(095) COVID-19 Lockdown and the Osteoporosis Treatment Among US Medicare Enrollees Who Initiated Romosozumab

Background: COVID-19 induced substantial healthcare changes, disproportionately affecting economically disadvantaged communities. The impact of the lockdown on the osteoporosis treatment of patients treated with romosozumab (romo), an injectable medication given by healthcare providers monthly, has not been evaluated.

Objectives: To evaluate potential disparities related to the impact of COVID lockdown on romo users among US Medicare enrollees who had an interruption in treatment within 12 months after initiation.

Methods: US female Medicare beneficiaries, aged 65+, initiating romo between Apr. 1, 2019, and Sep. 30, 2020, were studied. Exclusions applied for metastatic cancer or Paget’s disease. Tertiles of county-level socioeconomic (SES) theme of Social Vulnerability Index (SVI) classified social determinant, assigned based on patients’ county FIPS codes. Romo discontinuation was defined as 30-day absence of romo injection, plus a 30-day grace period. Patients were not considered for discontinuation after they completed 12 doses or 12 months of treatment. Primary outcome was the treatment restart (switch or romo resumption) within 6 months post-last romo injection before discontinuation. Binary lockdown indicator (anchored at Mar. 13, 2020) was introduced. Proportional subdistribution hazard models, with SES tertiles, lockdown, and interaction term, assessed association. Covariables included demographics, comorbidities, romo treatment duration pre-discontinuation, health care utilization, etc.

Results: The study involved 6777 romo new users, with 3372 meeting discontinuation criteria. Among them, 1929 (57.2%) restarted treatment within 6 months, with 713 (40.0%) switching to other OP medications. The mean duration (95% CI) of the treatment gap (days) were 106.6 [102.6, 110.5], 115.2 [111.7, 118.8], and 113.6 [110.6, 116.6] for SES tertiles 1, 2, and 3, respectively (p=0.003). The proportional subdistribution hazard models showed that the hazard ratios (95% CI) of treatment switch for SES tertiles 1 and 2 compared with tertile 3 were 0.74 [0.45, 1.23], and 0.96 [0.61, 1.50], respectively, before lockdown, but 1.27 [1.04, 1.57], and 1.00 [0.82, 1.22], respectively, after lockdown, while resumption of romo did not vary by SES.

Conclusions: Among female Medicare enrollees initiating romo treatment, those interrupting or stopping therapy during the pandemic were more likely to switch osteoporosis treatment in the least vulnerable SES counties. This contributed to a shorter osteoporosis treatment gap. Timely assistance for vulnerable patients, especially in low SES areas, is crucial during public health emergencies like COVID-19.