The COVID-19 pandemic disrupted routine healthcare delivery. Understanding how dispensing patterns of cardiovascular disease (CVD) medications changed during this period is essential to evaluating the continuity of chronic disease management. We examined weekly patterns in both prescriptions written by providers and prescriptions dispensed to patients for CVD medications before and after the onset of the COVID-19 pandemic in March 2020 using data from the U.S. Department of Veterans Affairs (VA).
Methods
We identified adults with CVD diagnosed between 3/1/2018 and 3/1/2020 using national VA claims and electronic health record data. The index date was 3/1/2020, corresponding to the national emergency declaration for COVID-19. We analyzed 39 weeks before (6/1/2019 to 2/29/2020) and 39 weeks after (3/1 to 12/1/2020) the index date. We included individuals who had at least one cardiovascular medication fill between 6/1/2019 and 3/1/2020 and excluded those who died before 12/1/2020 or were lost to follow-up. We assessed weekly prescription patterns across 12 therapeutic classes, including ACE inhibitors (ACEi), angiotensin receptor blockers (ARB), angiotensin receptor-neprilysin inhibitors (ARNI), beta-blockers (BB), calcium channel blockers, hydralazine, loop diuretics, mineralocorticoid receptor antagonists, nitrates, thiazides, sodium-glucose cotransporter-2 (SGLT2) inhibitors, and statins. Prescriptions were evaluated as provider orders and prescriptions dispensed to patients. For both, we measured weekly frequency (prescriptions/wk), quantity (total days supplied/week), and duration (mean days supplied/prescription). Interrupted time series (ITS) analysis was used as the primary analytic approach to evaluate immediate level changes (β2) and slope changes (β3) in weekly patterns following the COVID-19 onset. Prepost t-tests were conducted as secondary analyses comparing the pre and post-COVID-19 periods. Statistical significance was set at p < 0.05. All analyses were conducted using SAS Enterprise Guide, version 8.3 (SAS Institute Inc.). The study was approved by the VA Greater Los Angeles Institutional Review Board.
Results
A total of 2,201,805 Veterans were included (mean [SD] age, 69.5 [11.1] years; 2,103,526 males [95.5%]).
Prescriptions written by providers
ITS analysis demonstrated significant increasing pre-COVID-19 trends in both frequency (β1 = +2,034 prescriptions/wk, p = 0.0002) and quantity (β1 = +147,544 d/wk, p = 0.0006), followed by significant negative slope changes post-COVID-19 onset (β3 = −3,499 prescriptions/wk, p = 0.014; β3 = −268,227 d/wk, p = 0.007) ( Figure 1 A). In contrast, duration showed no significant level (β2 = −2.02 days; p = 0.27) or slope change (β3 = +0.0003 d/wk, p = 0.996) after COVID-19 onset. In secondary analyses, overall prescribing frequency remained largely unchanged between the preand post-COVID-19 periods (385,084 vs 391,409 prescriptions/wk; p = 0.68), whereas mean duration declined modestly (77.7 vs 74.5 days; p < 0.0001) ( Figure 1 C ). Total quantity showed no significant difference (29,920,457 to 29,147,953 d/wk; p = 0.51).
Cardiovascular medication prescription patterns during COVID-19 pandemic. Interrupted time series (ITS) analysis of cardiovascular medication prescriptions during the COVID-19 pandemic. (A) ITS analysis of prescriptions written with counterfactual projection (dashed line) representing the expected trend in absence of the pandemic. (B) ITS analysis of prescriptions dispensed with counterfactual projection (dashed line). Blue data points and lines represent the pre-pandemic period (39 weeks before March 1, 2020), red data points and lines represent post-pandemic period (39 weeks after March 1, 2020). The vertical line indicates March 1, 2020, corresponding to the U.S. COVID-19 National Emergency Declaration. Analysis includes 2,201,805 veterans across 12 cardiovascular medication classes. (C) Weekly number of prescriptions written by providers (frequency). (D) Weekly number of prescriptions dispensed to patients (frequency).
Prescriptions dispensed by patients
ITS analysis confirmed significant increasing pre-COVID-19 trends in both frequency (β1 = +533 prescriptions/wk, p < 0.0001) and quantity (β1 = +45,588 d/wk, p < 0.0001). However, after COVID-19 onset, the slope for quantity declined significantly (β3 = −80,008 d/wk, p = 0.0001), whereas no significant change was observed for frequency (β3 = +44 prescriptions/wk, p = 0.88) ( Figure 1 B). For duration, ITS showed a significant immediate increase (β2 = +1.36 days; p = 0.009), followed by a sustained negative slope (β3 = −0.209 d/wk, p < 0.0001), indicating an initial adaptation followed by a longer-term shift toward shorter durations. In secondary analyses, dispensing frequency increased (372,408 vs 395,058 prescriptions/wk; p < 0.0001), whereas duration declined (79.2 vs 76.8 days; p < 0.0001) ( Figure 1 D). Total quantity increased modestly (29,497,177 vs 30,296,981 d/wk; p = 0.038).
Drug class-specific trends
Trends varied substantially by drug class. ARBs showed the most remarkable change: frequency increased from 30,386 to 50,634 prescriptions/wk for prescriptions written (p < 0.0001), and from 26,580 to 45,696 prescriptions/wk for prescriptions dispensed (p < 0.0001). ITS analysis similarly demonstrated post-COVID changes in ARB dispensing patterns. Despite these increases, duration decreased sharply from 70.9 to 40.4 days per prescription written (p < 0.0001), and from 80.9 to 52.4 days per dispensed prescription (p < 0.0001). This pattern of increased frequency with decreased duration was unique to ARBs and substantially contributed to the overall decline in duration observed across all cardiovascular medications. Relatively newer agents such as ARNI and SGLT2 inhibitors showed increases in frequency, quantity, and duration.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
