The Stroke Prevention Programme: a programme of research to inform optimal stroke prevention in primary care

Article history

The research reported in this issue of the journal was funded by PGfAR as project number RP-PG-0606-1153. The contractual start date was in August 2007. The final report began editorial review in October 2014 and was accepted for publication in October 2015. As the funder, the PGfAR programme agreed the research questions and study designs in advance with the investigators. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The PGfAR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Richard McManus reports grants from the Department of Health, National Institute for Health Research, Career Development Fellowship and Professorship during the conduct of the study; blood pressure machines from OMRON and Lloyds Pharmacy were received for research outside the submitted work.

Copyright statement

© Queen’s Printer and Controller of HMSO 2016. This work was produced by Fletcher et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

Background

Stroke accounts for about 10% of deaths internationally, and for over 4% of direct health-care costs in developed countries. 1 If other costs, such as lost productivity, benefits payments and informal care costs, are taken into account, the total costs double; for example, in the UK annual care costs are around £4.4B per annum, but total costs are £9B per annum. 2 Over 20% of strokes are recurrent events,3 and if one also takes into account prior history of transient ischaemic attack (TIA), this figure rises to about 30%. 1

The majority of patients survive a first stroke, often with significant disability. Impact of stroke is likely to increase because the prevalence of stroke in the population is likely to rise as a consequence of the ageing of the population and improvements in stroke survival. 4 The National Audit Office estimates that preventing just 2% of strokes in England would save care costs of over £37M in 1 year. 5

Blood pressure (BP), smoking status and serum cholesterol concentration are the most important modifiable risk factors for cardiovascular disease (CVD) and more deaths are attributable to these than any other major disease risk factors. 6 BP and cholesterol lowering both offer enormous potential for stroke risk reduction. Each reduction of 5 mmHg in diastolic BP is associated with a 38% reduction in stroke risk7 and a 1-mmol/l reduction in low-density lipoprotein (LDL) cholesterol is associated with a 20% reduction in risk. 8 The efficacy of cholesterol and BP lowering in both primary and secondary prevention9–11 has been demonstrated. Both national and international guidelines promote CVD prevention through pharmacological control of BP and lipid levels, plus encouragement of behavioural change and lifestyle modification. 12–18

Despite guidelines, previous studies across Europe19–23 and North America24 have repeatedly identified undertreatment of patients for CVD prevention in specific at-risk groups. This may be because of concerns about the applicability of the evidence base for primary care, a particular issue with regard to BP lowering following stroke,25 or it may be because better implementation strategies are needed. Implementation research suggests that improving professional education, feedback via audit and patient-level prompts and greater involvement of patients including education can all lead to greater adherence to guidelines. The Quality and Outcomes Framework (QOF), a system for incentivising practices to achieve specific targets, has had some impact, but QOF targets for managing BP and cholesterol in people at high risk of CVD or after stroke are less intensive than guideline recommendations. Most effective guideline implementation programmes utilise multiple strategies. 26 A third barrier to optimal stroke prevention is that the existing evidence base raises further questions. What is the optimal target for BP following stroke? Is it better to focus on a population strategy (i.e. make sure that all people are treated with something)27 or to focus on achieving the intensive targets that have been set in current guidelines for BP and cholesterol?28,29

The research programme

The aim of this programme is to provide an evidence base to inform optimal stroke prevention in primary care with regard to both BP and cholesterol lowering, and to develop tools to support implementation by both users and health-care professionals. Within the general theme of generating an evidence base in the setting in which the evidence will actually be used,30 this programme focused on three areas: the potential for a polypill, the potential for self-management, and BP lowering following stroke/TIA in primary care.

A series of studies involving different approaches (surveys of practice information systems; screening of patients; interviews with patients and doctors and nurses; economic modelling; and clinical trials) were carried out to answer the following research questions:

• Is it more cost-effective to titrate treatments to target levels of cholesterol and BP, or to use fixed doses of statins and BP-lowering agents (polypill strategy)?

• Will telemonitoring and self-management improve BP control in people on treatment for hypertension, and in people with a history of stroke/TIA in primary care?

• In people with a history of stroke/TIA, can intensive BP-lowering targets be achieved in a primary care setting, and what impact will this have on health outcomes?

In addition, a number of service developments were developed:

• tools to support optimal implementation of guidelines for primary and secondary prevention of stroke in primary care, including training programmes

• methods to optimally involve patients in self-management of their stroke risk.

The research was carried out between April 2008 and March 2014.

The inter-relationships between the three work programmes are summarised in Figure 1. The epidemiological data collected in work programme 1 contributed to the cost-effectiveness analyses in all three work programmes, and to aspects of trial design in work programmes 2 and 3. There was shared methodology in the health economics across the three work programmes.

FIGURE 1.

Overview of the programme. GP, general practitioner.

Evidence from practice information systems

The aim of our data analysis of practice information systems was to provide data of relevance to any future polypill trial in terms of numbers of potential participants, and data to go into our modelling studies of the potential role of a polypill in different populations. Our analyses were packaged into three separate publications. 33,43,44

The design for all three of these analyses was a cross-sectional retrospective study of electronic health records from all patients aged ≥ 40 years registered with 19 general practices in the West Midlands. 43

Evidence from a screening study

Our analysis of general practitioner (GP) information systems identified that a substantial proportion of the population aged 40–74 years were of unknown CV risk. We therefore carried out a screening study in which we invited those individuals of unknown risk to a ‘health check’ to assess their CV risk. We restricted these invitations to people aged 50–74 years, as mean CV risk in people below this age is not high enough to warrant blanket population treatment with a polypill. This study provided key data for our modelling studies (see Modelling studies); see Appendix 1.

A total of 2642 people were screened (39% of those invited). Of these, 1.3% were found to have existing CVD that had not been apparent from the electronic searches; 28.7% were found to have a > 20% 10-year risk of CVD, and 69.2% had a < 20% 10-year risk (in 0.8%, a valid serum cholesterol was not taken, so overall risk remained uncalculated). This indicates that screening in the over 50 years age group identifies a substantial proportion (approximately 1 in 3) of people screened who would require pharmacological treatment according to current guidelines. As the age threshold is raised, so the proportion increases, again raising the possible rationale of switching to a population strategy of treating everyone over a certain age. The recent change in the NICE lipid guidelines45 that now recommend offering statins to all people with a > 10% 10-year CV risk further raises the possibility of treatment purely on the basis of age, rather than on an assessment of CV risk, as even higher proportions of people over a given age will exceed the risk threshold.

Qualitative studies

A key consideration in whether or not a polypill strategy might be useful is its acceptability to health-care professionals and to patients. This was the focus of two qualitative studies which have been published. 52,53

Modelling studies

We identified three different scenarios in which a polypill might be used: as an alternative to ‘health checks’ (i.e. offering a polypill rather than a health check); as an alternative to ‘treat to target’ management of BP and cholesterol for people at high risk of CVD; and as an alternative to ‘treat to target’ for secondary prevention of vascular events. The third scenario was subsequently funded as a separate Stroke Association/British Heart Foundation programme grant to conduct a trial, so we focused on the first two scenarios. We had also originally intended to conduct a subanalysis of whether or not it might be cost-effective to introduce a QOF indicator for assessment of CV risk. However, the roll-out of the ‘health checks’ programme (the national introduction of this programme post-dated the application for this research programme) made this question redundant, so we did not pursue it.

Conclusions

In this programme we have explored the potential role of a polypill for improving prevention of CVD. There are a number of concluding observations to make:

1. Our analysis of current patterns of use of pharmacological therapy for modifying CV risk in general practice shows that, for many patients, current care does not adhere to national guidelines. This raises the question of whether the way forward is to develop better mechanisms to implement the evidence base, or whether an entirely different approach should be adopted. In this section, we explored the entirely different approach, namely using a polypill. In Work programme 2: will telemonitoring and self-management improve blood pressure control in people on treatment for hypertension or with a history of stroke/transient ischaemic attack in primary care? we explore the potential for a different mechanism (using patient self-management for the control of hypertension).

2. Our second analysis found that CV risk-modifying therapy, particularly statins, was used less in older populations (over the age of 75 years for statins). Again, this raises the question of whether or not a simple strategy such as use of a polypill might lead to higher uptake of these treatments in this age group compared with conventional strategies of targeting treatment on the basis of risk.

3. Our screening study, which was primarily used to inform our cost-effectiveness analyses, found that about one-quarter of people aged ≥ 50 years who attend screening are found to have 10-year CV risk of > 20%. These are key data for informing the relative value of treating a population without screening compared with current practice of screening and treating those at high risk.

4. Our qualitative work found that health-care professionals raised concerns about using a polypill for primary prevention, although they felt that it might have a role in secondary prevention.

5. This view was echoed in our qualitative work with patients, who voiced misgivings about people taking treatments that were not ‘necessary’.

6. Our cost-effectiveness analysis suggested that using a polypill strategy was likely to be highly cost-effective when compared with the current practice of screening people and treating those found to be at high risk. However, this is in the absence of any empirical evidence that a polypill would be used by this population, and so this conclusion has to remain speculative.

7. Furthermore, our cost-effectiveness analysis of using a polypill in people known to be at high risk (the more acceptable strategy from our qualitative research) found that such a strategy was unlikely to be cost-effective against better implementation of guidelines.

8. Finally, we collected helpful pilot data on the feasibility of a trial of polypill against screening and treating. Although the response rate was low, we did have sufficient interest (about 50% of respondents) indicating that a trial of this strategy would be feasible, at least in terms of patient recruitment.

On the basis of these observations, we conclude that there is a need for further evidence through RCTs to determine what role the polypill should play in CV prevention in first world countries such as the UK. There are two types of trial that need to be performed. One type of trial is to compare a polypill with usual care in people with high CV risk or existing CVD. Such trials are now under way in several countries. Indeed, our own research group is leading a trial on use of a polypill for secondary prevention of stroke, funded by the Stroke Association. The second type of trial is to compare offering a polypill rather than a CV screen. Our outputs from this programme suggest that such a trial will be challenging, but possible, and with the potential for leading to important health gains.

Research recommendations

1. A RCT of offering a polypill against offering health checks in people aged ≥ 50 years. Our economic modelling suggests that use of a polypill for all people aged > 50 years is likely to be highly cost-effective compared with current practice (screening for CV risk and treating those at high risk). However, our qualitative work and our patient survey suggest that uptake to such a trial might be low. Therefore, preliminary feasibility and pilot work should be carried out prior to such a trial.

2. Qualitative research should be carried out to better understand the attitudes of older people (aged ≥ 80 years) to taking drugs to reduce the risk of CVD. We demonstrated that the uptake of CV risk-reducing therapy falls in people over the age of 80 years. It is not clear whether or not this is appropriate. One possible explanation is that it reflects patient choice. There is little evidence on the attitudes of this age group towards taking such drugs and, therefore, qualitative work is required to understand whether this reflects informed underuse or possible inequity in access to such treatment. This research would also be relevant to understanding issues around polypharmacy and multimorbidity in this older population. There is little research to understand how older people regard strategies such as withdrawal of antihypertensives and lipid-lowering agents in older age. Therefore, such qualitative research could inform interventions to facilitate initiation/up titration of these drugs where appropriate, and also to inform withdrawal and down titration.

3. A RCT of statin therapy for prevention of CVD in people aged ≥ 80 years. Although there is a good evidence base, thanks to the Hypertension in the Very Elderly Trial (HYVET)55 that BP lowering is effective in people aged over the age of 80 years, there is a lack of evidence with regard to lipid lowering. Therefore, we do not know whether or not the low uptake in this age group that we observed is appropriate. This is an important question to answer, as the absolute benefits of statins in the very elderly are potentially very high. The qualitative work that we propose (see above) will also give us an indication of the likely attitude of the target population to this intervention.

Reflections on work programme 1

Over the timespan of the programme, there have been parallel developments in CV prevention in primary care. The QOF has influenced practice to some extent, but the targets used in the QOF are below those recommended in guidelines. Guidelines have changed, so that in effect NICE now promotes offering statins to everyone with a 10-year CV risk of > 10%, which is similar to offering them to everyone over a certain age (given that age is the strongest determinant of CV risk). Therefore, implicitly, NICE has endorsed one aspect of the polypill concept. NICE has also supported a second component of the polypill concept: the idea of not measuring cholesterol levels in people on statins. Our research suggests that primary care does not widely accept these two ideas, as evidenced by our interviews with health-care professionals.

A polypill approach runs counter to the growing interest in personalised medicine: the idea that drug therapy can be increasingly targeted, through better assessment of the genotype and phenotype, at patients who will directly benefit from them. It may be, therefore, that the role of the polypill will be limited to developing countries who cannot afford personalised medicine. However, it remains an open question whether or not a polypill approach might nevertheless be more cost-effective, even in first-world countries.

In all three work programmes, we used cost-effectiveness analysis in which we extrapolated beyond the trial data to estimate the impact of polypill, self-management and intensive BP lowering on long-term CV risk reduction. We had to select a number of study parameters, and any of these can be criticised, though we used what we felt were the best available. For example, we used the Framingham Risk Scores, even though they are US based, rather than the UK-based QRISK2, as we recognised that the latter underestimates the true impact of cholesterol and BP as it relies on electronic general practice data (which is often missing for the former and unreliable for the latter). To explore the potential impact of the parameters that we had used, we tested the impact of plausible variations in them using sensitivity analysis. Reassuringly, these sensitivity analyses did not suggest that parameter selection was a critical issue in any of the economic analyses. For work programme 1, the economic analyses were weaker than for work programmes 2 and 3, in that they were not based on RCT data on polypill impact.

Notwithstanding this limitation, our cost-effectiveness analysis suggested that a polypill would be cost-effective, at the right price, but in the absence of direct empirical data such a conclusion needs to remain speculative. At the time of writing it has been 12 years since the polypill concept was launched, but still it has not been taken up in first-world countries. An alternative approach to the two extremes of personalised medicine and polypill pharmacy is to look at ways of better treating people with existing risk factors. This is where the second work programme, on self-management of hypertension, comes to the fore.

Meta-analysis of trials of self-monitoring

This has been published as Bray et al. 65 For a full-text version of this paper, see Appendix 2.

Our aim in this systematic review was to determine the extent to which self-monitoring on its own led to reductions in BP, and to seek to understand through meta-regression the reasons for previously observed heterogeneity in the results of self-monitoring trials. We found that self-monitoring was associated with a 4 mmHg reduction in systolic BP. There was significant heterogeneity, but we could not explain this by any of the six factors that we explored through meta-regression (age, sex, length of follow-up, diastolic BP, adjustments made for self-monitored readings and use of additional cointerventions). However, the presence of cointerventions seemed to explain some of the heterogeneity with regard to one of our secondary outcomes, namely whether or not an office target BP was achieved. We concluded that self-monitoring does have a small but significant effect on reduction of office BP when compared with usual care.

Randomised controlled trial of Telemonitoring and Self-Management in the Control of Hypertension (TASMINH2)

Our systematic review had demonstrated that self-monitoring per se has a beneficial impact on BP. A logical component to add to self-monitoring is self-management: leaving the decision to up-titrate medication to the patient on the basis of their home BP readings. There has also been increasing interest in telemonitoring, whereby readings at home are relayed to a health-care professional. Therefore, we conducted a trial that tested an intervention based on these three components: patient self-monitoring, patient self-titration and telemonitoring.

The protocol of this trial and the principal results have been published. 66,67 The full text of the Lancet paper is given in Appendix 2.

In our trial, we randomised 527 patients, with a mean age 66 years and whose BP at baseline was > 140/90 mmHg, to self-management with telemonitoring or to usual care, and followed them up for 1 year. Our primary outcome was change in mean systolic BP, for which we obtained data for > 90% of participants. We found that in the self-management arm, mean BP reduced by 17.6 mmHg at 1 year, and in the usual care arm by 12.2 mmHg [difference between groups 5.4 mmHg, 95% confidence interval (CI) 2.4 to 8.5 mmHg]. Furthermore, there were no differences in quality of life or important differences in side effects between the two groups. A major contributor to the lower BP is likely to have been the increased use of antihypertensive medications in the self-management arm. In terms of applicability, self-management is not suitable for everyone, but even if only 20% of people with hypertension engaged in self-management, this would still equate to around 2 million individuals. Therefore, we concluded that self-management represented an important new addition to the control of hypertension in primary care.

How to design a trial of self-management with or without telemonitoring for secondary prevention

Our original intention was to follow on the TASMINH2 (Telemonitoring and Self-Management in the Control of Hypertension) trial with a trial in people with a history of cerebrovascular disease. However, we noted in the subgroup analysis of the TASMINH2 trial that people with high-risk conditions (diabetes and chronic kidney disease) appeared to derive less benefit from self-management (1.3 mmHg reduction, compared with 5.8 mmHg without a condition). This difference was not significant, but, nevertheless, we thought it would be of most value to test self-management in high-risk conditions, including these conditions as well as stroke. We also noted that the telemonitoring aspect of the intervention in TASMINH2 had little impact on the management, so we decided to drop that aspect of the intervention. Therefore, we decided to plan a trial of self-management in people with a high-risk condition. To inform the design of this trial, we carried out three further studies based on the TASMINH2 population: a cost-effectiveness analysis, and qualitative studies of patients’ and health-care professionals’ experiences of patient self-management. These three studies have been published. 68–70

What is the effect of self-monitoring and medication self-titration on systolic blood pressure in hypertensive patients at high risk of cardiovascular disease?

The protocol for this trial and the main results have been published. 71,72

We performed a RCT to determine the effect of self-management of hypertension compared with usual care in people with CVD, diabetes or chronic kidney disease. We recruited 552 patients, who had a mean age of 69 years and whose BP at baseline was > 130/80 mmHg (i.e. above the target for BP control in national guidelines at the time the trial was performed). We followed them up for 1 year, using a primary outcome of difference in systolic BP between the intervention and the control groups. We obtained primary outcome data on 81% of participants. The mean systolic BP in the self-management arm at 1 year was 128.2 mmHg, compared with 137.8 mmHg in the usual care arm, a difference between arms of 9.2 mmHg after adjustment for baseline BP. As in TASMINH2, there were no differences in adverse events or quality of life between arms. 9.2 mmHg is an important difference in BP, equating to approximately a 30% reduction in stroke risk. Its importance is magnified in that this trial was in a high-risk group, who had more to gain in absolute terms than the typical hypertension population. We had lower follow-up for the primary outcome in this trial than in TASMINH2, reflecting the greater comorbidity in the TASMIN-SR (Targets and Self-Management in the Control of Blood Pressure in Stroke and at Risk Groups) population. Nevertheless, multiple imputation for missing values did not make any important differences to the overall results. We concluded that self-management is an important management option for the control of hypertension in people with CVD, diabetes or chronic kidney disease.

Conclusions

The self-management work programme has presented evidence from a systematic review and two trials (TASMINH2 and TASMIN-SR) with embedded qualitative and economic studies that self-monitoring of hypertension with self-titration of antihypertensives is both effective and cost-effective using standard criteria and that, in comparison with the review, self-management appears to result in greater reductions of BP than seen in self-monitoring alone. The trial in high-risk groups (TASMIN-SR) benefited from running directly on from the trial in uncomplicated hypertension and resulted in significantly greater reductions in BP with expected linked improvement in cost-effectiveness (9.2 mmHg vs. 5.4 mmHg). TASMIN-SR was conducted in a frailer group and, subsequently, follow-up was reduced (81% vs. 91% after 1 year) but there was no evidence of increased side effects or adverse events. TASMIN-SR found the intervention effective even without the telemonitoring used in TASMINH2, but this may reflect the difficulties of integrating home monitored BP into electronic medical records. Patients in both trials were enthusiastic about the increased involvement and knowledge gained by self-management but there were understandable anxieties about actually self-titrating. Professionals were interviewed in TASMINH2 only and they were also keener on self-monitoring than self-management. The cost-effectiveness studies required a number of assumptions but both showed that, provided the interventions provided sustained benefits for at least 1 year (TASMIN-SR) or 2/5 years (men/women in TASMINH2), they are likely to be cost-effective. Overall, this body of work includes around 90% of patients in the published literature on self-monitoring and self-titration in hypertension and provides strong evidence that self-management should be offered to patients with poorly controlled hypertensions, with or without other CV comorbidities.

Research recommendations

Reflections on work programme 2

In contrast to the polypill, self-management as a strategy seems to be becoming widely adopted in the NHS. Although in some areas the evidence is lacking, in terms of BP control we have provided consistent evidence that it is a potentially very important approach. Why does it work? A major factor that we have demonstrated is that it increases the likelihood that up-titration will occur when BP is above target. Patients who self-managed were on more drugs than patients that did not self-manage. This suggests that health-care professionals may overestimate the risks of side effects on antihypertensive treatment; this is relevant to the third work programme of exploring the value of intensive BP lowering after stroke. It is interesting to speculate as to whether or not self-management leads to better adherence to existing therapy. Unfortunately, we did not have sufficiently robust measures of adherence in this work programme and so cannot directly comment on this. One reason we did not is that the more accurate measures of pill taking may also influence pill-taking behaviour. Therefore, had we measured adherence, we might unwittingly have reduced the potential impact of self-management.

Does setting a lower systolic blood pressure target for people with stroke in primary care lead to a lower blood pressure?

The protocol74 and trial report75 have been published. A version of this paper (prior to responding to referee comments) is provided in Appendix 1.

In this RCT, we recruited 529 patients (mean age 72 years) with a history of stroke or TIA from 99 general practices. Patients were randomised to an intensive systolic BP target (< 130 mmHg or a 10 mmHg in BP if baseline BP was < 140 mmHg) or a standard systolic BP target (< 140 mmHg). Apart from the systolic BP target, the management of BP was the same in both groups, with the frequency of review determined by whether or not the BP was above target (therefore, there was more health-care professional contact with participants in the intensive arm). The primary outcome was systolic BP at 1 year. Systolic BP went down by 16.1 mmHg in the intensive target arm and by 12.8 mmHg in the standard target arm. After adjustment for baseline BP, a 2.9-mmHg greater reduction in BP was achieved in the intensive target arm. The most striking finding in this study was the size of the reduction in BP in both arms, which was larger than would have been expected because of regression to the mean or accommodation effects. Therefore, our overall conclusion was that although aiming for a 130 mmHg or lower target for systolic BP in people with cerebrovascular disease in primary care can lead to small additional reductions in BP, the more important lesson for primary care is that active management of BP in this population will lead to clinically relevant improvements in BP control.

What are patients’ and health-care professionals’ views on blood pressure targets after stroke?

See Appendix 1 for the full report.

In this study, we sought to understand what barriers there might be to more intensive BP lowering in people who had had a stroke. We interviewed 13 patients who had taken part in the PAST-BP (Prevention After Stroke – Blood Pressure) trial (seven in the intensive target arm and six in the standard target arm) and eight health-care professionals (five GPs and three practice nurses). Key concerns (expressed by both patients and health-care professionals) were that the medication needed to be long term and that there was a risk of side effects. Health-care professionals in particular were concerned about side effects of hypotension, and therefore questioned the value of intensive BP lowering and whether or not the extra effort was worthwhile. Our own trial suggested that these concerns were probably misplaced, but we acknowledge that this may reflect a recruitment bias in the trial. However, it is of note that in the TASMINH2 and TASMIN-SR trials of self-management (see Work programme 2: will telemonitoring and self-management improve blood pressure control in people on treatment for hypertension or with a history of stroke/transient ischaemic attack in primary care?), greater use of antihypertensive therapy was not associated with any extra side effects. Our conclusion was that if more intensive targets were to be introduced into routine practice, there would be some resistance to this from health-care professionals, which would require specific intervention (e.g. educational initiative) to overcome.

Is a lower blood pressure target likely to be cost-effective?

See Appendix 1 for the full report.

We carried out a modelling study to determine what the long-term costs and benefits of intensive BP lowering would be in people with cerebrovascular disease in primary care, compared with using a standard target using the results of the PAST-BP trial. We extrapolated from the trial results using existing evidence from the literature to estimate what impact the different achieved BP in the two arms would have on CV events. We found that intensive BP lowering was dominant: that is, it lowered cost and improved outcomes. However, this conclusion was based on the assumption that a difference in BP was maintained between the arms. We noted that the difference between arms in the trial dropped between 6 months and 1 year, and so were concerned that the effect would become attenuated over time. Therefore, we are unable to make firm conclusions about whether or not more intensive targets would be cost-effective.

Conclusions

The overall aim of this work programme was to determine whether or not a trial of intensive BP lowering in people with cerebrovascular disease in primary care was warranted and feasible. In the trial that we carried out, we found that recruitment was difficult, requiring nearly 100 practices to randomise just over 500 patients. Therefore, to scale it up to an end-point trial would need to involve 1200 practices [e.g. PROGRESS (The perindopril protection against recurrent stroke study) had over 6000 patients followed up for 4 years]. 10 This might be potentially feasible with automated collection of end-point data (e.g. using the Clinical Practice Research Datalink with Hospital Episode Statistics in practices when the data were linked), but it is likely from our qualitative research that continued input would be required to practices to maintain the intervention. Thus, we concluded that we had not demonstrated that such a trial would be feasible. Furthermore, we did not feel that a further trial was warranted, given the impressive reduction in BP achieved in the standard care arm in the PAST-BP trial. We drew the overall conclusion from this work programme that the focus should be on more active management of BP to a standard target (< 140 mmHg) rather than to an intensive target. This would result in many more strokes being prevented than would be by aiming for more ambitious targets, which our qualitative research suggests would alienate patients and health-care professionals alike.

Research recommendation

Reflections on work programme 3

This work programme perhaps encapsulates better than the other work programmes the potential conflict between primary care and secondary care. Specialists in general are keen to promote intensive BP lowering after stroke. Primary care remains sceptical. Our research suggested that it would not be worthwhile to do an end-point powered trial to determine the impact of more intensive BP lowering in primary care. This raises the interesting question as to what should be policy if a secondary care-based trial indicates that intensive BP lowering does improve outcomes.

Service developments

We developed a hypertension management algorithm to support the PAST-BP trial (see Appendix 3). This took the form of a web-based decision support tool. Screen shots to illustrate how this tool supported GPs in their decision-making over which antihypertensive to use (and at what dose) are provided in Appendix 3. The algorithm was based on the NICE and British Hypertension Society guidelines in operation at the time,18,29 and drug information was taken from the British National Formulary. 76 Informal feedback suggested that the algorithm was very positively regarded by participating GPs.

We developed a training course on the primary and secondary prevention of stroke aimed at GPs and practice nurses (see Appendix 3 for details). This was to support the management of people identified as being at high risk of CVD in work programme 1 and to underpin ‘usual’ care in the RCTs (although attendance at the course was not mandatory for participating practices). Feedback was positive.

We took the opportunity of carrying out screening for CV risk factors in work programme 1 to carry out a before-and-after study of the impact of our screening programme (which was similar in concept to the NHS Health Checks programme), by comparing the prevalence of ‘high CV risk’ and the use of CV risk-lowering treatments before and after the screening programme, and in comparison with data taken over the same time period from control practices in which screening was not carried out. The main finding was that, between 2008 and 2012, there was a significant drop in people with uncontrolled hypertension in both intervention and control practices. There was no difference in prevalence of ‘high CV risk’ or use of antihypertensives or lipid-lowering agents between screening and control practices either in 2008 or in 2012. The main weaknesses of this study (apart from its non-randomised nature) were that we had no information on patients who were no longer registered with a practice after 2008, and we could not quantify the extent to which NHS Health Checks had been introduced into control practices. Our conclusion was that it was unlikely that, against the backdrop of NHS Health Checks, practice-based screening initiatives would have a major impact on better management of CV risk.

As part of the intervention development for the TASMINH2 trial (see Work programme 2: will telemonitoring and self-management improve blood pressure control in people on treatment for hypertension or with a history of stroke/transient ischaemic attack in primary care?), we developed a training programme to ensure that patients were confident and accurate in their self-management of BP. Over 90% of participants randomised to self-management successfully completed the training package, which included an assessment of their own ability to measure their BP and understand what they should do depending on the readings obtained. After 12 months’ participation, 72% of patients randomised to self-management continued to self-monitor and self-manage, although only 55% of recommended medication changes were implemented (sufficient to lead to significant reductions in BP in the whole group compared with usual care; see Appendix 2).

Research recommendations

1. Further develop web-based hypertension management algorithms to support clinical decision-making in general practice, and evaluate in a RCT. Informal feedback in our programme suggested this approach was very popular among GPs. Implementation of evidence-based guidelines for management of hypertension remains suboptimal, so there is potential value in developing and evaluating new interventions to support implementation of guidance.

2. Further develop self-management for BP to adapt and simplify procedures with the aim of improving adherence to the self-management schedule. Self-management has been shown to be effective in both TASMINH2 (see Appendix 2) and TASMIN-SR (see Appendix 4). The service development work suggests this is despite some attrition of adherence to the self-management guidelines over time. Therefore, further simplification and adaptation is likely to be important, to both maximise effectiveness and facilitate roll-out of the intervention.

Contributions of authors

Kate Fletcher compiled the original full report that was submitted to NIHR.

Jonathan Mant was lead and Richard McManus and Richard Hobbs were coapplicants on the original programme grant.

Jonathan Mant wrote the first draft of this pilot report, with contributions from Richard McManus on work programme 2.

Richard McManus, Kate Fletcher and Richard Hobbs contributed to the second draft of the pilot report.

Data sharing statement

Requests for access to data should be addressed to the corresponding author.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, CCF, NETSCC, PGfAR or the Department of Health. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the PGfAR programme or the Department of Health.

Work programme 1

Sheppard JP, Singh S, Fletcher K, McManus RJ, Mant J. A cross-sectional study investigating the impact of age and sex on primary preventative treatment for cardiovascular disease in the West Midlands, UK. BMJ 2012;344;e4535. URL: www.bmj.com/content/345/bmj.e4535

Virdee SK, Greenfield SM, Fletcher K, McManus RJ, Hobbs FDR, Mant J. Would primary healthcare professionals prescribe a polypill to manage cardiovascular risk? A qualitative interview study. BMJ Open 2013;3:e002498. URL: http://bmjopen.bmj.com/content/3/3/e002498.full

Sheppard JP, Fletcher K, McManus RJ, Mant J. Missed opportunities in prevention of cardiovascular disease in primary care: cross sectional study. Br J Gen Pract 2014;64:e38–46. URL: http://bjgp.org/content/64/618/e38

Sheppard JP, Fletcher K, McManus RJ, Mant J. Prevalence and costs of treating uncomplicated stage 1 hypertension in primary care: a cross sectional analysis. Br J Gen Pract 2014;64:e64–8. URL: http://bjgp.org/content/64/627/e641

Virdee SK, Greenfield SM, Fletcher K, McManus RJ, Mant J. Patients’ views about taking a polypill to manage cardiovascular risk: a qualitative study in primary care. Br J Gen Pract 2015;65:e447–53. URL: http://bjgp.org/content/65/636/e447

Work programme 2

McManus RJ, Bray EP, Mant J, Holder R, Greenfield S, Bryan S, et al. Protocol for a randomised controlled trial of telemonitoring and self-management in the control of hypertension: telemonitoring and self management in hypertension. BMC Cardiovasc Dis 2009;9:6. URL: http://bmccardiovascdisord.biomedcentral.com/articles/10.1186/1471-2261-9-6.

Bray EP, Holder R, Mant J, McManus RJ. Does self monitoring reduce blood pressure? Meta-analysis with meta regression of randomised controlled trials. Ann Med 2010;42:371–86. Not available open access from the publisher. See Appendix 2.

McManus RJ, Mant J, Bray EP, Holder R, Jones MI, Greenfield S, et al. Telemonitoring and self-management in the control of hypertension (TASMINH2): a randomised controlled trial. Lancet 2010; 376:163–72. Not available open access from the publisher. See Appendix 2 for post-acceptance version of the paper.

Jones MI, Greenfield SM, Bray EP, Baral-Grant S, Hobbs FDR, Holder R, et al. Patients’ experiences of self-monitoring blood pressure and self-titration of medication: the TASMINH2 trial qualitative study. BJGP 2012;62:90–1. URL: http://bjgp.org/content/62/595/e135

Jones M, Greenfield S, Bray E, Hobbs F, Holder R, Little P, et al. Patient self-monitoring of blood pressure and self-titration of medication in primary care: the TASMINH2 trial qualitative study of health professionals’ experiences. Br J Gen Pract 2013;63:378–85. URL: http://bjgp.org/content/63/611/e378

O’Brien C, Bray EP, Stirling Bryan S, Greenfield SM, Haque MS, Hobbs FDR, et al. Targets and self-management for the control of blood pressure in stroke and at risk groups (TASMIN-SR): protocol for a randomised controlled trial. BMC Cardiovasc Dis 2013;13:21. URL: http://bmccardiovascdisord.biomedcentral.com/articles/10.1186/1471-2261-13-21

Kaambwa B, Bryan S, Jowett S, Mant J, Bray EP, Hobbs FDR, et al. Telemonitoring and self-management in the control of hypertension (TASMINH2): a cost-effectiveness analysis. Eur J Prevent Cardiol 2014;21:1517–30. URL: http://cpr.sagepub.com/content/21/12/1517.full

McManus RJ, Mant J, Haque MS, Bray EP, Bryan S, Greenfield SM, et al. Effect of self-monitoring and medication self-titration on systolic blood pressure in hypertensive patients at high risk of cardiovascular disease: the TASMIN-SR randomized clinical trial. JAMA 2014;312:799–808. URL: http://jama.jamanetwork.com/article.aspx?articleid=1899205

Penaloza-Ramos MC, Jowett S, Mant J Schwartz C, Bray EP, Haque S, et al. Cost-effectiveness of self-management of blood pressure in hypertensive patients over 70 years with suboptimal control and established cardiovascular disease or additional cardiovascular risk disease (TASMIN-SR) [published online ahead of print 24 November 2015]. Eur J Prevent Cardiol 2015.

Work programme 3

Fletcher K, Mant J, McManus RJ, Campbell S, Betts J, Taylor C, et al. Protocol for Past BP: a randomised controlled trial of different blood pressure targets for people with a history of stroke of transient ischaemic attack (TIA) in primary care. BMC Cardiovasc Dis 2010;10:37. URL: http://bmccardiovascdisord.biomedcentral.com/articles/10.1186/1471-2261-10-37

Mant J, McManus RJ, Roalfe A, Fletcher K, Taylor CJ, Martin U, et al. Different systolic blood pressure targets for people with a history of stroke or transient ischaemic attack, the PAST-BP (Prevention After Stroke – Blood Pressure) study: randomised controlled trial. BMJ 2016;352:i708.

Polypill acknowledgements

Self-management acknowledgements

Intensive blood pressure targets acknowledgements

The PAST-BP study is registered as International Standard Randomised Controlled Trial Number ISRCTN29062286.

Acknowledgements for the service development studies

Conference proceedings

Baral S. Who monitors blood pressure and/or glucose and how do they compare? Psychol Health 2008;23(Suppl. 1):61.

Jones MI. Training patients to self-manage hypertension in the TASMINH2 trial: blood pressure monitoring, telemonitoring and adjusting medication. Society for Academic Primary Care (SAPC), 2008, Galway, Ireland.

Vince EP. Doctor, is my blood pressure ok? A comparison of routine and systematically measured blood pressure. Society for Academic Primary Care (SAPC), 2008, Galway, Ireland.

Greenfield S. Blood Pressure Monitoring and the Development of the Home as a Formal ‘Health Space’. Society for Academic Primary Care South West Region (SWSAPC), 2009, Winchester, UK.

Jones M. ‘I Don’t Want to Increase My Medication’: A Qualitative Study of Patients’ Experiences of Self Management of Hypertension (TASMINH2 trial). Society for Academic Primary Care South West Region (SWSAPC) Conference, 2009, Winchester, UK.

Mant J. RCT of pharmacy delivered behavioural intervention to reduce BP after stroke/TIA in primary care (poster). National School for Primary Care Research National Conference, 2009, London, UK.

Bray EP. Blood Pressure and Beliefs: Does Self Management of Hypertension Affect Beliefs about Medicines? Society for Academic Primary Care (SAPC), 2010, Norwich, UK.

Bray EP. Blood Pressure and Beliefs: does self management of hypertension affect beliefs about medicines? Society for Academic Primary Care South West Region (SWSAPC), 2010, Oxford, UK.

De Simoni A. RCT of pharmacy delivered behavioural intervention to reduce BP after stroke/TIA in primary care. Society for Academic Primary Care (SAPC), 2010, Norwich, UK.

Mant J. Quality of blood pressure control in people with cerebrovascular disease in primary care: insights from the Prevention After Stroke – Blood Pressure (PAST-BP) randomised controlled trial. Society for Academic Primary Care (SAPC), 2010, Norwich, UK.

McManus RJ. A randomised controlled trial of Telemonitoring And Self-Management In The Control of Hypertension: Telemonitoring And Self-Management In Hypertension (TASMINH2). European Society of Hypertension, 2010, Oslo, Norway.

McManus RJ. A randomised controlled trial Of Telemonitoring And Self-Management In The Control Of Hypertension: Telemonitoring And Self-Management In Hypertension (TASMINH2). Society for Academic Primary Care (SAPC), 2010, Norwich, UK.

Jones MI. Training patients to self-manage hypertension in the TASMINH2 trial: blood pressure monitoring, telemonitoring and adjusting medication. Society for Academic Primary Care (SAPC), 2010, Norwich, UK.

Jones MI. A qualitative study of health professionals’ views of patient self-management of hypertension (TASMINH2 trial). Society for Academic Primary Care (SAPC), 2010, Norwich, UK. Also presented at Society for Primary Care South West Region (SWSAPC), 2010, Oxford, UK.

Jones MI. A qualitative study of health professionals’ views of patient self-management of hypertension (TASMINH2 trial). Society for Primary Care South West Region (SWSAPC), 2010, Oxford, UK.

Jones MI. Training patients to self-manage hypertension in the TASMINH2 trial: blood pressure monitoring, telemonitoring and adjusting medication. Society for Primary Care South West Region (SWSAPC), 2010, Oxford, UK.

Sheppard J. NHS Health Checks: why not treat those we know about first? Society for Academic Primary Care (SAPC), 2010, Norwich, UK.

Shroufi A. To what extent are NICE guidelines for the use of statins in primary prevention being implemented in general practice? John Glynn Trainees Prize Presentation. Royal Society of Medicine, 2010.

Shroufi A. The rule of halves is applicable to statins for the primary prevention of cardiovascular disease. Faculty of Public Health Annual Conference, 2010.

Virdee S. Healthcare attitudes to using a ‘polypill’ to manage cardiovascular risk – an interview study. Society for Academic Primary Care (SAPC), 2010, Norwich, UK.

Khan H. Beliefs about medicines among patients with cardiovascular disease – do socioeconomic factors and degree of risk influence beliefs? Society for Academic Primary Care (SAPC), 2012, Glasgow, UK.

Virdee SK, Greenfield SM, Fletcher K, Mant J. Patient attitudes towards the use of a polypill to manage cardiovascular risk. Society for Academic Primary Care (SAPC), 2012, Glasgow, UK.

Mant J. Intensive Blood Pressure Lowering: Randomised controlled trial of different systolic blood pressure targets for people with a history or stroke or transient ischaemic attack: the PAST-BP (Prevention After Stroke – Blood Pressure) study. Society for Academic Primary Care (SAPC), 2013, Nottingham, UK.

Mant J. Intensive Blood Pressure Lowering: Randomised controlled trial of different systolic blood pressure targets for people with a history or stroke or transient ischaemic attack: the PAST-BP (Prevention After Stroke – Blood Pressure) study. Primary Care Research Network National Conference, 2013.

Books/chapters

Hobbs R, McManus RJ, Taylor CJ, editors. Cardiovascular Disease in Primary Care. London: RCGP Publications; 2010. [Includes chapters on hypertension and other cardiovascular prevention (McManus) and stroke (Mant).]