New Persistent Opioid Use Among Patients With Cancer After Curative-Intent Surgery

Abstract

Purpose The current epidemic of prescription opioid misuse has increased scrutiny of postoperative opioid prescribing. Some 6% to 8% of opioid-naïve patients undergoing noncancer procedures develop new persistent opioid use; however, it is unknown if a similar risk applies to patients with cancer. We sought to define the risk of new persistent opioid use after curative-intent surgery, identify risk factors, and describe changes in daily opioid dose over time after surgery. Methods Using a national data set of insurance claims, we identified patients with cancer undergoing curative-intent surgery from 2010 to 2014. We included melanoma, breast, colorectal, lung, esophageal, and hepato-pancreato-biliary/gastric cancer. Primary outcomes were new persistent opioid use (opioid-naïve patients who continued filling opioid prescriptions 90 to 180 days after surgery) and daily opioid dose (evaluated monthly during the year after surgery). Logistic regression was used to identify risk factors for new persistent opioid use. Results A total of 68,463 eligible patients underwent curative-intent surgery and filled opioid prescriptions. Among opioid-naïve patients, the risk of new persistent opioid use was 10.4% (95% CI, 10.1% to 10.7%). One year after surgery, these patients continued filling prescriptions with daily doses similar to chronic opioid users ( P = .05), equivalent to six tablets per day of 5-mg hydrocodone. Those receiving adjuvant chemotherapy had modestly higher doses ( P = .002), but patients with no chemotherapy still had doses equivalent to five tablets per day of 5-mg hydrocodone. Across different procedures, the covariate-adjusted risk of new persistent opioid use in patients receiving adjuvant chemotherapy was 15% to 21%, compared with 7% to 11% for those with no chemotherapy. Conclusion New persistent opioid use is a common iatrogenic complication in patients with cancer undergoing curative-intent surgery. This problem requires changes to prescribing guidelines and patient counseling during the surveillance and survivorship phases of care.

Fig 1.

Study cohort and sample criteria. Patients with cancer who underwent curative-intent surgery from January 1, 2010 to June 30, 2014 were included if they met the following criteria: age ≥ 18 years, continuous insurance enrollment from 1 year before surgery to 1 year after surgery, filled an opioid prescription attributable to surgery (between 30 days before surgery and 14 days after discharge). We also specifically excluded patients with an additional operation within 180 days, hospital length of stay > 30 days, and those who died during their index admission or were discharged to home hospice.

Fig 2.

(A) Trajectory of daily opioid dose stratified by perioperative opioid use. Mean daily opioid dose for each group was calculated every 30 days from 1 year before surgery to 1 year after surgery, while adjusting for preoperative opioid prescriptions, initial opioid prescribed, procedure type, adjuvant and neoadjuvant therapy, and patient characteristics. One year after surgery, patients who developed new persistent opioid use continued filling opioid prescriptions with daily doses similar to intermittent and chronic opioid users (P = .05). (B) Trajectory of daily opioid dose stratified by timing of chemotherapy. Patients who developed new persistent opioid use and received adjuvant chemotherapy had higher daily opioid doses compared with those who received no chemotherapy (P = .002). All groups, however, continued filling prescriptions with high daily doses, equivalent to five to six tablets per day of 5-mg hydrocodone. OME, oral morphine equivalent.

Fig 3.

Risk of new persistent opioid use stratified by procedure and timing of chemotherapy. We calculated the risk of new persistent opioid use by procedure with patients stratified by timing of chemotherapy. This figure shows 95% CIs with the risk of new persistent opioid use adjusted for preoperative opioid prescriptions, initial opioid prescribed, timing of chemotherapy, and patient characteristics. Patients receiving adjuvant chemotherapy had a higher risk of new persistent opioid use (15% to 21%), but those who received no chemotherapy were still at risk (7% to 11%). HPB, hepato-pancreato-biliary.

Fig A1.

Risk of new persistent opioid use stratified by procedure and timing of chemotherapy (excluding patients with preoperative opioid prescriptions within 30 days before surgery). Patients who received preoperative opioid prescriptions (within 30 days before surgery) were excluded from this analysis. We calculated the risk of new persistent opioid use by procedure with patients stratified by timing of chemotherapy. This figure shows 95% CIs with the risk of new persistent opioid use adjusted for initial opioid prescribed, timing of chemotherapy, and patient characteristics. Patients receiving adjuvant chemotherapy had a higher risk of new persistent opioid use (14% to 20%), but those who received no chemotherapy were still at risk (6% to 11%).