In an interview with Pharmacy Times, Seth Martin, MD, MHS, professor and preventive cardiologist at Johns Hopkins Medicine and director of the Advanced Lipid Disorders Program and Digital Health Lab at the Johns Hopkins Ciccarone Center for the Prevention of Cardiovascular Disease, discussed a new machine learning version of the Martin-Hopkins equation published in JAMA Cardiology that matches the accuracy of the original while making implementation easier for laboratories.
Martin explained that older equations, including Friedewald and Sampson-NIH, tend to underestimate low-density lipoprotein cholesterol (LDL-C)—potentially leaving high-risk patients short of guideline goals and ineligible for therapies that reduce cardiovascular events. He urged pharmacists to know which equation their laboratory uses and to advocate for more accurate calculations.
Pharmacy Times: What does this simplified machine learning version of the Martin-Hopkins equation mean for pharmacists and other health care professionals who rely on LDL-C results to guide treatment decisions?
Seth Martin, MD, MHS: I think it means they're going to have more accurate LDL for their clinical practice. LDL is front and center in our clinical practice, and it helps guide decisions around medicines that reduce ASCVD—atherosclerotic cardiovascular disease—events and reduce death. These are literally decisions that we're making that could be a matter of life or death for our patients, and we want to have the most accurate results. I'm thrilled that by providing this new machine learning version of my equation, we can scale up accurate results even more widely.
To be clear, this machine learning version is not going to be a dynamic model that's changing over time. It's a very transparent, fixed equation that'll get coded into the lab, so year after year you'll continue to get the most accurate results. I think with the original equation and now with this equation, we've basically optimized LDL-C calculation. What was really cool is that the new machine learning version gives results that are similar in accuracy to the original equation that we published over a decade ago, which was developed in more of a manual fashion.
For pharmacists practicing, they'll be able to rely on more accurate LDL results. Pharmacists and other clinicians are very busy, so I think it's quite important that we get results that are automated in the lab. The whole point of this new equation is to make it easier for labs to automate the calculation so that when the results get to you, they're trustworthy.
Pharmacy Times: How could broader adoption of this updated equation help clinicians identify patients who may benefit from additional LDL-C-lowering therapy sooner?
Key Takeaways for Pharmacists
- Older LDL cholesterol equations, including Friedewald and Sampson-NIH, tend to underestimate LDL-C, which can produce falsely reassuring results and leave patients below the threshold that would qualify them for evidence-based lipid-lowering therapy.
- The greatest risk of misclassification occurs in patients with low LDL-C and elevated triglycerides—for example, an LDL-C near 50 mg/dL with triglycerides in the 200 to 300 mg/dL range in a patient with very high-risk ASCVD.
- Pharmacists should confirm which equation their local laboratory uses when interpreting lipid panels and can serve as local champions for adopting more accurate calculations.
Martin: It's a great question. The key clinical issue is that the original Friedewald equation, as well as other equations that have been developed, tend to have an issue of underestimating LDL-C levels. They may give a result that's falsely reassuring because it seems like the patient's LDL is really low, but in reality it's 10 or 20 mg/dL higher, and they're actually above a clinical threshold where they would qualify for evidence-based lipid-lowering therapy—including statins and non-statin therapies such as ezetimibe (Zetia; Organon), bempedoic acid, and proprotein convertase subtilisin/kexin 9 (PCSK9) inhibitors.
This change to a more accurate equation corrects for that underestimation and will allow those patients to get a better representation of their LDL-C levels and to qualify for these therapies that reduce the risk of cardiovascular disease. We're talking about preventing heart attacks and strokes and saving lives.
Pharmacy Times: Are there specific patient populations where pharmacists should pay especially close attention to the accuracy of LDL-C calculations when reviewing lipid panels?
Martin: This is a great question. The stress test of LDL accuracy is in patients who have lower LDL levels and higher triglycerides. If you see a lipid panel, for example, in someone you're treating with very high-risk ASCVD and their LDL level is coming back on the lower side—let's say your goal is less than 55, and you get an LDL of around 50, and their triglycerides are in the 200 or 300 range—that's a type of situation where I'd be very wary of which LDL-C equation was used.
If it is the Friedewald equation or the Sampson-NIH equation, I would be concerned that the LDL level is underestimated and they're actually not at goal. If you were to use the original version of my equation or the new machine learning equation, you would find that they're actually above that 55 goal, and they would still qualify for additional therapies to reduce risk of heart attacks and strokes and to reduce mortality.
Pharmacy Times: As more laboratories adopt this updated equation, what key takeaway would you like pharmacists and other health care professionals to keep in mind when interpreting LDL cholesterol results and managing cardiovascular risk?
Martin: It's important to keep in mind what we just said about knowing which equation was used to calculate the LDL result. If you're getting a lot of results from your local laboratory, hopefully it's clear what equation is being used. Otherwise, it's worth checking in on that. If you're getting results from Quest Diagnostics, it is using my equation, but other labs are using different equations. Being cognizant of that is really important—and serving as a local champion for change, if it is needed to make the switch.
I also think it's important to be cognizant of that issue as it pertains to LDL goal achievement. In our 2026 dyslipidemia guideline, we have very clear LDL goals of less than 100, less than 70, and less than 55, based on risk status. So knowing where a patient is, making sure you have the most accurate LDL results in relation to those goals, and then being mindful that in our guideline we go beyond LDL to also have non-high-density lipoprotein and apoliprotein B goals that can be used to further confirm that therapy has been optimized.
Pharmacy Times: Is there anything else that you would like to add?
Martin: I'd like to thank you for covering this, because this is such a key issue. Cardiovascular disease still is the leading killer in the US and around the world. What I would like to do is acknowledge the very important role that pharmacists play. My wife is a pharmacist. I recognize the very important role that pharmacists play in clinical care, and we're all in this together on the front lines of clinical care, addressing the leading killer of people.
We're at a time where we have very effective therapies, and if we implement these well and we use accurate LDL equations, I think we can continue to build momentum in reducing the burden of mortality and morbidity due to cardiovascular disease. I'm very hopeful for the future, and I want to recognize the very key role that pharmacists play in delivering this evidence-based care and saving a lot of lives.
