A population management dashboard for direct oral anticoagulants was associated with reduced off-label prescribing and fewer adverse clinical events in a nationwide analysis of the Veterans Health Affairs.

In the study, published in the Journal of the American Heart Association, investigators evaluated the implementation and effectiveness of a direct oral anticoagulants (DOAC) population management tool (PMT) dashboard across 123 Veterans Health Affairs (VHA) sites between August 2015 and December 2019. The dashboard facilitated pharmacist oversight, allowing them to review and correct off-label prescriptions.

The analysis included 128,652 patients prescribed DOACs for atrial fibrillation or venous thromboembolism (VTE). The VHA sites were grouped based on when they achieved moderate-high usage of the DOAC PMT dashboard, defined as 2 to 5 login days per month.

The investigators used data from the VHA Clinical Data Warehouse and Centers for Medicare and Medicaid Services. Off-label DOAC dosing was determined according to VHA criteria for use guidelines. Clinical adverse events were identified using ICD-9 and ICD-10 codes.

Effectiveness was assessed by measuring monthly rates of off-label DOAC prescribing and clinical adverse events (bleeding, composite of stroke, or VTE). Implementation was evaluated as the percentage of off-label DOAC prescriptions changed within 7 days.

The study population had a median age of 73 years (interquartile range [IQR] = 67–78), was predominantly male (96.2%), and received care in urban settings (92.7%). The median CHA2DS2-VASc score was 2 (IQR = 1–4), HAS-BLED score was 2 (IQR = 1–2), and Charlson comorbidity index was 3 (IQR = 2–4).

Among the key findings were:

• Off-label DOAC prescribing decreased from 8.7% to 7.6% after dashboard implementation across all sites.
• Sites adopting moderate-high dashboard use before July 2018 showed statistically significant declines in off-label DOAC prescribing.
• All groups experienced reductions in composite VTE and stroke rates following dashboard adoption.
• The group adopting moderate-high use between May and December 2017 demonstrated a significant decline in bleeding events.

DOAC prescribing patterns were:

• Apixaban: 55.7%
• Rivaroxaban: 30.1%
• Dabigatran: 13.4%
• Edoxaban: 0.9%.

Off-label DOAC prescribing rates by implementation group: preimplementation mean (standard deviation [SD])/postimplementation mean (SD):

• On/before April 2017: 7.9% (0.6%)/6.9% (0.7%)
• May to December 2017: 8.0% (0.4%)/7.5% (0.4%)
• January to June 2018: 8.1% (0.4%)/7.2% (0.3%)
• On/after July 2018: 10.2% (0.9%)/10.2% (0.4%).

Bleeding event rates (per 100 patient-years) by implementation group: preimplementation/postimplementation:

• On/before April 2017: 3.7 (0.6)/3.1 (0.3)
• May to December 2017: 2.9 (0.3)/2.8 (0.3)
• January to June 2018: 2.8 (0.3)/2.6 (0.3)
• On/after July 2018: 3.0 (0.5)/3.3 (0.4).

Composite VTE and stroke event rates (per 100 patient-years) by implementation group: preimplementation/postimplementation:

• On/before April 2017: 2.7 (0.2)/2.2 (0.3)
• May to December 2017: 2.7 (0.1)/2.1 (0.2)
• January to June 2018: 2.5 (0.1)/2.0 (0.2)
• On/after July 2018: 3.0 (0.2)/2.6 (0.2).

Linear regression analysis showed statistically significant declines in off-label DOAC prescribing trends for all groups after achieving moderate-high dashboard use. The earliest adopting group (before April 2017) demonstrated the most notable decline, beginning in September 2017.

Composite VTE and stroke rates declined significantly for all groups following dashboard adoption. The change was smallest for the earliest adopting group (change in slope = −0.01, 95% confidence interval [CI] = –0.03 to –0.01, P = .021) and largest for the latest adopting group (change in slope = −0.43, 95% CI = –0.05 to –0.03, P < .001).

The implementation analysis found no significant change in the rate of off-label DOAC prescriptions resolved within 7 days before and after dashboard adoption.

Sensitivity analyses using "very high" DOAC PMT dashboard use thresholds (≥ 10 days and ≥ 15 days with at least 1 login per month) yielded results largely consistent with the primary analysis for both bleeding and stroke/VTE outcomes.

The limitations included that the study population was predominantly male and urban, with higher dabigatran use compared with non-VHA populations. VHA pharmacists have unique prescribing authority that may not apply to other health systems. The analysis was conducted at a population level without adjustment for patient-specific factors at each site. As a retrospective study, causation could not be established.

The lead study author reported grants from Boston Scientific and consulting for several pharmaceutical companies, while another study author reported speaking for AstraZeneca; both are on the Anticoagulation Forum Board. The remaining authors have no disclosures.