Introduction

Changes to this year's indicator set

Descriptions of data methodologies

NHS case studies from last year's indicator set

Press release

The NHS Performance Assessment Framework was published in April 1999 and the first set of indicators published in June 1999. This introduced a new, broad-based approach to assessing performance in the NHS. Relevant, comprehensive and transparent information on performance is vital to enable clinicians and managers to drive up standards but it is equally important for patients and the public to have access to the same robust information. This expanded and updated set of performance indicators - published in easy to use graphs – has been developed to reflect more closely the key priorities for the health service.

These indicators also illustrate the need for change, not only in raising the performance of the average, but also in reducing the variations in service across the country. These indicators show that while there are many areas where the NHS has improved its performance over the past year there remain unacceptable variations in performance achieved across the country and across hospitals. We should note however that the effect of the NHS's response to the June 1999 indicator set will only begin to feed into the next set of data to be published.

Many of the variations exist in key policy areas and include:

The variation between Health Authorities is unacceptable and includes:

• A twofold variation in death from all causes between the worst and best H.A.s
• A twofold variation in the death rate from all circulatory disease between the worst and best H.A.s
• A more than twofold variation in the death rate from suicide and the death rates from accidents between the worst and best H.A.s

Performance also varies across the country in terms of how long patients wait for treatment.

• There are nearly five times as many people waiting in the worst performing H.A.s as compared to the best
• In some areas every emergency case in A&E in seen within 2 hours, but in others it is less than 40%
• In some Health Authorities, 9 out of 10 people are seen as outpatients within 13 weeks of GP referral, but in other Health Authorities this falls to 6 out of 10

In the key priority areas of heart disease and cancer there are major differences in performance. For example:

• While there was a 9% increase between 1997/98 and 1998/99 in the elective surgery rates for CABG/PTCA, one third of Health Authorities saw no increase at all
• While over half of all Health Authorities ensure that 98% of women are seen in outpatients within 2 weeks of referral by a GP for suspected breast cancer, in some areas this falls below 80%.

The inefficient use of resources prevents the NHS from treating as many patients as it should. There are major differences across the country in efficiency, including:

• Day case rates vary between 57% and 73% by Health Authority
• Casemix adjusted length of stay in hospital varies by over 60% from the worst to the best performing
• Generic prescribing rates vary from 53% to 77% across the country
• While the number of cancelled operations fell in over half of all Health Authorities, it actually rose in the rest

These variations have been driven by seven systemic weaknesses in the NHS that need to be addressed. These weaknesses are:

• Capacity: The NHS has suffered from historic under-funding. Average annual real terms growth rates of just 3% have left the NHS with too few doctors and nurses, with too few up to date buildings, equipment and IT systems. Real terms year on year increase averaging 6% over the next four years will include a major expansion in all of these areas including in the number of nurses, medical students and consultants.
• An absence of national standards: The NHS has never operated according to clear national standards. The National Plan will therefore, include a commitment to work with doctors, nurses, therapists and managers to develop clear national standards so that patients everywhere get the first class service they deserve. This process is already underway with the program of National Service Frameworks laying out standards in key disease areas and with the National Institute for Clinical Excellence providing guidance on a range of treatments.
• A lack of information both for clinicians and for patients: relevant, high quality information is a pre-requisite to identify and correct weak performance and to help clinicians and managers in their day-to-day work. The publication of this new indicator set is a major step in improving information and it complements work on reference costs that will help Trusts and H.A.s make meaningful comparisons. The National Plan will bring forward further proposals to improve the NHS information base.
• Few incentives to improve performance and a proliferation of perverse incentives: the National Plan will establish a clear link between performance and reward so that the best performers – both individuals and organisations – have the freedom and the extra resources they need to expand and improve their services to patients. These changes will build upon the important steps already taken, notably the creation of the unified budget that has reduced perverse incentives for cost-shifting across the health service.
• Too little inspection and assistance in re-designing services: frontline services need expert external advice, support and where necessary intervention to redesign services from the patients point of view. The work of the National Patients Access Team, Commission for Health Improvement and the Primary Care Development Team have demonstrated that effective change can be spread across the NHS.
• Rigid structural and professional demarcations: Rigid partitions between primary care, social care and hospital care need to be removed. Equally, inflexible demarcations between professions have reinforced outmoded working practices and held staff back from reaching their full potential. PMS pilots and the introduction of Primary Care Trusts are both examples of where old demarcations are breaking down and new flexibilities are being introduced. The National Plan will take this process further and include new roles for staff and new means for integrating health and social care services.
• Too little focus on tackling the root causes of disease and the prevention of illness: The NHS has traditionally focused on treatment rather than prevention. New initiatives are needed to reduce the burden from major killers such as heart disease and cancer. This requires improving the NHS' ability to work with other organisations and local communities to tackle the root causes of disease and to raise the profile and impact of preventative services. This new focus is being spearheaded by a number of innovations – prevention is a key theme in the NSFs and the Health Action Zones as well as in renewed efforts in smoking cessation.

It is clear that in some cases action has already begun to tackle the underlying causes of the variations in performance across the NHS. However, it is also clear that in many other areas the National Plan will need to fundamentally address the underlying weaknesses of the NHS in order to reduce these variations for the future.

Health Improvement area - Cancer registrations. The data simply is not timely, it had been hoped that publication in last year's set would act as a stimulus to more timely data from the cancer registries but this has not proved to be the case. Difficult to draw any meaningful conclusions from data for 1993 in the context of current NHS performance.

Fair access area - Children registered with an NHS dentist. Too much emphasis on dentistry in Fair Access area last year (2 out of 5 indicators). This year only Adult registrations retained – chiefly as a marker indicator as current need is to develop indicator to measure progress on PM's pledge on dental access at Autumn 99 party conference.

Effective Delivery of Appropriate Healthcare – Surgery rates and Early detection of cancer. Last year these two indicators appeared in both the Fair Access area and Effective Delivery area. This year rather than duplicate they appear in Fair Access only but the text acknowledges they have a Effective Delivery dimension.

Patient/Carer experience area - Percentage of those on waiting list waiting 12 months or more. This indicator has been changed to show the number of people waiting 18 months or more. This is because NHS Trusts and HAs have a performance target that no one should wait more than 18 months for treatment. Figures on those waiting more than 12 months are still published.

Health Outcomes area – Avoidable deaths – This indicator was a composite indicator and was difficult to understand. More importantly the two main elements of the composite CHD deaths, hypertensive and cerebrovascular deaths, which accounted for the vast majority of deaths, are captured in the Deaths from Circulatory diseases indicator which is in the set. In effect we had a form of double counting.

Health improvement – Deaths from accidents (changed). Last year the data populating this indicator was a proxy as the correct data was not available in time. This year the correct data populates the indicator.

Health improvement – Serious injury from accidents (new). This is the fifth OHN indicator – to be used to monitor reduction in rate of serious injury from accidents (OHN target).

Fair access – Cancer waiting times (new). Indicator showing 2 week breast cancer wait to monitor standard.

Fair access – Number of GPs (new). Indicator shows no. of GP WTEs per head of weighted population. To assess differential access to primary care.

Fair access – GP practice availability (new). Indicator based on patient survey to assess access to primary care. This indicator shows the proportion of patients who put off a visit to the GP in the last 12 months because of inconvenient opening hours

Fair access – Elective Surgery rates (changed). Last year this composite indicator had four elements, CABG/PTCA, hip replacements, knee replacement and cataract removals. This year it has just the last three elements. It still only covers Elective surgery, as all such activity should be planned rather than undertaken as an emergency.

Fair access - Surgery rates (CHD) (new/changed) CABG/PTCA is now a separate indicator. This separate indicator now looks at both Elective and non-elective surgery, as though most of the activity should be planned some proportion of the work will take place as an emergency. The indicator captures this and is consistent with the CHD NSF.

Effective Delivery – Chronic care management (changed). Previously this indicator looked at admissions for asthma, diabetes and epilepsy. The indicator now looks at asthma and diabetes only. This is in line with the two conditions for which GPs receive chronic disease management payments.

Effective delivery – Discharge from hospital (changed). Last year this indicator composited two of the clinical indicators, discharge home following treatment for stroke, discharge home following fractured hip. This year these two indicators appear as separate indicators with these titles - Returning home following treatment for a stroke, Returning home following treatment for a fractured hip.

Patient/carer experience – Patients satisfaction (new). This indicator shows the proportion of patients that complained, or felt like complaining, about GP surgery staff in the last 12 months.

Health Outcomes – Conceptions below age 18 (changed). Last year this indicator was conceptions below age 16. This year the indicator is Conceptions below age 18 in line with the Social Exclusion Unit report on Teenage Parenthood.

Health Outcomes – Cancer survival (new/changed). Last year this was a composite indicator of breast cancer 5 yr survival and cervical cancer 5yr survival. This year the breast and cervical cancer 5yr survivals are presented as separate indicators and with a revised methodology. Two further cancer 5 year survival indicators have also been added to the set – lung cancer and colon cancer.

Health Outcomes – Deaths in hospital following surgery (changed). The list of operations or procedures excluded from this indicator (emergency and non-emergency) has been refined to exclude a number of medical interventions that Trusts record using a procedure code. These interventions are not usually carried out in a theatre and are therefore inappropriate to measure. There is a modification to the specification for patients who are transferred between Trusts. The patient is now included in the count for the Trust where surgery occurred, not the Trust of first admission ie if surgery occurs after transfer to another Trust, the patient is counted against the second Trust. This improves the accuracy of the Trust-level data.

Health outcomes – Death in hospital following a heart attack (changed). Last year this indicator related to ages 50+. The age group has been revised this year to 35-74 for consistency with the Coronary Heart Disease NSF.

Other - There have been a number of minor changes to the definition of the HES based indicators which improve their robustness and validity, but have only a small impact in terms of actual number of cases.

All of our data is comparative information. Wherever possible we aim to show "like with like" comparisons. For example showing percentages rather than numbers can often produce fairer comparisons. Similarly standardisation is a process which enables better comparisons to be made between organisations (Trusts or HAs).

In this set of performance indicators for those indicators, based on Hospital Episode Statistics, which were standardised standardisation took account of variations which may have emerged through populations having age and sex structures which differed from the norm. For population here we mean population relevant to the particular indicator. Therefore for some indicators we are looking at an in-hospital population for other indicators we are looking at a resident population (irrespective of whether they are in or out of hospital).

The standardisation has not taken account of casemix (the varying severities of condition of the cases being dealt with), comorbidities (other co-existing illnesses or conditions) or deprivation, all of which may be explanatory factors in the indicator value for an organisation. These aspects may be things we want to take account of in the future though at present the data could not support these additional degrees of sophistication.

Last year the standardisation was Direct Standardisation. This year the standardisation is Indirect Standardisation.

Indirect standardisation was chosen as it is a more robust and appropriate method for the data we are dealing with. It handles the difficulties associated with small numbers more readily than direct standardisation. It will readily accommodate future refinements.

NB Some indicators, the 5 indicators from Our Healthier Nation remain Directly Standardised and for age only. The targets, milestones etc from OHN were fixed prior to methodological review and clearly can not now be changed. In practice because the OHN indicators are at HA level, pool data across 3 years and focus on big killers the small numbers issue becomes irrelevant. Similarly because the OHN targets are long term and set in stone the flexibility to future refinements offered by indirect standardisation is also irrelevant. Therefore either method for these indicators would be highly likely to produce very similar results.

This is where two or more indicators are combined to give one overall indicator. For example with screening rather than look at the cervical screening and breast screening separately we look at the overall performance on these two elements of the screening performance in one indicator.

For the composite indicators we report the performance in a locality and compare it to the national average performance. For the purposes of these indicators we have chosen a scale where the national average performance is 100. The difference between a locality value and 100 is the difference per component of the composite.

The confidence intervals assess the level of uncertainty caused by chance occurrences in the indicator value. The 95% confidence interval gives the range in which we would expect the true indicator value to fall 95 times out of 100. In interpreting these types of graphs it should be noted that if a Trust's or Health Authority's confidence intervals do not overlap with the England rate, it is likely that their indicator values are genuinely different from the national rate. In some graphs in order to keep a reasonable scale the end of the confidence intervals can not be shown. In the cancer survival indicators the confidence intervals are not symmetrical as the underlying distribution of the data is non-symmetrical.

The NHS Performance Indicators were published for the first time in June 1999. This enabled NHS Trusts and Health Authorities to look at their comparative performance on a range key health services. Some examples of how the NHS has responded locally to challenges highlighted by last year's indicators are outlined below.

We would not therefore expect to see the full year effect of the NHS responding to last year's indicators until the next set is published.

Berkshire Health Authority examines its performance against the Indicators, compared to the regional and national averages. Areas that have improved, stayed the same or got worse are investigated with consultant support. The HA identifies the factors contributing to an above average position, and looks in detail at areas of below average performance, checking these against local development/investment plans and applying further 'in year' monitoring to look for improvements. It also compares individual local Trust performance against the Indicators, and uses this information to highlight areas for improvement and investment.

West Sussex Health Authority formed a group to analyse the clinical indicators in detail, and presented the results to the HA Board and then individually to the PCG Boards. This was done with the intention of fully informing the board members in order that they understood the CI's with the further intention of using them as a performance management tool.

The HA is also using the data to produce benchmarking information for HA's within the same cluster and the local surrounding HAs. A similar exercise is underway for the HA's main provider trusts, comparing them to like sized hospitals. The HA and the PCG's are currently in the process of organising joint presentations to relevant trusts.

The information from these analyses has also been shared with the Primary Care Groups Governance leads at joint meetings, and has stimulated discussion and informed plans for future joint working. For example, some joint work is planned around re-admissions and the management of complex cases across the health/social care interface.

Liverpool Health Authority used last year's Indicators to note their relatively low access rate to elective hip and knee surgery. The HA and the local Trust (RLBUHT) have been in discussion on how to improve access rates. The Royal Liverpool & Broadgreen University Hospital Trust is now in a good position to make in-roads into the Orthopaedic waiting lists and increase the level of hip and knee replacements undertaken because:

• There are new consultants on the establishment.
• The workload from primary care is not increasing based on recent input reports.
• A large share of the additional waiting list investment in 1999/2000 went into Broadgreen which is becoming an elective Orthopaedic centre.
• Broadgreen work does not suffer from the pressures of emergency demand in the same that the Royal Liverpool does.
• Further investment has been put into waiting list reduction for 2000/2001 which should allow a further growth in the level of output.

This extra investment is expected to lead to an improvement in 2000/01 and will be monitored closely.

Nottingham City Hospital was prompted by last year's clinical indicator on 'deaths following surgery' to investigate this issue and develop initiatives to ensure risks of death following surgery are as low as possible.

A key development in Nottingham City Hospital's general surgery directorate has been the introduction of a robust risk assessment system called POSSUM (Physiological and Operative Severity Score for enumeration for Mortality and Morbidity). This is designed to quantify each patient's vulnerability by taking into account factors of their condition and the procedure they are to undergo.

This accurate assessment allows doctors and nurses to design care packages which best meet their individual needs - patients with high POSSUM scores are targeted for higher levels of support and patients with medium scores are monitored more closely to ensure they do not deteriorate.

The plan is to 'roll out' the system to other surgical directorates and POSSUM is clearly expected to improve the outcomes for more and more patients following surgery at Nottingham City Hospital.

East Lancashire Health Authority was already aware that its inappropriately used surgery indicator was very high (poor), reflecting rates of both dilation and curettage (D&C) operations and grommet insertions. It had taken steps to address this in previous years. When the indicator was published it was clear that, although the numbers of D&Cs in particular were declining in East Lancashire, this reduction had been more rapid elsewhere. The value of the data was to emphasise this difference and to add impetus to a more rapid change within the trusts locally.

Sheffield Health Authority - High cancer mortality rates – including those for breast cancer - were highlighted by publication of the indicators last year in Sheffield. Sheffield health community has pledged to take action to reduce these rates and dramatic changes have been and continue to be made. In relation to breast cancer:

• Management of specialists and specialist services have been consolidated and multi-disciplnary teams have been developed
• Sheffield's rapid diagnostic outpatient services are now top of the table for delivering appointments within two weeks of referral
• There has been improved up-take of screening – now in the top ten nationally
• A cross-city management board for breast cancer services has been developed
• This year, £200,000 plus capital investment will be invested in a new centralised breast cancer service.

The Health Authority is confident that these changes have already improved survival rates for breast cancer, yet to be reflected in published figures, and will improve further with current investment and improved service.

Salford Royal Hospitals NHS Trust is a designated NHS Learning Site, and uses the European Foundation for Quality Management (EFQM) Excellence Model to perform routine self-assessment. Self-assessment is undertaken across the Trust on an annual basis, with six monthly reviews, and directly involves the Chief Executive. The indicator for deaths in hospital within 30 days of emergency admission with a heart attack is included as part of the Cardiac Medicine Directorate's self-assessment, and has also been identified as a Clinical Governance Indicator by the Directorate. They also effectively have ownership of a Trust indicator, as AMI will encompass specialties other than purely Cardiology. Information concerning this indicator is also provided to senior management on a monthly basis, as part of the Trust's continuous Corporate Performance monitoring reporting system.

Good Hope Hospital Trust identified as an issue its performance against the clinical indicator "death in hospital within 30 days of emergency admission with a hip fracture ". Subsequent developments include;

• a "fast track" to emergency theatre for patients with hip fractures
• development of a multidisciplinary approach across all wards and departments
• audit and review of the patient pathway.

East Riding and Hull Health Authority has been working in partnership with Local Authorities to identify key indicators to share from both NHS and Social Services Performance Assessment Frameworks. For 2000/01 28 indicators have been identified as requiring joint monitoring and action.

An early trial of indicators in 1999/00 has led the Older Peoples Project Group to focus its attention on improving discharge arrangements between health services and local authorities (using the delayed discharge indicator) and providing greater support to people discharged from hospital e.g. improved catheter care in the community, enhanced intermediate care provision. The work on discharge arrangements has had an impact in 1999/00 with fewer delayed discharges than in 1998/99.

The Health Authority has sought to invest £468,000 in providing such support in the community to avoid the need for emergency admission to hospital. This resource will be targeted at keeping admission rates at the 1999/00 level, not increasing in line with general trends over the past few years. The Health Authority is to provide Local Authorities with regular reports on the rate of emergency admissions.

Barnsley Health Authority - Last year's high level performance indicators showed that GPs in the town prescribe twice the national average of tranquillisers such as Valium, Mogadon and certain sleeping pills. These benzodiazepine drugs can be highly addictive and modern medical practice is not to use them long-term, but to maximise use of other drugs, therapies and treatments.

Barnsley health community's 'Health Improvement Programme' has set a target to reduce benzodiazepine prescribing levels by 15 per cent as soon as possible. Primary Care Groups are leading multi-agency work on a strategy to tackle this priority. Already:

• new guidelines on prescribing have been agreed
• a specialist nurse has worked with GP practices to help re-assess current healthcare packages for patients prescribed benzodiazepine
• a new Mental Health Board has been established to oversee the development and expansion of mental health services in the community

The health community is now confident that benzodiazepine prescribing rates will be significantly reduced through the use of evidence-based policies and increase in accessible high quality non-drug therapies and support for both patients and doctors.

Calderdale & Kirklees Health Authority - In the previous publication of HLPIs, the HA was a national outlier on all three mental health HLPIs. Although there was a local realisation that performance could be improved, the extent of the difference between C&K HA and other HAs was not realised until the publication of HLPIs. There was agreement across the patch to address the issues raised by the mental health HLPIs and the resulting action was incorporated in the Health Improvement Plan and the HA's response to the Mental Health National Service Framework.

The HA has made significant progress on the suicide rates and on psychiatric readmission rates and is no longer an outlier for these two indicators. The actions included sharing lessons learned from suicides of people known to mental health and ensuring that practice changed as a result of the recommendations made; the introduction of assertive outreach teams; setting up a multi-agency 'Staying Out ' project in North Kirklees.

Lincolnshire Health Authority held the lowest overall generic prescribing rate in England last year. From this low point, Lincolnshire GPs have achieved unprecedented growth in generic prescribing in 1999 / 2000 and Lincolnshire is confident it will achieve the national target of 72 per cent by April 2002.

Progress is being driven forward by the co-ordinated efforts of Lincolnshire's 7 PCGs and the HA puts this forward as incontrovertible evidence that the new PCG structure is able to deliver change and modernisation across primary care.

Northamptonshire: tackling deaths from accidents. This is a topic of continuing attention because of the county's acknowledged poor record. They have made good use of this finding from last year's HLPIs to develop further accident prevention work.

The biggest causes of death from accidents remain: transport (38.2%), falls (30.3%), poisoning (10%) and fire and flames (4%). These four account for 82.5% of all deaths from accidents. None of these factors can be influenced directly by health service programmes on their own. However, the NHS in Northamptonshire is a major and active participant in multi-agency programmes aimed at reducing lives lost through accidents. These programmes include:

Launched in July 1999, the strategy aims to reduce accidents by one third by the year 2010. The strategy is underpinned by three main elements - enforcement of driving regulations, engineering to make roads safer, and raising public awareness.

Northamptonshire schools are participating in this national programme which includes developing safer routes to school, and encouraging families to walk children to school rather than using cars.

Healthcare workers are trained to equip them with the skills necessary to undertake opportunistic fire risk checks when they go into people's homes. Another initiative encourages managers and developers of residential establishments to install fire sprinkler systems.

This initiative includes a falls assessment and prevention clinic for follow-up work with patients who come to hospital after a fall; and a falls prevention programme under which a multi-disciplinary team will target over 65 year olds, seeking to identify people at risk of falling and to propose remedial action.

Patients, the public and NHS staff can see how well their local NHS performed last year thanks to figures published today. The NHS Performance Indicators set out NHS performance against 56 different measures, including cancer survival, patient complaints and deaths following surgery.

The indicators highlight wide variations in access to and outcomes of NHS care around the country. For example:

• In some areas virtually all emergency cases in A&E are seen within 2 hours, but in others it is only 60%
• A twofold variation in the death rate from all circulatory disease between the worst and best health authorities
• There are five times as many people waiting for treatment in the worst performing health authorities as compared to the best.
• In some health authorities 5% of operations are cancelled at the last minute, while in others almost none are.

Welcoming publication of the indicators, Health Minister John Denham said:

"These indicators highlight the variations in performance which we will need to tackle in the National Plan for the NHS. Whether the variations reflect underlying health inequalities, lack of capacity in the local health service, or poorly organised services, they must be tackled.

"We have made it clear that the NHS needs more doctors and more nurses. We are investing in additional critical care beds and the development of more comprehensive primary care services.

"At the same time, these performance indicators enable the local health service to identify where their performance falls below that of the best and we expect measures to be taken to improve poorly performing services.

"Last year was the first time the NHS could use this information to improve standards and services. Many of the issues highlighted by the figures in different parts of the country are long term ones which cannot be solved overnight - or even over a single year.

"Our determination to publish clear information also highlighted problems with the quality of information in the NHS. This year the indicators are both clearer and more robust. The standard of information that the NHS has been able to supply has been driven up, giving us and the health service itself a sound base to identify good and bad practices."

The indicators have been revamped in a single, clearer document so that they are easier to understand, while the standard of data they are based on has improved greatly since last year. NHS Trusts are grouped by type and size, while more account has been taken of population differences between health authorities.

1. Full details of the NHS Performance Indicators are available on the internet at the following web site:

http://www.doh.gov.uk/nhsperformanceindicators. If you are prompted for a username and password please enter "NHSPI" as the username and "s93x88" as the password.

2. Printed media copies of the indicators are available from the Department of Health Media Centre on 0207 210 5222 or 0207 210 4984.

4. The indicators fall into two groups. Clinical Indicators are based on data from NHS trusts and measure standards in hospital and community health services, while theHigh Level Performance Indicatorsare based on data from health authorities and illustrate wider public health issues.

5. NHS Performance Indicatorswere published for the first time last year (Press Notice 1999/0355). They replaced the older Patients Charter "league tables" and are directly geared to improving performance across the health service. They are designed to reflect key Government priorities for healthcare, such as cancer and heart disease.

[ENDS]