IT for Primary Care

At the end of September 2010, eHealth Insider hosted a roundtable on the future of IT in primary care, sponsored by ISOFT.

The debate took place just a few weeks after the launch of the white paper, ‘Equity and excellence: Liberating the NHS’, but before the publication of the information strategy that will support it. So there was much discussion of the many details of future policy that are still unknown.

Despite this, the participants agreed that information for both clinical decision making and commissioning will become ever more important - and that more information will need to be shared with other providers and with patients as well.

They also agreed that online services and telehealth will become more widely used and influential. And that future IT developments will need to be moulded around the needs of clinicians and the patients they serve.

Known unknowns

Dr Glyn Hayes opened the debate as chair by arguing that the agenda for primary care IT is linked inextricably with the white paper, ‘Equity and excellence: Liberating the NHS’, with its emphasis on improving patient outcomes and GP-led commissioning.

The aim of the debate, Dr Hayes said, was to come up with some conclusions about how commissioning will impact on primary care services.

However, he added: “One problem is that we do not know what the future landscape will look like. The information strategy has not yet come out and we do not understand how commissioning groups will work.”

Ian Herbert, vice-chair of the British Computer Society’s health informatics forum, outlined the challenges as he saw them; and, again, his thoughts were very much dominated by what we do not know.

Among them were what he called the “mechanics of referral” – how to refer, how to receive proof of the care provided, and paying for it. “This will be a huge increase in traffic compared to Payment by Results,” he predicted.

The notions of commissioning services based on patient needs and of providing choice were other areas in which Herbert identified big gaps in knowledge. “The white paper wants to see GPs commissioning care pathways. We do not know how to deal with this,” he said. “Some cases are simple – for example an acute episode of care - but others are much more complex.

“For example, a stroke patient may require a dozen care providers along a pathway. This is new. How do you do it? How do you manage a care pathway? How do you attribute outcomes to different providers?”

How, indeed, do you define and measure outcomes? And are GPs skilled and motivated to do this work? John Lockley, a GP at Oliver Street Surgery in Ampthill, Bedfordshire, and chair of the ISOFT primary care user group, said most GPs will not be interested in taking part personally in commissioning – but a small percentage would.

GPs, he said, were accustomed to managing a business and appointing people to manage on their behalf. “There is a group of committed GPs who will drive management,” he said.

Losing expertise?

The next big unknown is the extent to which expertise now based in primary care trusts will remain in the NHS. Certainly, this expertise will be needed to support GPs, the table agreed. But Jon Hoeksma, EHI’s editor, highlighted the despondency felt by PCT staff, according to a survey of strategic health authority and PCT staff conducted by EHI Primary Care.

“We had 400 responses and the point that came across most strongly was the huge uncertainty and the real risk of losing hard won skills, particularly around IT management and IT governance. We risk seeing clinical and IT skills being lost at exactly the time we need them most,” he argued.

Perhaps, said Dr Lockley. “Quite a proportion of managers in the NHS are not as good as we would like them to be,” he said. He referred to a survey by GP Magazine, in which 85% of GPs rated their PCTs as fair or poor - and only 15% rated them as good or excellent.

Nick Harte, ISOFT solutions director, said the white paper represented a major opportunity for GPs to shape services and dramatically impact the health of their populations. “The question for interested GPs is how to take advantage of the opportunity this offers to really start managing the population’s health status.” There was, he said, an opportunity to move from reactive healthcare provision to predictive and preventive delivery on a much greater scale.

Herbert agreed that this would motivate GPs. “I did some work with a practice based commissioning group in the South East of England and asked GPs about their commissioning ideas,” he said. “For the most part they said they wanted to commission services that are social.”

The problem of data sharing

The challenge in this scenario would be enabling data sharing, suggested Ian Manovel, associate director of analytics at Bupa Health Dialog.

“Currently, there is an over reliance on referring from primary to the acute hospital,” he said. “It should be possible with QIPP [the Quality, Innovation, Productivity and Prevention initiative] to establish a far more competitive tariff for providing high quality care in the community.”

But this would require data sharing between primary and community care. And Manovel suggested that there were some “very real challenges in the regulatory environment that need to be tackled to make this happen.”

He said: “This is very challenging for the owners of data and the controllers of data. There is no forward thinking regulatory authority to say that transfer of data and knowledge should be facilitated. We need to change the way that data is used and communicated.”

Herbert agreed. One perceived problem was protecting patient identities, especially when sharing large longitudinal records such as exist for people with long-term conditions. “This is very difficult to get away from because you need information about post codes and so on to do your work, and this risks identifying people.” But, he added: “There is sufficient privacy enhancing technology now to enable people to do this safely.”

The issue of data collection was dear to Dr Lockley. Too often, he told the debate, he had been asked by a PCT to provide data that was clinically meaningless and difficult or time consuming to extract from clinical systems. GP-led commissioning promised an end to this nonsense, he hoped. “We will be in a position to decide which data to collect and cleanse,” he said.

And of measuring outcomes

This brought the discussion back to outcomes. Mark Thornton, ISOFT’s director of operations, said: “People tend to manage best on information they can easily collect. There is a danger that because something is easy to measure it becomes an important driver for the system. We need to go back and ask: ‘what do we need to measure’?

“The other danger is that we collect too much data. We need to collect data that helps us to improve outcomes.”

Glen Griffiths, vice chair of the BCS Primary Health Care Specialist Group, said that too much data collection had been driven by targets that were not always clinically meaningful. “That’s why we have to move away from a target driven culture,” he said. “For me the acid test is whether you can justify a target to a clinician who then might be quite happy to provide the dataset.”

Dr Lockley said that as a GP he was concerned with the patient in front of him. “To my mind, the most important thing is whether I can get the information about the patient that has already been acquired. Can I get it quickly and in a form that can be assimilated, so that I can make a decision about patient care?”

One example might be hospital lab results out of hours. “It’s there but you cannot get it because there are barriers. The smooth flow of information is very important, but it has to be accurate information,” he said.

Putting on a commissioning hat, Dr Lockley made similar points. “As a commissioner, if I do not have information at my fingertips or I do now know how accurate it is, I cannot make an accurate diagnosis in economic terms. I need reliable clinical and economic data.”

Harte concluded: “GPs do not have the information to manage populations today. You need demographic information, utilisation and diagnostic. You need more information than you are currently getting and of a different type.”

How will this play out in future, Dr Hayes asked. Again, the table was back to the unknowns. Hoeksma pointed out that the information strategy was due, while the NHS Information Centre is expected to have some role in commissioning, including developing a market in analytical services.

This notion of intermediaries analysing their data could have quite important consequences, he said. “This is a very different world to one where someone installs their own version of software themselves. It is a dramatic acceleration.”

Taking charge of GP IT

This led again to another unknown: who will have responsibility for GP IT systems in future? Will this be local? Dr Hayes said: “This is important because it will be a cost. Will consortia be sophisticated enough to say it wants a major investment in systems rather than IT?”

Dr Lockley was comfortable with the current state of flux. “I think structures will evolve,” he said. “What we need to avoid right now is ossifying processes and structures but instead to allow them to be dynamic.”

Dr Hayes then moved on to outcomes. He stated his case quite clearly: the coalition government had demanded an outcome driven regime – but had no idea what it meant by this.

Patient reported outcome measures, PROMs, worked quite well for conditions or interventions where there was a start point, a treatment and an end point, for example hip replacement. He asked: “What I am interested in is whether this kind of approach could ever work in general practice?” And do cards with smiley faces to rate patient experience really add anything?

Griffiths said the focus on outcomes had to be aspirational, pointing out the varying rates of cancer survival around Europe – and Britain’s poor ranking. “We want to be the envy of the world,” he said. “It is important to have objective assessments but there are subjective assessments that can contribute to taking the right approach.”

This was a trap, said Dr Lockley, who used the example of a heroin addict. “Deprive him of heroin and he will feel lousy about the service provided. Give him heroin and he will say the service was wonderful even though you have done the wrong thing.”

Simon Gill, solution architect at ISOFT, said there were lots of things that could be measured that would make a “phenomenal difference” to care but currently do not happen. For example, having a clinician who knows your history, or not having to repeat your name and address to different care providers.

There were other tools for measuring, for example websites where patients can give opinions about doctors. But Hoeksma warned of a new practice in the US, in which professional data cleansers are challenging negative comments and cleaning up doctors’ online reputations – for a fee.

Linked to patient outcomes is the issue of measuring quality. Manovel suggested capitalising on GPs’ competitive nature. “We measure effective care measures for patients with long term conditions,” he said. “These are all objective measures from GP records. It is anonymised and fed back to GPs so that they can compare their performance. Being competitive, they quite like it.”

Thornton added that clinical outcomes would not be the only ones. “We will need a collection of different measures including value for money, service quality and activity. These numerical ones will be the most important to the manager looking after the delivery of the service but for clinicians they may be less meaningful.”

More issues in data sharing

The next theme for debate was sharing information. Technically it is possible to move information around the NHS, said Dr Hayes. The question is whether it is desirable or useful and what it might be used for once it has been divorced from the original purpose for which it was collected.

Manovel addressed the desirability issue. It all depends who you ask. “Ask a relatively young, healthy person of they want someone else to have access to their personal information and they say ‘no’,” he said.

“But ask someone with a long term condition who has been in and out of hospital and has to repeat their medical history 100 times and they will say ‘yes’. If it means that I or a person I care for will get better quality of care or outcomes, I am happy for it to be shared.”

Dr Lockley had no objections to data sharing in principle but made a plea that data should only be used for the purpose for which it was collected.

“I am happy for people to know that I measured a patient’s blood pressure as x on day y if it is for the benefit of the patient – whom I might immediately arrange to be admitted to hospital with a coronary. What I am not happy about is someone asking why I did not measure it the next day (when actually the patient was still in hospital and out of my direct care) and making value judgements about my practice based on that ‘missing’ data.”

Griffiths said there was a case to be made for providing patients access to their own records. “I have had experience of this with hundreds of patients. It is empowering and it is safe,” he said. “It is a complex area but does provide, for example, an opportunity for a market in second opinion.”

The jury was out on whether patient access offered any economic benefits, he said. But there was good anecdotal evidence of the clinical benefits of patients who knew more about their health. “Patients are our greatest untapped resource,” he said.

Adrian Stevens, ISOFT’s managing director, turned to patient held records. “One of the big drivers in cost and quality is the use of personal records and home monitoring devices,” he said. “It’s a subject that always comes up with diabetes.

“Take a rural area: home monitoring is a perfect solution for managing that patient. But how? If you gave a device to one of my relatives who also has senile dementia you would have no idea what was coming back in terms of results. We talk about cost – but at what cost?”

Our current model was based on available resources – for the most part, we have enough GPs and we can make our way to see them face-to-face so they hold the records. But in China, he said, where distances between populations are vast and doctors few, patients consult with their GP via television over a Skybox.

In Malaysia, added Griffiths, patients who live so remotely that they only see a doctor when he/she flies in, patients hold their own records and hold them dearly.

The white paper puts a big emphasis on patients controlling how and with whom their data is shared and wants to see patients holding their own health records.

This was all very well, said Dr Lockley, but in current practice, where doctors do not expect their records to be seen by patients, they will use them to record suspicions and thoughts – for example tentative thoughts about cancer or queries about possibly abusive relationships. Could this be continued in a system where patients hold their records?

Yes, was the answer from the technologists around the table. “Most GP systems do provide sealed envelopes,” said Herbert. “Synergy can do it down to individual level.”

Hoeksma pointed out that other providers are developing ways of giving patients a view of their records that made them understandable and accessible, for example EMIS and TPP. “It is almost as though what NHS Connecting for Health intended has become a reality through provider response,” he said. “It is almost as though ISOFT is the one out of step.”

Not so, said Simon Gill. “We are talking to EMIS about this very thing. We have chosen not to make an announcement before there is something concrete. As an organisation, interoperability is core to our business.”

Looking ahead ten years

Finally Dr Hayes moved the debate on to the future of primary care services. What will they look like in ten years time, he asked.

Herbert predicted consultations via electronic chat and video. Already the US healthcare giant Kaiser Permanente suggests that without electronic chat, its services would be unsustainable.

Manovel saw the further roll out of telehealth, with home monitoring and telephone coaching for people with long-term conditions a commonplace.

Dr Lockley reminded the table of GPs’ role in referring to secondary care, a role that requires them to have a generalist’s knowledge at a clinical level, and also knowledge about care pathways.

“In another ten years we may well be moving towards artificial intelligence systems where we can put in symptoms and get a diagnosis,” he said. “That will mean GPs having access to a whole lot of information.”

But if the NHS does fragment then there will be a role in primary care for the service broker, helping patients to navigate their way through the system.

Griffiths, meanwhile, imagined intelligent booking services that would challenge patients about why they needed an appointment and whether there were other more appropriate services or actions they could take.

Dr Hayes wrapped up. We do not know what commissioning will look like, he said. Measuring patient outcomes is complex. Sharing information – including with patients – will be necessary and that will mean overcoming some technical issues as well as some rather larger human issues.

Returning to IT – and speaking directly to ISOFT – he said that GP IT had been so successful because its development was driven by clinicians and the people who use the systems.

“So I would continue to plead: keep taking into account not only the desire of the patient, which is extremely important and will have a greater impact in future, but also the knowledge, experience and needs of the clinicians.”

 

Go to top of this page | Go to E-Health Insider homepage