Charting a new course: data analysis of the NHS across the UK

Article on data analysis in the NHS, first published in Health Service Journal, 5 July 2005

Two years ago, Sheila Leatherman, research professor at University of North Carolina’s school of public health, and Kim Sutherland, a senior research associate at University of Cambridge’s Judge institute of management, wrote ‘The Quest for Quality in the NHS’ for the Nuffield Trust, comparing England’s NHS to other developed countries.

The authors noted the lack of a ‘shared robust information base that provides a common understanding of the NHS’s strengths and weaknesses’. Now, Prof Leatherman and Dr Sutherland have attempted to show that such an information base, using independent and routinely-reported data, can and should be compiled – by doing it themselves, through compiling more than 100 charts from numerous sources into a single chartbook of NHS quality.

Two years ago in England, they found a positive trend in various areas of quality – and Prof Leatherman said that remains the case, despite some weaknesses: ‘We would again revalidate our assertion that the NHS in England is not broken, and that it did have a predictable capacity to improve.’

However, the chartbook’s international comparisons show the UK performing worse than other rich countries, despite some relative improvement (and much data being several years old). In 1998, the UK’s mortality rate for under 75s from conditions either amenable to medical treatment or preventable was 173.1 including ischaemic heart disease (IHD) and 87.4 excluding it (all figures per 100,000 of population), higher than Australia, France, Germany and the Netherlands, and the highest by far if IHD is included. However, separate data for England and Wales shows this measure falling steadily in both countries between 1993 and 2003.

On cancer, the UK’s mortality rate dropped 12% between 1990 and 1999, compared with 9% in Australia, 7% in the US and Germany and 5% in Sweden and France. That still left the UK with the highest rate, although with the gap much-narrowed. More recent data looks encouraging: a government target to cut cancer deaths by 20%, between 1995-7 to 2010, was halfway met by 2000-2. This may have much to do with changes in lifestyle, however: male lung-cancer deaths have fallen hardest, from 87.9 per 100,000 in 1990 to 58.3 in 2002, while the number of women dying from the same condition is almost unchanged over the period.

Prof Leatherman said such changes have several causes, including patient behaviour. ‘One is the improvement in healthcare services and the capacity of the system, so for instance increasing the imaging and scanning equipment is important, and that is taking place.’ It needed to: in 2001, Germany and the US both had more than twice as many CT scanners per capita than the UK.

Targets and heart attacks

Between 2000 and 2004, England met its NHS Plan targets on numbers of GPs (up 2,622 to 31,215) and nurses (whole-time equivalents increased by 48,453 to 315,440, more than double the 20,000 target). However, a target of 7,500 more consultants was not met, with the number rising 6,249 to 30,650.

In some areas, the chartbook finds England’s targets causing problems, such as with treatment for acute myocardial infarctions (heart attacks). ‘The standard of success here has been largely focused on thrombolitic therapy,’ said Prof Leatherman of England, and with apparent success.

The Department of Health has tracked English hospitals’ performance through the percentage of patients receiving thrombolysis (injection of clot-dissolving drugs) within 30 minutes of arriving at hospital (‘Door-To-Needle 30’) and 60 minutes of calling for an ambulance (‘Call-To-Needle 60’).

Hospitals have responded: in England, the number of patients treated within ‘Call-To-Needle 60’ rose from 32% in March 2002 to 54% in September 2004.6 No Welsh hospital managed to reach the English set of performance targets for Call-To-Needle 60, compared with 47% of English ones, although the principality managed better on secondary measures such as use of statins.

However, ‘one could ask whether the standard that was set for thrombolysis maybe back around 2000 has been sufficiently updated to be refocused on revascularisation,’ said Prof Leatherman: in other rich countries, the latter is the standard of care.

The UK’s performance on revascularisation has been very poor. According to research from 2000 – the most up-to-date information available, with Dr Sutherland describing the lack of newer data in this area as ‘very frustrating’ – France, Australia, Sweden and the US performed more coronary revascularisation procedures than the national number of deaths from IHD. In both France, with a low IHD mortality rate of 48 (per 100,000 of population) and the US (with a similar mortality rate to the UK), there were nearly four times as many procedures as deaths.

But in the UK, IHD deaths at 146 exceeded coronary revascularisation procedures at 80 – compared with the US, which performed 568 procedures (all per 100,000). ‘The United States, everyone would agree, has too high a rate, but it is recognised that, here in the UK, that the revascularisation rate is too low,’ said Prof Leatherman.

Deprivation

Despite the NHS’s aim of providing the same levels of healthcare to all, the chartbook shows a widening health gap between rich and poor. In England and Wales, infant mortality dropped from 5.6 per thousand live births in 1997-9 to 5.1 in 2001-3. But among routine and manual groups, the figure has hardly moved, from 6.0 to 5.9. In 2002, the Treasury set a target of reducing this difference by 10% by 2010 against the 1997-9 figure. So far, the gap has increased by 100%.

Research from 2002, using a location-based deprivation index, backs this up: 9.6% of babies born to parents living in the most deprived fifth of English electoral wards weighed less than 2.5 kilos, compared with 6.6% of those with parents in the least deprived fifth.

Some evidence suggests that lifestyles are to blame: across the UK, the lower the socio-economic group, the more likely a pregnant woman is to smoke during pregnancy. Admissions to hospital with asthma and emergency diabetic amputations – both chronic, but normally controllable, conditions – get worse in order of the five English deprivation ward groups, with those from the most deprived fifth of wards 76% more likely to be admitted to hospital with asthma than those from the least deprived.

But Prof Leatherman sounds a warning. ‘In a number of areas, it’s both a sociological, socio-economic factor, such as on asthma and diabetes, and it is patient behaviour. But in some of the areas, the healthcare system can make a difference.’ This includes diabetes: ‘So if you see an increasing trend in amputations or ketoacidosis or emergency admissions to hospital, you do legitimately get concerned about optimising the out-patient care.’

On the bright side, international comparisons between the UK and the US, Canada, Australia and New Zealand show deprivation as little barrier to primary healthcare. When asked if they had avoided visiting a doctor in the last year because of cost, only 4% of Britons said yes, with 6% of those on below-average income. Both figures were the lowest of the five countries: by contrast, 29% of all Americans and 44% of below-average income had avoided the doctor.

The UK also recorded the lowest figures on skipping a medical test and avoiding the dentist when needing dental care for cost reasons, with only two or three percentage points separating everyone and those on below-average income.

But prescriptions were an exception. 45% of all Britons said they had avoided filling a prescription or skipped doses because of cost, rising to 54% of those on below-average incomes, putting the UK second highest on both.

‘Often, because the NHS is equitable and does have universal coverage, these particular issues do not get highlighted,’ said Prof Leatherman. ‘But the implication of that chart in particular, I think, is that there is a policy issue and it is a serious one.’

Waiting lists

Many of the differences between the UK’s nations (see below) can be put down to history, genetics and lifestyle factors. But differences on waiting lists are likely to arise from the different policies the nations have adopted since devolution in 1999.

The chartbook suggests success for England. On official outpatient appointment figures, England has improved, while the other nations got worse. On the proportion of completed waits lasting more than 13 weeks, England’s figure dropped by 7.8 percentage points between 1999/2000 and 2004/5, while Scotland’s rose by 8.1 points and Northern Ireland’s by 7.9. Wales, which measures waits differently, also recorded an increase.

England, along with Scotland, achieved even greater things on hospital admissions: 21% of its list waited for more than six months in 1999, but only 9% in September 2004, while waits longer than a year went from 5% to zero. Scotland also ended waits of more than a year, with 7% waiting six months. But 36% on Wales’ list waited more than six months and 11% more than a year.

However, another measure makes England’s achievement appear more modest: its median waiting times peaked in 1998 at 14.2 weeks, falling to 10.2 weeks by 2004.

‘One of the reasons why we have been working on the chartbook is because we know, having worked here since 1997, that there is a general problem of data not being believed in the country,’ said Prof Leatherman of such differences. ‘I don’t think we would have any reason to believe that the government was manipulating targets per se, but we do think it’s very critical to have independent reports on performance.’

‘What I would say on the median wait is, they did get it down to around ten weeks,’ added Dr Sutherland. ‘That’s a reasonable amount of time to be waiting,’ by international standards.

In another area of apparent English success, access to GPs, the official data is at odds with other research. In 2004, the Healthcare Commission found 23% of English patients claiming to wait more than two days for an appointment, at odds with the official figure of 0.8% in November of that year.17 ‘There’s a discrepancy in the data. What we feel is the balanced way to report data is to report both sources,’ said Prof Leatherman.

Such problems underline the authors’ main point. ‘We are interested in stirring up the constructive debate of how to do this in the future, what are the fair and balanced set of measures and indicators that should be routinely reported,’ said Prof Leatherman.

‘The government has three roles here: the financier, the provider and the evaluator of the healthcare system,’ she added. ‘That’s just a very difficult set of roles to do simultaneously, no matter how well intentioned. There needs to be an independent report.’

Round Britain quiz

The chartbook reveals a number of differences within the UK…

* In 2002/3, Scotland spent £1,262 per person on health, 16% more than England’s £1,085. Northern Ireland spent £1,214 and Wales spent £1,186.
* Scotland has, per person, significantly more GPs, consultants, nurses and hospital beds than other parts of the UK. In 2003 there were 70 GPs for each 100,000 Scots, compared with 53 in England, 56 in Wales and 59 in Northern Ireland.
* In 2001-3, the average Englishman lived to the age of 76 years, 3 months – two years and nine months more than a Scotsman. The life expectancy of an Englishwomen was 80 years, 9 months, one year 10 months more than a Scotswoman. For both sexes, Welsh and Northern Irish life expectancy was between the two, but closer to England’s.
* In 2002, England’s mortality rate from coronary heart disease was 192 male deaths (per 100,000) and 65 female deaths, both the lowest in the UK. Scotland’s was 246 male and 96 female, both the highest in the UK.
* In 1999-2001, Wales had the highest male mortality rate for cancer of 427.8 (per 100,000), while Northern Ireland had the lowest, 392.1. For women, Scotland was highest at 370.1, and England lowest at 339.1.
* Londoners are less satisfied with their GPs than the rest of the UK, across which opinions were fairly similar. In 2004, 46% of Londoners felt their GPs always spent enough time with them, compared with 57% in Scotland and 60% in England outside the capital.
* Waits for GPs were longer in London than in the provinces, with 52% of Londoners getting an appointment the same day or next, compared with 62% in England outside the capital – although the worse figure was Scotland’s, at 45%.

Copyright SA Mathieson 2005