Researchers Retract Ambulatory BP Monitoring Paper in NEJM Due to Data Problems

A study that bolstered the value of ambulatory blood pressure monitoring (ABPM) was retracted by the investigators Wednesday due to “inaccuracies in the analytic database and data analyses.”

Published in April 2018 in the New England Journal of Medicine, the study showed that mortality was more strongly associated with readings obtained with ABPM than with those measured in the clinic. At the time, lead author José Banegas, MD (Universidad Autónoma de Madrid, Spain), told TCTMD that “our study is the largest worldwide and provides unequivocal evidence that ABPM is superior to clinic pressure at predicting total and cardiovascular mortality.”

But earlier this week, the researchers sent a one-line letter to the editor saying they wished to retract the article.

Senior author Bryan Williams, MD (University College London, England), elaborated on the reasons for the retraction in an email to TCTMD. He said the original database was created and managed by Banegas and that Banegas, along with a statistician from the Universidad Autónoma de Madrid, performed the statistical analyses.

“Responding to an enquiry about the results to the journal, we discovered an error in the original statistical analysis and this prompted some of us, working with an independent statistical team, to focus attention on the original database used for the study,” said Williams. “Working with this independent statistical team, I along with two of my author colleagues conducted an investigation into the original database linking ABPM data with mortality and the subsequent statistical analysis for the original paper. We found errors in the original database and compared the original database with a new database independently created from the source data.”

They became concerned about the inconsistencies between the original database and the new database and asked the Universidad Autónoma de Madrid to investigate and provide an explanation.

The other study authors are not involved in the ongoing process, Williams said, adding that any explanation that comes out of the investigation will be made available. “In the meantime, working with the independent statistical team, we are now undertaking a new analysis of the data and plan to publish these results as soon as they are available as the questions this study set out to address remain important,” he said.

Asked to comment on the retraction, Jan Staessen, MD, PhD (University of Leuven, Belgium), said in an email, “We knew the very day the Banegas paper was published that it was a nonevent and that the statistical computations were not performed professionally.” Staessen and two of his colleagues raised some of their concerns in a letter to the editor published in NEJM in September 2018.

“Our suspicion was confirmed when the Spanish investigators kindly provided us a subset of their data with the intention to replicate findings in IDACO [International Database on Ambulatory Blood Pressure in Relation to Cardiovascular Outcomes],” Staessen said. “We could not reproduce what they had published. We thanked the Spanish investigators and also explained why we could not use their data.”

The situation with the Banegas paper, however, should not take away from the importance of ABPM, Staessen indicated. “ABPM remains the state-of-the-art method to diagnose and manage hypertension,” he said, pointing to a study his group published in August 2019.

Suzanne Oparil, MD (University of Alabama at Birmingham), also said the retraction should not affect how ABPM is viewed. “There is abundant evidence supporting the importance of out-of-office BP measurement,” she told TCTMD in an email.