Economic burden and clinical impact of preoperative opioid dependence for patients undergoing lower extremity bypass surgery.

Objective: Surgeons' prescription practices and the opioid epidemic have received significant attention in the media. Limited data exist, however, on the impact of prior or coexistent opioid use on vascular surgery outcomes. This study aimed to quantify the incidence, economic burden, and clinical impact of pre-existing opioid dependency in patients undergoing lower extremity bypass (LEB) surgery.

Methods: Data were collected from 1,132,645 weighted (230,858 unweighted) patient admissions for LEB in the National Inpatient Sample for the years 2002 to 2015. Patients with a concomitant diagnosis of opioid abuse or dependency were identified using International Classification of Diseases, Ninth Revision codes. Matched cohorts of patients with (n = 606 unweighted) and without (n = 32,343 unweighted) opioid dependence were created using coarsened exact matching to control for patient demographics. Linear regression was used to control for hospital-level factors and to identify differential outcomes for patients with opioid dependency. Our primary end points were hospital cost and length of stay. Our secondary end points were surgical complications and in-hospital mortality.

Results: There were 1,132,645 (230,858 unweighted) patient admissions for LEB in the National Inpatient Sample during 2002 to 2015. There were 3190 (0.3%) patients (643 unweighted) who had a diagnosis of pre-existing opioid dependency. The incidence of opioid dependency rose over time (2002, 0.13%; 2015, 0.63%; R = 0.90; P < .001). Before matching, opioid-dependent patients were younger (53.9 ± 12.3 years vs 66.7 ± 12.1 years; P < .001) and more likely to be male (65.2% vs 61.9%; P < .001), to be nonwhite (37.9% vs 24.1%; P < .001), to pay with Medicaid (29.6% vs 7.4%; P < .001), and to fall in the lowest income quartile based on ZIP code (39.6% vs 27.5%; P < .001). After matching, opioid-dependent patients (n = 606 unweighted vs n = 32,343 unweighted nonopioid-dependent patients) were at increased risk of surgical site infections (odds ratio [OR], 1.61; P = .006), major bleeding (OR, 1.56; P = .04), acute kidney injury (OR, 1.46; P = .02), and deep venous thrombosis (OR, 2.53; P = .005). Linear regression of matched cohorts revealed that opioid-dependent patients had an increased length of hospital stay (11.76 days vs 9.80 days; P < .001) and an increased mean inflation-adjusted in-hospital cost of U.S. $7032 ($37,522 vs $30,490; P < .001).

Conclusions: The incidence of pre-existing opioid dependency in patients undergoing LEB continues to rise. Patients with opioid use disorder undergoing LEB surgery have substantial increases in length of hospital stay and costs. These findings highlight the importance of early preoperative recognition of this disorder in vascular surgery patients and open the opportunity for early intervention in that cohort.