OUP user menu

Is there a Failure to Optimize theRapy in anGina pEcToris (FORGET) study?

D.H.J. Elder , M. Pauriah , C.C. Lang , J. Shand , I.B.A. Menown , B.D.W.C.K. Sin , S. Gupta , S.G. Duckett , W. Foster , D. Zachariah , P.R. Kalra
DOI: http://dx.doi.org/10.1093/qjmed/hcq011 305-310 First published online: 24 February 2010


Background: In the management of chronic stable angina, percutaneous coronary intervention (PCI) provides symptomatic relief of angina rather than improvement of prognosis. Current guidelines recommend optimization of medical therapy prior to elective PCI. It is not clear if these guidelines are adhered to in clinical practice.

Aim: The aim of this multi-centre study was to determine the extent to which these treatment guidelines are being implemented in the UK.

Design: This was a multi-centre study involving six hospitals in the UK.

Methods: The medical treatment and extent of risk factor modification was recorded for consecutive patients undergoing elective PCI for chronic stable angina at each site. Data collected included anti-anginal drug therapy, lipid levels and blood pressure (BP). Data on heart rate (HR) control were also collected, since this represents a fundamental part of medical anti-anginal therapy. Target HR is <60 b.p.m. for symptomatic angina.

Results: A total of 500 patients [74% male; mean age ± SD (64.4 ± 10.1 years)] were included. When considering secondary prevention, 85% were receiving a statin and 76% were on an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker. In terms of medical anti-ischaemic therapy, 78% were receiving beta-blockers [mean equivalent dose of bisoprolol 3.1 mg (range 1.25–20 mg)], 11% a rate limiting calcium antagonist, 35% a nitrate or nicorandil and one patient was receiving ivabradine. The mean total cholesterol (95% confidence interval) was 4.3 mmol/l (4.2–4.4), mean systolic BP of 130 ± 24 mmHg and mean diastolic BP of 69 ± 13 mmHg. Serum cholesterol was <5 mmol/l in 77% and <4 mmol/l in 42% of the patients, 62% of the patients had systolic BP < 140 mmHg and 92% had diastolic BP < 90 mmHg. Considering European Society of Cardiology targets, 50% had systolic BP < 130 mmHg and 76% had diastolic BP < 80 mmHg. A large proportion of patients did not achieve target resting HR; 27% of patients had a resting HR of ⩾70 b.p.m., 40% had a resting HR between 60 and 69 b.p.m. and 26% had a resting HR between 50 and 59 b.p.m. The resting HR was not related to the dose of beta-blocker.

Conclusions: A significant proportion of the patients with chronic stable angina undergoing elective PCI did not achieve therapeutic targets for lipid, BP and HR control. Over 50% of patients did not receive adequate HR lowering anti-anginal therapy to achieve recommended target resting HR.


Percutaneous coronary intervention (PCI) can be an effective means of relieving symptomatic angina. While PCI is associated with improved prognosis in patients presenting with acute coronary syndromes,1 this is not necessarily the case for patients with stable angina pectoris.2 As such, guidelines from the European Society of Cardiology recommend that in stable angina, PCI should be considered after a trial of anti-anginal therapy.3 The guideline recommends initial use of a beta-blocker unless contraindicated.

Heart rate (HR) and systemic blood pressure (BP) play a key role in the development of myocardial ischaemia. Adequate control of these physiological variables is fundamental to optimizing medical therapy in patients with angina. The American College of Cardiology/American Heart Association Guidelines for the management of stable angina recommend a target resting HR of 55–60 b.p.m.4 With respect to Scotland, the SIGN guidelines does comment that resting HR of <60 b.p.m. is an indication of adequate beta-blockade. However, a recent survey of medical practitioners in the UK has suggested that HR is commonly ignored in patients with angina pectoris.5

Patients selected for elective PCI have manifest significant atherosclerotic disease. Appropriate secondary prevention is required to complement anti-anginal therapies and thereby reduce future cardiovascular risk. There are few data available that detail anti-anginal therapy and level of risk factor modification in patients undergoing elective PCI in the UK.

The decision to subsequently perform PCI in patients with stable angina is most commonly made at the time of diagnostic coronary angiography. Fast access chest pain clinics have significantly enhanced the investigation of patients with presumed stable angina (http://www.dh.gov.uk/en/Publicationsandstatistics/ Publications/PublicationsPolicyAndGuidance/DH_4094275); patients are seen within 2 weeks of referral. At this point, where appropriate, patients are referred for diagnostic angiography. Historically, there were long waits for these procedures. The advent of the 18-week pathway in England (http://www.dh.gov.uk/en/Publicationsandstatistics/Publications /PublicationsPolicyAndGuidance/DH_4134668) and marked reduction, in general, in the waiting times for angiography mean that the majority of patients now have their routine coronary angiogram within a few weeks of referral. While in general this will enhance the patient journey, it is plausible that decisions are made with respect to PCI prior to a trial of adequate medical therapy.

We hypothesized that medical treatment of patients with angina pectoris undergoing elective PCI in the UK would be suboptimal, and that this would be particularly so in relation to control of resting HR. To establish current use of anti-anginal therapy and the degree of risk factor modification, we undertook a cross-sectional study of patients undergoing elective PCI across a number of sites in the UK. It was not the intention of the study to assess the appropriateness or otherwise of the PCI.


The ‘Is there a Failure to Optimize theRapy in patients with anGina pecToris’ (FORGET) study was a multi-centre study that evaluated patients undergoing elective PCI for chronic stable angina to assess if they were on optimal medical therapy. It involved six hospitals across the UK, and included district general hospitals as well as tertiary referral centres that offered PCI. Local audit committee and/or Caldicott guardian approval was obtained for each participating site. Full Ethics Committee approval was obtained to permit amalgamation of the anonymized data sets.

Study population

Patients undergoing elective PCI for chronic stable angina were identified at each site, by reviewing the catheter laboratory databases to ascertain the reason for referral. Only patients identified as ‘stable angina’ were included in the study. Patients with acute coronary syndromes or those in whom a diagnosis of chronic stable angina was not clear were excluded. Each site identified between 40 and 100 consecutive patients.

Data collection

Baseline demographical data collected included age, sex and Canadian Cardiovascular Society (CCS) angina Class (Appendix 1). Details of medication included drug name and dose. To allow comparison between different drugs within the same class, the equivalent dose of beta-blocker and angiotensin converting-enzyme (ACE) inhibitor was calculated using bisoprolol and ramipril as the reference drugs, respectively.6,7

Non-invasive assessment of BP and resting HR were obtained from assessment at pre-clerking or admission. Catheter laboratory HR or BP was not utilized since these are likely to be influenced to a greater extent by enhanced sympathetic drive. Total serum cholesterol was obtained from the respective hospital laboratory computer systems.

Data analysis

Statistical analyses were performed using SPSS software package version 17 (SPSS Inc., Chicago, IL, USA). Data were analysed for normal distribution by looking at the mean, median and by visual inspection of the distribution. Comparisons between different groups were performed by one-way analysis of variance (ANOVA), and differences between groups were determined by using Bonferroni correction. For non-normally distributed data, a logarithmic transformation was applied to achieve a Gaussian distribution and ANOVA with Bonferroni correction was applied. Results for normally distributed data are presented as mean ± SD, while those that are not normally distributed data are presented as geometric mean [95% confidence interval (CI)]. For categorical data, Fisher’s exact tests were used to determine differences in proportions. All values are two-tailed and P < 0.05 was considered statistically significant.


Data were available for 500 patients across all hospitals. Seventy-four per cent of the patients were male and the mean age was 64.4 ± 10.1 years. The majority of patients were in CSS angina Class I (31%) with 51% Class II, 15% Class III and 3% in Class IV (Table 1).

View this table:
Table 1

Clinical characteristics and drug therapy according to CCS score

VariablesAll groupsClass IClass IIClass IIIClass IVP-value
Total number (%)500 (100)3151153
Age64.4 ± 10.163.3 ± 9.564.7 ± 10.264.1 ± 10.570.3 ± 10.20.13
Gender (% male)74727480670.12
Drug therapy (%)
    No anti-anginal therapy9127880.38
    One anti-anginal51574848420.25
    Two anti-anginals32243632500.04
    Three anti-anginals7671100.55
    Four anti-anginals111100.87
    Equivalent bisoprolol dose (mg)a3.1 (2.9–3.3)3.2 (2.8–3.6)3.0 (2.8–3.3)3.5 (3.0–4.0)3.7 (2.3–5.9)0.19
    DHP calcium antagonists1716172300.25
    Non-DHP calcium antagonists1111121180.98
    ACE inhibitors/ARB76657676580.08
Haemodynamic data
    Resting HR (b.p.m.)64 ± 1166 ± 1164 ± 1162 ± 1262 ± 110.14
    Systolic BP (mmHg)130 ± 24130 ± 25130 ± 24132 ± 24126 ± 260.17
    Diastolic BP (mmHg)69 ± 1371 ± 1469 ± 1369 ± 1365 ± 130.13
Blood results
    Cholesterol (mmol/l)a4.3 (4.2–4.4)4.3 (4.1–4.5)4.3 (4.2–4.5)4.2 (4.0–4.4)4.3 (3.8–4.8)0.88
  • DHP: dihydropyridine; ARB: angiotensin receptor blocker.

  • aData presented as geometric mean (95% CI).

Anti-anginal therapy and achievement of target resting HR

In terms of anti-anginal therapy, 51% were receiving a single agent, 32% were receiving two agents, 7% were receiving three agents and 1% of the patients were receiving four agents. Despite the presence of symptoms, 9% of the patients were not receiving specific anti-anginal therapy. Beta-blockers were the most commonly prescribed therapy (78%), followed by nitrates (21%). Dihydropyridine calcium channel blockers were prescribed for 17% of patients, while 11% received a rate-limiting calcium channel blocker.

When resting HRs were considered, 27% of patients had a resting HR of ⩾70 b.p.m., 40% had rate between 60 and 69 b.p.m. and 26% had rate between 50 and 59 b.p.m. A greater proportion of the patients with lower resting HR were receiving a beta-blocker. HR, however, was unrelated to the dose of the agent: geometric mean equivalent doses of bisoprolol were 2.9 mg (2.3–3.7; HR <50 b.p.m.); 2.8 mg (2.5–3.2; HR 50–59 b.p.m.); 3.4 mg (3.1–3.7; HR 60–69 b.p.m.); and 3.2 mg (2.8–3.7; HR ⩾ 70 b.p.m.) (P = 0.07). There was no correlation between HR and mean equivalent dose of bisoprolol (r = −0.6, P = 0.23). There was no difference in the use of rate-limiting calcium channel blockers across the HR groups (Table 2).

View this table:
Table 2

Clinical characteristics and drug therapy categorized by resting HR

Variable0–49 b.p.m.50–59 b.p.m.60–69 b.p.m.≥70 b.p.m.P-value
Percentage of patients7264027
Mean age66 ± 1165 ± 964 ± 1064 ± 110.82
Gender, male (%)867475710.15
CCS Class (%)
    Class I202732350.26
    Class II514956480.50
    Class III262011150.43
    Class IV34120.50
Drug therapy (% receiving)
    Equivalent bisoprolol dosea2.9 (2.3–3.7)2.8 (2.5–3.2)3.4 (3.1–3.7)3.2 (2.8–3.7)0.07
    DHP calcium antagonists171916170.93
    Non-DHP calcium antagonists17711130.29
    ACE/ARB inhibitor816873740.39
    Equivalent ramipril dosea3.1 (2.3–4.3)4.9 (4.2–5.7)3.5 (3.1–4.1)3.9 (3.1–5.0)0.04***
Haemodynamic data
    Systolic BP (mmHg)128 ± 25130 ± 24128 ± 22134 ± 270.22
    Diastolic BP (mmHg)67 ± 1469 ± 1369 ± 1371 ± 140.17
Blood results
    Total cholesterol (mmol/l)a4.5 (4.1–4.9)4.0 (3.8–4.2)4.3 (4.2–4.5)4.5 (4.3–4.7)0.001**
  • Unless otherwise stated, all results are presented as mean ± SD; **significant across Groups 2 and 3, 3 and 4. ***not significant across any two groups.

  • aData presented as geometric mean (95% CI).

Adherence to guidelines on cardiovascular risk management

Sixty-two per cent of patients had systolic BP < 140 mmHg and 92% had diastolic BP < 90 mmHg. With respect to the achievement of the European Society of Cardiology target for optimal BP control in patients with coronary artery disease,8 50% had systolic BP < 130 mmHg and 76% had diastolic BP < 80 mmHg. BP was similar across the different HR groups (Table 2). The geometric mean total cholesterol (95% CI) was 4.3 mmol/l (4.2–4.4) for the whole population; 42% had a total cholesterol concentration of <4 mmol/l and 77% had <5 mmol/l. There was high usage of statins (85%) and 76% were receiving an ACE inhibitor or angiotensin receptor blocker.


Our study shows that in various centres in the UK, patients with stable angina undergoing elective PCI are suboptimally treated. Despite the high usage of statins and ACE inhibitors or angiotensin receptor blockers, approximately one quarter of patients had a total cholesterol >5.0 mmol/l and systolic BP > 140 mmHg. HR received little attention with 67% of patients having a resting HR > 60 b.p.m. Beta-blocker dosage was no different in the groups divided according to HR, despite having similar levels of BP, suggesting that factors other than hypotension were involved in the prevention of the up-titration of beta-blockers, the mainstay of HR control in patients with angina.

To our knowledge, no such data have been published previously. The inclusion of real world data from a number of UK sites and regions is likely to enhance the applicability of the data as compared with that gathered from a single site.

The process by which a decision is made to undertake PCI in an individual patient is complex with a number of likely contributory factors. These will include patient age, co-morbidities such as anaemia, chronic kidney disease and presence of diabetes, the extent and anatomy of coronary artery disease, the cardiologist and patient preferences. The presence and severity of symptoms on medical therapy should be a key consideration according to the published data and guidelines.2,3 The European Society of Cardiology recommends that revascularization should be considered for persistent symptoms after a trial of dual anti-anginal therapy with optimized dosages.

These data show that in the UK a significant number of patients receiving elective PCI have relatively fewer symptoms, with just under a third (CCS Class I) remaining asymptomatic during ordinary activity. This is in keeping with data from the COURAGE study, where ∼12% of subjects were in CCS Class 0 and 30% in CCS Class 1 at baseline.2 Anti-anginal therapy appears underutilized with the majority of patients receiving a single agent and a significant minority receiving no therapy. While 78% of patients received a beta-blocker, many patients did not achieve the recommended HR control; 67% of patients had resting HRs >60 b.p.m. Resting HR of ⩾70 b.p.m., a level shown to be associated with adverse cardiovascular outcome in patients with ischaemic heart disease,9 was present in 27% of individuals. A previous survey highlighted that the majority of clinicians (primary and secondary care) felt that control of HR was important in the management of angina.5 Yet, the majority did not specifically target therapy towards it. While we do not have data regarding HR at initial presentation with symptoms, our study suggests that therapy is not routinely titrated to lower HR even in patients referred for PCI.

There may be a number of reasons why anti-anginal therapy is underutilized. Patients may be intolerant of certain drugs or develop side effects with higher doses. This study provides a snap shot of care, and as such it is plausible that some patients had been tried on more intense anti-anginal drug regimes prior to consideration of PCI. The timing of angiography is likely to be of substantial influence. The decision to refer for elective PCI is generally made at the time of characterization of coronary anatomy. Historically, waits for diagnostic angiography were long in the UK, giving ample opportunity for intensification of drug therapy in those patients who remained symptomatic. While a marked reduction in waiting times for angiography has provided major benefits in the clinical care of patients with cardiovascular disease, it may actually negatively impact on the opportunity to up-titrate anti-anginal therapy. This is particularly pertinent in the increasing trend to immediately follow on to PCI at the time of a diagnostic angiogram (i.e. a single procedure).

Secondary prevention on the whole was suboptimal with the exception of ACE inhibitor/angiotensin receptor blocker use. Only 42% of the study population had a total cholesterol concentration of <4 mmol/l. Arguably, it is likely that further changes in medication would be made following PCI. It should, however, be noted that the geometric mean cholesterol (95% CI) [4.3 (4.2–4.4) mmol/l] and BP (130 ± 24/69 ± 13 mmHg) were very similar to those found at randomization in the COURAGE study (mean cholesterol ∼4.5 mmol/l, BP ∼130/74 mmHg). Recent data supporting potential benefits of aggressive cholesterol reduction on plaque stability may influence management of this risk factor around the time of PCI in due course.

This study provides a cross-sectional overview of the quality of medical care for patients who receive elective PCI for stable symptoms. It was not designed to comment on the nature of disease or look at the outcome. It did not assess whether other anti-anginal therapies had been tried. While this study suggests that HR control is suboptimal, it was not designed to identify the reasons why this happens. It is, however, clear that factors other than hypotension were involved. However, we have not assessed whether patients were intolerant to increasing doses of beta-blockers (e.g. due to fatigue or concomitant medical problems). We have not assessed whether left ventricular dysfunction could have limited the use of other negatively inotropic agents. Additionally, neither have we examined the role of non-invasive tests in referrals for PCI nor did we determine the original source of referral, although all these cases were performed at the request of a cardiologist. The fact that ivabradine use was very low would seem to suggest that HR control was genuinely given little attention. This would be in keeping with the survey of medical practitioners in the UK.5


Patients undergoing elective PCI in the UK receive suboptimal secondary prevention and anti-anginal therapy is underutilized according to the current guidelines. This is particularly so in respect of HR control. While this may be consequent of a number of factors, changes in health policy such as introduction of the 18-week pathway in England may play an important role. A trial of titrated medical therapy may be appropriate for many lower risk patients with stable angina.

Conflicts of interest: C.C.L., I.B.M., S.G. and P.R.K. are part of a National HR Research Group which has been funded by Servier UK. No funding has been received for this study.

Appendix 1

CCS class definitions

CCS classSymptoms
Class 0Asymptomatic
Class 1Angina with strenuous exercise
Class 2Angina with moderate exertion
Class 3Angina with mild exertion i.e. walking 1–2 level blocks at normal pace and climbing a flight of stairs at normal pace
Class 4Angina at any level of physical exertion


View Abstract