Precision Prescribing of SGLT2 Inhibitors for Primary Prevention of Heart Failure in Type 2 Diabetes
Heart failure is a major and often under-recognized complication of type 2 diabetes. Even in people who do not yet have established atherosclerotic cardiovascular disease, chronic kidney disease, or prior heart failure, diabetes itself raises the long-term risk of developing heart failure. Sodium-glucose cotransporter 2 inhibitors, or SGLT2 inhibitors, have changed diabetes care because they lower blood glucose and also provide cardiovascular and kidney benefits. However, the key clinical question is not simply whether these drugs work, but which patients with type 2 diabetes are most likely to benefit from them for preventing heart failure before it starts.
This study addressed that question by developing and validating a prediction model called SABRE, short for SGLT2i Absolute Benefit Response. The model combines an individual’s baseline risk of heart failure with the known relative benefit of SGLT2 inhibitors to estimate the absolute reduction in heart failure risk over 5 years. In practical terms, that means the model tries to answer: how much heart failure risk can be prevented for this specific person, not just for the average patient in a trial?
Why This Study Matters
Current guidelines strongly support SGLT2 inhibitor use in people with type 2 diabetes who already have atherosclerotic cardiovascular disease, heart failure, or chronic kidney disease. The difficulty is that most people with type 2 diabetes do not fall into those high-risk groups. For them, guidelines are less specific about who should receive an SGLT2 inhibitor first, especially when cost, pill burden, side effects, and patient preferences must be considered.
That gap creates a need for individualized decision-making. A patient with a very low baseline risk of heart failure may gain only a small absolute benefit, while someone with higher baseline risk may gain much more. SABRE was designed to capture that difference and help clinicians identify patients for whom SGLT2 inhibitor treatment is likely to provide the greatest preventive value.
How the SABRE Model Was Built
The investigators developed SABRE by combining two pieces of evidence. First, they used a validated heart failure risk model, QDiabetes-HF, to estimate each person’s absolute 5-year risk of developing heart failure. Second, they applied the relative effect of SGLT2 inhibitors on heart failure hospitalization from a meta-analysis of randomized trials, which showed a hazard ratio of 0.63. In other words, in trial populations, SGLT2 inhibitors were associated with a 37% relative reduction in heart failure hospitalization.
By integrating these two elements, SABRE estimates the expected absolute benefit for an individual patient. Absolute benefit is especially useful clinically because it translates statistical risk reduction into a more concrete estimate of how many events may actually be prevented over a given period.
Data Source and Validation
The model was validated using U.K. primary care data linked to hospital and death records from 2013 to 2020. The study included 57,368 people starting an SGLT2 inhibitor and 111,673 comparison patients starting either a dipeptidyl peptidase 4 inhibitor or a sulfonylurea. These comparators are commonly used glucose-lowering drugs and provided a clinically relevant reference group for evaluating real-world outcomes.
This design allowed the researchers to compare predicted risk and observed outcomes in a large population typical of routine practice, rather than relying only on clinical trial participants. That is important because trial populations often differ from the broader diabetes population seen in day-to-day care.
What the Study Found
In real-world practice, SGLT2 inhibitor use was associated with a 30% lower risk of new-onset heart failure compared with the comparator group, with a hazard ratio of 0.70 and a 95% confidence interval of 0.63 to 0.78. This result was consistent with prior trial evidence and supports the role of SGLT2 inhibitors in preventing heart failure in type 2 diabetes.
Interestingly, the relative benefit did not vary meaningfully according to baseline absolute heart failure risk. This means that the proportional reduction in risk was fairly similar across risk groups. However, the absolute benefit differed a great deal because people with higher starting risk can prevent more events in absolute terms. That is the central advantage of individualized prediction.
The SABRE model estimated 5-year absolute heart failure benefit ranging from less than 0.1% to 14.1%, with a median benefit of 1.0% and an interquartile range of 0.6% to 1.8%. In plain language, many patients would have only a small benefit, but a meaningful subset could have a much larger benefit. The model also calibrated well against observed outcomes, meaning its predictions aligned closely with what actually happened in the validation dataset.
Clinical Interpretation
These findings support a more personalized approach to prescribing SGLT2 inhibitors in people with type 2 diabetes who do not already have heart failure, chronic kidney disease, or atherosclerotic cardiovascular disease. Rather than treating all such patients the same, clinicians may be able to use an individualized risk tool to identify those with the greatest likely absolute gain.
For example, two people may have the same diagnosis of type 2 diabetes, but one may be older, have hypertension, obesity, or other features that increase heart failure risk. That person may derive more benefit from an SGLT2 inhibitor than someone with a much lower baseline risk. SABRE is intended to make that distinction clearer.
This does not mean SGLT2 inhibitors should be reserved only for very high-risk patients. They remain valuable glucose-lowering agents with broader cardio-renal benefits, and treatment decisions should still account for kidney function, comorbidities, adverse effects, cost, and patient preference. But the model provides a more evidence-based way to prioritize therapy when the expected benefit is uncertain.
Implications for Practice
In routine diabetes care, precision prescribing can help balance benefit and burden. SGLT2 inhibitors are generally well tolerated, but they are not appropriate for everyone. Potential adverse effects include genital mycotic infections, volume depletion, and rare but important complications such as diabetic ketoacidosis. Therefore, predicting who is likely to benefit most can improve the benefit-risk ratio.
The SABRE approach may be especially helpful in primary care, where most people with type 2 diabetes are managed. A simple clinical tool that estimates heart failure prevention benefit could support shared decision-making. Patients may better understand why a medication is recommended if clinicians can explain the likely personal benefit in absolute terms.
From a health system perspective, targeting treatment to those most likely to benefit may also improve cost-effectiveness. That is particularly relevant because SGLT2 inhibitors can be more expensive than older glucose-lowering medications in some settings. Precision prescribing may help allocate resources more efficiently without denying therapy to patients who stand to gain the most.
Strengths and Limitations
A major strength of this work is the integration of randomized trial evidence with real-world validation in a very large cohort. The study also focused on a clinically important population: people with type 2 diabetes who are commonly encountered in primary care but are not automatically captured by existing high-risk indications.
As with all prediction models, there are limitations. The model’s performance depends on the quality of the underlying risk estimates and the assumption that the relative effect of SGLT2 inhibitors is reasonably stable across patient subgroups. Treatment effects may differ in special populations not fully represented in the data. In addition, real-world prescribing patterns, adherence, and follow-up can influence actual benefit. A prediction tool helps guide decisions but does not replace clinical judgment.
Bottom Line
The SABRE model represents an important step toward precision medicine in diabetes care. By estimating the absolute heart failure prevention benefit of SGLT2 inhibitors for individual patients, it helps move beyond one-size-fits-all recommendations. The study suggests that SGLT2 inhibitors can prevent heart failure in type 2 diabetes even before cardiovascular or kidney disease develops, but the magnitude of benefit varies from person to person. A model like SABRE may help clinicians choose treatment more accurately, support shared decision-making, and improve the targeting of preventive therapy.
In the future, tools like SABRE could be incorporated into electronic health records or clinical decision support systems, making individualized prescribing more practical in everyday care. If implemented well, this approach may improve outcomes while ensuring that treatment is directed where it is most likely to help.
