Some, perhaps many, previously unrecognized cases of atrial fibrillation (AF) will come to light in a screening program aimed at older asymptomatic adults. The key question is whether the challenges of such systematic but age-restricted AF screening in the community, with oral anticoagulation (OAC) offered to those found to have the arrhythmia, is worthwhile in preventing events such as death or stroke.
Now there is evidence supporting such a clinical benefit from a large, prospective, randomized trial. A screening program restricted to people 75 or 76 years of age in two Swedish communities, which called on them to use a handheld single-lead ECG system at home intermittently for 2 weeks, was followed by a slight drop in clinical events over about 7 years.
The 4% decline in risk (P = .045) in the STROKESTOP trial’s “intention-to-treat” (ITT) analysis yielded a number needed to treat of 91; that is, that many people had to be targeted by the screening program to prevent one primary-endpoint clinical event.
Those included ischemic stroke, systemic thromboembolism, hospitalization for severe bleeding, and death from any cause, investigators reported April 23 during the virtual European Heart Rhythm Association (EHRA) 2021 Congress.
If that benefit and its significance seem marginal, some secondary findings might be reassuring. Half the population of the target age in the two communities – 13,979 randomly selected people – were invited to join the trial and follow the screening protocol, comprising the ITT cohort. The other half, numbering 13,996, was not invited and served as control subjects.
However, only 51% of the ITT cohort accepted the invitation and participated in the trial; they represented the “as-treated” cohort, observed Emma Svennberg, MD, PhD, Karolinska Institute, Danderyd Hospital, Stockholm, who presented the analysis at the EHRA sessions.
The screening protocol identified untreated AF, whether previously known or unknown, in about 5% of the 7,165 as-treated screening participants; OAC was initiated in about three-fourths of those cases.
The as-treated group, on their own, benefited with a 24% drop in the prospectively defined secondary endpoint of ischemic stroke, compared with the entire control group.
The clinical benefit in the ITT population was “small but significant,” but over the same period in the as-treated cohort, there was a highly significant drop in risk for ischemic stroke, Dr. Svennberg said in an interview.
The trial’s lead message, she said, is that “screening for atrial fibrillation in an elderly population reduces the risk of death and ischemic stroke without increasing the risk of bleeding.”
Caveats: As-treated vs. ITT
But there are caveats that complicate interpretation of the trial and, Dr. Svennberg proposed, point to the importance of that interpretation of both the ITT and as-treated analyses.
“We detected significantly more atrial fibrillation in the group that was randomized to screening. A major strength of our study was that we referred all of those individuals for a structured follow-up within the study,” she said. “Although the focus of the follow-up was oral anticoagulant therapy, other risk factors were also assessed and managed, such as hypertension and diabetes.”
It’s possible that increased detection of AF followed by such structured management contributed to the observed benefit, Dr. Svennberg proposed.
However, the exclusion of those in the prespecified ITT population who declined to be screened or otherwise didn’t participate left an as-treated cohort that was healthier than the ITT population or the control group.
Indeed, the nonparticipating invitees were sicker, with significantly more diabetes, vascular disease, hypertension, and heart failure, and higher CHA2DS2VASc stroke risk scores than those who agreed to participate.
“We took a more difficult route in setting up this study, in that we identified all individuals aged 75 to 76 residing in our two regions and excluded no one,” Dr. Svennberg said in an interview. “That means even individuals with end-stage disease, severe dementia, bedridden in nursing homes, et cetera, were also randomized but perhaps not likely or eligible to participate.”
Therefore, some invitees were unable to join the study even as others might have declined “out of low interest” or other personal reasons, she said. “We believe that this mimics how a population-based screening program would be performed if done in our country.”
In the ITT analysis, screening successfully identified previously unknown or untreated cases of AF, which led to expanded OAC use and intensified risk-factor management, “which was key to a successful outcome.”
In the as-treated analysis, Dr. Svennberg said, “I think a combination of the intervention and the population being overall more healthy was driving the secondary endpoint.”