The Survey Says: BP Measurement In SPRINT Was All Over The Place

–BP measurement question limits the trial’s ability to inform clinical practice.

A long-promised report from the SPRINT investigators offers important new information about how blood pressure was monitored in the trial. The report is unlikely to satisfy critics or resolve the larger controversy of how the trial should be interpreted.

Drs. Johnson and Bangalore

At the American Heart Association meeting in Anaheim Karen Johnson (University of Tennessee), Vice Chair of the SPRINT Steering Committee, presented for the first time the most detailed information yet about one of the major sources of contention about the trial, which involves the methods used at the trial sites to obtain blood pressure measurements. The key finding of the new report is that although blood pressure measurements were obtain using an automated measurement device there were substantial differences in the methods used by the different SPRINT centers. The finding confirms revelations about about these differences that first emerged in 2016.

The blood pressure in SPRINT was defined as the average of 3 measurements taken with an automated measurement device. But until now it has not been entirely clear whether these measurements were obtained in the presence or absence of medical staff. Previous research suggests that unattended blood pressure is significantly lower than attended blood pressure.

Johnson presented the results of a survey that was conducted after the trial was completed. Her data showed that patients were completely alone during the measurement period at 43% of the sites and were alone for part of the time at 29% of the sites. According to Dr. Johnson there were a number of baseline differences between the groups, but, she emphasized, the achieved BP levels were similar in both treatment groups no matter the BP measurement category. But, she said “there was a significant heterogeneity of treatment effect for the primary outcome of the study.”

Sripal Bangalore (NYU) was the discussant for the AHA presentation. He expressed concern about the reliability of the classification scheme used in the SPRINT survey. He pointed out that the methodology of the survey did not appear to take into account that sites may well have had significant internal variation in their BP measurement techniques.

Bangalore summarized his perspective in an email: “To put it simply- the majority of sites in SPRINT measured BP with the patient being alone at some part of the BP measuring process. Clinical practices should take this into consideration if they are to adopt a SPRINT-like strategy. Using the same BP target is absurd if your BP measurement technique is different since studies have clearly shown a difference in BP readings with different techniques.”

“What a messy situation!” said Sverre Kjeldsen (Oslo, Norway), an early critic of SPRINT. He said the study helped inform the new US hypertension guideline but he expressed satisfaction that the guideline did not adopt the lower 120 mm Hg target supported by the SPRINT investigators.

“The plot thickens!” commented Bryan Williams (University College London), chairman of the ESC Council on Hypertension. “There has been intense interest in the BP measurement technique in SPRINT because the interpretation of the absolute BP values has a bearing on decisions by guideline developers with regard to optimal BP targets in treated patients. The first thing that this data shows is that the protocol for BP measurement was inconsistently followed during the trial which would be regarded as a protocol violation because the measurement of BP was so fundamental to the trial procedures in terms of treatment escalation in the intensively treated patients.”

However, Williams said, “I am confident we should be recommending a lower systolic BP target than <140mmHg for most patients.”

Previous SPRINT Coverage:



  1. My little corner of science isn’t half as glitzy as this exciting field of medicine. But I do think we’d tend to be careful to use internally consistent measurement methods.

Speak Your Mind