Heart failure prevention in type 2 diabetes in primary care: STRONG-DM

This transcript was generated using an AI-based transcription tool and may contain errors. It reflects the spoken content of a live webinar and has been lightly edited for clarity.

Speaker: Ambarish Pandey, M.D., M.S.C.S. Associate Professor at UT Southwestern Medical Center

Can you briefly describe the aim and the design of the STRONG-DM study?

The STRONG-DM trial is focused on evaluating if a structured electronic health record-based clinical decision support intervention can improve the screening of patients with type 2 diabetes by their primary care doctors using NT-proBNP to identify high-risk patients for heart failure, and also if we can pair that with an aggressive and intensive strategy to get them on guideline-directed medical therapy for prevention of heart failure. So in this randomized controlled trial, we are enrolling patients with type 2 diabetes who are being seen by primary care doctors across four health systems. And then the primary care doctors are randomized to an intervention arm where they get a best practice alert when they see these patients in the clinic, suggesting them to consider testing for natriuretic peptide for screening for heart failure. And the natriuretic peptide test is NT-proBNP. And if the NT-proBNP levels are elevated, or if they have elevated clinical risk based on the WATCH-DM risk score, they also get as a second phase of the intervention, virtual consult recommendations for initiating these high-risk patients on evidence-based therapies like SGLT2 inhibitors, finerenone, and GLP-1 receptor agonists. And the outcome of interest in this study is the uptake of these evidence-based therapies by one year follow-up, and also looking at incidence of heart failure hospitalization and mortality in the intervention versus the control group. The control group is usual care, and they do not get any best practice alert recommendation or virtual consult guidance. But they are free to use best practices that they have based on the guideline recommendations.

What are the main findings from the pilot phase of STRONG-DM? 

We conducted a pilot phase of the STRONG-DM trial to better understand the process and the implementation framework of the study. This was a single-center study that was done at UT Southwestern Medical Center and we enrolled seven physicians — primary care physicians who were taking care of patients with type 2 diabetes. And we provided them with the same intervention: the BPA alert, paired with the virtual consult recommendations for evidence-based therapies for those who are high risk based on the NT-proBNP testing or the WATCH-DM risk score. And the outcome that we were looking at was a short-term, up to six-month uptake of evidence-based therapies like SGLT2 inhibitors, finerenone, and GLP-1 receptor agonists. And the idea was: can we feasibly implement such an intervention? That is, best practice alert-based screening with recommendations for initiation of evidence-based therapies. And we also had a control group which was a historical control of patients who were seen in the clinic by the PCP for diabetes during the same time period and were age-, sex-, BMI-matched. So the idea was that we can compare it to a control group that was not randomized, but rather coming from the same time period from different PCPs who were not part of the pilot study. We observed that there was a significant increase in the screening — so up to 87 or 88% of the times that the BPA alert fired, prompting the PCP to do NT-proBNP testing, it led to a successful ordering of the NT-proBNP test. In the pilot study we enrolled around 545 patients, of which 36.9% were identified as high risk for developing heart failure, and NT-proBNP screening identified 18.3% of these patients. So almost 50% of all high-risk patients that were identified were identified thanks to NT-proBNP and would have been missed if we had just used a risk score because they were deemed low risk based on the risk score. We also noted that in response to our intervention, pairing the screening with the virtual consult recommendation, there was a significant increase in utilization rates of SGLT2 inhibitors and finerenone, which was much greater and statistically significantly higher than what we observed in the control group that was matched from the same time period, from different clinicians who were not part of this pilot phase. So we got a lot of good, encouraging results from the pilot study regarding the implementation, and regarding the workflow integration, and regarding the potential efficacy of this approach, and that has informed the setup and the conduct of the larger trial, the randomized controlled multicenter trial, which is the STRONG-DM study. We also had a negative control of looking at the prescription of statins, which was not part of the heart failure screening guideline recommendations for prevention. And there was no meaningful change in the use of statins from baseline to follow-up in the intervention arm, highlighting that the screening strategy focused on heart failure risk, and the intervention of virtual consul guideline recommendations for heart failure preventive therapies, was meaningfully associated with improvement in use of these intended therapies.

What were the key success factors for such a high compliance rate with primary care providers?

So we saw a very high rate of compliance to our best practice alert recommendation amongst the patients in the pilot study. Of all the patients in whom the best practice alerts fired, requesting the PCP to consider NT-proBNP-based heart failure screening in patients with diabetes, up to 93% of the time the order was placed. So there was a very high rate of success in terms of translating the best practice alert-based guidance into successful ordering of the NT-proBNP screening. The key factors underlying our success, with very high levels of uptake that we saw in terms of prescription, in terms of screening and testing with NT-proBNP was related to the participation of PCPs in the study design. We met with the PCPs before the launch of the study. We got their input on the challenges that they faced with doing screening and testing in a general primary care population. We educated them about the role of NT-proBNP and the importance of heart failure risk assessment early, and the benefits of evidence-based therapies like SGLT2 inhibitors, finerenone, and GLP-1 receptor agonists for preventing heart failure in high-risk patients. And then we worked with them to design a workflow that was seamlessly integrated into their clinical practice, and that allowed them to do the screening study without interrupting their clinical workflow too much. So I think that active engagement and participation of the PCPs was a key factor in the implementation success of the screening. We also took intentional effort to make sure that there was no additional blood draw, and that the testing was being done in patients who were already getting another blood test for HbA1c. So our best practice alert prompted patients to be tested for NT-proBNP screening if they were going to get a blood draw for hemoglobin A1c for routine surveillance. That allowed us to avoid additional needle sticks and additional blood draws, and also made it more efficient for the primary care doctors to order all the blood tests at one time.

What were the main challenges or concerns raised by PCPs? 

The main challenge that the PCPs highlighted and that we addressed during the study were related to, one: alert fatigue. And as you know, the primary care doctors get a lot of best practice alerts for different conditions and different diseases. So making sure that our best practice alert for heart failure screening was not adding extra burden. And to minimize that, we made sure that there's only a two-click testing order. So they only had to click twice on the BPA alert side to get to the NT-proBNP order. And also combining it with hemoglobin A1c testing reduced the burden on the patients, which was also a concern raised by PCPs. It is also an important point from the PCP standpoint to make sure that the patients understand the concept of risk and the concept of prevention — that testing for intervention did not mean that they had heart failure, but it meant that we were being proactive and taking steps to prevent development of heart failure. So we had to act and educate the patients as well as the PCPs regarding what the testing was done for and what the results would mean for their heart failure care. So I think education and comprehensive awareness about the need for prevention, and the need for screening among both PCPs and patients, was a critical step in the implementation of the pilot phase. Another concern that was raised by PCPs during the planning phase was regarding the coverage of NT-proBNP screening tests in patients with diabetes by insurance. And we worked with the PCPs and we worked with our electronic health record and informatics team to make sure that we were linking the test order with the right diagnosis code for screening and for identifying high-risk patients with heart failure, to ensure smooth coverage by insurance companies for the NT-proBNP screening order. And we did not see any major issues in the running of the pilot phase with the insurance coverage for the NT-proBNP test.

What are the prerequisites to implement in clinical practice the prevention strategy proposed in STRONG-DM?

I think there are several important considerations as we think about scaling this to clinical practice and having it at a larger scale. First and foremost is the buy-in from the primary care doctors and making sure that we have communication with the primary care doctors, and that they are aware of the need for screening and the challenges with implementation of evidence-based therapies and how the screening can facilitate. So education of PCPs is an important first step. Partnership with PCPs to make sure that the workflow is integrated seamlessly into their day-to-day clinical practice and is not disruptive to that routine clinical workflow is also important. Having a robust cardiometabolic support program led by endocrinologists or cardiologists, nephrologists, or a multidisciplinary team that can be available to respond to the PCP screening results — so that they don't feel like the onus is on them to act on the results, but they feel supported in terms of having the subspecialists' recommendations and guidance with regards to what to do once the results are abnormal for preventing progression of heart failure. So I think having a robust cardiometabolic specialist support, PCP awareness and education are really critical for the success of this program in the real world. I think dedicated effort needs to be put in place for developing the content that is tailored towards educating PCPs, and that is providing them the right resources, evidence-based resources,  for them to buy into the concept of screening and invest in screening strategies during their clinical encounters.

What clinical and economic benefits do you expect from this approach? 

We think early screening and identification of high-risk patients for heart failure with early initiation of evidence-based therapies can have immense clinical benefits. We know that therapies like SGLT2 inhibitors, non-steroidal MRAs, and GLP-1 receptor agonists are very effective in preventing development of heart failure and other atherosclerotic cardiovascular outcomes. And the sooner we initiate these patients on these therapies, the bigger the returns are in terms of preventing these cardiovascular complications. So we do think that early screening and early initiation of these therapies will lead to reduction in the future risk of heart failure development in these patients. And I think, as we all know, heart failure is a costly problem for the health system. Heart failure hospitalizations and subsequent readmissions are associated with a high cost burden on the health system, on the patients, and on society. So we do think if we can be more proactive and identify patients who are high risk early and get them on the therapies that can prevent progression to overt clinical heart failure, that can reduce heart failure hospitalizations, we will cut down the burden of healthcare expenditure in our society, and it will be a great economic benefit for our health systems as well. And we do agree that while the initial screening, virtual consultant skills and preventive therapies have a cost, they are far less expensive than a single heart failure hospitalization. So our strategy for preventing heart failure aligns well with value-based care models, by optimizing specialist time to only see patients who truly need it, and by making sure that we can leverage virtual models and EHR-based resources for a lot of the preventive strategies that can be implemented at scale without necessarily needing specialist face time. So I think that is a really key component in terms of the economic proposition and economic value of our screening strategy.