Background: Buprenorphine and other medications for opioid use disorder (OUD) can reduce opioid-related morbidity and mortality. It is unknown whether state Medicaid prescription cap policies that restrict the monthly number of covered prescription fills affect the duration of buprenorphine use.

Objective: To identify trajectories of buprenorphine use and determine the association of caps with trajectory group membership among individuals with OUD.

Methods: Using 10 states' Medicaid claims data from 2010 to 2015, we employed group-based trajectory models to identify patterns of buprenorphine fills over 12 months. We conducted multinomial logistic regression to estimate the association of cap policies with buprenorphine trajectory group membership, adjusting for individual- and state-level covariates.

Results: Among 69,306 Medicaid enrollees with OUD who initiated buprenorphine, 16.9 % resided in states with caps. The mean age was 36.2 (SD = 9.8) years and 59.2 % were female. We identified five trajectories: consistent use (40.9 %), delayed discontinuation (14.5 %), early discontinuation (26.4 %), gradually declining use (9.5 %), and rebounding use (8.8 %). Caps were associated with greater risk of membership in the early discontinuation group (adjusted relative risk ratio = 1.47, 95%CI = 1.36,1.59, referent = consistent use). Younger age, male sex, Black race, Hispanic ethnicity, non-opioid substance use disorder, history of acute care utilization were also positively associated with early discontinuation.

Conclusions: Medicaid cap policies were associated with increased likelihood of early discontinuation and other trajectories of inconsistent buprenorphine use relative to states without these policies. Medicaid's prominence as a payer for OUD treatment and 12 states' continued implementation of caps warrant safeguards to ensure cap policies do not undermine buprenorphine access.

Keywords: Buprenorphine; Group-based trajectory modeling; Medicaid; Opioid use disorder.

Copyright © 2025. Published by Elsevier Inc.

背景： 丁丙诺啡和其他用于阿片类药物使用障碍（OUD）的药物可以减少与阿片相关的发病率和死亡率。尚不清楚限制每月覆盖处方数量的状态医疗补助计划政策是否会影响丁丙诺啡使用的持续时间。

目的： 确定丁丙诺啡使用的轨迹，并评估上限政策与OUD患者使用轨迹组归属之间的关联。

方法： 我们利用了2010年至2015年间10个州的医疗补助索赔数据，采用基于群体的轨迹模型来识别一年内丁丙诺啡处方填充模式。通过多项逻辑回归分析，调整个人和州层面的协变量后，估计上限政策与丁丙诺啡使用轨迹组归属之间的关联。

结果： 在69,306名开始使用丁丙诺啡的医疗补助计划登记人员中，16.9%的人居住在实施了限制政策的州。平均年龄为36.2岁（标准差=9.8），59.2%为女性。我们识别出五种轨迹：持续用药（40.9%）、延迟停药（14.5%）、早期停药（26.4%）、逐渐减少使用（9.5%）和反复使用的患者（8.8%）。上限政策与加入早期停药组的风险增加有关（调整后的相对风险比=1.47，95%CI = 1.36, 1.59，参考值为持续用药）。年轻年龄、男性性别、黑人种族、西班牙裔身份、非阿片类物质使用障碍和急性护理利用历史也与早期停药呈正相关。

结论： 医疗补助上限政策相对于没有这些政策的州而言，增加了早期停止使用和其他不一致丁丙诺啡使用的轨迹的可能性。鉴于医疗补助作为OUD治疗主要支付方的重要性以及12个州继续实施限制措施的情况，有必要采取保护措施以确保上限政策不会影响到丁丙诺啡的可及性。

关键词： 丁丙诺啡；基于群体的轨迹建模；医疗补助；阿片类药物使用障碍。