Principal Investigator Defends SPRINT Against Critics

–Blood pressure experts disagree about the NIH ‘Landmark’ Trial.

Since the first announcement of its preliminary results the SPRINT trial has been the focus of intense controversy. One major focus of contention revolves around the precise technique used to measure blood pressure in the trial. As in all recent large outcome trials in hypertension, blood pressure was obtained from multiple measurements taken in the office with the use of an automatic blood pressure machine. In earlier trials the measurements were obtained with a healthcare professional in the room. The SPRINT investigators believed that the presence of a healthcare professional led to higher measurements, so they specified that the healthcare professional should leave the room while the measurements were being obtained.

However, at the ESC meeting last August SPRINT’s critics, many of them Europeans, contended that BP measurement in SPRINT had been inconsistent, with some centers taking unattended measurements and others attended measurements. These critics maintained that this issue cast further doubt on the reliability and usefulness of the trial. At the time, and as far as I am aware, until now, the SPRINT investigators had not responded to this question.

At the ACC meeting last month SPRINT principal investigator William Cushman (University of Tennessee) revealed during a debate that the SPRINT investigators would publish a paper digging into the topic of BP measurement in SPRINT. I subsequently wrote Cushman and asked him to respond to some of the charges that have been levelled at the trial.

Cushman explained why he earlier had been unable to respond to the criticism regarding blood pressure measurement: “At the time of ESC last year, the Steering Committee was in the midst of discussing whether and what might be done to further assess BP measurement. As you know we are now developing a paper and cannot discuss anything further about it since that could jeopardize acceptance.”

Although unable to shed more light on the specific questions about what happened in SPRINT, Cushman emphasized that the measurement technique in SPRINT was largely similar to the technique used in other trials:

“In SPRINT we used the standard BP measurement technique, as in all hypertension outcome trials and as recommended by all hypertension guidelines that I am aware of, and when SPRINT was designed there was very little discussion/awareness in the literature about observed vs unobserved with an automated manometer. It was a moot issue in most prior major hypertension trials that used mercury manometers. The most authoritative BP measurement guidelines at the time SPRINT was designed were the 2005 AHA guidelines and there was no mentioned of observed vs unobserved. There was no concern about BP measurement in SPRINT prior to our publishing the primary results. We still don’t believe it changes the interpretation that a systolic BP goal of <120 saves lives and reduces cardiovascular events. My suspicion is at the lower systolic BP levels achieved in the SPRINT intensive group, there may be little difference between observed and unobserved BPs when the same careful technique is used. But that doesn’t matter if either method leads to benefit.”

Cushman argued that the fact that blood pressure is not measured accurately in clinical practice should not invalidate the better method used in SPRINT. In fact, he urged practitioners to upgrade their blood pressure measurement techniques. “This methodology is not mysterious or difficult and it is unfortunate that it is not followed in most practice settings … but it can and should be. Labs would be shut down if they used similar sloppy technique to measure cholesterol, glucose or creatinine.”

Cushman predicted that this issue will be addressed in upcoming guidelines: “I believe new guidelines will emphasize proper measurement technique even more going forward. The answer is not continuing with the status quo because one cannot predict with accuracy what the BP would be if taken properly by taking it incorrectly and assuming some fudge factor – improper technique doesn’t just introduce a fixed higher BP level, it injects unpredictable variability. The SPRINT investigators will continue to emphasize the importance of proper BP measurement technique – we hope our subsequent publications will help draw further attention to this. I can’t answer for the entire hypertension community, whoever that is, but I assume guidelines and organizations will place renewed emphasis on the importance of proper BP measurement technique — SPRINT can help in that discussion. I certainly try to emphasize the importance of proper BP measurement technique in every presentation I make.”

SPRINT Critics Still Unsatisfied

But the debate over SPRINT is unlikely to be resolved any time soon. Several SPRINT critics expressed continued disagreement with the trial.

Franz H. Messerli (Mount Sinai) said that “the fact that it [attended versus unattended measurements] was a a moot issue in most prior major hypertension trials does not exculpate the authors of SPRINT since they used for the first time unobserved automated BP measurements.” Messerli cast doubt on the ability of clinicians to apply the SPRINT findings to their own practices:

“My suspicion is at the lower systolic BP levels achieved in the SPRINT intensive group, there may be little difference between observed and unobserved BPs when the same careful technique is used. This may be a valid point, most likely the lower the BP the less the difference.” But, said Messerli, the difference does matter “for the practicing physician who has neither time, nor space nor manpower to measure BP similarly as it was done in SPRINT.”

Sverre Kjeldsen (Oslo, Norway) said that his skepticism regarding SPRINT “is unchanged.” Going beyond the BP measurement issue he said “there is doubt whether there are any true reductions in endpoints.” He pointed to the small difference in the number of primary endpoint events between the two groups. He said that half of the difference of 78 events were heart failure “which could be expected because of the up and down titration of diuretics.” The difference in mortality of 55 patients, 28 for CV death and 27 for non CV death, “indicates an unspecific effect of generally better follow-up of the intensive arm.” Further, he asked, “what about the 245 patients who were lost to follow up? Maybe they should count as dead? Or the quite high number of patients with renal failure in the intensive arm?”

Kjeldsen said he “cannot see SPRINT as a ‘landmark’ trial but rather as a study aborted prematurely on loose ground. And not well designed.”

Previous SPRINT Coverage:



  1. What a waste of time and resources!
    I would address all concerned to this thoughtful examination the nature of research:

Speak Your Mind