State opioid prescribing laws and associations with radiotherapy delays and radiotherapy completion.

Abstract

122 Background: In July 2018, Missouri introduced a 7-day limit for initial opioid prescriptions, though there was an exemption for cancer-related pain. However, some worry that the exemptions may be inadequate for patients with cancer who experience cancer-related pain. Since some courses of radiotherapy can cause pain (eg, mucositis, esophagitis), appropriate opioid access is important for helping patients complete radiotherapy in a timely manner, which is associated with oncologic outcomes. Methods: Patients with cancer of all ages treated with radiotherapy from 2018-2019 at a large academic institution in Missouri were identified. More recent data were excluded due to the COVID-19 pandemic and Missouri Medicaid expansion. Trends over time in the proportion of missed/delayed radiotherapy fractions, incomplete radiotherapy courses, and a composite measure of excessive delays (≥5 days) or incomplete course were evaluated in linear and linear probability models. Additional analyses included patients from Illinois, an adjacent state without opioid prescribing legislation, and we subsequently performed difference-in-differences analyses comparing the outcomes from pre- to post- Missouri opioid legislation between the states. We conducted unadjusted analyses as well as analyses adjusted for age, sex, site, radiotherapy modality and fractions, as well as other covariates that could impact access to radiotherapy (ZIP code deprivation, distance from the facility, insurance status, and social work referral for unmet needs). Subgroup analyses focused on patients with head and neck cancer and lung cancer receiving definitive radiotherapy, who are at high risk of radiotherapy-related pain. Results: A total of 2613 and 1518 patients from Missouri and Illinois were identified, respectively. 4.7% of fractions were missed/delayed, 3.7% of patients did not complete treatment, and 8.1% of patients had either an incomplete course or excessive delays. There were no statistically significant increases (ie, worsening) in any outcome in any quarter after opioid policy implementation, and there were significant decreases in the composite outcome in multiple post-policy quarters in both unadjusted and adjusted analyses. In difference-in-differences analyses comparing trends over time with patients from Illinois, there were no significant associations of policy implementation with any of the outcomes. Results were similar in subgroup analyses for patients with head and neck and lung cancers. Conclusions: Missouri legislation limiting opioid prescribing for non-cancer-related pain was not associated with significant changes in a patient’s ability to undergo radiotherapy. Limitations of the study include evaluation of a single center in a single state and the lack of other intermediate outcomes important to patients including access to opioid medication itself, which we will examine in future work.