The Secondary User Service is a vital conduit of NHS data. NHS Information Centre chief executive Tim Straughan tells Andy Cowper how his organisation is setting its sights on improving SUS and eliminating some historical problems.

Nobody ever said it would be easy to give the NHS the information it will need to deliver high quality patient services in the 21st century. Yet pretty much everybody agrees that it's essential.

Whenever any national organisation undergoes reform of its information services, it usually involves certain sets of data being collected and held nationally. In the case of the NHS, the system for doing this is the Secondary Uses Service - or SUS - which is the data warehouse jointly implemented by The NHS Information Centre and NHS Connecting for Health.

According to the 2008-9 Operating Framework, SUS should be the standard data repository for activity, performance management and monitoring, payment and reconciliation by April 2009.

An annex to the Framework makes clear the national expectation that SUS will offer "support for NHS providers to deliver initially coded datasets weekly and comprehensively coded datasets monthly".

Yet many are aware that SUS has its critics. Tim Straughan, The NHS IC chief executive and senior responsible officer for SUS, candidly outlines some of the major challenges found with the SUS system to date, which have focused around three principal areas:

  • Timeliness and completeness of data

  • Technical performance

  • User experience and confidence

Mr Straughan points out that for these three problem areas, the first two are empirically measurable, while the third is about perception. "Trust and confidence - the 'soft' elements - are very important," he says.

"These perception factors set us a key test around user confidence and fitness for purpose: if we get the first two right, the third will follow.

"We put great emphasis on standards on data and interoperability which in lay terms, is having a common language and set of definitions. The purpose of that is to ensure that whether SUS data is being provided to the NHS or other bodies, we're counting, measuring and reporting in a consistent way to a common definition.

"The high-level, overall vision for SUS is to be the main source of comprehensive data and to support a wide range of activities - from reporting and analysis through to planning, commissioning, management, research, audit and public health."

Major services produced by The NHS IC, such as the 18-week tracker and NHS Comparators, are based on SUS data. So its importance to The NHS IC is both emblematic and paramount.

Mr Straughan says: "SUS is fundamental and critical to what we do in The NHS IC. If SUS succeeds in its goals, we'll succeed in ours. It's intertwined."

Timeliness and quality

Mr Straughan knows that the issues of data timeliness and quality have been a stick to beat SUS with. But he explains: "A key information principle is that what comes out can only be as good as what goes in. If the service wants really good quality outputs, SUS needs good and regular inputs."

Timeliness of SUS data has been a key dissatisfaction. Yet Mr Straughan adds that the frequency of data input into SUS is the real issue. "To date, because SUS was used for payment and billing, the system's been running on the quarterly 'flex-and freeze' inputs," he says. "Which means data goes in and so comes out in very big doses around those payment / performance times.

"So if people have tried to take data extracts between the quarterly 'F&F' dates, the quality will have been poor as so little up-to-date data has gone in. If people want more 'real-time' data out of SUS, it needs to be put in to SUS collection much more frequently than quarterly."

An incentive for change

How will that change be led? Mr Straughan says: "People will get better data out, for themselves and others, benefits which should incentivise change. But we've also looked at trying to measure and report back to SHAs on individual trusts on what data is inputted and when, so it becomes director-level issue to know how often and how well coders and people down the tree submit data. As an Operating Framework target, it's linked to pay for performance, and national visibility ups the stakes."

The NHS IC's work with SUS's service provider BT around data quality has examined the data deletion facility, used to remove duplicate records. Mr Straughan says: "Multiple submission and doubling cause big issues around quality so we're making it easier to delete. As well as the 'bulk and net' submissions into SUS, trusts can send in extra data to previous work, or resubmit the whole of a previous period as well, with refinements and changes to more accurate coding."

Money talks

Ultimately, the link to payment is providing trusts with an incentive to get the quality of coding right. SUS software detects missing fields or incorrect formatting, helping to automate the quality process.

"Trusts may not find it comfortable to have their SUS returns bounced back to them, but it's important for everybody to get the data right," says Mr Straughan.

"The technical service performance of SUS is already improving how quickly well-entered data is processed. BT's processing is measured through service-level agreements and metrics with ourselves and NHS Connecting For Health to monitor SUS data processing. We're working on reducing turnaround times.

"User trust will sit on the back of those improvements. If we get good inputs and good processing, we also need a good data extraction and analysis function so that users can log on and get information in the format they want on a system that doesn't time them out or have lots of down-time.

"We're working on all those things, putting a lot of effort into user experience monitoring - whether the system responds and gives them what they want. We're linking this to NHS Comparators, which is one of a number of ways of getting information and data analysis out of SUS."

Myths to bust

Mr Straughan addresses some of the negative perceptions about SUS many of which he believes are being fixed.

"I think people respect the fact we are honest about the issues, frustrations and problems that exist with quality, and that we're well enough plugged-in to user groups to know the problem areas to address.

"We take the problems facing users very seriously and we're working on actions to resolve them. The NHS IC's goal has to be to gain the whole system's confidence and trust to meet not the only short-term Operating Framework requirements, but the longer-term goals.

"However, we're creating a huge and complex system. SUS is one of biggest databases in world, with a huge number of transactions and users."

The Department of Health's health informatics review gave trusts the go-ahead to purchase interim systems while Connecting For Health procured products. Does this cause Mr Straughan concerns about potential SUS data compatibility issues? "I'm reasonably comfortable that it's a pragmatic way forward. Trusts have to do it. The real issues are more for local service providers," he says.

What more can the service do to play its part? Mr Straughan says he'd welcome more frequent data inputting into SUS, and also wants more user feedback. "Accuracy of coding is another very important are for attention," he adds.

"The other key theme of the informatics review was about what the service is doing with data. SUS is more than a method for payment - it can shift into the commissioning and quality agendas, so commissioners aren't just paying for numbers, but rewarding providers for how well they're doing and how much value they're adding. The future challenges for The NHS IC are how to modify and build SUS to support that approach."

In terms of quality metrics to affect future payment in the NHS, Mr Straughan quotes NHS medical director Bruce Keogh's dictum that we will only make progress on quality when we can measure it. "So we need a metric or indicator," he says. "There are lots of soft perception quality indicators, but hard measurable quality indicators are rare.

"The move towards measuring quality will be good for the NHS and there's a clear role for The NHS IC in defining common standards and establishing definitions of quality. We already have work under way to develop quality metrics, tying other quality indicators into SUS. For example, clinical audit data is already taken for clinical purposes by practitioners, many of whom take pride in the quality of their data. Such data can feed in, and be linked across.

"The key question for trusts and commissioners, is simply, 'are you using the data'? If data is used, people pay attention to its quality, and make comparisons with others' performance. There's lots of emphasis in the Darzi review on scorecards and dashboards. Trusts should be comparing themselves with others, and also against norms of high-medium-low, be it for mortality or morbidity rates. Measurement and comparison aren't just about finance as in income, turnover and profit - it's about reputation, which will drive patient choice. Quality measurement is increasingly going to be a key focus for the NHS.

Going forwards, Mr Straughan says: "The whole area of syndication of information is a huge opportunity for us in signposting resources and in developing standards and metrics. Information hooks in to everybody's agenda: commissioning, finance, performance, workforce, clinical, public health, and social care. It's the glue that sticks all of these things together - the common thread supporting all of these agendas."