The NHS is institutionally attuned to the big things – what’s coming down from the big beast in Whitehall or the latest missive from the Chief Executive. This stacks up into institutional agendas that focus on things like Should we be shutting the casualty unit? Do we need another Consultant surgeon? How are we going to implement the Darzi report?

And when that isn’t occupying the operational brain space then it’s all the middle range stuff – how do we reduce the number of expensive bank nurses? When is the next data return due? How do we implement the latest policy on C difficile?

But the experience of the patient is made up of the micro – Was I washed gently? Did I feel included in my care? Was the place clean? Was it too noisy to sleep at night? Such things are visible to professionals who hopefully care a lot about getting them right, but they are more or less invisible to the institution except when they result in a complaint.

Patient Opinion takes thousands of these comments about micro aspects of care and makes them more visible. But what happens then? Well often not a lot. Being focused on the strategic or the middle-range must-do’s means that busy staff all too often see patient feedback as something that just highlights irritating blemishes. At best its something to add to the already over-long To Do list. At worst something that can be easily ignored.

So how can Patient Opinion turn blemishes into blossoms? How can patient comments help set a thousand flowers blooming? Well the answer here is more than the technology. The latest guidance about how government should use Web 2.0 says that people should be prepared to respond quickly, listen and act. Which as Dan Herman points out on wikinomics blogs is great advice.

But our experience on Patient Opinion is that the only people who aren’t caught in the cross fire of conflicting incentives here are the patients. So the trick is using their energy, enthusiasm, and ideas to drive micro change hundreds of times across the NHS. One early example is allowing people to comment on how the trust has responded – see what one less than pleased punter thinks of the hospital's response to their posting.

No one thought that thousands of people would step up to the mark and help create Wikipedia. Or that there were millions of people just waiting for an on-line auction site. The trick is to find ways to galvanise people’s real enthusiasm for the NHS into a process that involves staff and causes thousands or service improvements to blossom. A good analogy perhaps as creating this culture is probably more like gardening than meeting policy targets or implementing Lord Darzi’s report. Slow, gentle persistent work to get the conditions right. And resisting the temptation to pull everything up by the roots once a day to see how they are doing.

Does it matter who runs web-based public feedback sites? There’s been some pretty heated discussion in the past with lots of people saying that citizens will only trust sites if they are independent of government (and we’ve certainly added our share of heat to support this argument!).  Meanwhile the government and people in the health service often think that Joe Public just doesn’t care and the NHS is a trusted brand that people will be happy to give feedback to. 

Up to now this argument has been almost exclusively assertion-based on both sides because web-based feedback at scale is all so new that there has been little hard evidence either way. But now that NHS Choices has been collecting feedback for more than a year we’re beginning to get some results from this natural experiment in citizen democracy and feedback.

The data so far is interesting. In terms of the numbers of postings both sites generate about equal volumes of stories across England.  Thanks to a Freedom of Information request made by someone on MySociety’s fabulous FOI site we now know that:

·        NHS Choices reject around 24% of all stories submitted compared with a 5% rejection rate for Patient Opinion.

·        Comparable hospitals respond more frequently to stories on Patient Opinion than on Choices. Comparing like with like across both sites  42% of all postings on NHS Choices generate a response from the hospital compared with about 65% on Patient Opinion.

·        Length of story and response are about equal on both sites although stories in general are a bit more positive on Patient Opinion. 

For Patient Opinion we also know that stories are reasonably balanced across all classes: the number of stories from the less affluent 50% of postcodes more or less equals the number from the most affluent. Comparable data is not available for NHS Choices.   So no firm conclusions yet but some interesting straws in the wind.

And of course we’re hardly unbiased in all this. What is needed is a quick study of both sites using full access to internal data by someone nice and independent. So we’ve suggested to the Central Office of Information in Whitehall that this might be a really interesting thing to do before we spend lots more on huge government sites like NHS Choices or – conversely – perhaps waste (rather less) money on independent organisations like us. And happily they seem quite interested. We’ll keep you posted…. 

There's a lot of talk around about using technology to improve healthcare - making more information available, giving patients a voice, helping them manage their care by making patient records accessible online, and it's thowing up some real key issues.  Empowering patients to have more knowledge about their conditions, more information about the services they receive, to have more choice, appears to a step in the right direction, but what about the implications?  A lot of the issues are around power and who is in control - the public has access to more information than ever before, and this all works against the paternalistic model of care.

We are starting to see a real power struggle, between those who have traditionally controlled healthcare (clinicians, GPs) and those receiving it.  For a number of years there's been a recognition that expert patients, those with a good understanding of their condition, visit their GP less often - in 2005 the Department of Health launched the Expert Patient Initiative.  At the time only 21% of doctors were in favour of this approach to supporting those with long term conditions like diabetes, arthritis and Parkinson's, there's still a considerable number out there who think patients "do not have wisdom but follow prejudice, hearsay, and urban mythology" (Ray Jones).

Patients are being encouraged by the government to take control of their care, the internet has supported a new phenomenon, patient opinion leaders, and expert patients are becoming more and more common, as this post in HSJ indicates "People with chronic conditions are sharing their stories with each other, not just for emotional support, but for the clinical knowledge they gain in an online community." (Jane Sarasohn-Kahn of the think tank THINK-Health and author of a report on health 2.0).

This power struggle is becoming a major issue in the US.  According to Dr. Robert Lamberts, an internal medicine physician and medical blogger in America “Doctors used to be the only source for information on medical problems and what to do, but now our knowledge is demystified.”  “When patients come in with preconceived ideas about what we should do, they do get perturbed at us for not listening. I do my best to explain why I do what I do, but some people are not satisfied until we do what they want.”

Sharing information is driving the shift in power.  Patients have a voice and have access to others stories about their care, but is anyone really listening to them?  Here's the real issue - technology is just an enabler, it supports change and improvement and innovation, but it can't make it happen.  Without the ability for people to make it happen, it won't.  Throwing money at technology is not the answer.  Technology can only go so far, it needs a real, adult conversation between patient and doctor and the ability of staff to make changes without being stuck in a bureaucratic loop.  Patients are changing and starting to take control, those working in the NHS need, in my humble opinion, to release some.

Now that the Department of Health has decided it will be spending £60-80m over three years to develop the NHS Choices web site, it's worth having a rethink about how our cash might best be spent.

In both the UK and the US, new thinking is rapidly emerging on how governments should be on the Internet. Some are arguing, to put it bluntly, that governments shouldn't be building web sites at all. Instead, the first priority of government should be to make its data available on net in ways which are open, standards-compliant, and re-usable by third parties - whether they be commercial or third sector organisations - because others will innovate around the data far faster and more freely than government ever can.

And if governments do find a need to build their own web sites, they should do so using these same data services that they have exposed for public use (providing a nice incentive to make sure they work).

Interestingly, this view seems to be striking a chord with our own Power of Information taskforce, which has recently succeeded in opening up more public sector information for public reuse (great work!).

Not wanting to be left out, and in an enlightened spirit of public service, Patient Opinion has itself "mashed up" the public feedback which government publishes on NHS Choices, with public feedback submitted through our own site. Why? To make it easier both for patients and staff to access all the feedback about their local services, and to increase the chance that public feedback will generate real improvements in services.

So, coming back to that £60-80m about to be spent on "a web site"... wouldn't it be interesting to discuss the possibility that, first, they build a set of data services for public use. (This thinking is already underway in government.) And only after the data has been shared, build the government site (on top of the data services) to present it. Or, at that point, save the taxpayer some money and just link to sites which are already doing a good job with the same data.