Reply

We commend Tuminello and colleagues for their excellent original investigation¹ and subsequent letter² in response to our recently published work.³ The authors report an additional analysis of opioid prescriptions with a negative binomial regression, finding that patients who underwent video-assisted thoracic surgery (VATS) had fewer opioid prescriptions 6 months after surgery in their Surveillance, Epidemiology, and End Results (SEER)-Medicare opioid-naïve stage I primary lung cancer population (βadj = −0.27, p <0.0001).²

We subsequently performed an additional analysis to investigate the total dosage prescribed after VATS and open lung resections in our previously reported Truven population. While we recognize the utility in analyzing number of opioid prescriptions in the first 6 months after surgery in order to capture continued fills, number of prescriptions only indirectly accounts for total dosage prescribed. We thus assessed total oral morphine equivalents (OME) prescribed to assess whether: 1) opioid prescription size after VATS and thoracotomy differed and 2) total OME prescribed confounded predictors of new persistent opioid usage in our regression model.

We found that despite our previously reported new persistent usage rates of 17.1% (294/1722) after thoracotomy versus 9.4% (117/1243) after VATS (p<0.001),³ mean opioid prescription size did not differ for patients undergoing thoracotomy versus VATS (580 ± 2582 OME vs. 549 ± 790 OME, p=0.68). Adding prescription size in log OME to our multivariable regression for new persistent opioid usage (odds ratio, 0.99; 95% confidence interval, 0.88–1.11; p=0.80) did not affect the odds ratios for the two strongest predictors of new persistent usage: adjuvant therapy (odds ratio, 2.19; 95% confidence interval, 1.74–2.75; p<0.001) and thoracotomy (odds ratio, 1.58; 95% confidence interval, 1.24–2.01; p<0.001).

We conclude that size of opioid prescription for those undergoing thoracotomy and VATS did not differ between surgical approach, whereas new persistent usage rates were nearly twice as high after thoracotomy. In addition, adjuvant therapy and thoracotomy remain the 2 strongest predictors of new persistent usage in our population, even after adjusting for total prescription size.