Page 38 - BH-2-3

Brain & Heart Digital tools for stroke and bleeding risk in AF

7.2. Ability to export results for physician among patients with AF with anticoagulation subtracted
documentation by the same defined risk without anticoagulation and the
Most of the tools also generate a dynamic text output that risk of developing a stroke or a bleeding complication in
can be used for electronic health record documentation. patients solely with AF. Absolute risk reduction (ARR) is an
These statements generally involve an explanation of important measure of the effect of a particular intervention.
what the tool has calculated and what the associated risk The actual reduction is formed by taking the difference
is (expressed as a percentage). They furthermore provide between the risk of stroke and bleeding complications
distinction on recommendations for what is required based in patients with AF without any anticoagulation and the
on the digital tool output. With patients having access to risk with anticoagulation in this population, represented
82
their records, such statements are vital to understanding as an actual value. The net clinical benefit (NCB) is a
quantification of the relationship between benefit and
provider decision-making and influencing patient self- harm. It is the difference between the benefit received
advocacy.
and the risk of harm, both based on a specifically defined
7.3. Augmentation/mitigation of risk based on outcome. The derivation of the NCB is typically based on
anticoagulant/antiplatelet used the specific outcomes studied. Among the various digital
tools analyzed in this review, the measures of benefit and
An advantage of these digital tools is that they allow for risk differ depending on the tool’s focus. In general, the
the consolidation of risk data derived from different benefit is defined as the reduction in the risk of bleeding or
validation studies beyond the original point-based score, stroke in patients with AF, whereas harm is defined as the
incorporating multiple types of risk estimation into a single risk of a specific complication associated with the particular
platform. While stroke and/or bleeding risk are essential agent used. For instance, the ACC AnticoagEvaluator app
outputs of these digital tools, it is especially valuable for reports on the annual probability of experiencing harm
physicians and patients to be able to see the impact of from pharmacological agents that modify either stroke or
anticoagulation or antiplatelet therapy on the mitigation or bleeding risk in AF. 83
augmentation of bleeding or stroke risk. The SPARCtool
and the ACC AnticoagEvaluator Application present Each of these statistical measures produces
risks based on each specific pharmacological agent being a comparative conclusion on the benefit of an
considered. The SPARCtool additionally provides a intervention. While the predictive risk scores described
graphical representation of the comparative risk of various in Tables 1 and 2 estimate bleeding or stroke risk, digital
agents. tools incorporate data from other studies that estimate
the modification of this risk from different interventions.
7.4. Statistical contextualization of risk These statistical representations are important aids in
While most of the digital tools evaluated present the communicating risk to both patients and other providers
exact proportion of risk, the ACC AnticoagEvaluator and should be critical features in the further development
Application, the SPARCtool, and the GARFIELD-AF of digital.
risk calculator include statistical effect estimates of their 8. Validation studies for digital tools
outcomes. These features commonly include the relative
risk (RR), the relative risk reduction (RRR), the absolute To determine if these digital tools were clinically validated,
risk reduction, and the net clinical benefit. The RR is aimed we performed searches on Pubmed and Google Scholar
at measuring the strength of the association between the and directly contacted the primary creators of each tool
risk of developing bleeding or stroke on anticoagulation to inquire about any published studies validating their
compared to being without anticoagulation. Any RR >1 clinical performance. Out of the five authors contacted,
suggests a higher risk within the anticoagulation group three returned requests for literature: The Mayo Clinic
compared to the no anticoagulation group, while the Anticoagulation Decision Aid, the GARFIELD-AF Risk
inverse, meaning a ratio < 1, indicates a lower risk in the Calculator, and the SPARCtool. However, we did not
anticoagulation group compared to the no-anticoagulation receive a response or find any validating studies for the
group. For both the risk of stroke and the risk of bleeding, ACC AntiCoagEvaluator and the CardioSmart Atrial
an RR <1 is desired. RRR is a measure of how much an Fibrillation and Bleeding Risk Calculator.
intervention reduces the risk of an outcome compared to a
baseline or control condition, expressed as a percentage. In 8.1. Studies associated with the Mayo Clinic
the context of the output from the digital tools discussed in Anticoagulation Decision Aid
this study, the relative risk reduction is the ratio between The Shared Decision Making for Atrial Fibrillation
the risk of developing a stroke or bleeding complication (SDM4AFIB) consortium published an initial whitepaper

Volume 2 Issue 3 (2024) 15 doi: 10.36922/bh.3068