POLYPILL

These studies are now complete.

Over 100,000 people in England have a stroke every year. Many die, or if they survive they may suffer with significant disability. Stroke management (e.g. ongoing rehabilitation, social care) represents a major component of NHS expenditure and also imposes substantial costs on the wider economy due to lost productivity and disability.

Optimal management of blood pressure and blood cholesterol could reduce the number of strokes occuring, and the devastating consequences of stroke.

More information

Treatment of blood pressure and cholesterol in people who have already had a stroke is particularly important, because they are at higher risk of having another one. Most management of blood pressure and cholesterol in stroke patients is the responsibility of the GP, yet much of the evidence on which GPs need to base their treatment is derived from studies carried out in people who are under specialist  care, not GP care.

Surveys show that treatment by GPs is generally below the standard recommended. This may be because GPs are uncertain as to whether advised treatment is applicable to general practice patients, or because better methods are needed to help achieve targets.

One way in which care might be improved is to use a fixed dose combination ‘‘polypill’’ rather than the current practice of giving different doses for different blood pressure and cholesterol targets. This project will explore the potential role of a ‘polypill’ that combines drugs that lower blood pressure and cholesterol. This will provide the data needed to design a pilot randomised controlled trial that will test the effectiveness and cost-effectiveness of the ‘polypill’ stratgey versus the treating-to-target strategy.

Design

The research will comprise a series of linked projects, conducted over 6 years:

The results from this study will complement data available from the Health Survey for England (HSE) but unlike the HSE, which was informed by patient input, the present study will reflect actual information available from practices providing data on the whole population aged 40+. The results will also complement analyses such as those performed on routine general practice databases such as QRESEARCH by providing data in a geographically discrete area where it is possible to link the data to individual patients.

The specific objectives are to obtain:

1. Data on prevalence of raised CV risk in primary care, using data on risk factors, in individuals aged 40+, from GP computerised information systems;

2. Data on current adherence of primary care to national guidelines for stroke prevention (cholesterol and blood pressure management; anti-thrombotic therapy);

3. Data on suitable individuals to be invited to participate in the screening and qualitative interviewing aspects of the programme.

4. Data to inform the modelling study

5. Data on current adherence of primary care to national guidelines for stroke prevention (cholesterol and blood pressure management; anti-thrombotic therapy) after the implementation of the national cardiovascular screening programme.

Project 1 - Computerised searches of GP records

Computer searches of practice clinical computer systems will be run to collect data on those risk factors used to calculate CV risk.

Twenty practices will be recruited. Ten of the practices will take part in the screening study (see Project 2).

In all practices, CV risk will be calculated, using the Framingham risk calculator, and the study population will then be classified into the following risk groups: 

  1. Unknown CV risk;
  2. Existing CVD;
  3. High CV risk
  4. Low/Intermediate CV risk.

Using the data collected from searches, patients who have unknown or high CV risk will be identified. Those who might benefit from changes to their current medication (i.e identified as not being treated optimally according to current guidelines) will be offered an appointment with their GP to consider treatment options. Any treatment recommendations described by the GP will be recorded and returned to the research office. In the screening practices, people aged 50+ who have unknown CV risk will be invited to attend a study CV screening appointment at their practice.

Practices will be notified of the identifying codes of all patients aged 40-49 who are identified either as receiving sub-optimal care in line with current guidelines, or who are in the unknown CV risk group. Practices will invite the latter to attend for screening in due course as part of the national screening programme.

In all twenty participating GP practices, a second computer search, using the same criteria outlined above, will be run on completion of screening appointments. This second search will occur at least one year after the initial search. The data will be categorised as detailed above. The proportions of patients in each risk category will be compared with the results from the initial search, and used to assess the impact of the national cardiovascular screening programme.

Project 2 - Screening study

Patients aged 50+ identified as having unknown CV risk, including those on treatment for either raised cholesterol or raised blood pressure in whom there is not sufficient information to calculate a risk score (to judge whether they should receive additional treatment) will be invited to attend their practice for a screening assessment.

Using the results of the screening, patients will be allocated into one of the three risk categories as detailed above. Again, people identified as being at high CV risk will be offered an appointment with their GP to consider treatment options. Any treatment recommendations described by the GP will be recorded and returned to the research office.

Estimated sample size

Patients aged >50 represent 33% of the population and therefore 10 practices with an average list size of 5,000 should generate 16,500 patients in the target age group. Estimates assume that of those patients in the target age group:

  • About 15% will have a past history of CV disease;
  • About 33% will be at high CV risk;
  • About 52% will be at low/intermediate CV risk.

 

Inclusion criteria

Patients aged 50 years or over

Exclusion criteria

Patients known to be terminally ill; housebound patients

Project 3 - Qualitative Interview Study

The aim of the interview study is to determine patient and health professional attitudes to using a ‘polypill’ to manage CV risk, as opposed to treatment to individual targets for blood pressure and cholesterol.

The understanding of factors that make people and practitioners keen or reluctant to initiate a ‘polypill’ strategy are critical to the success of an RCT comparing the two treatment strategies in question. People on medication find combination tablets convenient and easier to remember, but do not like the reduced flexibility and the potential problems with identifying the source of side effects. Beliefs and practices of GPs with regard to hypertension treatment vary and patient behaviour with regard to CV risk is heavily influenced by subjective factors. Therefore, qualitative interviews will be conducted, which are considered to be the most appropriate way to elicit people’s beliefs, knowledge and the meanings they ascribe to their experiences. A grounded theory approach will be used to guide sampling, data collection and data analysis.

Participants and Sampling

20 patients and 30 healthcare professionals (nurses and GPs) will be selected for interview. A further theoretical sample within each group will then be sought out, with selection guided by emerging data analysis in order to extend and challenge earlier data and interpretation, and test the integrity and credibility of developing analysis.Interviewing will continue until new concepts are no longer being generated and theoretical saturation (repetition of themes) is reached.

Patients Sampling Strategy

The researcher will identify a list of all the patients in the existing CVD and high risk groups in the anonymised dataset and then send a list of the identifying codes of all these patients to the practice. The practice staff will then de-anonymise these patients and will send them a Beliefs about Medicines Questionnaire (BMQ-General). Patients who return a completed BMQ questionnaire will be eligible for interview.

Sampling will be carried out on the basis of BMQ score group, CV risk status (existing CVD, high CV risk and on treatment or high CV risk and no treatment) and number of medications. A researcher will randomly select patients from each of these categories, ensuring that similar numbers of patients in each category are included. Other criteria that will be taken into consideration when selecting patients will be their age, gender, ethnicity and socio-economic status. The sample will attempt to reflect a range of these criteria.

Health Care Professionals Sampling Strategy

GPs and practice nurses from each practice will be asked to complete a BMQ - General questionnaire at the study outset. They will also be asked to complete a section regarding their gender, date of qualification and ethnicity for sampling purposes. Those returning the BMQ will be eligible for interview. 

Sampling will be carried out on the basis of BMQ score group, practice size and practice socio-economic status (reflected in the Index of Multiple Deprivation (IMD) score of the practice location). A researcher will randomly select practitioners from each of these categories ensuring that similar numbers of practitioners in each category are included. Other criteria that will be taken into consideration when selecting practitioners will be the date they obtained their professional qualification, gender and ethnicity. The sample will attempt to reflect a range of these criteria.

Practice staff will then be sent an invitation letter and information sheet, inviting them to take part in an interview.

Interviews

The data will be collected by confidential, face-to-face, in-depth interviews using a topic prompt developed from the literature and then modified and refined during the first interviews. The interviews will follow broad topic areas based upon the study objectives, but encourage respondents to discuss their perceptions and experiences freely. Patient interviews will investigate understanding and beliefs about the relationship between CV risk factors and outcomes and the acceptability of a ‘polypill’ strategy compared with current individual treatments. GP and practice nurse interviews will explore the prescribing decision and perceived personal, patient, clinical and financial barriers to prescribing. Experience of CV risk factor management will be explored from patient and professional perspectives, as will issues around concordance.

Project 4 - Questionnaire Study

The aim of this study is to determine patient’s willingness to participate in a clinical trial of the polypill. Patients will be given information about a proposed Randomised Controlled Trial (RCT) and asked if they would consider participating in such a trial in the future. They will be given the opportunity to provide suggestions for improvement of the information sheet and any aspect of the trial. Demographic data will also be collected in the information sheet/questionnaire. This data will inform whether such a trial would be acceptable to patients and their willingness to participate.

Participants

Patients will be identified by computer searches, as previously performed for Project 1. Five research active practices in the West Midlands that have practice based research nurses will be included in the study.

Study design

Patients aged 50 years or over with unknown CV risk from 5 practices will be identified by searches of clinical practice systems. The questionnaire will include questions such as:

  • Would you take part in a clinical trial such as this?
  • Can you suggest any improvements to this information sheet?
  • Do you have any other comments regarding a polypill trial?

The questionnaire will also request some demographic data, such as patient gender, age and ethnicity. Patients will be asked to complete the questionnaire and post back to the research team in the prepaid envelope provided. All responses will be anonymous and data will be entered into a password protected study database, with access restricted to study team.

Inclusion Criteria

Patients aged 50 years or over at run date with unknown CV risk. Unknown risk is defined as:

  • No BP and/or total cholesterol measurement in the last year AND not on a anti-hypertensive or cholesterol lowering agent

OR

  • On an anti-hypertensive but not on a cholesterol lowering and with no total cholesterol measurement available for the last year

OR

  • On a cholesterol lowering agent, but not on a BP lowering agent, with no BP measurement available for the last year

Project 5 - Mathematical Modelling Study

In this study a mathematical model will be constructed to explore the potential impact of using a ‘polypill’ strategy in different categories of CV risk as compared to: current practice; and full implementation of current guidelines. The model will determine which approach to CV management will best meet the needs of people in the different CV risk categories. Impact in terms of CV related events, costs and quality of life of different potential uses of the ‘polypill’ and guideline implementation will be explored for the following people:

Age >50 not already on BP or cholesterol lowering treatment.

  • A comparison of predicted outcomes of:
  • doing nothing
  • prescribing a ‘polypill’ to all
  • inviting to a screening clinic and treating as per guidelines.

Age >50 already on BP or cholesterol lowering treatment.

  • A comparison of predicted outcomes of:
  • maintaining current medication
  • switching to a ‘polypill’
  • optimising current medication to achieve BP and cholesterol targets as per guidelines.

History of CV disease as per those already on primary preventative treatment.

A comparison of predicted outcomes of:

  • maintaining current medication
  • switching to a ‘polypill’
  • optimising current medication to achieve BP and cholesterol targets as per guidelines.

 

Methods

An individual sampling model will treat each patient as a separate case to allow for the complex nature of each individual’s medical history. This generates a large number of patient histories to ensure a fair representation of what could be expected were any given policy to be applied. The model will enable a comparison to be made in terms of how many CV events are prevented and the cost of the different patterns of service provision. Following this a cost utility analysis will be conducted using data on costs of managing CV events and the impact of CV events on quality adjusted life years (QALYs). The model will be populated with risk factor prevalence data derived from projects 1 and 2.

Definition of treatments

Data relating to current treatment will be obtained from project 1, the prevalence study. For primary prevention, the ‘polypill’ intervention will comprise: simvastatin 40mg, hydrochlorothiazide 12.5mg, lisinopril 5mg and amlodipine 2.5mg. For secondary prevention, the ‘polypill’ will include a beta-blocker instead of the amlodipine. Different doses will be modelled for different side effects. It is assumed that aspirin 75mg daily will be prescribed in addition to the ‘polypill’ if the patient has existing CV disease. We will model adding or not adding aspirin to the ‘polypill’ where  10 year CV risk of >20%. ‘Optimal’ care will be based upon the Joint British Societies’ guidelines for BP and cholesterol, NICE Hypertension and Lipid guidelines and QOF.

Estimation of costs

The cost data will be principally derived from National Reference Costs and published literature and will cover a broad perspective to include costs incurred by:

  • Primary Care consultations including staff and surgery costs
  • Out patients
  • Hospital admissions
  • Drug costs
  • Investigation costs
  • Patient costs

Sensitivity analyses will be used to explore the effects of different plausible prices of the ‘polypill’. The numbers and classes of drugs required to achieve guideline targets will be estimated from the literature.

Estimation of outcomes for each treatment strategy

The effect of the ‘polypill’ and guideline strategies on BP and cholesterol levels will be derived from the literature with sensitivity analyses and a prediction of impact on CV events and mortality.[i] The impact of different CV events (myocardial infarction, stroke and TIA) on QALYs will be determined through literature reviews.

Cost effectiveness and outcomes analysis

Costs (from NHS and societal perspectives to include costs incurred by patients) and outcomes (CV events averted) will be compared for the 3 different patterns of care for each population sub-group mentioned above. Outcomes will be predicted based on epidemiology of stroke and myocardial infarction and the effectiveness of therapeutic interventions. The results will be initially reported using a cost consequence approach for which a full description of all important results relating to costs and consequences will be provided. An incremental cost utility analysis will be conducted that will incorporate the impact of stroke and myocardial infarction on cost and QALYs. The robustness of the results will be explored using sensitivity analysis. This will explore uncertainties in the study based data itself, the methods used to analyse the data and the generalisability of the results to other settings.

This model will explore for each of the 3 subgroups whether there is a dominant treatment strategy and the extent to which this dominance is maintained following sensitivity analyses. These analyses will inform the entry criteria for an RCT.