Could an App Replace Statin Prescribing by Doctors?

Technology may one day supplant physicians when it comes to prescribing statins for cardiovascular disease prevention, a new study suggests. Researchers have found that a web-based decision aid is just as good at identifying moderate-risk patients who are appropriate candidates for statin therapy as well as those who shouldn’t be taking the lipid-lowering drugs.

Use of the decision aid, which asks questions about family history, LDL-cholesterol levels, and blood pressure, among other risk factors, may pave the way for statins being available without the need for physician oversight. Experts believe this may increase their uptake.  

“Only half of patients eligible for statins are actually taking them,” lead investigator Steven Nissen, MD (Cleveland Clinic, OH), told TCTMD. “Some people are in underserved communities where they have less access to healthcare. Imagine being in Wyoming where the nearest doctor is 50 miles away or being in the inner city where some people don’t have good access to healthcare. Our hope is that this can help bridge that gap.”

Efforts to make statins available without a prescription have been going on for more than 20 years. In the United States, the makers of pravastatin, lovastatin, and atorvastatin have all made over-the-counter (OTC) bids, with both Merck and Bristol Myers Squibb making their case to US Food and Drug Administration advisory committees in the hopes of persuading the agency that allowing patients to access medication OTC would lead to better cholesterol control. No effort to date has been successful.

To TCTMD, Nissen said that these prior attempts “failed miserably,” with no study yet demonstrating that patients could appropriately self-select statin therapy. The pravastatin and lovastatin programs were rejected by FDA committees because there were concerns about this aspect. The atorvastatin program didn’t even reach the advisory committee, abandoned because individuals failed to comply with LDL-testing requirements and demonstrated poor adherence to warnings in the drug’s label.  

“These are people who are eligible for a moderate-intensity statin,” said Nissen, referring to those who’d be candidates for nonprescription statin use based on this latest study. “Rosuvastatin 5 mg can typically get as much as a 40% reduction in LDL cholesterol. It’s very effective. A 30% to 40% reduction is going to have a major health benefit for these people.”

The study by Nissen et al was published September 6, 2021, in the Journal of the American College of Cardiology. 

Others Have Failed: Will This Be Different?

The new nonprescription statin effort—which was conceived and developed by AstraZeneca—includes a web-based application that uses the 2018 American College of Cardiology/American Heart Association (ACC/AHA) cholesterol treatment guidelines and the proposed drug facts label for rosuvastatin to assess eligibility for treatment. As part of the app, patients answer questions at home about their medical history, medication use, blood pressure, triglycerides, and levels of total, HDL, and LDL cholesterol. They are also asked questions, if it’s needed for the assessment, about waist circumference, high-sensitivity C-reactive protein (CRP) levels, and coronary artery calcium (CAC) score.

Patients are ineligible for the nonprescription program if their 10-year risk of atherosclerotic cardiovascular disease (ASCVD) is low (< 5%) or high (≥ 20%). They also are ineligible if they have a 10-year risk of ≥ 5% or < 7.5% without a risk-enhancing factor, such as elevated CRP or a high CAC score. Based on the assessment, there are three possible outcomes from the web-based program: rosuvastatin is “OK to use,” “not right for you,” or “ask a doctor.”

To test how effectively the technology could screen patients eligible for treatment when compared with a physician, Nissen and colleagues conducted the CREST trial, a 500-participant implementation study, which included 83 people with limited literacy. The mean age of those in CREST was 59 years (62.2% women). Roughly one-third of patients were Black.

Concordance between the self-selection of statin therapy and the doctor’s assessment was seen in 481 (96.2%), of whom 23 (4.6%) were appropriate candidates for nonprescription statin therapy and 91.6% were not appropriate. For those with limited literacy, concordance was 96.4%.

Of the three participants told the statin was “OK to use” but who were considered inappropriate candidates by the physician, two were not eligible for statins based on their family history. The other participant underestimated their 10-year risk of ASCVD. For the 14 cases where participants were told the treatment was “not right for you” but who were considered eligible by the doctor, discordance was attributed to participants inputting the wrong lab values or failing to recognize family history of heart disease, among other things.

The researchers point out that while the incorrect rejections are a lost opportunity, at least this poses no inherent risks from treatment. In the current study, no participants at increased risk for adverse events self-selected rosuvastatin.

“This was really different from prior efforts,” said Nissen. “We think it meets the highest FDA standards. The FDA is looking for 90% [accuracy] and we got 96%. It’s a major step towards making statins more widely available. We chose rosuvastatin 5 mg because it’s known to be safe, to be effective, and it has very few drug-drug interactions. That takes a certain amount of risk off the table. It was the right choice of the right drug at the right time.”

Motivation and Compliance

Despite the high rate of concordance between the web-based technology and physician decisions, there are concerns about making it easier for patients to access drugs without physician involvement.

In the same issue of JACC, Alan Jacobsen, MD (Johns Hopkins Hospital, Baltimore, MD), and colleagues published a research letter questioning whether aspirin should be available OTC. Senior author John McEvoy, MBBCh (National University of Ireland, Galway), discussing both the aspirin and statin papers with TCTMD, said the concept of nonprescription statin therapy has merit because there are a lot of people with untreated or undertreated LDL cholesterol. If OTC statin therapy led to greater access and uptake, he’d be all for it; he just isn’t sure technology is the answer.

“Access to healthcare providers isn’t the problem, in my opinion,” said McEvoy. “The problem is the patient’s motivation to take and comply with statin therapy. This app isn’t going to help with any of that.”

Still, he does think the technology could benefit those in lower- and middle-income countries who want to see a physician but can’t because of limited access. On the whole, he suspects nonprescription statin therapy would only play a niche role in higher-income countries, that being for motivated, health-conscious patients who can’t get to a doctor. He added that the new clinical prevention guidelines recommend treating patients to a specific LDL target based on their 10-year risk of ASCVD, and that some patients would require intensification of treatment. Nonprescription statin therapy doesn’t allow for that type of escalation, which would be a limitation to its implementation, he said. 

Understanding Risk and Benefit

For McEvoy, however, nonprescription statin therapy makes more sense than OTC aspirin in primary prevention, given that the former have a better risk-benefit profile than the latter—something the public may not fully appreciate.

In their survey of 300 primary-prevention patients in the US and Ireland, Jacobsen, McEvoy, and colleagues evaluated the use of low-dose aspirin and how the drug is sourced. Nearly all (96%) of the Irish patients obtained aspirin from a physician, while 84% of the American group bought aspirin OTC, yet patients in both countries were unable to quantify the relative benefits and risks of aspirin. Nearly half incorrectly believed that aspirin was indicated for primary prevention. Almost half of all US patients without CVD were taking aspirin compared with 26% of Irish primary-prevention patients.

“It’s very clear, and no surprise in terms of my own clinical experience, that patients often misunderstood their own CVD risk and often overestimate the benefits of aspirin and underestimate its bleeding harms,” McEvoy said.  

In an editorial, Neha Pagidipati, MD, MPH (Duke Clinical Research Institute, Durham, NC), and Eric Peterson, MD, MPH (University of Texas Southwestern Medical Center, Dallas), write that the two new studies raise the important question of whether “lay individuals” can make informed decisions about CVD preventive therapies without a doctor.

While there is a need for broader study, CREST suggests that a technology-assisted decision aid does translate into an accurate assessment for nonprescription statin therapy. Like shopping, investing, and doing taxes with self-guided, computer-aided programs, medicine may be going the same way, they suggest.

“To date, medicine has paternalistically controlled prescription-writing by clinicians as one of the last human holdouts against technology,” they write. “Yet, the cost of clinicians’ control over their patients’ medication access has been an overall failure to deliver high-quality, preventive care at scale.”

McEvoy, on the other hand, was more critical of the technology-assisted self-selection program, noting that it was funded and developed by industry. While the research group’s credentials are excellent, he would like to see the technology validated in an independent study.

However, McEvoy’s biggest criticism of the study was that clinicians in CREST used the same web application as the participants to input lab and blood pressure data to help make the decision about statin use. A more-rigorous assessment of concordance/discordance would compare the technology against decisions made by physicians in the clinic, he said. Given that they used the same app, and put in the same information, the high rate of concordance is not a surprise. Moreover, McEvoy noted that the study screened 1,500 individuals to include just 500 patients, which suggests a degree of selection bias.

“The people who responded to this online questionnaire were probably motivated and health-conscious, and I’m not sure that’s the market we need for getting patients access to statins,” said McEvoy. “The people who need access to statins are the ones who aren’t health literate, who aren’t motivated, who aren’t going to their doctor. I don’t see how this technology changes that, at least in Western countries where patients mostly have access to healthcare providers. I don’t see how this technology will change the main problem with statins where patients won’t take them, where they don’t like what they’ve heard about them, or where they don’t want to go to the doctor to get refills.”

‘It’s the Future’

To TCTMD, Nissen said that if AstraZeneca’s current nonprescription program goes forward, patients eligible for rosuvastatin 5 mg would not purchase the medication OTC at the pharmacy. Instead, it would be shipped to them, thereby eliminating the possibility that consumers have the potential to unnecessarily take the drug or if might cause them harm. 

“It’s the future,” said Nissen. “Self-empowerment of patients is something that’s been going on for a long time and this is just the next step in the development.”

In CREST, the majority of patients were ineligible for rosuvastatin 5 mg, which investigators say was expected given that they included a wide range of individuals, including those with cardiovascular disease. This type of implementation study is important, said Nissen, as it needs to show that the web-based app can prevent the treatment of ineligible patients, which was a failing of past attempts at making statins available without a prescription. The next step will include a study of patients eligible for statin therapy to show that the eligibility criteria are effective, he said.