significant predictors in univariate Cox regression were considered in multivariate analysis. Three significant predictors were identified. Visual-analog pain score at the time of opioid cessation predicted a prolonged time to sleep recovery (HR=0.458, p=0.0213). One question of the Screener and Opioid Assessment for Patients with Pain (“How often have you had a problem getting along with the doctors who prescribed your medicines?”) also predicted a prolonged time to sleep recovery (HR=0.182, p=0.0119). Finally, pain interference with sleep prior to surgery predicted shortened time to sleep recovery (HR=6.981, p=0.0003). Surgery type and pre-operative anxiety, depression and post-traumatic stress disorder symptoms were not significant predictors. Higher pain scores at the time of opioid cessation after surgery would reasonably be expected to indicate continued pain interference with sleep. Also, elevated pre-operative pain interference with sleep should predict faster sleep recovery merely given the elevated baseline sleep disruption. It is plausible that frustration with prescribing physicians may be indicative of underlying personality or psychological traits that may predict prolonged time to sleep recovery, but the association is not easily explained. This exploratory analysis underscores the complexity of the relationship between pain, sleep and recovery from surgery. (1. Carroll I, Barelka P, Wang CK, Wang BM, Gillespie MJ, McCue R, et al, Anesth Analg, 2012.)