Chief Medical Informatics Officer Illumination Health, United States
Background: The impact of the SARS-Cov-2 pandemic on adherence to parenteral therapies, in particular to the anabolic osteoporosis biologic romosozumab (romo) administered once monthly by healthcare providers, has not previously been evaluated.
Objectives: To evaluate the impact of COVID lock-down on the adherence of romo treatment.
Methods: Using fee-for-service Medicare data from 1/1/2017 to 9/30/2021, women age ≥65 years newly initiating romo between 4/1/2019 and 6/30/2021 were identified. Patients were excluded if they had metastatic cancer or Paget’s disease before romo initiation. Patient demographics (age, race, geographic region), provider specialty, and baseline comorbidities were identified. Romo dispensations were grouped into (5) 6-month periods based on the dispensing date from FDA licensure (4/2019) to end of the data (Period 1 to 5). As the nationwide COVID emergency was declared on Mar 13, 2020, the period immediately before lockdown (Period 2: 2019-10-1 ~ 2020-3-30) was set as referent group. “Missed dose” was defined as any interval gap between 2 dispensations > 60 days. The numbers of Missed Doses were aggregated per period per patient. Mixed effect Poisson regression model with patient level random effects was performed to evaluate the association between the time period of dispensing and Missed dose, adjusting for number of prior doses, demographic, comorbidities, and socioeconomic factors. We then evaluated an interaction term between Period and number of prior doses.
Results: There were 12216 romo new users identified, contributing 85561 pairwise romo dose intervals separated by a mean of 34.3±23.0 days. A total of 2724 Missed dose events were identified among 2229 romo users. After adjustment, the IRRs (95% CI) of Missed dose for Period 1, 3, 4, and 5 were 0.51 (0.34, 0.77), 3.10 (2.73, 3.52), 2.70 (2.37, 3.08), and 2.90 (2.54, 3.31), respectively (p < 0.01 for all IRRs), compared to the pre-lockdown referent period. Receipt of more doses was associated with a lower rate of a subsequent Missed dose (IRR=0.71, p< 0.01). After adding the interaction term, per 1 prior dose increase, the IRRs (95% CI) of Missed dose for pre-lockdown periods (Period 1 and 2) were 0.04 (0.02, 0.08) and 0.44 (0.39, 0.49), respectively, while for post-lockdown periods (Period 3, 4, and 5) were 0.70 (0.69, 0.71), 0.76 (0.75, 0.78), and 0.71 (0.69, 0.73), respectively (p < 0.01 for all IRRs).
Conclusions: Compared to the pre-COVID period, the lockdown substantially and negatively impacted romo adherence among Medicare beneficiaries who initiated romosozumab. Prioritizing in-time assistance for patients receiving a provider-administered parenteral therapy is critical when patient’s in-person access to their provider is compromised.
