BackgroundEnhanced Recovery After Surgery programs have been shown to effectively reduce opioid prescriptions at discharge after their implementation in several institutions, but little is known regarding the sustainability of this effect. Understanding opioid prescribing patterns after long-term implementation of Enhanced Recovery After Surgery initiatives may help guide further opioid prescription reduction and improvements. Our group aimed to determine whether reductions in opioid prescriptions at discharge are sustained in an Enhanced Recovery After Surgery program for thoracic surgery. MethodsThis retrospective cohort included 2,081 patients undergoing thoracic surgery within a 4-year Enhanced Recovery After Surgery program from March 2016 through April 2020. Our Enhanced Recovery After Surgery protocol included a standardized multimodal analgesic regimen (ie, preoperative gabapentin, tramadol, intercostal nerve block with liposomal bupivacaine, and intraoperative acetaminophen, and ketorolac) and the rest of the interventions recommended by the Enhanced Recovery After Surgery society guidelines. Our primary outcomes were the presence of opioid prescriptions at discharge (hydrocodone, hydromorphone, and oxycodone) and the total opioid amount prescribed (morphine equivalent daily dose). Multilevel generalized linear models were used to account for surgeon variabilities and types of thoracic resection. ResultsOver the study period, the rate of opioid prescriptions at discharge reduced from 35% (Mar 2016) to 25% (Apr 2020), and the amount of opioid prescribed declined from 184 ± 321 morphine equivalent daily dose to 94 ± 251 morphine equivalent daily dose. In multilevel generalized linear models, there was a sustained downward trend in opioid prescriptions over the study period (β -11.8 morphine equivalent daily dose per year, P = .048), which was also directly correlated with the use of minimally invasive surgery (β -84.9 morphine equivalent daily dose for video-assisted thoracoscopic surgery, P < .001; β -139.2 morphine equivalent daily dose for robotic-assisted thoracic surgery, P < .001), intraoperative opioid administration (β -1.4 morphine equivalent daily dose per 1 morphine equivalent dose, P = .026), and the amount of postoperative acetaminophen (β -18.2 morphine equivalent daily dose per 1 g, P = .026). The sustained reduction of opioid prescriptions at discharge did not impact hospital readmission rates within 30 days (odds ratio 1.17, 95% confidence interval 0.86–1.59, P = .306). Subgroup analysis showed a significant, sustained decrease in hydromorphone (β -10.9 morphine equivalent daily dose per year, P = .004), but not for hydrocodone prescriptions (β -5.7 morphine equivalent daily dose per year, P = .168) or oxycodone (β +4.78 morphine equivalent daily dose per year, P = .183). ConclusionOur Enhanced Recovery After Surgery program for thoracic surgery contributed to a sustained reduction of opioid prescriptions at discharge, which positively correlated with the duration of its implementation and the use of minimally invasive surgical techniques but was negatively impacted by the amount of intraoperative opioid administration.