Paul Westerby, CVD Prevention and Screening Manager with Sandwell PCT, discusses the use of a CVD risk-assessment tool, improvements to service, and potential cost savings

Regardless of whoever is commissioning, the stages required remain very similar (see Figure 1, below).1–3 This article describes the way in which the cardiovascular disease (CVD) risk-assessment service in Sandwell has sought to improve health, reduce health inequalities, and provide measurable, high-quality services within the resources available. The methods and models used are described in this article.

Background to the Sandwell study

The CVD risk-assessment service in Sandwell began with a study of the efficiency and effectiveness of a software tool developed to identify accurately those people most at risk of developing CVD.4 The study was undertaken in Sandwell partly because of the challenge that the area presents: standardised mortality rates (2003–2005) for coronary heart disease (CHD) were 138 in males <75 years of age and 165 in females <75 years of age; for stroke the ratios were 177 and 156, for males and females <75 years of age, respectively.5 The software identifies probable CVD risk, based on data where present, and uses default values based on the Health Survey for England6 where data are missing (>80% of the registered eligible population has one or more default values used). This enabled stratification and invitation of those at highest risk when the study was set up, and still underpins the work today. The tool was accurate in identifying those who, on attendance for CVD risk assessment, were calculated to be at high risk (?20% risk of CVD event over subsequent 10 years).

The pilot work was found to be acceptable to both GP practices and patients, and addressed a workforce capacity issue within some GP surgeries. An advantage of the study system was, and still is, the ability to predict accurately, in the majority of cases, the type of expertise required in the workforce. Patients with different degrees of CVD risk required access to different members of staff:

  • Highest CVD risk—CVD nurse and GP for assessment and medical treatment, respectively
  • Moderate CVD risk—health trainer
  • Lower CVD risk—healthcare assistant assessment (within a GP locally enhanced service plan) and signposting (see Figure 2, below).

Guidelines to support appropriate escalation underpin this model. These are based on JBS2: Joint British Societies’ guidelines on prevention of cardiovascular disease in clinical practice.7

Birmingham Collaboration for Leadership in Health Research and Care is currently evaluating the CVD risk-assessment service.

Understanding the local population
The starting place for the CVD risk-assessment service was understanding of the local population and its health needs, and deciding the priorities.

Sandwell is an area of the West Midlands; there are approximately 280,000 residents who are served by 68 general practices, managing between them a registered population of 340,000 people. Half of the practices have fewer than 5000 patients on their list. Many GPs are long serving and know their population well. The area is characterised by high rates of unemployment and 17 of the 25 wards fall into the most deprived 25% nationally, with 26% of adults being smokers and 29% who are clinically obese.8 Sandwell is ranked bottom in the West Midlands region in terms of its rates of physical activity.9 Early CVD mortality is 98 per 100,000. It is improving, but is some way behind national figures. The inequalities gap remains, with male life expectancy 2006–08 being 3.6 years below the national average.10

Sandwell is also ethnically diverse, having well-established Asian, Black, and Irish populations, as well as African, Arabic, and Eastern European resident groups. Sandwell Health Development workers support numerous distinct populations.

In terms of planning, awareness of the population is essential, as is an understanding of the differences in recorded prevalence of key cardiovascular diseases:11

  • local CHD prevalence is 3.6%—range approximately 2% to 6%
  • frequency of stroke and transient ischaemic attack is 1.6%—range approximately 0.5% to 2.2%
  • occurrence of diabetes is 5.2%—range approximately 3% to 9%.

Differences in demography and ethnicity account for some of these variations, but they are, in part, due to a stoical trait in the population in the borough, which results in varying likelihood of people accessing healthcare services such as the GP practice.

After introduction of the Department of Health (DH) Vascular Checks,12 further analysis, via the software, of the population aged between 40 and 74 years (120,000 people) found that 80,000 did not have CHD, stroke, chronic kidney disease, diabetes, or hypertension (hypertension accounting for approximately 15% of the list) and were eligible for CVD risk assessment.

Although the GP practice was an appropriate place to access the majority of people at high risk, working with other service providers was necessary in order to include those who would otherwise not access the CVD risk-assessment service. This included a third-sector organisation (Healthy Hearts/Sandwell Medical Research Unit), the local acute Trust (Sandwell Hospital), the Mental Health Trusts (acute and community), and the lifestyle services teams.



Figure 1: The commissioning cycle 2,3

graph

Adapted from:

  • Department of Health. Commissioning framework for health and well-being. London: DH, 2007
  • Royal Society for Public Health. World class commissioners: promoting health and well-being—reducing inequalities. DH/RSPH, 2010.

 

Case finding
Screening for high risk of CVD (?20%) has been taking place in GP surgeries supported by PCT-employed nurses. From one nurse in 2005 during the initial study, the team has grown to four nurses (since January 2009); 2300 people have been brought into medical treatment via their GP, including 550 who have been signposted into weight management, and 500 into smoking cessation programmes. Excluding lifestyle change and emergency admissions via rapid access chest pain clinic (1%), cancers (1%), specialist alcohol services (1%), and long-term conditions (6%–mostly diabetes), prophylactic medication with cholesterol-lowering agents and antihypertensive therapies are predicted to result in about 230 fewer heart attacks or strokes, or 23 lives saved over the subsequent 10 years. This represents a reduction in secondary care usage of £1.4 million at 2010 prices, or £140,000 per year (pers. comm from internal Sandwell PCT document). The service therefore represents a cost saving.

At the end of 2008, the DH published modelling data around the cost and cost–benefit of CVD risk assessment,13 and the Sandwell team was able to implement an ambitious rollout plan, with some 90% of the annual cohort eligible for screening to be offered CVD risk assessment in 2011–2012. The DH model however, is based on all patients screened and does not prioritise those to be invited for CVD risk assessment. Modelling would suggest that the most cost-effective intervention would be implemented for the high-risk population, while there will be a lower short-term gain from intervening in the low-risk group. In Sandwell, where the support required to effect lifestyle/behaviour change is significant, the inclusion of very low risk CVD risk assessment creates a dilemma about service provision where resources are limited. A recent BMJ editorial supports a risk-stratified approach.14



Figure 2: CVD risk assessment—Sandwell model

graph

CVD=cardiovascular disease

Patients with diabetes
As we have worked our way down the list of those at ranked CVD risk, generated via the software, the prevalence of previously undetected diabetes and ‘high-risk’ diabetes has increased. The Association of Public Health Observatories suggested that Sandwell should have a prevalence level for diabetes of 9.4% in 2010.15 The lower CVD risk population has, in the main, had little contact with primary care, particularly the male population. A challenge for Sandwell is to meet the need via service provision, which is problematic for diabetes and ‘high-risk’ diabetes. Specialist services, such as X-PERT (www.xperthealth.org.uk), saw only 3% of the population with type 2 diabetes during 2010–11, so access or uptake is limited and ‘high-risk diabetes’ has so far not had a service to refer patients to, although one is now being developed. In addition to those patients already identified, Sandwell is planning a parallel algorithm to run alongside the CVD risk algorithm to stratify the population by both CVD risk and diabetes risk; the two algorithms identify different people. A suitable tool is likely to be the Q diabetes score (QDScore®).16 This will enable identification of those individuals at risk of developing diabetes and will facilitate treatment/lifestyle change in an equally targeted manner.

Preventive work and designing services

Work has been undertaken to augment the skills of the workforce; all lifestyle service staff and a number of third- and voluntary sector staff have been trained in blood pressure measurement using an L2 Open College Network-accredited course, working to an agreed protocol. This means that practices can rely on results that are reported from outside GP settings, and people who are having their blood pressure taken can be confident that they are being given the right advice. The lifestyle teams have also been given CVD/cancer/confidence, and well-being awareness training and all areas are discussed as a matter of course during a consultation. The core services can also deliver advice on weight/physical activity and smoking so that a patient’s individual lifestyle issues are not treated in isolation. Protocols and guidelines providing agreed actions for referral and escalation are used across the teams involved. This allows for audit and quality measurement.

Health trainers can now deliver a group session for those at high risk of developing diabetes, which has been developed in conjunction with diabetes specialist services. This has not yet been evaluated, but 50% of patients initially identified as being at high risk for diabetes attended the group session, and 50% of them joined the health trainer programme—the drop out rate over the six one-to-one sessions has been <10%. Early indications are that patient physical parameters will be consistent with a reduction in risk of diabetes.

Pathways

It is important to channel people into the most appropriate service and, where possible, support holistic treatment, because the determinants of CVD and the support required to make changes are not all to do with health services.17 The process needs to be sufficiently robust that information can be moved around, metrics recorded to evaluate efficacy, and still be simple and straightforward for the patient to access. To this end we have linked all our key services and have the ability to refer internally.

Future improvements

Sandwell operates a process of seeking patient feedback via questionnaires to attendees and a patient representative who sits on the CVD prevention steering group, and outcome measures are monitored. Horizon scanning has been implemented to identify new opportunities, and reviews of service provision are constantly being undertaken to ensure the best outcome is being achieved.

Practically, the service can show that it brings people into treatment, is cost effective, and detects and manages disease at an early stage. Practices are central to the service, but do not operate in isolation. Each practice has access to a broad range of services for its patients and can benefit from contact with specialist groups and training opportunities. The patient receives a high-quality service that meets need and provides the opportunity to influence future service development.

Commissioning
From a commissioning perspective, to implement an evidence-based, cost-effective programme of CVD risk assessment and prevention, the provider must:

  • identify and support those most at risk of dying prematurely18
  • support behaviour change.19

In light of the Government 2010 White Paper,20 and subsequent changes in the recent Government response to the future forum report21 there is a need to develop infrastructures to ensure services are available and accessible outside the GP practice, and that partnerships exist which will seek to address the complex and interlinked needs of those at highest CVD risk.

References

  1. The Health Foundation website. Commissioning. www.health.org.uk/areas-of-work/topics/commissioning/ (accessed 27 July 2011).
  2. Department of Health. Commissioning framework for health and well-being. London: DH, 2007. Available at: www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_072604
  3. Royal Society for Public Health. Guide for world class commissioners: promoting health and well-being—reducing inequalities (funded by the DH). London: RSPH, 2009. Available at: www.rsph.org.uk/en/about-us/policy-and-projects/projects/commissioning-tool-for-health-promotion.cfm
  4. Marshall T et al. The Sandwell project: a controlled evaluation of a programme of targeted screening for prevention of cardiovascular disease in primary care. Biomed Central Public Health 2008; 8: 73 doi 10.1186/1471-2458-8-73.
  5. National Centre for Health Outcomes Development website. Compendium of clinical and health indicators. NHS clinical and health outcomes knowledge base. www.nchod.nhs.uk/NCHOD/HomeDb2R6.nsf/089de59e1cffbe8f65256cd100209bf0/26a48783df209d2580256cf5007145d0!OpenDocument (accessed 10 August 2011).
  6. The Information Centre website. Health survey for England 2009. www.ic.nhs.uk/statistics-and-data-collections/health-and-lifestyles-related-surveys/health-survey-for-england (accessed 10 August 2011).
  7. JBS 2: Joint British Societies’ guidelines on prevention of cardiovascular disease in clinical practice. Heart 2005; 91 (Suppl 5): 1–52.
  8. Department of Health. NHS health profiles: local health. Available from: www.localhealth.org.uk/GC_preport.php?lang=en&s=109&selId0=60&nivgeo=uaut&dyn=0
  9. Sport England website. Active people survey 1: regional results. www.sportengland.org/research/active_people_survey/active_people_survey_1/regional_results.aspx (accessed 5 August 2011).
  10. Association of Public Health Observatories. Health inequalities intervention toolkit life expectancy tool for spearhead areas, 2010. Key results for Sandwell. Available at: www.lho.org.uk/LHO_Topics/Analytic_Tools/HealthInequalitiesInterventionToolkit.aspx
  11. The Information Centre website. Quality and outcomes framework: online GP practice results database. www.qof.ic.nhs.uk/search/index.asp (accessed 10 August).
  12. Department of Health. Putting prevention first—vascular checks: risk assessment and management. London: DH, 2008. Available at: www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_083822
  13. Department of Health. Putting prevention first—vascular checks; risk assessment and management. Impact assessment. London: DH, 2008b. Available at: www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsLegislation/DH_090351
  14. Marshall T. Targeted case finding for cardiovascular prevention. BMJ 2010; 340: c1376.
  15. Association of Public Health Observatories, Diabetes Health Intelligence. APHO diabetes prevalence model: key findings for England. AHPO, 2010. Available at: www.yhpho.org.uk/default.aspx?RID=81090
  16. ClinRisk Ltd. QIntervention®. Available at: qintervention.org/
  17. National Audit Office. Tackling inequalities in life expectancy in areas with the worst health and deprivation. London: The Stationery Office, 2010. Available at: www.nao.org.uk/publications/1011/health_inequalities.aspx
  18. National Institute for Health and Care Excellence. Reducing the rate of premature deaths from cardiovascular disease and other smoking-related diseases: finding and supporting those most at risk and improving access to services. Public Health Guidance 15. London: NICE, 2008. Available at: www.nice.org.uk/guidance/PH15
  19. National Institute for Health and Care Excellence. Behaviour change at population, community and individual levels. Public Health Guidance 6. London: NICE, 2007. Available at: www.nice.org.uk/nicemedia/pdf/PH006guidance.pdf
  20. Department of Health. Equity and excellence: liberating the NHS. London: DH, 2010. Available at: www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_117353
  21. Department of Health. Government response to the NHS future forum report. London:
    The Stationery Office, 2011. Available at: www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_127444 G
  • Screening for CVD is most cost efficient when targeted at individuals at high risk; these can be identified through software systems and prioritised to receive screening first
  • Stratifying probable CVD risk in advance also helps direct the skill mix required to screen the local population—commissioners should ensure that they obtain expert advice from local public health representatives when designing a screening programme
  • Special measures may be needed to screen people who fail to engage with general practices or are socially excluded, thereby helping to reduce health inequalities
  • Commissioners would be wise to make sure their local formulary for cholesterol-lowering, antihypertensive and antidiabetic medications is agreed and shared with all prescribers in advance and uses cost-effective agents to avoid major cost pressures to budgets
  • Risk-assessment programmes for CVD are likely to increase prescribing costs for relevant drugs, and allowances for these should be made in practice prescribing budgets.