When a prestigious medical journal challenged scientists to analyse data from a pivotal blood-pressure study in search of new findings, hundreds of researchers around the world rushed to sign up.
The contest, sponsored by the New England Journal of Medicine, offered scientists a rare opportunity to access detailed trial data that otherwise might have remained proprietary for another year — if not indefinitely. But the competition, whose winners were announced on 7 March, also illustrates the tension between speeding access to data and protecting the interests of those who laboured to collect them.
Jackson Wright, a pharmacologist at Case Western Reserve University in Cleveland, Ohio, spent nearly a decade designing and carrying out the blood-pressure trial featured in the data challenge. The landmark study, funded largely by the US National Institutes of Health (NIH), revealed a surprising benefit of lowering blood pressure below the usual targets — and had the potential to reshape the treatment of millions of patients.
Wright’s team, which grew to include researchers at 102 institutions, published preliminary results in the New England Journal of Medicine in late 2015. In November 2016, the NIH and the journal made the data from the trial available for the data-challenge competition; Wright says the researchers were not given a vote in the decision because they conducted the trial under a contract with the NIH.
Now one-third of the 60 papers that Wright’s team had planned to publish are in jeopardy of being scooped. “I think the incentives to do these trials will be dramatically lessened if this is going to be the expectation going forward,” he says. “It’s a huge time commitment.”
But others favour making data from trials publicly available as soon as possible. Doing so, they argue, opens up the possibility of a wide range of additional analysis, and speeds up analyses that can yield important clinical insights. “Clinical trial data are quite valuable, but usually they’re kept locked away,” says Sandosh Padmanabhan, a participant in the competition who researches cardiovascular genomics at the University of Glasgow, UK. “Everybody who does clinical trials needs to open up their data for everybody to use.”
The Systolic Blood Pressure Intervention Trial (SPRINT) studied 9,361 people who had elevated blood pressure and an increased risk of cardiovascular disease.
The goal was to find out whether there was any benefit — or harm — of choosing a systolic blood-pressure goal lower than 140 mm Hg. About half of the participants received standard therapy pegged to the 140 mm Hg limit; the other half were treated more intensively, with a goal of forcing their blood pressure below 120 mm Hg.
SPRINT, which began enrolling patients in 2010, was halted in August 2015 when an interim analysis showed that patients receiving intensive therapy were 43% less likely to die from cardiovascular causes such as heart attack or stroke than those on the standard regimen. Three months later, SPRINT investigators published their results in the New England Journal of Medicine1.
Although SPRINT’s blood-pressure intervention had been halted, the team continued to collect data until July 2016, and is still validating those data before carrying out final analyses.
The SPRINT investigators expected to have two years after the final data were collected to conduct those analyses, says Wright. Instead, the NIH and the journal made the data available for its competition in November 2016.
Wright worries that hundreds of researchers are now picking through the data while the SPRINT investigators are still busy closing down the trial. “Others who had nothing to do with the trial are able to publish a lot faster than we are,” he says. “The return on investment is dramatically reduced for the investigators in SPRINT, no question.”
The team that won the data competition was led by Noa Dagan, chief data officer at Clalit Research Institute in Tel Aviv, Israel. The researchers used the data to develop a program to help clinicians determine the risk of intensive blood-pressure treatment to patients, on the basis of their age, gender and other parameters.
In second place was a team of medical students at Boston University in Massachusetts that looked at the impact of blood-pressure treatment on patients with chronic kidney disease — a particular concern because intensive therapy has been thought to place stress on the kidneys.
The 143 participants who advanced to the final stages of the data challenge embarked on a wide range of analyses. Padmanabhan tried to determine whether patients who combined more than four blood-pressure medicines were at higher risk of dying, and found a hint that they might be. He says that clinicians have written him to say that the analysis has influenced how they approach treating their patients.
All of the analyses submitted as part of the challenge should be considered preliminary, says New England Journal of Medicine editor-in-chief Jeffrey Drazen. The studies generate new hypotheses to test, he says, but are working with the earlier, incomplete data set. As a result, the SPRINT team’s analyses, which will be based on the final data, are not entirely at risk, he adds.
“I’m very grateful to the SPRINT team,” says Drazen. “A lot of interesting ideas have come from this.”