Post by Mr. Jon Donnis on Jan 11, 2011 22:54:59 GMT
Medical science will benefit from the research of crowds
Monday was a great day for public health research. It was also a scary day for researchers. Scary because on Monday, with a minimum of fanfare, the paymasters of public health research put the scientists they fund at the frontline of the data-sharing revolution. We are a reluctant fighting force.
Chivvied along by the UK's biggest charity, the Wellcome Trust, science funders from across the industrial world issued a joint statement that essentially said they expect the data generated by studies they fund to be shared. It might not sound scary, but it could change the face of health research.
Look at what happened with genetics. In the early 1980s, geneticists worked away in their different labs, racing to sequence genes and patent them before the neighbouring lab could. The result: duplication, very slow progress and a huge bill. This infuriated the US National Institutes of Health and the Wellcome Trust, who between them footed much of that bill. So in 1986 they knocked some heads together, and decided they'd only fund geneticists who were willing to make their data available immediately. Nowadays, gene sequences get posted on the web daily and scientists build on one another's work. The pace of discovery has increased exponentially and, as a result, so have diagnostics and cures.
Now those funders, and another 15 including the Bill and Melinda Gates Foundation, the World Bank and the national research councils of the UK, France, Germany, Canada, Australia and New Zealand, want to do the same for any research they pay for that involves collecting data from people specifically for research (samples collected during provision of health services are excluded). They reason that taxpayers and charities deserve to get the maximum public health bang out of each research buck. That means allowing researchers to trawl though one another's data, combining the results to answer new questions. An early experiment combining data on malaria is showing how powerful sharing can be.
The statement is carefully worded to minimise legal obligations or financial commitments – and it doesn't actually say it will require researchers to share data. But I have been involved in the three years of discussions that led to this statement, and feel sure that the major players are committed to changing the culture of data-hoarding. In a commentary published in the Lancet, the heads of the Wellcome Trust and the US-based Hewlett Foundation thundered: "We call on researchers and funders to mend their ways."
As a sometime researcher myself, I'm ambivalent. I do believe that shared data will mean more and faster progress. It will also mean better quality data – a good thing, of course, until someone asks me about the six data records with no gonorrhoea tests (I spilled the urine samples over my shoes). The fear of exposing the dirty underside of our data – the mistakes that are inevitable but rarely admitted to – is a major reason for not wanting to share. There are also the visceral feelings of ownership that come with months spent trekking samples through the red light district at midnight. Worst of all is the thought that some data monkey with no grant applications to write or urine samples to spill might crunch "my" data and publish the results. Since published papers are virtually the only measure of success in my job, that's like giving away my future.
The joint statement deals with this by "aiming" to recognise data-sharing as a professional achievement. If the funders do this, and academia follows suit, they will remove the major barrier to sharing data. Then it will be up to researchers to hold the funders to the other commitments implied in the statement – support for data management, development of infrastructure, resources for curation of data. If funders put their money where their statement is, researchers will be more pleased than scared.
www.guardian.co.uk/commentisfree/2011/jan/11/medical-research-data-sharing
Monday was a great day for public health research. It was also a scary day for researchers. Scary because on Monday, with a minimum of fanfare, the paymasters of public health research put the scientists they fund at the frontline of the data-sharing revolution. We are a reluctant fighting force.
Chivvied along by the UK's biggest charity, the Wellcome Trust, science funders from across the industrial world issued a joint statement that essentially said they expect the data generated by studies they fund to be shared. It might not sound scary, but it could change the face of health research.
Look at what happened with genetics. In the early 1980s, geneticists worked away in their different labs, racing to sequence genes and patent them before the neighbouring lab could. The result: duplication, very slow progress and a huge bill. This infuriated the US National Institutes of Health and the Wellcome Trust, who between them footed much of that bill. So in 1986 they knocked some heads together, and decided they'd only fund geneticists who were willing to make their data available immediately. Nowadays, gene sequences get posted on the web daily and scientists build on one another's work. The pace of discovery has increased exponentially and, as a result, so have diagnostics and cures.
Now those funders, and another 15 including the Bill and Melinda Gates Foundation, the World Bank and the national research councils of the UK, France, Germany, Canada, Australia and New Zealand, want to do the same for any research they pay for that involves collecting data from people specifically for research (samples collected during provision of health services are excluded). They reason that taxpayers and charities deserve to get the maximum public health bang out of each research buck. That means allowing researchers to trawl though one another's data, combining the results to answer new questions. An early experiment combining data on malaria is showing how powerful sharing can be.
The statement is carefully worded to minimise legal obligations or financial commitments – and it doesn't actually say it will require researchers to share data. But I have been involved in the three years of discussions that led to this statement, and feel sure that the major players are committed to changing the culture of data-hoarding. In a commentary published in the Lancet, the heads of the Wellcome Trust and the US-based Hewlett Foundation thundered: "We call on researchers and funders to mend their ways."
As a sometime researcher myself, I'm ambivalent. I do believe that shared data will mean more and faster progress. It will also mean better quality data – a good thing, of course, until someone asks me about the six data records with no gonorrhoea tests (I spilled the urine samples over my shoes). The fear of exposing the dirty underside of our data – the mistakes that are inevitable but rarely admitted to – is a major reason for not wanting to share. There are also the visceral feelings of ownership that come with months spent trekking samples through the red light district at midnight. Worst of all is the thought that some data monkey with no grant applications to write or urine samples to spill might crunch "my" data and publish the results. Since published papers are virtually the only measure of success in my job, that's like giving away my future.
The joint statement deals with this by "aiming" to recognise data-sharing as a professional achievement. If the funders do this, and academia follows suit, they will remove the major barrier to sharing data. Then it will be up to researchers to hold the funders to the other commitments implied in the statement – support for data management, development of infrastructure, resources for curation of data. If funders put their money where their statement is, researchers will be more pleased than scared.
www.guardian.co.uk/commentisfree/2011/jan/11/medical-research-data-sharing