Report of the Third Working Party of the British Hypertension Society

Journal of Human Hypertension (1999) 13, 569–592  1999 Stockton Press. All rights reserved 0950-9240/99 $15.00 http://www.stockton-press.co.uk/jhh BHS GUIDELINES Guidelines for management of hypertension: report of the third working party of the British Hypertension Society

LE Ramsay, B Williams, GD Johnston, GA MacGregor, L Poston, JF Potter, NR Poulter and G Russell for the British Hypertension Society

¼ Use non-pharmacological measures in all hyperten- as first-line therapy for the majority of hypertensive sive and borderline hypertensive people. people. In the absence of compelling indications for ¼ Initiate antihypertensive drug therapy in people with beta-blockade, diuretics or long acting dihydropyrid- sustained systolic blood pressures (BP) у160 mm Hg ine calcium antagonists are preferred to beta-blockers or sustained diastolic BP у100 mm Hg. in older subjects. Compelling indications and contra- ¼ Decide on treatment in people with sustained systolic indications for all antihypertensive drug classes are BP between 140 and 159 mm Hg or sustained diastolic specified. BP between 90 and 99 mm Hg according to the pres- ¼ For most hypertensives, a combination of antihyper- ence or absence of target organ damage, cardio- tensive drugs will be required to achieve the rec- vascular disease or a 10-year coronary heart disease ommended targets for blood pressure control. у (CHD) risk of 15% according to the Joint British ¼ Other drugs that reduce cardiovascular risk must also Societies CHD risk assessment programme/risk chart. be considered. These include aspirin for secondary ¼ In people with diabetes mellitus, initiate antihyperten- у prevention of cardiovascular disease, and primary sive drug therapy if systolic BP is sustained 140 prevention in treated hypertensive subjects over the mm Hg or diastolic BP is sustained у90 mm Hg. у ¼ age of 50 years who have a 10-year CHD risk 15% In non-diabetic hypertensive people, optimal BP treat- and in whom blood pressure is controlled to the audit ment targets are: systolic BP Ͻ140 mm Hg and dias- standard. In accordance with existing British rec- tolic BP Ͻ85 mm Hg. The minimum acceptable level of ommendations, statin therapy is recommended for control (Audit Standard) recommended is Ͻ150/Ͻ90 hypertensive people with a total cholesterol mm Hg. Despite best practice, these levels will be dif- у5 mmol/L and established vascular disease, or 10- ficult to achieve in some hypertensive people. у ¼ In diabetic hypertensive people, optimal BP targets year CHD risk 30% estimated from the Joint British are; systolic BP Ͻ140 mm Hg and diastolic BP Ͻ80 Societies CHD risk chart. Glycaemic control should mm Hg. The minimum acceptable level of control also be optimised in diabetic subjects. ¼ (Audit Standard) recommended is Ͻ140/Ͻ90 mm Hg. Specific advice is given on the management of hyper- Despite best practice, these levels will be difficult to tension in specific patient groups, ie, the elderly, achieve in some people with diabetes and hyperten- ethnic subgroups, diabetes mellitus, chronic renal sion. disease and in women (pregnancy, oral contraceptive ¼ In the absence of contraindications or compelling use and hormone replacement therapy). indications for other antihypertensive agents, low ¼ Suggestions for the implementation and audit of dose thiazide diuretics or beta-blockers are preferred these guidelines in primary care are provided.

Keywords: BHS; management of hypertension

Introduction patients3–7; treatment of isolated systolic hyperten- sion in the elderly8; comparison of the antihyperten- These guidelines update previous reports by work- sive efficacy and tolerability of different classes of ing parties of the British Hypertension Society in drug9–11; and the role of non-pharmacological meas- 1 2 1989 and 1993. Since the 1993 guidelines much ures in the prevention12–15 and treatment16 of hyper- new evidence has emerged, notably on optimal tension. There has been a vigorous debate about the blood pressure targets during antihypertensive treat- 17,18 3 safety of dihydropyridine calcium antagonists, ment ; management of hypertension in diabetic but also new evidence from randomised controlled trials for their efficacy and safety, particularly in iso- lated systolic hypertension in the elderly.8 These Correspondence: Professor Bryan Williams, Cardiovascular important additions to an already formidable body Research Institute, University of Leicester, Sir Robert Kilpatrick Building, PO Box 65, Leicester Royal Infirmary, Leicester LE2 of evidence are very welcome and provide the basis 7LX, UK for these new recommendations from the British Received and accepted 7 July 1999 Hypertension Society. Guidelines for management of hypertension British Hypertension Society 570 On the whole, physicians report that they adhered to previous recommendations by the British Hyper- BOX 1. Categories of Strength Used in Statements 19 (based on North of England evidence based guidelines, tension Society, but with some important excep- BMJ 1998)53 tions. They are less aware of, or less inclined to implement, recommendations to treat mild hyper- Strength of evidence tension and isolated systolic hypertension in the la-Evidence from meta-analysis of randomised controlled 19,20 trials elderly. National and international surveys con- Ib-Evidence from at least one randomised controlled trial tinue to reveal that there is under-diagnosis of IIa-Evidence from at least one controlled study without hypertension, that those diagnosed as hypertensive randomisation often do not continue on treatment, and that those IIb-Evidence from at least one other type of quasi- treated are often not controlled satisfactorily.21–23 experimental study III-Evidence from descriptive studies, such as The situation has improved in recent years but in comparative studies, correlation studies, and case- general the management of hypertension in the controlled studies United Kingdom remains suboptimal.23 IV-Evidence from expert committee reports or opinions Also disturbing is further evidence that conven- or clinical experience of respected authorities, or both tional management of hypertension leaves patients Strength of recommendation at an unacceptably high risk of cardiovascular com- A-Directly based on category I evidence plications and death, particularly from coronary B-Directly based on category II evidence or extrapolated heart disease (CHD) but also from stroke.24–29 In part recommendation from category I evidence this is a consequence of suboptimal blood pressure C-Directly based on category III evidence or extrapolated 30 recommendation from category I or II evidence control, but other factors are also important. In a D-Directly based on category IV evidence or extrapolated recent study, the persistent excess of CHD events in recommendation from category I, II or III evidence treated hypertensive subjects was predicted by three factors; (i) evidence of target organ damage before treatment, (ii) a history of cigarette smoking before treatment, and (iii) the serum cholesterol values Societies ‘Cardiac Risk Assessor’ computer pro- before and during treatment.29 These observations gramme. Consequently, whilst acknowledging that support the concept that effective management of CVD prevention is the proper focus of hypertension hypertension requires the identification of those at management, the levels of CHD risk quoted in these highest cardiovascular risk and the adoption of guidelines, appropriately precipitate intervention multifactorial intervention, targeting not only blood for those at higher CVD risk. pressure levels, but also associated cardiovascular These guidelines are intended for general prac- risk factors. These new guidelines embrace this con- titioners, practice nurses, and generalists in hospital cept and provide detailed guidance on the manage- practice, and aim to present as clearly as possible, ment of hypertension and associated cardiovascular the best currently available evidence on hyperten- risk factors. sion management. The evidence supporting the rec- The recent trials of statins31–34 and aspirin3,35 for ommendations contained in these new British the prevention of CHD are important steps forward Hypertension Society guidelines is graded using the for hypertension management since the previous North of England Group Criteria53 (Box 1). The guidelines. A significant proportion of hypertensive guidelines should be applied with due regard to patients will benefit from aspirin and statin treat- local circumstances and policies, and with appropri- ment, even if these treatments are only targeted at ate clinical judgement as regards the needs of indi- those with a high level of CHD risk.36,37 Formal esti- vidual patients (Box 2). mation of CHD risk has been proposed as an aid to 38–45 treatment decisions in hypertension, and Blood pressure measurement (Box 3) debated.46–49 This estimation ideally entails coun- ting and weighting major cardiovascular risk factors All adults should have blood pressure measured in addition to blood pressure itself.50 Mindful of the routinely at least every 5 years until the age of 80 strong relationship between blood pressure and the years. Those with high-normal values (135–139/85– risk of stroke, the Society acknowledges that tar- 89 mm Hg) and those who have had high readings geting cardiovascular disease risk (CVD) rather than at any time previously should have blood pressure CHD risk is preferable. However, in order to be con- re-measured annually. The British Hypertension sistent with three existing National guideline rec- Society recommendations for measuring blood ommendations,50–52 we recommend formal esti- pressure,54 now available on CD-ROM (see Appen- mation of 10-year CHD risk using the computer dix 2), should be followed (Table 1). Seated blood programme ‘Cardiac Risk Assessor’ or the CHD risk pressure recordings are generally sufficient, but chart issued by the Joint British Societies in their standing blood pressure should be measured in eld- recommendation for the prevention of CHD.50 It is erly or diabetic patients to exclude orthostatic hypo- reasonable to make this pragmatic recommendation tension. When assessing CHD/CVD risk the average because CHD risk is a good predictor of CVD risk of several measurements at separate visits is more which can be estimated by multiplying the esti- accurate than measurements taken at a single visit.55 mated 10-year CHD risk level by 4/3 (eg, 30% CHD In uncomplicated mild hypertension the average of risk Ϸ40% CVD risk) [see Appendix 1]. Moreover, two readings per visit at monthly intervals over individualised estimates of 10-year stroke risk as 4–6 months should be used to guide the decision to well as CHD risk are provided by the Joint British treat. In more severe hypertension prolonged Guidelines for management of hypertension British Hypertension Society 571 observation before treatment is not necessary or war- BOX 2. Guidelines ranted. The average blood pressure is only one factor ¼ These guidelines, which are an update of previous determining cardiovascular risk in uncomplicated reports in 1989 and 1993 (B), are based on collective mild hypertension. Formal estimation of CHD/CVD expert interpretation of current clinical evidence as risk needs consideration of age, sex, smoking habit, assessed by members of the British Hypertension diabetes, total:HDL cholesterol ratio, and family his- Society. tory in addition to blood pressure,50 as described ¼ Important new inclusions are optimal blood pressure later. targets (A), management of hypertension in diabetic patients (A), treatment of isolated systolic hypertension in the elderly (A), comparison of Systolic or diastolic blood pressure? different classes of drug therapy (A) and the role of non-pharmacological treatment (A). There has been much debate on the relative impor- tance of systolic and diastolic blood pressure, but in ¼ In view of clear evidence for other measures to reduce practice systolic blood pressure should be regarded cardiovascular risk; lowering cholesterol (A), use of aspirin (A), treatment of diabetes (A) and as the more important. In general, systolic and dias- discontinuation of smoking (B), formal estimation of tolic blood pressure correlate highly, and in epide- risk and risk thresholds for treatment have been miological studies both are important risk factors for included (B) see Appendix 1. cardiovascular disease.56 Outcome trials of antihy- ¼ pertensive treatment based on thresholds of dias- There is evidence that current guidelines are not being 57,58 8,59 implemented and the detection and management of tolic or systolic blood pressure have shown hypertension remains sub-optimal (B). There is a need similar reductions in cardiovascular events. This to improve the quality of care by better screening, has important implications with regard to blood protocols to support clinical decision making and pressure thresholds for the treatment of hyperten- increased education of doctors, nurses and patients. sion and blood pressure targets during treatment. When treatment is recommended at a blood pressure threshold of 140/90 mm Hg, this means 140 mm Hg systolic or 90 mm Hg diastolic. An optimal blood pressure target of Ͻ140/85 mm Hg means Ͻ140 BOX 3. Blood Pressure Measurement mm Hg systolic and Ͻ85 mm Hg diastolic.

¼ Use the British Hypertension Society recommendations (C). Ambulatory blood pressure (ABPM)

¼ Use a device with validated accuracy that is properly All outcome trials in hypertension have been based maintained and calibrated. on surgery or clinic blood pressure, not ABPM, and it is therefore difficult to provide firm guidance ¼ Patient should be seated with the arm at the level of the heart. The bladder size should be adjusted for the based on evidence for use of this technique. Never- arm circumference, the cuff deflated at 2 mm/sec and theless, ABPM is widely used and may be valuable the blood pressure measured to the nearest 2 mm Hg. in special circumstances. ABPM provides numerous Diastolic pressure is recorded as disappearance of the measurements over a short time, and so reduces sounds (phase V). variability when compared to the average of a lim- 60,61 ¼ At least two measurements should be made at each ited number of surgery or clinic readings. Blood visit and four visits to determine blood pressure pressure by ABPM correlates more closely with evi- thresholds. Possible indications for ‘home’ or dence of target organ damage,60–63 presumably in ambulatory blood pressure monitoring include the part because of reduced variability and measure- diagnosis of ‘white coat hypertension’, suspected 60,61 64 hypotension, excessive blood pressure variability and ment error. In one controlled trial treatment resistance to drug therapy (C). based on ABPM rather than surgery or clinic read- ings resulted in less drug treatment, but at the expense of slightly but significantly higher ABPM and clinic blood pressures. There was no difference in well-being or cost-effectiveness. Incorporation of Table 1 Blood pressure measurement ABPM measurements in formal CHD/CVD risk esti- mation has little effect on risk, because blood press- ¼ Follow BHS guidelines on technique.55 ure is only one of several risk factors for cardio- ¼ Use device with validated accuracy, that is properly vascular complications.65 maintained and calibrated. ¼ Measure sitting BP routinely: standing BP in elderly or ABPM may be indicated in the following circum- diabetic patients. stances: ¼ Remove tight clothing, support arm at heart level, ensure hand relaxed. • when blood pressure shows unusual variability; ¼ Use cuff of appropriate size. • in hypertension resistant to drug therapy, defined ¼ Lower mercury slowly, by 2 mm per second. as blood pressure Ͼ150/90 mm Hg on a regimen ¼ Read BP to the nearest 2 mm Hg. ¼ of three or more antihypertensive drugs; Measure diastolic as disappearance of sounds (phase 5). • ¼ Take two measurements at each visit. when symptoms suggest the possibility of hypo- ¼ Use the average for several visits when estimating tension; cardiovascular risk in mild hypertension. • to diagnose white coat hypertension.66 The term ‘white coat hypertension’ is widely used to Guidelines for management of hypertension British Hypertension Society 572 describe consistent hypertension in the clinic tensive treatment, and normal blood pressure values with consistent normotension by ABPM. There is by ABPM may alter the treatment decision. However a systematic clinic-ABPM difference in the popu- a decision to withhold treatment in such patients lation that is related to the level of clinic blood should be based on appropriately-adjusted normal pressure,66,67 and white coat hypertension is con- values for ABPM (see below), and should be con- sidered to be present only when the clinic-ABPM firmed by a second ABPM record because of within- difference exceeds the population average differ- patient variability and limited reproducibility.67 ence. Furthermore patients left untreated on the basis of ABPM will need to be followed up, with reassess- It is not necessary or feasible to perform ABPM to ment of blood pressure and cardiovascular risk at exclude white coat hypertension in all hypertensive least once a year. The annual reassessment may patients. It is not indicated in patients who are at require repeated ABPM measurement. high CHD/CVD risk. This includes patients who Some important points in interpreting the results already have target organ damage or cardiovascular of ABPM records need emphasis. The average day- complications (Table 2), and those who have an esti- time blood pressure should be used for treatment mated 10-year CHD risk of 15% or higher. In these decisions, not the average 24-h blood pressure. patients treatment decisions should be based on sur- Blood pressure measured by ABPM is systematically gery or clinic pressures rather than ABPM, as was lower than surgery or clinic measurements in hyper- the case in outcome trials of hypertension treatment. tensive and normotensive people.67,68 Because of ABPM is also unnecessary in patients with mild this, treatment thresholds and targets must be hypertension (140–159/90–99 mm Hg) with no tar- adjusted downwards when making decisions based get organ damage, no cardiovascular complications, on ABPM data. Precise adjustment is complex, but and an estimated 10-year CHD risk is Ͻ15%.65 These the average difference between clinic and daytime patients may be left untreated without using ABPM mean pressures determined by ABPM is approxi- but must be followed up. mately 12/7 mm Hg.65,67,68 Thus an ABPM average ABPM may alter management when the average daytime blood pressure of 148/83 mm Hg is approxi- clinic blood pressure is у160/100 mm Hg, there is mately equivalent to a surgery blood pressure of no target organ damage or cardiovascular compli- 160/90 mm Hg, and this may require treatment in cations, and the estimated 10-year CHD risk is some patients. Recommended targets for ABPM Ͻ15%. Here elevated blood pressure is the only measurements are given in Table 3. These are indication of high CHD/CVD risk, and for antihyper- extrapolated from clinic or surgery thresholds and targets derived from controlled trials using conven- Table 2 Initial evaluation of the hypertensive patient tional blood pressure measurements.

¼ Causes of hypertension: Blood pressure measurement at home –drugs (NSAID’s, oral contraceptive, steroids, liquorice, sympathomimetics, ie. some cold cures). This technique is less expensive and more con- –renal disease (present, past or family history: palpable venient for patients than ABPM. Evidence on the kidney(s) – polycystic, hydronephrosis or neoplasm). –renovascular disease (abdominal or loin bruit). role of self-measurement of blood pressure is less –phaeochromocytoma (paroxysmal symptoms). extensive than for ABPM, but many of the same con- –Conn’s syndrome (tetany, muscle weakness, polyuria). siderations apply. In particular, measurements made –coarctation (delayed or weak femoral pulses). at home need to be ‘adjusted’ upwards by approxi- –Cushings (general appearance). mately 12/7 mm Hg for equivalence to surgery or ¼ Contributory factors: clinic measurements when making treatment –overweight decisions.69 Information on the validity of different –excess alcohol (Ͼ3 units/day) home blood pressure monitoring devices can be –salt intake obtained from the British Hypertension Society –lack of exercise Information Service (see Appendix 2). ¼ Complications of hypertension/target organ damage: –stroke, TIA, dementia Table 3 Suggested target blood pressures during antihypertensive –LVH, heart failure treatment. Systolic and diastolic should both be attained, eg –myocardial infarct, angina, CABG or angioplasty Ͻ140/85 mm Hg means less than 140 systolic and less than 85 –peripheral vascular disease diastolic –fundal hemorrhages or exudates –proteinuria Clinic BP Mean day-time ABPM or –renal impairment (mm Hg) home BP ¼ Cardiovascular risk factors: –smoking No diabetes Diabetes No diabetes Diabetes –diabetes –total – cholesterol:HDL-cholesterol ratio Optimal BP: Ͻ140/85 Ͻ140/80 Ͻ130/80 Ͻ130/75 –family history Audit Ͻ150/90 Ͻ140/85 Ͻ140/85 Ͻ140/80 –age Standard:a –sex The Audit Standarda reflects the minimum recommended levels ¼ Contraindications to drugs: of blood pressure control. See Table 5 Despite best practice, the Audit Standard will not be achievable in all treated hypertensives. Guidelines for management of hypertension British Hypertension Society 573 Accuracy of sphygmomanometers Established hypertension A properly maintained mercury sphygmoman- Recent controlled trials12,13,15,16,78–83 have confirmed ometer is the traditional accurate and robust gold that changes in diet and lifestyle do lower blood standard for routine clinical practice, but for health pressure and may also reduce cardiovascular risk. and safety reasons the use of mercury devices is Clear verbal and written advice on the measures declining and will eventually cease completely. below should be provided for all hypertensive There will be increasing use of alternative methods patients and also for those with high-normal blood such as aneroid, semi-automated and automated pressure or a strong family history. They may lower devices. There is evidence that many alternatives to blood pressure as much as drug monotherapy; the mercury sphygmomanometer are unacceptably reduce the need for drug therapy12,16; enhance the inaccurate when they are subjected to formal vali- antihypertensive effect of drugs; reduce the need for dation.70 Those purchasing or using alternatives to multiple drug regimens; and favourably influence the mercury sphygmomanometer should insist on overall cardiovascular risk. Conversely, failure to evidence from the manufacturer that the device has adopt these measures may attenuate the response to been validated formally, and proved accurate, antihypertensive drugs.84 according to the standards of the British Hyperten- (1) Measures that lower blood pressure: sion Society protocol71 or American National Stan- – weight reduction dards.72,73 Information on the validity of different – reduced salt intake devices can be obtained from the British Hyperten- – limitation of alcohol consumption sion Society Information Service (see Appendix 2). – physical exercise – increased fruit and vegetable consumption Non-pharmacological measures (Box 4) – reduced total fat and saturated fat intake. (2) Measures to reduce cardiovascular risk: Primary prevention of hypertension – stop smoking; The current strategy for preventing cardiovascular – replace saturated fat with polyunsaturated complications associated with hypertension is and monounsaturated fats; unsatisfactory in that it requires the detection and – increase oily fish consumption. lifelong drug therapy of a large proportion of the – reduce total fat intake. adult population including about one half of all eld- In patients with mild hypertension, but no cardio- erly people. Moreover, such management as cur- vascular complications or target organ damage, the rently practised does not reduce the cardiovascular response to these measures should be observed dur- risk of hypertensive patients to that of normotensive ing the initial 4–6 month period of evaluation. subjects, as discussed earlier. There is increasing When drug therapy has to be introduced more evidence that a population strategy could prevent quickly, for example in patients with severe hyper- the rise in blood pressure with age, reduce the tension, non-pharmacological measures should be prevalence of hypertension and need for drug ther- instituted in parallel with drug treatment. apy, and reduce overall cardiovascular risk.74–77 The requirements for this strategy have been summarised Weight reduction by calorie restriction is appropri- by Stamler77 as a diet high in fruit and vegetables; ate for the majority of hypertensive patients because high in legumes and whole grains; high in fat-free most are overweight, and results in blood pressure and low-fat dairy, poultry, fish, shellfish, and meat reduction12,16 of about 2.5/1.5 mm Hg for each kilog- products; high in all essential nutrients; reduced in ram lost. salt; reduced in total fat, saturated fat and choles- Salt reduction from an average of 10 to 5 grams (5 terol; with no more than 2–3 units of alcohol per grams Ϸ1 teaspoon) daily lowers blood pressure by day; and controlled in calories to prevent or cor- about 5/3 mm Hg,16,85,86 with larger blood pressure rect obesity. falls in the elderly and those with higher initial blood pressure levels.86 All hypertensive patients should have clear verbal and written advice to BOX 4. Non-Pharmacological Measures reduce salt intake to 5 grams per day. Many will ¼ Non-pharmacological measures; weight reduction (A), have already discontinued adding salt at the table reduced salt intake (A), reduced fat intake (A), limited and even when cooking, but few are aware of the alcohol consumption (A), dynamic exercise (A) and large amounts of salt in processed foods, such as increased fruit and vegetable consumption (A) are bread (one slice contains 0.5 grams of salt), some effective in lowering blood pressure. breakfast cereals, and flavour enhancers such as ¼ Alone or in combination these interventions can stock cubes or manufactured sauces. Patients, and reduce the need for drug therapy and enhance the those who cook for patients, should be provided effect of antihypertensive agents (A). A favourable with specific written advice (see Appendix 2). effect on cardiovascular outcome is assumed but not proven. Alcohol intake above 21 units per week is associated with blood pressure elevation that is reversible by ¼ To reduce overall cardiovascular risk, patients should reducing the intake.87 Binge drinking is associated stop smoking (B), reduce their saturated fat intake and 88 increase consumption of poly-unsaturated, mono- with an increased risk of stroke. Hypertensive unsaturated fats and oily fish (B). patients should be advised to limit their alcohol intake to 21 units per week for men, and 14 units Guidelines for management of hypertension British Hypertension Society 574 per week for women. However in a recent study this useful to summarise the aims, which are to elicit had a limited effect as a single intervention, because and document: 14 alcohol intake was little reduced. Consumption of • smaller amounts of alcohol, up to the recommended causes of hypertension, eg, renal disease, endo- 89 crine causes; limit, may protect against CHD and should not • be discouraged. contributory factors, eg, obesity, salt intake, excess alcohol intake; Exercise should be regular; dynamic (eg, brisk • complications of hypertension, eg, previous walking) rather than isometric (eg, weight training); stroke, left ventricular hypertrophy; and tailored to the individual patient.13 For • cardiovascular risk factors, eg, smoking, family example, three vigorous training sessions per week history; may be appropriate for fit younger patients82 or brisk • contraindications to specific drugs, eg, asthma walking for 20 min per day for older patients.81,90 (beta-blockers), gout (thiazides). Increased fruit and vegetable consumption is now These aspects are summarised in more detail in supported by controlled-trial evidence showing that Table 2. an increase from two to seven portions daily low- Routine investigation should be limited to: ered blood pressure significantly in hypertensive patients, by 7/3 mm Hg.12 This effect of fruit and • urine strip test for protein and blood; vegetables on blood pressure may be a consequence • serum creatinine and electrolytes; of increased potassium intake.78,91 Hypertensive • blood glucose; patients should have clear advice on increasing fruit • serum total: HDL cholesterol; and vegetable intake.92 When this is combined with • ECG. an increase in low-fat dairy products and reduction Note that chest X-ray, urine microscopy and cul- of saturated and total fat, blood pressure falls may ture, and echocardiography are not required rou- be larger, averaging 11/6 mm Hg in hypertensive tinely. An echocardiogram is valuable to confirm or patients and 4/2 mm Hg in those with high-normal refute the presence of left ventricular hypertrophy blood pressure.12 when the ECG shows ‘high’ left ventricular voltage Cigarette smoking increases cardiovascular risk without T-wave abnormalities, as is often the case more than mild hypertension, and smoking was one in young patients. When the clinical evaluation or factor related to the persistent increase in coronary results of these simple investigations suggests a mortality in men with treated hypertension.29 need for further investigation it is usually best to Hypertensive patients who smoke should be given refer for specialist advice, because the additional advice and help to stop smoking. The use of nicotine investigations needed are often difficult to arrange replacement therapies approximately doubles smok- from general practice. Indications for referral for ing cessation rates.93 specialist advice or treatment are suggested in Table 4. Serum cholesterol before and during the treatment of hypertension is also an important predictor of car- diovascular disease.29 All patients should be advised to reduce saturated fat and cholesterol intake, and to substitute polyunsaturated and mono- Table 4 Suggested indications for specialist referral unsaturated fats. These diet changes will reduce ¼ 94 Urgent treatment needed: serum cholesterol by an average of 6%, but it is –accelerated (malignant) hypertension important to recognise the difficulties in –severe hypertension (eg, Ͼ220/120 mm Hg) implementing and sustaining these measures.95 –impending complications (eg, TIA, left ventricular failure) Many hypertensive patients are at very high risk of ¼ 36,37 Possible underlying cause: CHD and will need aspirin and statin treatment –any clue in history or examination of a secondary cause in addition to non-pharmacological measures if they –hypokalaemia/increased plasma sodium (Conn’s are to have adequate coronary prevention. syndrome?) –elevated serum creatinine Effective implementation of these non-pharmaco- –proteinuria or haematuria logical measures requires enthusiasm, knowledge, –recent onset or worsening of hypertension patience, and considerable time spent with patients –resistant to a three drug regimen –young age (any hypertension Ͻ20 years; needing treatment and other family members. It is best undertaken by Ͻ30 years) well-trained health professionals, eg, practice or clinic nurse, and should be backed up by simple ¼ Therapeutic problems: clear written information (see Appendix 2). –treatment resistance –multiple drug intolerance –multiple drug contraindications –persistent non-compliance Evaluation of hypertensive patients –treatment declined (the reluctant hypertensive) All hypertensive patients should have a thorough ¼ Special situations history and physical examination, but need only a –unusual BP variability limited number of routine investigations. It is –possible isolated clinic hypertension beyond the scope of these guidelines to discuss –hypertension in pregnancy every detail of the clinical evaluation, but it may be Guidelines for management of hypertension British Hypertension Society 575 Thresholds for antihypertensive drug be used to calculate stroke risk over the same period. therapy (Box 5) The use of a computer to estimate CHD risk may not be practical in all clinical settings. In such circum- Absolute cardiovascular risk stances, we recommend the use of the ‘coronary risk The importance of absolute cardiovascular risk was chart’ issued by the Joint British Societies (Figure 1). recognised in the previous British Hypertension This chart is also based on the Framingham risk Society guidelines,2 which advised early drug treat- function and uses colour-coded bands to specify three levels of 10-year CHD risk; у30%, у15% and ment of patients with more severe hypertension р (у200/110 mm Hg), treatment of sustained blood 15%. The practicalities of how to measure CHD pressure у160/100 mm Hg, and treatment of risk using the Joint British Societies computer pro- patients with diastolic pressure 90–99 mm Hg who gramme or risk chart are detailed in Appendix 1 had particularly high cardiovascular risk. Patients and figure 2. with cardiovascular complications (eg, previous Targeting antihypertensive treatment at absolute stroke or coronary disease) or target organ damage CHD/CVD risk is underpinned by evidence from (eg, left ventricular hypertrophy) were recognised as meta-analyses of outcome trials which show that the having cardiovascular risk sufficiently high to war- relative risk reduction by antihypertensive treat- rant treatment of even mild hypertension, for ment is approximately constant, with a 38% example 140/90 mm Hg. These recommendations reduction in stroke and 16% reduction in coronary events.57,58 Benefit from treatment can be expressed remain sound, and are not altered. 99 A difficulty acknowledged in the previous guide- as the number needed to treat (NNT) over 5 years. lines concerned the treatment decision for patients In patients with mild hypertension treatment with ‘mild’ hypertension, averaging 140–159/90–99 reduces cardiovascular complications by approxi- mately 25%,44,55 and treatment of patients at a 10- mm Hg, who were at variable risk depending on у у other risk factors. Advice to treat, or to leave year CHD risk of 15% (CVD risk of 20%) corre- untreated and observe, was based on the presence of sponds to a NNT for 5 years of 40. This means treat- additional cardiovascular risk factors including age, ment of 40 patients for 5 years to prevent one cardio- male sex, smoking, serum lipids, and family history. vascular complication. Formal estimation of However no formal method to estimate cardiovascu- CHD/CVD risk also informs decisions on the treat- lar risk using these risk factors was provided. Intuit- ment of hypertensive people with aspirin or statins, ive estimates of absolute risk are very inaccurate.96,97 as discussed below. Risk estimation is improved when additional risk It is recommended that all patients with average factors are simply counted,98 but is significantly blood pressure 140–159 or 90–99 mm Hg should be more accurate when all major risk factors are coun- offered antihypertensive drug treatment if: (i) there ted and weighted using risk functions derived from is any complication of hypertension or target organ epidemiological studies, most commonly the Fram- damage, or diabetes (Table 2) and/or (ii) The 10-year ingham risk function. The Joint British Societies recently issued recommendations on preventing CHD and included a computer programme; the ‘Car- diac Risk assessor’ and a CHD risk chart both of which are based on the Framingham risk function.50 The BHS recommends the use of either of these methods to estimate 10-year CHD risk and thereby help to rationalise treatment decisions for hyperten- sive people. The Joint British Societies computer programme ‘Cardiac Risk Assessor’ is the preferred method for estimating 10-year CHD risk and can also

BOX 5. Thresholds and Treatment Targets for Antihypertensive Drug Therapy

¼ Drug therapy should be started in all patients with sustained systolic blood pressures у160 mm Hg or sustained diastolic blood pressures у100 mm Hg despite non-pharmacological measures (A).

¼ Drug treatment is also indicated in patients with sustained systolic blood pressures 140–159 mm Hg or diastolic blood pressures 90–99 mm Hg if target organ damage is present, or there is evidence of established cardiovascular disease, or diabetes, or the 10-year CHD risk is у15% (B).

¼ For most patients a target of р140 mm Hg systolic and р85 mm Hg is recommended (A). For patients with diabetes a lower target of р140/80 mm Hg is Figure 1 *Estimated by the Joint British Societies ‘Cardiac Risk recommended (A). Assessor’ computer programme or CHD risk chart.51 **Repeated measurements (see text). Guidelines for management of hypertension British Hypertension Society 576 Guidelines for management of hypertension British Hypertension Society 577 Guidelines for management of hypertension British Hypertension Society 578 CHD risk is у15% (Ϸ20% CVD risk) despite advice pressure was below 150/90 mm Hg. Reduction of on non-pharmacological measures. blood pressure below the optimal level caused no Decisions on treatment at lower levels of harm. An important practical point is that the opti- CHD/CVD risk will be influenced by the patient’s mal blood pressure was attained, by titrating treat- attitude to treatment, and the benefit anticipated ment in stepped-care fashion aiming for diastolic from treatment. The benefit expected for an individ- blood pressures of р90, р85 and р80 mm Hg. With ual patient can be calculated using the absolute risk this systematic method of treatment the final dias- of a cardiovascular event and the relative risk tolic blood pressure was above 90 mm Hg in only reduction by treatment (25%). 7% of patients. Using an intention-to-treat analysis, When a decision is reached not to treat any patient in hypertensive patients with diabetes there with mild hypertension, it is essential to continue appeared to be a significant advantage in lowering observation and monitoring of blood pressure, at blood pressure to below the optimal level for non- least once per year. Blood pressure will rise within diabetic patients. Titration of treatment in diabetics 5 years to levels clearly requiring treatment in about aiming for a diastolic pressure р80 mm Hg halved 10–15% of patients.100 In addition CHD/CVD risk the incidence of major cardiovascular events when will increase with age, and risk should be reassessed compared to treatment aiming for diastolic blood at yearly intervals. These patients should all be pressure р90 mm Hg. encouraged to continue with non-pharmacological Considering prospective observational data and measures to lower blood pressure and cardiovascu- the findings of the HOT trial,3 recommendations for lar risk. target blood pressures during treatment are shown Thresholds for intervention are summarised in Table 3. below and in Figure 1. • Accelerated (malignant) hypertension [papilloed- Choice of antihypertensive drug (Box 6) ema, fundal hemorrhages and exudates] or For each major class of antihypertensive drug there impending cardiovascular complications; admit are compelling indications for use in specific patient for immediate treatment. • у groups, and also compelling contraindications. Blood pressure 220/120 mm Hg; treat immedi- There are also indications, contraindications, and ately. • cautions that are less clear-cut, and which are given Blood pressure 200–219/110–119 mm Hg: confirm different weight by different doctors. These indi- over 1–2 weeks, then treat. • cations, contraindications, and cautions for each of Blood pressure 160–199/100–109 mm Hg: the drug classes are summarised in Table 5. When –cardiovascular complications/target organ dam- none of the special considerations listed in Table 5 age (Table 2) or diabetes (type I or II), present— apply, the least expensive drug, with the most sup- confirm over 3–4 weeks, then treat. portive trial evidence; a low dose of a thiazide –cardiovascular complications/target organ dam- diuretic, should be preferred. age or diabetes (type I or II) absent: Since the previous guidelines2 three long-term –non-pharmacological advice, re-measure weekly double-blind studies have compared the major initially, and treat if blood pressure persists at classes of antihypertensive drug (thiazide, beta- these levels over 4–12 weeks. • blocker, calcium antagonist, ACE-inhibitor, and Blood pressure 140–159/90–99 mm Hg: alpha-blockers,9–11 and overall showed no consistent –cardiovascular complications/target organ dam- or important differences as regards antihypertensive age (Table 2) or diabetes (type I or II) present— efficacy, side effects, or quality of life. There were, confirm within weeks and treat; however, differences in average response between –cardiovascular complications/target organ dam- drug classes related to age and ethnic group.10 Few age or diabetes absent: trials have compared different classes of drugs –non-pharmacological advice, re-measure at directly as regards reduction in cardiovascular monthly intervals. events,101 and none is entirely satisfactory, but they –If mild hypertension persists, estimate 10-year CHD risk formally using the Joint British Societies ‘Cardiac Risk Assessor’ computer programme or BOX 6. Choice of Antihypertensive Drug Treatment the Joint British Societies CHD risk chart50 (Figure 2). Treat if the estimated 10-year CHD ¼ Use a low dose of thiazide as first-line treatment у15% (Ϸ20% CVD risk). unless there is a contraindication or a compelling indication for another drug class (A). Treatment goals (Box 5) ¼ Long acting dihydropyridine calcium antagonists are a suitable alternative for isolated systolic hypertension The HOT trial3 has provided the best evidence to in the elderly when low-dose thiazide is not tolerated or contraindicated (A). date on optimal blood pressure targets during anti- hypertensive treatment. In patients with diastolic ¼ Choice of drug will depend on relative indications and pressures of 100–115 mm Hg, the optimal blood contra-indications in the individual patient (Table 5). pressure based on an on-treatment analysis for ¼ Less than half of all hypertensives will be controlled reduction of cardiovascular events was reported to on monotherapy and one-third will require three or be 139/83 mm Hg. However, hypertensive patients more drugs (A). were apparently disadvantaged little provided blood Guidelines for management of hypertension British Hypertension Society 579 Table 5 Compelling and possible indications, contraindications and cautions for the major classes of antihypertensive drug

Class of drug Compelling indications Possible indications Possible Compelling contraindications contraindications

Alpha-blockers Prostatism Dyslipidaemia Postural hypotension Urinary incontinence

ACE-inhibitors Heart failure Chronic renal Renal impairmenta Pregnancy LV dysfunction diseasea PVDb Renovascular disease Type 1 diabetic Type 2 diabetic Nepropathy Nephropathy

All-antagonists ACE inhibitor-induced Heart failure PVDb Pregnancy Cough Intolerance of other Renovascular disease antihypertensive drugs

Beta-blockers Myocardial infarction Heart failurec Heart failurec Asthma/COPD Angina Dyslipidaemia Heart block PVD

Calcium antagonists Elderly ISH Elderly – – (dihydropyridine) Angina

Calcium antagonists Angina Myocardial infarction Combination with Heart block (rate-limiting) Beta-blockade Heart failure

Thiazides Elderly Dyslipidaemia Gout a ACE-inhibitors may be beneficial in chronic renal failure but should only be used with caution, close supervision and specialist advice when there is established and significant renal impairment. b Caution with ACE-inhibitors and AII-antagonists in peripheral vascular disease because of association with renovascular disease. c Beta-blockers may worsen heart failure, but in specialist hands may be used to treat heart failure. COPD, chronic obstructive pulmonary disease; ISH, Isolated systolic hypertension; PVD, peripheral vascular disease. have shown no consistent differences between regi- risk of coronary and stroke death,24–30 as discussed mens based on different drug classes.55 With the before. These drug classes cause metabolic disturb- exception of Syst-Eur,8 Syst-China102 and the CAPPP ances, including changes in lipids and glucose toler- study,103 most evidence from outcome trials is for ance, that are dose-dependent.104 Beta-blockers may treatment based on thiazide and/or beta-block- also promote weight gain6 and in the CAPPP ers.58,59,104 Indirect comparison between SHEP,59 study,103 treatment based on beta-blockers and thiaz- based on diuretic treatment, and Syst-Eur based on ides resulted in significantly more patients (+21%) a dihydropyridine calcium antagonist,8 suggests that developing diabetes over 5 years when compared to the outcome with these regimens was similar quali- treatment based on ACE-inhibition. Nevertheless in tatively and quantitatively. these studies, body weight and metabolic changes Overall these outcome trials have shown signifi- did not adversely influence the efficacy of antihy- cant reductions in stroke, by 38%, in coronary pertensive therapy at reducing cardiovascular mor- events, by 16%, and in cardiovascular mortality, by bidity and mortality. The question as to whether 21%.57,58 The absolute benefit from treatment is metabolically neutral classes of antihypertensive smaller in women than men, but this is compatible therapy will improve the outcome with regard to with their lower cardiovascular risk.58 The coronary and stroke reduction will only be answ- reduction in coronary events observed in all trials ered by the results of large comparative trials now was less than the 20–25% reduction predicted from in progress, such as ALLHAT107 and ASCOT.108 epidemiological observations.24,105 The reduction in There have been suggestions that dihydropyridine coronary events in trials based on low-dose thiaz- calcium antagonists may increase the risk of coron- ides has been significantly larger, at 28%, than those ary events, cancer, bleeding, depression, suicide, in trials of regimens based on high-dose thiazide or and other adverse events.17,18 These signals of poss- beta-blocker.8,59,105 Low-dose thiazide based regi- ible harm have emerged from case-control studies, mens also reduced cardiovascular and all-cause or from subgroup interim analyses or secondary mortality significantly.8,59,105 The larger benefit on end-point analyses in randomised controlled trials. coronary events observed in these trials with low- Evidence of this type cannot be ignored, but it is dose thiazides is not necessarily related to the dose invariably open to bias and confounding and is of thiazide per se. It may be related to differences in therefore less robust when evaluating cause and age,102 to more effective potassium conservation in effect. There is little biological plausibility for some these trials,106 or to the play of chance. However of the adverse effects proposed. Controlled trials of higher doses of thiazide diuretics (bendrofluazide dihydropyridine calcium antagonists such as PRA- Ͼ2.5 mg or hydrochlorthiazide Ͼ25 mg daily) are ISE,109 STONE,110 and particularly Syst-Eur8 have unnecessary and should no longer be used. not supported these concerns about safety. Any Hypertensive patients controlled adequately on remaining doubts may be resolved by large compara- thiazides and beta-blockers have a persistent excess tive trials in progress such as ASCOT and ALLHAT. Guidelines for management of hypertension British Hypertension Society 580 Nifedipine in capsule form should no longer be pre- scribed,111 but otherwise the evidence available now BOX 7. Other Measures to Reduce Cardiovascular Risk strongly suggests that the benefits of dihydropyrid- ¼ Patients with established cardiovascular disease or at ine calcium antagonist treatment clearly exceed any high risk according to the Joint British Societies risks, when they are prescribed for appropriate indi- ‘Cardiac Risk Assessor’ computer programme or CHD cations. risk chart should be considered for aspirin (A) and statin therapy (A) as follows: Dosage ¼ For primary prevention, 75 mg aspirin is recommended for hypertensive patients aged 50 years or more who The drug or formulation used should ideally be have satisfactory control of their blood pressure and effective when taken as a single daily dose. An inter- either target organ damage or diabetes. val of at least 4 weeks should be allowed to observe ¼ For primary prevention, statin therapy is indicated the full response, unless it is necessary to lower when the serum total cholesterol is у5.0 mmol/l and blood pressure more urgently. The dose of thiazide the 10 year CHD risk is у30%. diuretic should not be titrated up,104 whereas other ¼ For secondary prevention, statin therapy is indicated drug classes should be titrated according to the when the total serum cholesterol is у5.0 mmol/L and manufacturers’ instructions. When the first drug is there is evidence of cardiovascular disease ie. well tolerated but the response is insufficient, as is angina/myocardial infarction etc. Aspirin is also the case in about one half of all hypertensive indicated when there is evidence of cardiovascular patients, the options are to substitute an alternative disease. drug or to add a second drug. Substitution of an alternative drug is appropriate when hypertension is mild and uncomplicated and the response to the cardiovascular events by 15% and myocardial initial drug was small. In more severe or compli- infarction by 36%, but had no effect on fatal events. cated hypertension it is safer to add drugs stepwise In the Thrombosis Prevention Trial of aspirin 75 mg until blood pressure control is attained. Treatment daily for primary prevention,35 26% of those studied can be stepped down later if the blood pressure falls had treated hypertension. The outcome was similar substantially below the optimal level. to that of the HOT trial; a 16% reduction in all car- In the HOT study, less than one-third of hyperten- diovascular events, 20% reduction in myocardial sive patients were controlled by monotherapy and infarction, and no effect on fatal events. In both more than one-third required three or more drugs trials the number of clinically-significant bleeding in combination to achieve optimal blood pressure episodes caused by aspirin was similar to the num- control.3 Similar conclusions were reported from the ber of cardiovascular events prevented by aspirin, UKPDS study of hypertension in type 2 diabetes.6 suggesting that the margin between benefit and harm Thus, the majority of hypertensive people will was narrow. The average cardiovascular event risk require combinations of antihypertensive therapy to in the two trials was 1.0–1.5% per year, and the achieve optimal blood pressure control. The major threshold for aspirin treatment for primary preven- classes of drug generally have additive effects on tion in hypertension should therefore probably be blood pressure when they are prescribed together.112 set slightly higher than this. The HOT trial studied Submaximal doses of two drugs result in larger well-controlled hypertensives, and in the Throm- blood pressure responses and fewer side effects than bosis Prevention Trial aspirin was withheld when maximal doses of a single drug. Rational drug com- blood pressure was above 170/100 mm Hg. Further- binations combine drugs with different modes of more, those who developed cerebral haemorrhage in action that are additive. Such combinations include; the Thrombosis Prevention Trial had significantly a diuretic with beta-blocker, diuretic with ACE higher systolic blood pressures before the adverse inhibitor, beta-blocker with calcium antagonist, cal- event (158 mm Hg vs 135 mm Hg in those with no cium antagonist with ACE inhibitor. For the third- stroke). Hypertension must therefore be controlled line drug therapy commonly used combinations are satisfactorily (to the Audit standard, Ͻ150/90 diuretic, ACE inhibitor and calcium antagonist or mm Hg) before starting aspirin treatment for primary diuretic, beta-blocker and calcium antagonist. prevention of cardiovascular disease in hyperten- Fixed-dose combinations are not widely used in Bri- sive subjects. tain, but are convenient for patients and acceptable Given these considerations, aspirin 75 mg daily is provided they are used as second-line treatment recommended for hypertensive patients who have: when monotherapy is ineffective; the individual • no contraindication to aspirin drug components are appropriate; and there are no and major cost implications. Certain drugs should not be • Secondary Prevention: Cardiovascular compli- co-prescribed for the treatment of hypertension, for cations (myocardial infarction, angina, non-haem- example beta-blocker with verapamil or diltiazem; orrhagic cerebrovascular disease, peripheral vas- ACE-inhibitor with angiotensin II antagonist; or pot- cular disease or atherosclerotic renovascular assium-sparing diuretic with ACE-inhibitor. disease) Aspirin (Box 7) or In the HOT trial3 aspirin 75 mg daily in treated • Primary Prevention: hypertensive patients aged 50 years or more reduced blood pressure controlled to Ͻ150/90 mm Hg Guidelines for management of hypertension British Hypertension Society 581 and angina, CABG or angioplasty, non-haemorrhagic Age у50 years cerebrovascular disease, peripheral vascular dis- and ease, or atherosclerotic renovascular disease. target organ damage (eg, LVH, renal impairment, • Primary prevention: 10-year CHD risk у30% (Ϸ10 or proteinuria) year CVD risk of у40%) estimated formally by the or Joint British Societies cardiac risk assessor pro- A 10-year CHD risk у15% when estimated for- gramme or risk chart (Figure 2). mally by Joint British Societies computer pro- • Familial hypercholesterolaemia. gramme or risk chart. Based on current evidence, the upper age limits for or starting statin treatment suggested are 75 years for Type II diabetes mellitus. secondary prevention, and 70 years for primary pre- vention. Patients with an estimated 10-year CHD risk of у15% will have their cardiovascular risk reduced by 25% using antihypertensive treatment (see Follow-up above). The addition of aspirin further reduces 3,35 The frequency of follow-up for treated patients after major cardiovascular events by 15%, giving a adequate blood pressure control is attained depends NNT for 5 years of about 90 for one cardiovascular upon factors such as the severity of the hyperten- complication and 60 for one myocardial infarction sion, variability of blood pressure, complexity of the prevented by aspirin. This benefit is acceptable treatment regimen, patient compliance, and the given that aspirin treatment is simple and inexpen- need for non-pharmacological advice. Review every sive. 3 months is sufficient when treatment and blood pressure are stable, and the interval should not gen- HMG CoA reductase inhibitors (statins) erally exceed 6 months. Those who have been (Box 7) hypertensive in the past, or who have untreated mild hypertension and a low estimated 10-year In the last 4 years the results of several controlled CHD/CVD risk, should have their blood pressure outcome trials have shown that statin treatment for measured and their 10-year CHD/CVD risk estimated secondary31,33 and primary32,34 prevention reduces annually. The routine for follow-up visits should be major coronary events by 30%, reduces all-cause simple; measure blood pressure and weight; enquire mortality significantly, and is safe, simple and well about general health, side effects and treatment tolerated. Statin treatment also reduces the risk of problems; reinforce advice on non-pharmacological stroke substantially in patients who have coronary measures; and test urine for proteinuria annually. In heart disease.31,33 In subgroup analyses the benefits general practice and hospital clinics trained nurses were similar in hypertensive patients.33,34 Given the have an important role in the accurate measurement persistent high risk of coronary and stroke death in of blood pressure, and can advise and educate treated hypertensive patients,24–30 and the relation patients on aspects such as non-pharmacological of this risk to serum cholesterol before and during measures and possible side effects from drugs. A treatment,29 the implications of these trials for man- large proportion of hypertensive patients disappear agement of hypertension are large. There is general from regular follow-up for a variety of reasons. This acceptance that statin treatment should be targeted may be reduced by thorough education of the at a specified threshold of coronary risk, and not at patient about hypertension and its treatment, and thresholds of lipid values.113 Statin treatment could provision of written information (see Appendix 2). now be justified at a 10-year CHD risk of 6%,34 but A formal system of recall for those who miss routine this would entail treating approximately half of all appointments, using the practice computer, is desir- adults in Britain, and an even higher proportion of able. hypertensive patients. The main constraints on sta- tin treatment at present are the workload for general practitioners and the enormous potential cost. Special patient groups Our recommendations for statin therapy are Elderly hypertensives (Box 8) designed to be consistent with the Joint Recommen- Hypertensive people over the age of 60 years deserve dations of the British Hypertension Society, British special consideration for several reasons. Systolic Cardiac Society, British Diabetic Association and blood pressure rises steadily with increasing age, British Hyperlipidaemia Association,50 and with and the prevalence of hypertension including iso- guidance from a national Standing Medical Advis- lated systolic hypertension (у160/Ͻ90 mm Hg) is ory Committee (SMAC)51 and those from the Scott- more than 50% in those over 60 years.23 These ish Intercollegiate Guideline Network (SIGN).52 We people have a high risk of cardiovascular compli- emphasise that these are very conservative rec- cations when compared to younger hypertensives,114 ommendations and represent minimum acceptable and antihypertensive treatment of diastolic105 and levels of treatment on the basis of existing evidence. isolated systolic8,59 hypertension reduces this risk. We advise that statin treatment is prioritised to Recent evidence also shows that antihypertensive those at highest cardiovascular risk, as follows: therapy reduces the incidence of heart failure by • Serum total cholesterol у5.0 mmol/L; and 50%115 and possibly dementia,116 important compli- • Secondary prevention: myocardial infarction, cations of hypertension in this age group.117 The Guidelines for management of hypertension British Hypertension Society 582 pated on this basis. The treatment of choice, a low BOX 8. Elderly dose thiazide diuretic, is safe and effective in the ¼ Absolute benefit from treatment is greater in the elderly. On the other hand, direct evidence of bene- elderly (at least up to 80 years) than younger age fit from randomised controlled trials is lacking for groups (A). elderly people with blood pressures 140–159/Ͻ90 mm Hg and the high prevalence would mandate ¼ Optimum BP levels on treatment should be similar to those of younger patients if possible (A). treatment of a large proportion of the elderly popu- lation. A firm recommendation is not appropriate in ¼ Older patients tolerate antihypertensive treatment as the circumstances. Doctors should consider the well as younger age groups (A). benefit anticipated and the resource implications in the context of other health and quality of life issues ¼ Low-dose thiazides are the drug of choice for elderly hypertensive people. when reaching a treatment decision in individual elderly patients with borderline isolated systolic ¼ Dihydropyridine calcium antagonists are a suitable hypertension. alternative when thiazides are contraindicated or Low-dose thiazides are the accepted first-line poorly tolerated (A). treatment for the elderly.114 Beta-blockers were second-line agents in many of the outcome trials, but are less effective than thiazides as first-line treat- absolute benefit from treatment in the elderly is ment and meta-analyses suggest that beta-blockers much larger than that for younger hypertensives114 decrease stroke but no other cardiovascular events because of their higher absolute risk. Physicians in this age group.118 In Syst-Eur8 treatment of iso- may not be fully aware of this evidence,19,20 and lated systolic hypertension in elderly patients was hence treatment is not implemented fully. This is based on the dihydropyridine nitrendipine, and the not surprising, because before these outcome trials outcome was similar to that with diuretic-based it was a widely and incorrectly held view that a rise treatment in SHEP.59 Hence a dihydropyridine cal- in blood pressure with age was inevitable and harm- cium antagonist is a suitable alternative for elderly less, and that isolated systolic hypertension was of hypertensive patients when thiazides are ineffec- no consequence. There was also concern that elderly tive, contraindicated, or not tolerated. people might tolerate antihypertensive drugs poorly, a concern that has been dispelled by detailed analy- Hypertension in diabetes (Box 9) ses of adverse effects experienced in the outcome trials.114 Evidence for benefit from antihypertensive Hypertension is common in diabetes and plays a treatment extends until at least the age of 80 years, major role in the development of diabetic macrovas- and regular blood pressure screening should con- cular and microvascular complications.6,119 The tinue until this age. prevalence of hypertension differs in type I and type Once started, antihypertensive treatment should II diabetes. In type I diabetes, in the absence of be continued after patients reach the age of 80. nephropathy (microalbuminuria or proteinuria) the When hypertension is first diagnosed beyond the prevalence of hypertension is similar to that in the age of 80 there is no firm evidence to guide policy, non-diabetic population.120 In type 2 diabetes, but decisions should probably be based on biologi- hypertension (Ͼ140/90 mm Hg) is present in over cal rather than chronological age. Patients with 70% of patients.121,122 newly-diagnosed hypertension after the age of 80 should be considered for treatment provided they Type 1 diabetes without nephropathy: The thres- are generally fit and have reasonable life expectancy, hold for intervention with antihypertensive therapy particularly if they have hypertensive complications in patients without evidence of nephropathy is a or target organ damage. systolic BP of у140/90 mm Hg and the optimal Elderly hypertensives respond to non-pharmaco- blood pressure target is Ͻ140/80 mm Hg. The prin- logical measures to lower blood pressure at least as well as younger patients.13,78–81,86 Antihypertensive therapy is indicated and clearly beneficial in people BOX 9. Diabetes aged 60 years or more, when blood pressure aver- ages are Ͼ160 mm Hg systolic and Ͼ90 mm Hg dias- Type 1 Diabetes tolic.114 ¼ Threshold for starting antihypertensive treatment is у140/90 mm Hg (B). There is no firm evidence to guide policy for sys- ¼ Target blood pressure Ͻ140/80 mm Hg or lower if tolic blood pressure 140–159 mm Hg and diastolic proteinuria present (A). blood pressures Ͻ90 mm Hg (borderline isolated ¼ BP reduction and ACE inhibitors reduce the rate of systolic hypertension). Treatment is advised when decline in renal function (A). there are cardiovascular complications or evidence Type II Diabetes of target organ damage. Management of those with ¼ Threshold for starting antihypertensive treatment is borderline isolated systolic hypertension who do у140/90 mm Hg (B). not have these complications represents a difficult ¼ Target blood pressure Ͻ140/80 mm Hg (A). dilemma. Most will have an estimated 10-year CHD ¼ Optimal first-line therapy is not yet established but trial evidence supports the use of: ACE inhibitors, risk in excess of 15% because of the profound effect beta-blockers, dihydropyridine CCBs, alpha-blockers, of age on cardiovascular risk. A large absolute bene- and low-dose thiazide diuretics (B). fit from antihypertensive therapy would be antici- Guidelines for management of hypertension British Hypertension Society 583 ciples guiding the management of these patients is antihypertensive therapy as у140/90 mm Hg in type similar to that recommended for type 2 diabetes 2 diabetes. (see below). In the diabetic cohort of the HOT study, blood pressure lowering improved cardiovascular out- come and importantly there were 50% fewer cardio- Type 1 diabetes and diabetic nephropathy: Hyper- vascular end-points when diastolic blood pressure tension in type 1 diabetes often indicates the pres- was titrated beyond Ͻ90 mm Hg to Ͻ80 mm Hg.3 ence of diabetic nephropathy.123 Blood pressure Thus, for optimal blood pressure control, diastolic reduction and ACE-inhibitor treatment slow the rate blood pressure should be targeted to below 80 of decline of renal function in overt diabetic nephro- mm Hg. There is less evidence to guide optimal sys- pathy4 and delay progression from the microalbumi- tolic BP targets; thus we recommend systolic blood nuric phase to overt nephropathy.123–125 ACE-inhibi- pressure should be maintained at the levels tors may have a specific renoprotective action in observed in those randomised to р80 mm Hg in the patients with incipient or overt nephropathy and are HOT trial, ie, systolic blood pressure р140 mm Hg. recommended as first-line therapy. If ACE-inhibitor The aforementioned studies have highlighted that treatment has to be discontinued because of persist- at least two antihypertensive drugs are usually ent cough an angiotensin II receptor antagonist may required to achieve optimal levels of blood pressure be considered, although evidence for specific renop- control in type 2 diabetes.6 As regards the choice of rotection by this drug class is awaited from ongoing drug therapy for hypertension in type 2 diabetes, clinical trials. The ACE-inhibitor should be titrated one study suggested that regimens based on ACE- to the maximum dose recommended and tolerated. inhibition (captopril) and beta-blockade (atenolol) For renoprotection blood pressure control is most were equally effective at reducing macrovascular important,126 and combinations of antihypertensive complications but the study was too small to drugs are invariably required to achieve rec- exclude a difference.131 Subgroup analysis of out- ommended blood pressure targets. Thiazide come trials have shown that other classes of antihy- diuretics, calcium antagonists, cardioselective beta- pertensive drugs, ie, diuretics and dihydropyridine blockers, and alpha-blockers are all suitable. The calcium antagonists, also improve the prognosis of threshold for antihypertensive treatment in type 1 diabetics with hypertension.3,6,7,132 Thus, the opti- diabetes with nephropathy is у140/90 mm Hg. The mal first-line drug is yet to be established but there target blood pressure is Ͻ130/80 mm Hg,3,123 or is evidence from sub-group analyses of outcome lower (Ͻ125/75 mm Hg) when there is proteinuria trials in diabetic people for the safety and efficacy у1 g/24 h.126 Type 1 diabetic subjects with persist- of ACE-inhibitors, dihydropyridine calcium antag- ent microalbuminuria or proteinuria and ‘normal’ onists, low dose thiazide diuretics and beta-block- blood pressures, may also benefit from ACE-inhi- ers. The choice among these drug classes should be bition titrated to the recommended maximum determined using the criteria set out earlier for non- dose.4,123,127 It remains unclear whether this benefit diabetic patients. accrues from ACE-inhibition per se, or the associa- ted blood pressure reduction.128 Type 1 diabetic sub- Type 2 diabetes and diabetic nephropathy: There ject with evidence of nephropathy are at very high is much less evidence to guide practice in type 2 risk of cardiovascular disease and may be con- diabetic subjects with nephropathy. Hypertension sidered for statin therapy if their total cholesterol is accelerates the decline of renal function in type 2 у5 mmol/L and aspirin therapy if they if they satisfy diabetic patients with established nephro- the criteria set out above. pathy.123,133 Moreover, antihypertensive therapy slows the progression of nephropathy in patients Hypertension in type 2 diabetes: Hypertension is with type 2 diabetes.123 ACE-inhibitors have an anti- very common in type 2 diabetes. It is strongly proteinuric action and delay progression from related to but not fully accounted for by obesity121,129 microalbuminuria to overt nephropathy,123,125,134 but and is highly predictive of cardiovascular and it is less clear whether they have a specific renopro- microvascular complications.119,129 In the UKPDS tective action beyond blood pressure reduction in study, antihypertensive therapy was more effective overt nephropathy complicating type 2 diabetes. than tight glycaemic control6 in protecting against The high cardiovascular risk of type 2 diabetic microvascular and macrovascular disease, and was subjects and the fact that many have established car- the only intervention that improved survival of diovascular disease at diagnosis, means that thera- patients with type 2 diabetes. Dramatic survival peutic strategies other than blood pressure lowering benefits were also observed in the elderly diabetic are important. Aspirin should be offered to all cohort treated with antihypertensive therapy in the patients with diabetes and hypertension using the SystEur trial.7 Data from the UKPDS demonstrated criteria outlined above. Many type 2 diabetic people that many type 2 diabetic subjects have established with hypertension will also require statin therapy cardiovascular or microvascular disease at diagnosis for primary prevention because their estimated 10- and overall, have a 10-year cardiovascular event rate year CHD risk is у30%. These risk levels can be esti- (predominantly coronary events) of Ͼ30%.6,130 mated from the Joint British Societies computer pro- Since the presence of target organ damage or a 10- gramme which adjusts for the higher CHD risk of year CHD risk of у15% per year would mandate diabetic subjects, or the risk chart which contains treatment of a blood pressure of у140/90 mm Hg, boxes dedicated to CHD risk estimation in diabetic we recommend the threshold for intervention with subjects.50 This multifactorial approach should be Guidelines for management of hypertension British Hypertension Society 584 complemented by efforts to optimise glycaemic con- be ineffective in patients with renal impairment, trol and continued non-pharmacological advice and loop diuretics (ie, frusemide), often in high because many diabetic patients, particularly those dose, are frequently required. The dose of renally- with type 2 diabetes are overweight and have many excreted antihypertensive drugs may need to be other cardiovascular risk factors.121,122,129,135 adjusted. The threshold for antihypertensive treatment is у140 mm Hg systolic, or у90 mm Hg diastolic for Renal disease and hypertension patients with persistent proteinuria or renal impair- Renovascular disease (renal artery stenosis) is rela- ment. Optimal blood pressure control is Ͻ130/85 tively uncommon, but is the most frequent curable mm Hg126,141 and reducing blood pressure to cause of hypertension. Routine investigation of all Ͻ125/75 mm Hg may produce additional benefit in hypertensive patients is not justifiable, but doctors patients with chronic renal disease of any aetiology should be aware of important clues suggesting reno- and proteinuria of у1 g per 24 h.126,141 Patients with vascular disease. These are: renal failure have a very high risk of cardiovascular complications,142 and may need aspirin or statin • onset of hypertension before the age of 30; treatment in addition to non-pharmacological meas- • documented sudden onset of hypertension or ures to reduce their cardiovascular risk burden risk. recent worsening of hypertension in middle age; • accelerated (malignant) hypertension; • resistant hypertension (to a three drug regimen); Hypertension in pregnancy • renal impairment of unknown cause; • elevation of serum creatinine by ACE inhibitor or This has been reviewed elsewhere.143,144 It occurs in angiotensin II antagonist treatment; up to 10% of pregnancies, and may be the first sign • peripheral vascular disease or severe generalised of impending pre-eclampsia, a potentially more seri- atherosclerotic disease; ous condition of the second half of pregnancy and • recurrent pulmonary oedema or heart failure with the puerperium. Diastolic blood pressure should be no obvious cause. measured at the disappearance of sounds (phase V) and not at muffling (phase IV) as recommended in Patients with any of these features should be the past.145,146 referred for specialist advice because the investi- gations required to confirm or exclude renovascular disease are complex. Idiopathic hypertension in pregnancy Hypertensive patients with elevated serum creati- nine or proteinuria may have parenchymal or Care must be taken to distinguish between chronic obstructive renal disease, and should be referred for hypertension (defined as a blood pressure of specialist evaluation. Accelerated (malignant) у140/90 mm Hg before 20 weeks’ gestation) and hypertension requires immediate hospital treatment pre-eclampsia. Elevated blood pressure before 20 because it causes rapid loss of renal function that weeks’ gestation usually means that hypertension can be irreversible if untreated. Otherwise hyperten- preceded pregnancy. This will commonly be idio- sion per se does not cause renal failure,136 and the pathic but clinical evaluation is needed (Table 2) to corollary is that renal impairment in the absence of exclude other causes such as renal disease, Conn’s previous accelerated phase hypertension suggests syndrome or co-arctation. Phaeochromocytoma is primary renal disease or renovascular disease.136 In rare but may cause sudden death in pregnancy and patients with chronic renal impairment, hyperten- should be excluded by measuring urinary cat- sion accelerates the rate of loss of renal function, echolamines if there are suggestive features. An and good blood pressure control is essential to apparent onset of hypertension after 20 weeks’ ges- retard this process.137 Whether ACE inhibitors have tation may reflect hypertension that was undetected a specific renoprotective effect in non-diabetic renal prior to pregnancy, and disguised by the blood failure, over and above their antihypertensive pressure fall of early-mid pregnancy. All women action, remains uncertain.137,138 Meta-analysis of all with hypertension in pregnancy should have their controlled trials showed a 30% reduction in inci- blood pressure checked post-natally. dence of end-stage renal failure with ACE-inhibi- Meta-analysis of trials of antihypertensive drugs tors,138 but this may not all be explained by in pregnancy shows reduction in the risk of pro- additional blood pressure reduction.138 ACE-inhibi- gression to severe hypertension and fewer hospital tors reduce proteinuria, and are probably renopro- admissions.147 Firm evidence is not available on the tective in patients with proteinuria у3 g/day or who optimal threshold for treatment. There is consensus have rapidly-progressing renal failure139,140 but may that treatment is essential at у170/110 mm Hg but not be renoprotective in those with polycystic kid- many treat blood pressure at levels у140/90 mm Hg. ney disease.138 ACE-inhibitors may cause or worsen Women with essential hypertension are at increased renal impairment in patients with critical renovas- risk of pre-eclampsia and intra-uterine growth cular disease, thus they should be used with cau- restriction (IUGR). Management should therefore tion, preferably with specialist supervision, in include frequent blood pressure checks, preferably patients with advanced chronic renal impairment. once a week, urinalysis and an assessment of foetal Blood pressure is particularly salt sensitive in growth. Hospital referral should be made if there is patients with impaired renal function, and dietary poorly controlled hypertension, new onset pro- salt reduction is important. Thiazide diuretics may teinuria or suspicion of IUGR. Guidelines for management of hypertension British Hypertension Society 585 Pre-eclampsia Oral contraceptives and hypertension Criteria for the diagnosis of pre-eclampsia include: Limited trial data154 and other observational stud- a rise in blood pressure of Ͼ15 mm Hg diastolic or ies155,156 suggest that combined oral contraceptives Ͼ30 mm Hg systolic from early pregnancy, or dias- (OCs) have a small adverse effect on blood pressure tolic blood pressure of Ͼ90 mm Hg on two occasions with average increases of 5/3 mm Hg.157 In a small 4 h apart or Ͼ110 mm Hg on one occasion and pro- proportion of women (Ϸ1%) severe hypertension teinuria (1+ is a indication for referral and may be induced.158 The increase in blood pressure Ͼ300 mg/24-h is the criterion for diagnosis). It is appears to be idiosyncratic in that it may occur emphasised that 30% of eclamptic convulsions many months or years after first using a combined occur in the absence of either raised blood pressure OC. No particular subgroups of OC users have been or proteinuria. Risk factors for pre-eclampsia are: clearly identified as being susceptible to blood first pregnancy, change of partner, previous pre- pressure increases and the cause for the blood press- eclampsia, family history of pre-eclampsia, idio- ure has not been established.157 Moreover, the dose- pathic hypertension, chronic renal disease, diabetes, related effects of oestrogens and progestogens on SLE, multiple pregnancy and obesity. Women with blood pressure have not been clearly established. As pre-eclampsia generally have no symptoms and can the blood pressure response to any combined OC only be detected by routine screening. When preparation is unpredictable, and there is a small present, the most frequent symptoms are headache, increase in cardiovascular risk associated with OC visual disturbance (often ‘flashing lights’), vomiting, use159 blood pressure should be measured before epigastric pain, oedema (particularly facial oedema). starting OC use and then 6 monthly thereafter. These symptoms in conjunction with raised blood Observational data suggest that oral progestogen- pressure require urgent referral and treatment. only contraceptive pills (POPs) do not on average Women rarely present with a convulsion, but a first induce an increase in blood pressure.156,160 fit in the second half of pregnancy with no other Although, clinical trial evidence in support of this known cause is highly suggestive of eclampsia. conclusion is very limited, hitherto, the POPs have been recommended for use in women with a pre- vious history of combined OC-induced hyperten- sion, or those women with hypertension wishing to Choice of antihypertensive therapy in pregnancy use an oral contraceptive. Evidence underpinning the choice of antihyperten- Pending further evidence, the use of the combined sive therapy in pregnancy is inadequate to make OC is not absolutely contraindicated for women firm recommendations. Methyldopa remains the who are already hypertensive or those who develop antihypertensive drug of choice for idiopathic hypertension during use of combined OCs. We rec- hypertension or pre-eclampsia because of its long ommend, however, that other non-hormonal forms and extensive use without reports of serious adverse of contraception should be sought, particularly if effects on the foetus.148 Calcium antagonists other risk factors for cardiovascular disease (eg, (especially nifedipine) and the vasodilator hydrala- smoking or migraine) coexist. In those women for zine are commonly used as second-line drugs. Lab- whom other methods of contraception are unaccept- etolol (alpha and beta-blocker) is also widely used able, changing to a POP with careful monitoring of as a second-line agent, particularly for resistant blood pressure is recommended. It should be recog- hypertension in the third trimester.149 Other beta- nised that data on the impact of POPs on cardio- blockers are used less often, particularly before 28 vascular risk is limited and that POPs are less effec- tive contraceptives than combined OCs. This is most weeks’ gestation because of concerns that they may Ͻ inhibit foetal growth.150 Meta-analysis of controlled relevant for younger women ( 35 years) who are trials of diuretics has shown a reduced incidence of more fertile and in whom the risk of pregnancy asso- pre-eclampsia,151 although no benefit was shown on ciated adverse cardiovascular events outweighs that foetal outcome. In practice, diuretics are used little related to the use of the combined OC. Advice to for the management of hypertension on theoretical modify coexisting risk factors (eg, smoking, excess grounds that they have the potential to further alcohol consumption) should be supplied. If blood reduce the already decreased circulatory blood vol- pressure does not fall below 160/100 mm Hg, anti- ume in women with pre-eclampsia.152 However, hypertensive medication should be initiated. there is nothing to suggest that low-dose thiazide diuretics in women with pre-existing hypertension Hormone replacement therapy (HRT) and are harmful and they may be continued through hypertension pregnancy. ACE inhibitors should be avoided because they may cause oligohydramnios, renal fail- Observational data161 and a recent clinical trial ure, hypotension and intrauterine death in the foe- examining various HRT formulations,162 have sug- tus.153 Hypertensive women who plan pregnancy, or gested that in general, the use of HRT is not associa- who become pregnant whilst on antihypertensive ted with an increase in blood pressure. HRT is not treatment, should be advised to change their therapy contraindicated for women with hypertension and to one of the drugs recommended for the treatment women with hypertension should not be denied of hypertension in pregnancy. It is usual to switch access to HRT as long as BP levels can be controlled from such agents back to the previous antihyperten- by antihypertensive medication. Given the uncer- sive regimen after delivery. tainties regarding cardiovascular protection associa- Guidelines for management of hypertension British Hypertension Society 586 ted with ‘opposed HRT’ (containing oestrogen and • To increase the proportion of patients on antihy- progestogen163 it is prudent to monitor blood press- pertensive treatment who are controlled to opti- ure after the initiation of HRT, 2–3 times in the first mal blood pressure levels. 6 months, then 6 monthly. Rare idiosyncratic rises • To reduce the cardiovascular risk of treated hyper- in blood pressure in patients receiving HRT have tensive patients by non-pharmacological meas- been reported. Consequently, HRT should be tem- ures, and by appropriate use of aspirin and sta- porarily discontinued in women with resistant tin treatment. hypertension to assess the contribution of HRT to • To increase identification and treatment of the blood pressure. patients with mild hypertension who are at high cardiovascular risk, for example; – elderly patients Hypertension in ethnic groups – in those with isolated systolic hypertension Hypertension is particularly common in African- – diabetic people Caribbean blacks, with a prevalence as high as 50% – those with target organ damage or multiple over the age of 40 years.164 Among this group hyper- risk factors tension tends to be more severe and associated with • To promote continuation of drug treatment, and a higher risk of complications, particularly stroke, compliance with treatment, by optimising the renal failure, and left ventricular hypertrophy. The choice and use of drugs, minimising side effects, hypertension is often sensitive to dietary salt restric- and increasing information and choice for tion and in those without evidence of target organ patients. damage, a low salt diet may occasionally be suf- ficient to control blood pressure. Black subjects are Implementation more responsive to diuretic and calcium antagonist treatment.10 ACE-inhibitors and beta-blockers may Realisation of these objectives will depend largely be ineffective in monotherapy in black subjects on the efforts of doctors and nurses in general prac- because the renin angiotensin system is frequently tice. Surveys revealing incomplete detection, treat- suppressed.10 However, these patients may respond ment and control of hypertension21–23 indicate a to ACE-inhibition or beta-blockade when given in serious failure to implement the knowledge we combination with drugs that activate the renin-angi- have, although there has been some improvement in otensin system, ie, diuretics, calcium channel block- recent years.23 Current methods of delivering medi- ers or alpha-blockers. In resistant hypertension a cal care to hypertensive people are evidently unsat- combination of diuretic, calcium antagonist, ACE- isfactory, a problem that is not unique to Britain. inhibitor and/or alpha-blocker is particularly effec- A WHO-WHL audit project168 in several European tive. countries showed that no more than 29% of all British South Asians (from the Indian sub- treated hypertensive patients had attained normal continent) also have a high prevalence of hyperten- blood pressure. The same study revealed a failure to sion, are commonly insulin resistant, and have a implement non-pharmacological measures to reduce high prevalence of type 2 diabetes.165,166 They are at blood pressure and cardiovascular risk, and inad- increased risk of stroke and particularly high risk equacy of information and education for patients. A of coronary heart disease.167 The limited evidence study from Australia169 showed that hypertensive available suggests that the response to drug treat- patients were no more knowledgeable than normo- ment of hypertension in South Asian patients is tensive controls about hypertension, and 70% of similar to that in white Europeans. Important points patients expressed a desire for more information in management are particular emphasis on glucose about their condition. The quality of care of hyper- tolerance, lipids, and the increased coronary tensive patients might be enhanced170 by improved risk.165,166 Clear advice is needed to reduce fat and screening or case-finding; use of computer-based refined sugar intake and to increase exercise. Good systems and protocols to support clinical decision control of blood pressure is particularly important making; workshops on hypertension for general in those with diabetes, and aspirin and/or statin practitioners and practice nurses; clinical audit; treatment may be indicated for those at high risk of improved doctor-patient or nurse-patient communi- coronary heart disease. cation; and increased patient education. It is suggested that all practices or primary care Implementation and audit groups should develop a protocol for hypertension management that covers the screening policy; initial Challenges for the future evaluation and investigation; implementation of non-pharmacological measures; formal estimation of The British Hypertension Society believes that the cardiovascular risk; treatment policy for antihyper- following objectives should have high priority: tensive drugs, aspirin and statins; treatment targets; • To promote the primary prevention of hyperten- policy for follow-up; and methods for identifying sion and cardiovascular disease by changes in the and recalling patients who drop out of follow-up. diet and lifestyle of the whole population. Written information should be available for patients • To increase the detection and treatment of undiag- about hypertension and its treatment, and on non- nosed hypertension by routine screening and pharmacological measures to reduce blood pressure increase awareness of hypertension among the and cardiovascular risk (see Appendix 2). The prac- public. tice policy should detail those aspects of manage- Guidelines for management of hypertension British Hypertension Society 587 ment that are in the province of the practice nurse European Trial on Isolated Systolic Hypertension in and of the doctor, and the indications and procedure the Elderly. Lancet 1997; 360: 757–764. for passing management from nurse to doctor or vice 9 Neaton JD et al, for the Treatment of Mild Hyperten- versa. It is recommended that implementation of the sion Study Research Group. Treatment of Mild Hyper- practice policy should be audited periodically. tension Study. Final results. JAMA 1993; 270: 713– 724. 10 Materson BJ et al, for the Department of Veterans Affairs Cooperative Study Group on Antihypertensive Audit points Agents. Single-drug therapy for hypertension in men. Several aspects of hypertension management lend A comparison of six antihypertensive agents with themselves readily to audit procedures that are placebo. N Engl J Med 1993; 328: 914–921. objective and reasonably simple. For example: 11 Philipp T et al, on behalf of the HANE Trial Research Group. Randomised, double blind, multicentre com- • the proportion of all adults in the practice who parison of hydrochlorothiazide, atenolol, nitrendipine, have had a blood pressure measurement in the last and enalapril in antihypertensive treatment: results of 5 years; the HANE study. BMJ 1997; 315: 154–159. • 12 Appel LJ et al, for the DASH Collaborative Research the proportion of all hypertensives given non- Group. A clinical trial of the effects of dietary patterns pharmacological advice; • on blood pressure. N Engl J Med 1997; 336: 1117–1124. the proportion of all hypertensives given antihy- 13 Applegate WB et al. Nonpharmacologic intervention pertensive therapy; to reduce blood pressure in older patients with mild • the proportion of hypertensives receiving antihy- hypertension. Arch Intern Med 1992; 152: 1162–1166. pertensive therapy who have suboptimal control, 14 Cushman WC et al, for the PATHS Group. Prevention ie, blood pressure levels у150 mm Hg systolic and and Treatment of Hypertension Study (PATHS): у90 mm Hg diastolic BP; Effects of an alcohol treatment program on blood • the proportion of patients lost from follow-up; or pressure. Arch Intern Med 1998; 158: 1197–1207. of treated patients who have not been reviewed 15 The Trials of Hypertension Prevention Collaborative within the last 6 months; Research Group. Effects of weight loss and sodium • reduction intervention on blood pressure and hyper- the use of aspirin and statins by those who require tension incidence in overweight people with high-nor- secondary prevention; or their use when indicated mal blood pressure: the Trials of Hypertension Preven- for primary prevention, ie, when the estimated 10- tion, Phase II. Arch Intern Med 1997; 157: 657–667. year CHD risk is у15% (aspirin) or у30% per 16 Whelton PK et al, for the TONE Collaborative Research year (statins). Group. Sodium reduction and weight loss in the treat- ment of hypertension in older persons. A randomized controlled trial of nonpharmacologic interventions in the elderly (TONE). JAMA 1998; 279: 839–846. 17 Stanton AV. Calcium channel blockers. BMJ 1998; 316: References 1471–1472. 1 Swales JD et al. Treating mild hypertension. BMJ 1989; 18 Cutler JA. Calcium-channel blockers for hyperten- 298: 694–698. sion – uncertainty continues. N Engl J Med 1998; 338: 2 Sever P et al. Management guidelines in essential 679–681. hypertension: report of the second working party of 19 Dickerson JEC, Garratt CJ, Brown MJ. Management of the British Hypertension Society. BMJ 1993; 306: hypertension in general practice: agreements with and 983–987. variations from the British Hypertension Society 3 Hansson L et al, for the HOT Study Group. Effects of guidelines. J Hum Hypertens 1995; 9: 835–839. intensive blood-pressure lowering and low-dose 20 Dickerson JEC, Brown MJ. Influence of age on general aspirin in patients with hypertension: principal results practitioners’ definition and treatment of hyperten- of the Hypertension Optimal Treatment (HOT) ran- sion. BMJ 1995; 310: 574. domised trial. Lancet 1998; 351: 1755–1762. 21 Smith WCS, Lee AJ, Crombie IK, Tunstall-Pedoe H. 4 Lewis EJ, Hunsicker LG, Bain RP, Rohde RD, for the Control of blood pressure in Scotland: the rule of Collaborative Study Group. The effect of angiotensin- halves. BMJ 1990; 300: 981–983. converting-enzyme inhibition on diabetic nephro- 22 Mashru M, Lant A. Interpractice audit of diagnosis and pathy. N Engl J Med 1993; 329: 1456–1462. management of hypertension in primary care: edu- 5 Curb JD, Pressel SL, Cutler JA et al, for the Systolic cational intervention and review of medical records. Hypertension in the Elderly Program Co-operative BMJ 1997; 314: 942–946. Research Group. Advantage of diuretic-based antihy- 23 Colhoun HM, Dong W, Poulter NR. Blood pressure pertensive treatment on cardiovascular disease risk in screening, management and control in England, results older diabetic patients with isolated systolic hyperten- from the health survey for England 1994. J Hypertens sion. JAMA 1996; 276: 1886–1892. 1998; 16: 747–753. 6 United Kingdom prospective diabetes study group. 24 Collins R et al. Blood pressure, stroke, and coronary Tight blood pressure control and risk of macrovascular heart disease. Part 2, short-term reductions in blood and microvascular complications in type 2 diabetes. pressure: overview of randomised drug trials in their UKPDS 38. BMJ 1998; 317: 703–713. epidemiological context. Lancet 1990; 335: 827–838. 7 Tuomilehto J et al. Effects of calcium-channel block- 25 Isles CG et al. Mortality in patients of the Glasgow ade in older patients with diabetes and systolic hyper- blood pressure clinic. J Hypertens 1986; 4: 141–156. tension. Syst Eur Investigators. N Engl J Med 1999; 26 Clausen J, Jenson G. Blood pressure and mortality: an 340: 677–684. epidemiological survey with 10 years follow-up. 8 Staessen JA, Fagard R, Thijs L et al, for the Systolic J Hum Hypertens 1992; 6: 53–59. Hypertension-Europe (Syst-Eur) Trial Investigators. 27 Thu¨ rmer HL, Lund-Larsen PG, Tverdal A. Is blood Morbidity and mortality in the placebo-controlled pressure treatment as effective in a population setting Guidelines for management of hypertension British Hypertension Society 588 as in controlled trials? Results from a prospective 48 MacMahon S. Guidelines for antihypertensive therapy: study. J Hypertens 1994; 12: 481–490. a commentary. J Hypertens 1996; 14: 691–693. 28 Merlo J et al. Incidence of myocardial infarction in eld- 49 Durrington PN et al. Indications for cholesterol ower- erly men being treated with antihypertensive drugs: ing medication: comparison of risk-assessment population based cohort study. BMJ 1996; 313: 457– method. Lancet 1999; 353: 278–281. 461. 50 Joint British recommendations on prevention of coron- 29 Andersson OK et al. Survival in treated hypertension: ary heart disease in clinical practice. Heart 1998; 80 follow up study after two decades. BMJ 1998; 317: (Suppl 2): S1–S29. 167–171. 51 Standing Medical Advisory Committee. The use of sta- 30 Du X et al. Case-control study of stroke and the quality tins. Department of Health, London, 1997. of hypertension control in north west England. BMJ 52 SIGN Guideline. Scottish intercollegiate guideline net- 1997; 314: 272–276. work; Lipids and the primary prevention of coronary 31 Scandinavian Simvastatin Survival Study Group. Ran- heart disease, 1999. domised trial of cholesterol lowering in 4444 patients 53 Eccles M et al. North of England evidence based guide- with coronary heart disease: the Scandinavian Simvas- line development project. BMJ 1998; 316: 1303–1309. tatin Survival Study (4S). Lancet 1994; 344: 1383– 54 O’Brien ET et al. Blood Pressure Measurement: Rec- 1389. ommendations of the British Hypertension Society. 32 Shepherd J et al. Prevention of coronary heart disease BMJ Publishing Group, 3rd Edition, 1997. with pravastatin in men with hypercholesterolemia. 55 Ramsay LE, Haq IU, Yeo WW, Jackson PR. Interpret- N Engl J Med 1995; 333: 1301–1307. ation of prospective trials in hypertension: do treat- 33 Sacks FM et al. The effect of pravastatin on coronary ment guidelines accurately reflect current evidence? events after myocardial infarction in patients with J Hypertens 1996; 14 (Suppl 5): S187–S194. average cholesterol levels. N Engl J Med 1996; 335: 56 Stamler J, Stamler R, Neaton JD. Blood pressure, sys- 1001–1009. tolic and diastolic, and cardiovascular risks. US popu- 34 Downs JR et al. Primary prevention of acute coronary lation data. Arch Intern Med 1993; 153: 598–615. events with lovastatin in men and women with aver- 57 Peto R, Collins R. Antihypertensive drug therapy: age cholesterol levels. Results of AFCAPS/TexCAPS. effects on stroke and coronary heart disease. In: Swales JAMA 1998; 279: 1615–1622. JD (ed). Textbook of Hypertension. Blackwell: Oxford, 35 The Medical Research Council’s General Practice 1994, pp 1156–1164. Research Framework. Thrombosis prevention trial: 58 Gueyffier F et al. Effect of antihypertensive drug treat- randomised trial of low-intensity oral anticoagulation ment on cardiovascular outcomes in women and men. with warfarin and low-dose aspirin in the primary pre- A meta-analysis of individual patient data from ran- vention of ischaemic heart disease in men at increased domised controlled trials. Ann Intern Med 1997; 126: risk. Lancet 1998; 351: 233–241. 761–767. 36 Haq IU, Jackson PR, Yeo WW, Ramsay LE. Impli- 59 SHEP Cooperative Research Group. Prevention of cations of recent trials with lipid-lowering drugs for stroke by antihypertensive drug treatment in older per- hypertension management. J Hypertens 1996; 14: 1502. sons with isolated systolic hypertension. Final results 37 Hasnain S, Webster J, McLay JS. Cardiovascular risk of the Systolic Hypertension in the Elderly Program in hypertension: implications of the Sheffield table for (SHEP). JAMA 1991; 265: 3255–3264. lipid-lowering strategy. J Hum Hypertens 1998; 12: 60 Fagard R, Staessen J, Thijs L, Amery A. Multiple stan- 469–471. dardized clinic blood pressures may predict left ven- 38 Jackson R. Which hypertensive patients should be tricular mass as well as ambulatory monitoring. A treated? Lancet 1994; 343: 496–497. metaanalysis of comparative studies. Am J Hypertens 39 Kannel WB. Potency of vascular risk factors as the 1995; 8: 533–540. basis for antihypertensive therapy. Eur Heart J 1992; 61 Fagard RH, Staessen JA, This L. Prediction of cardiac 13 (Suppl G): 34–42. structure and function by repeated clinic and ambulat- 40 Alderman MH. Blood pressure management: individ- ory blood pressure. Hypertension 1997; 29: 22–29. ualized treatment based on absolute risk and the 62 Perloff D, Sokolow M, Cowan R. The prognostic value potential for benefit. Ann Intern Med 1993; 119: of ambulatory blood pressures. JAMA 1983; 249: 329–335. 2792–2798. 41 Davey Smith G, Egger M. Who benefits from medical 63 Verdecchia P et al. Ambulatory blood pressure: an interventions? Treating low risk patients can be a high independent predictor of prognosis in essential hyper- risk strategy. BMJ 1994; 308: 72–74. tension. Hypertension 1994; 24: 793–801. 42 Me´nard J, Chatellier G. Mild hypertension: the mys- 64 Staessen JA et al, for the Ambulatory Blood Pressure terious viability of a faulty concept. J Hypertens 1995; Monitoring and Treatment of Hypertension Investi- 13: 1071–1077. gators. Antihypertensive treatment based on conven- 43 Jackson RT, Sackett DL. Guidelines for managing tional or ambulatory blood pressure measurement. A raised blood pressure. Evidence based or evidence bur- randomized controlled trial. JAMA 1997; 278: 1065– dened? BMJ 1996; 313: 64–65. 1072. 44 Ramsay LE, Wallis EJ, Yeo WW, Jackson PR. The 65 Haq IU et al. The impact of ambulatory blood pressure rationale for differing national recommendations for on estimation of cardiovascular risk in mild hyperten- the treatment of hypertension. Am J Hypertens 1998; sion. J Hypertens 1997; 15: 1540–1541. 11: 79S–88S. 66 Mancia G, Zanchetti A. White-coat hypertension: mis- 45 Core Services Committee. Guidelines for the manage- nomers, misconceptions and misunderstandings. ment of mildly raised blood pressure in New Zealand. What should we do next? J Hypertens 1995; 9: 1049– Ministry of Health: Wellington, 1995. 1052. 46 Swales JD. Treating hypertension. J Hypertens 1996; 67 Parati G et al. Limitations of the difference between 14: 813–814. clinic and daytime blood pressure as a surrogate meas- 47 Simpson FO. Guidelines for antihypertensive therapy: ure of the ‘white-coat’ effect. J Hypertens 1998; 15: problems with a strategy based on absolute cardio- 23–29. vascular risk. J Hypertens 1996; 14: 683–689. 68 Staessen JA, O’Brien ET, Atkins N, Amery AK. Ambu- Guidelines for management of hypertension British Hypertension Society 589 latory blood pressure in normotensive compared with consumption and risk of a major coronary event. BMJ hypertensive subjects. J Hypertens 1993; 11: 1289– 1997; 314: 1159–1164. 1297. 90 Arroll B, Beaglehole R. Exercise for hypertension. 69 Stergiou GS, Skeva II, Zourbaki AS, Mountokalakis Lancet 1993; 341: 1248–1249. TD. Self-monitoring of blood pressure at home: how 91 Whelton PK et al. Effects of oral potassium on blood many measurements are needed? J Hypertens 1998; 16: pressure: meta-analysis of randomized controlled 725–731. clinical trials. JAMA 1997; 277: 1624–1632. 70 O’Brien E, Atkins N, Staessen J. State of the market. A 92 Williams C. Healthy eating: clarifying advice about review of ambulatory blood pressure monitoring fruit and vegetables. BMJ 1995; 310: 1453–1455. devices. Hypertension 1995; 26: 835–842. 93 Raw M, McNeill A, West R. Smoking cessation: 71 O’Brien E et al. The British Hypertension Society pro- evidence based recommendations for the healthcare tocol for the evaluation of blood pressure measuring system. BMJ 1999; 318: 182–185. devices. J Hypertens 1993; 11 (Suppl 2): S43–S62. 94 Tang JL et al. Systematic review of dietary inter- 72 Prisant LM et al. American National Standard for non- vention trials to lower blood total cholesterol in free- automated sphygmomanometers: summary report. Am living subjects. BMJ 1998; 316: 1213–1220. J Hypertens 1995; 8: 210–213. 95 OXCHECK Study. BMJ 1995; 310: 1099–1104. 73 White WB et al. National standard for measurement of 96 Grover SA et al. Do doctors accurately assess coron- resting and ambulatory blood pressures with auto- ary risk in their patients? Preliminary results of the mated sphygmomanometers. Hypertension 1993; 21: coronary health assessment study. BMJ 1995; 310: 504–509. 975–978. 74 National High Blood Pressure Education Program 97 Evans J St BT, Harries C, Dennis I, Dean J. General Working Group. National High Blood Pressure Edu- practitioners’ tacit and stated policies in the prescrip- cation Program Working Group report on primary pre- tion of lipid lowering agents. Br J Gen Pract 1995; 45: vention of hypertension. Arch Intern Med 1993; 153: 15–18. 186–208. 98 Grover SA, Coupal L, Hu X-P. Identifying adults at 75 Watt GCM. Prevention of high blood pressure and its increased risk of coronary disease. How well do the complications: strategic issues in a local context. current cholesterol guidelines work? JAMA 1995; J Hypertens 1995; 13: 377–383. 274: 801–806. 76 Whelton PK. Epidemiology of hypertension. Lancet 99 Cook RJ, Sackett DL. The number needed to treat: a 1994; 344: 101–106. clinically useful measurement of treatment effect. 77 Stamler J. Setting the TONE for ending the hyperten- BMJ 1995; 310: 452–454. sion epidemic. JAMA 1998; 279: 878–879. 100 Moser M. Hebert PR. Prevention of disease pro- 78 Fortherby MD, Potter JF. Potassium supplementation gression, left ventricular hypertrophy and congestive reduces clinic and ambulatory blood pressure in eld- heart failure in hypertension treatment trials. JAm erly hypertensive patients. J Hypertens 1992; 10: Coll Cardiol 1996; 27: 1214–1218. 1403–1408. 101 Psaty BM et al. Health outcomes associated with anti- 79 Geleijnse JM et al. Reduction in blood pressure with a hypertensive therapies used as first-line agents. A low sodium, high potassium, high magnesium salt in systematic review and meta-analysis. JAMA 1997; older subjects with mild to moderate hypertension. 277: 739–745. BMJ 1994; 309: 436–440. 102 Liu L et al. Comparison of active treatment and pla- 80 Fotherby MD, Potter JF. Effects of moderate sodium cebo in older Chinese patients with isolated systolic restriction on clinic and twenty-four-hour ambulatory hypertension. Systolic hypertension in China (Syst- blood pressure in elderly hypertensive subjects. J Hy- China) Collaborative group. J Hypertens 1998; 16: pertens 1993; 11: 657–663. 1823–1829. 81 Braith RW et al. Moderate- and high-intensity exercise 103 Hansson L et al. Effect of angiotensin-converting lowers blood pressure in normotensive subjects 60–79 enzyme inhibition compared with conventional ther- years of age. Am J Cardiol 1994; 73: 1124–1128. apy on cardiovascular morbidity and mortality in 82 Kokkinos PF et al. Effects of regular exercise on blood hypertension: the Captopril Prevention Project pressure and left ventricular hypertrophy in African- (CAPPP). Lancet 1999; 353: 611–615. American men with severe hypertension. N Engl J Med 104 Thijs L et al. A meta-analysis of outcome trials in eld- 1995; 333: 1462–1467. erly hypertensives. J Hypertens 1992; 10: 1103–1109. 83 McCarron DA et al. Nutritional management of cardio- 105 Siscovick DS et al. Diuretic therapy for hypertension vascular risk factors. A randomised clinical trial. Arch and the risk of primary cardiac arrest. N Engl J Med Intern Med 1997; 157: 169–177. 1994; 330: 1852–1857. 84 Singer DRJ et al. Reduction of salt intake during con- 106 Carlsen JE, Kober L, Torp-Pedersen C, Johansen P. verting enzyme inhibitor treatment compared with Relation between dose of bendrofluazide, antihyper- addition of a thiazide. Hypertension 1995; 25: 1042– tensive effect, and adverse biochemical effects. BMJ 1044. 1990; 300: 975–978. 85 Law MR, Frost CD, Wald NJ. III-Analysis of data from 107 Whelton PK et al. Experimental approaches to trials of salt reduction. BMJ 1991; 302: 819–824. determining the choice of first step therapy for 86 Midgley JP, Matthew AG, Greenwood CMT, Logan AG. patients with hypertension. The ALLHAT Research Effect of reduced dietary sodium on blood pressure. A Group Antihypertensive and Lipid Lowering Treat- meta-analysis of randomized controlled trials. JAMA ment to prevent heart attack trial. Clin Exp Hypertens 1996; 275: 1590–1597. 1996; 18: 569–579. 87 Kaplan NM. Alcohol and hypertension. Lancet 1995; 108 Dahlof B, Sever PS, Poulter NR, Wedel H. On behalf 345: 1588–1589. of the ASCOT steering committee. J Hypertens 1998; 88 Gill JS et al. Stroke and alcohol consumption. N Engl 16 (Suppl 2): S212. J Med 1986; 315: 1041–1046. 109 Packer M et al, for the Prospective Randomized 89 McElduff P, Dobson A. How much alcohol and how Amlodipine Survival Evaluation Study Group. Effect often? Population based case-control study of alcohol of amlodipine on morbidity and mortality in severe Guidelines for management of hypertension British Hypertension Society 590 chronic heart failure. N Engl J Med 1996; 335: artery disease in non-insulin dependent diabetes. 1107–1114. BMJ 1998; 316: 823–838. 110 Gong L et al. Shanghai trial of nifedipine in the eld- 131 UKPDS group 39. Efficacy of atenolol and captopril erly (STONE). J Hypertens 1996; 14: 1237–1245. in reducing risk of macrovascular and microvascular 111 Beevers DG, Sleight P. Short acting dihydropyridine complications in type 2 diabetes. BMJ 1998; 317: (vasodilating) calcium channel blockers for hyperten- 713–720. sion: is there a risk? BMJ 1996; 312: 1143–1145. 132 Curb JD et al. Effect of diuretic-based antihyperten- 112 Materson BJ, Reda DJ, Cushman WC, Henderson WG, sive treatment on cardiovascular risk in people with for the Department of Veterans Affairs Cooperative diabetes. Systolic Hypertension in the Elderly Pro- Study Group on Anti-hypertensive Agents. Results of gram (SHEP) Cooperative Research Group. JAMA combination anti-hypertensive therapy after failure of 1996; 276: 1886–1892. each of the components. J Hum Hypertens 1995; 9: 133 Gall MA, Hougaard P, Borch-Johnsen K, Parving HH. 791–796. Risk factors for development of incipient and overt 113 Haq IU et al. Lipid-lowering for prevention of coron- diabetic nephropathy in patients with non-insulin ary heart disease: what policy now? Clin Sc 1996; 91: dependent diabetes mellitus: prospective, obser- 399–413. vational study. BMJ 1997; 314: 783–788. 114 Lever AF, Ramsay LE. Treatment of hypertension in 134 Ravid M, Leug R, Rachmanni R, Lisner M. Long-term the elderly. J Hypertens 1995; 13: 571–579. renoprotecive effect of angiotensin converting 115 Kostis JB et al, for the SHEP Cooperative Research enzyme inhibitors in non-insulin dependent dia- Group. Prevention of heart failure by antihyperten- betes: a seven year follow-up. Arch Int Med 1996: sive drug treatment in older persons with isolated 156: 286–289. systolic hypertension. JAMA 1997; 278: 212–216. 135 Williams B. Insulin resistance: the shape of things to 116 Forette F et al. Prevention of dementia in randomised come. Lancet 1994; 344: 521–524. placebo controlled systolic hypertension in Europe 136 Madhaven S, Stockwell D, Cohen H, Alderman MH. (SystEur) trial. Lancet 1998; 352: 1347–1357. Renal function during antihypertensive treatment. 117 Levy D et al. The progression from hypertension to Lancet 1995; 345: 749–751. 137 Navis G, de Zeeuw D, de Jong PE. ACE-inhibitors: congestive heart failure. JAMA 1996; 275: 1557–1562. panacea for progressive renal disease? Lancet 1997; 118 Messerli FH, Grossman E, Goldbourt U. Are ␤-block- 349: 1852–1853. ers efficacious as first-line therapy for hypertension 138 Giatras I, Lau J, Levey AS, for the Angiotensin-Con- in the elderly? A systematic review. JAMA 1998; 279: verting-Enzyme Inhibition and Progressive Renal Dis- 1903–1907. ease Study Group. Effect of angiotensin-converting 119 Adler A et al. Risk of macrovascular and microvascu- enzyme inhibitors on the progression of nondiabetic lar complications of diabetes at different levels of renal disease: A meta-analysis of randomized trials. blood pressure—Observations from the UKPDS. Dia- Ann Intern Med 1997; 127: 337–345. betes 1999; 48: A16. 139 Gruppo Italiano di Studi Epidemiologici in Nefrolo- 120 Johnston CI, Cooper ME, Nicholls GM. Meeting report gia. Randomised placebo-controlled trial of effect of of the International Society of Hypertension confer- ramipril on decline in glomerular filtration rate and ence on hypertension and diabetes. J Hypertens 1992; risk of terminal renal failure in proteinuric, non-dia- 10: 393–397. betic nephropathy. Lancet 1997; 349: 1857–1863. 121 Colhoun HM et al. The scope for cardiovascular dis- 140 Joint National Committee on Prevention, Detection, ease risk factor intervention among people with dia- Evaluation and Treatment of High Blood Pressure betes in England. Diabetic Med 1999; 16: 35–40. and the National High Blood Pressure Education Pro- 122 Colhoun HM, Williams B. The cardiovascular risk gram Coordinating Committee. The sixth report of the burden of diabetes mellitus. Acta Diabetilogia 1999; Joint National Committee on prevention, detection, (in press). evaluation, and treatment of high blood pressure. 123 Cooper ME. Pathogenesis, prevention, and treatment Arch Intern Med 1997; 157: 2413–2446. of diabetic nephropathy. Lancet 1998; 352: 213–219. 141 Klahr S et al, for the Modification of Diet in Renal 124 Mogensen CE et al. Prevention of diabetic renal dis- Disease Study Group. The effects of dietary protein ease with special reference to microalbuminuria. restriction and blood-pressure control on the pro- Lancet 1995; 346: 1080–1084. gression of chronic renal disease. N Engl J Med 1994; 125 Parving HH. Initiation and progression of diabetic 330: 877–884. nephropathy. N Engl J Med 1996; 335: 1682–1683. 142 Wheeler DC. Cardiovascular disease in patients with 126 Lazarus JM et al, for the Modification of Diet in Renal chronic renal failure. Lancet 1996; 348: 1673–1674. Disease Study Group. Achievement and safety of a 143 Broughton Pipkin F. The hypertensive disorders of low blood pressure goal in chronic renal disease. pregnancy. BMJ 1995; 11: 609–613. Hypertension 1997; 29: 641–650. 144 Sibai BM. Treatment of hypertension in pregnant 127 The Euclid Study Group. Randomised placebo-con- women. N Engl J Med 1996; 335: 257–265. trolled trial of lisinopril in normotensive patients 145 Shennan A et al. Lack of reproducibility in pregnancy with insulin-dependent diabetes and normaoalbumi- of Korotkoff phase IV as measured by mercury sphyg- nuria or microalbuminuria. Lancet 1997; 349: 1787– momanometry. Lancet 1996; 347: 139–142. 1792. 146 Rubin P. Measuring diastolic blood pressure in preg- 128 Williams B. Lisinopril and albumin excretion in dia- nancy. Use the fifth Korotkoff sound. BMJ 1996; 313: betes. Lancet. 1997; 350: 663–664. 4–5. 129 The Hypertension in Diabetes Study Group. Hyper- 147 Collins R, Wallenburg HCS. Pharmacological preven- tension in Diabetes Study (HDS): I. Prevalence of tion and treatment of hypertensive disorders in preg- hypertension in newly presenting type 2 diabetic nancy. In: Chalmers I, Enkin M, Keirse MJNC (eds). patients and the association with risk factors for car- Effective Care in Pregnancy and Childbirth. Vol 1: diovascular and diabetic complications. J Hypertens Pregnancy. Ch 24, 1992, pp. 382–402. 1993; 11: 309–317. 148 Kyle PM, Redman CW. Comparative risk-benefit 130 UKPDS Group. UKPDS 23: risk factors for coronary assessment of drugs used in the management of Guidelines for management of hypertension British Hypertension Society 591 hypertension in pregnancy. Drug Safety 1992; 7: ischaemic heart disease and cerebrovascular disease 223–234. in England and Wales. BMJ 1991; 302: 560–564. 149 Naden RP, Redman CW. Antihypertensive drugs in 168 Wilhelmsen L, Strasser T, on behalf of the Study Col- pregnancy. Clinics in Perinatol 1985; 12: 521–538. laborators. WHO-WHL hypertension management 150 Butters L, Kennedy S, Rubin PC. Atenolol in essential audit project. J Hum Hypertens 1993; 7: 257–263. hypertension during pregnancy. BMJ 1990; 301: 169 Carney S, Gillies A, Smith A, Taylor M. Hypertension 587–589. education: patient knowledge and satisfaction. J Hum 151 Collins R, Yusuf S, Peto R. Overview of randomized Hypertens 1993; 7: 505–508. trials of diuretics in pregnancy. BMJ 1985; 290: 17– 170 Taylor FC, Ebrahim S. Prevention of cardiovascular 23. disease in hypertensive patients in general practice: 152 Sibai BM, Grossman RA, Grossman HG. Effects of improving the quality of care. Br J Cardiol 1997; 4: diuretics on plasma volume in pregnancies with 331–335. long-term hypertension. Am J Obstetrics Gynaecol 1984; 150: 831–835. Appendix 1 153 Hansenns M, Keirse MJ, Vankelecom F, Van Assche FA. Fetal and neonatal effects of treatment with angi- How to measure 10-year coronary heart disease otensin converting enzyme inhibitors. Obstetrics and risk Gynecol 1991; 78: 128–135. 154 Nichols M et al. Effect of four combined oral contra- The British hypertension society recommends that ceptives on blood pressure in the pill-free interval. the computer programme ‘Cardiac Risk Assessor’ Contraception 1993; 47: 367–376. based on the Framingham risk function and issued 155 World Health Organisation. The WHO multicentre by the Joint British Societies should be used wher- trial of the vasopressor effects of combined oral con- ever appropriate to estimate 10 year CHD risk. This traceptives. 1. Comparison with IUD. Contraception programme will also estimate 10-year CHD and 1989; 40: 129–145. stroke risk simultaneously. 156 Dong W, Colhoun HM, Poulter NR. Blood pressure The computer programme or risk chart should not in women using oral contraceptives: results from the be used in patients with: (i) established CHD or health survey for Endlagnd 1994. J Hypertens 1998; atherosclerotic disease, or (ii) familial hypercholes- 15: 1063–1068. terolaemia, because these patients are already at suf- 157 Wier RJ. Oral contraceptives, hormone replacement ficiently high CHD/CVD risk to justify intervention. therapy and hypertension. In: Swales JD (ed). Text- book of hypertension. Oxford Blackwell Scientific The following information is needed: Publications: Oxford, 1994, pp 904–922. (1) Name 158 Lim KG et al. Malignant hypertension in women of (2) Sex childbearing age and its relation to the contraceptive (3) Age pill. BMJ 1987; 294: 1057–1059. (4) Systolic BP (mm Hg) 159 World health organisation collaborative study of car- (5) Diastolic BP (mm Hg) diovascular disease and steroid hormonec contracep- (6) Smoking status tion. Ischaemic stroke and combined oral contracep- (7) Serum cholesterol (any units) tives: results of an international, multicentre, case (8) HDL cholesterol (same units as serum control study. Lancet 1996; 348: 498–505. cholesterol) 160 Wilson ESB, Cruickshank J, McMaster M, Weir RJ. A (9) Diabetic or non-diabetic propsective controlled study of the effect on blood (10) Presence of LVH on the ECG pressure of contraceptive preparations containing dif- The 10-year % probability of developing CHD ferent types of progestagen. Br J Ostet Gynaecol 1984: 91; 1254–1260. (fatal or non-fatal myocardial infarction) and stroke 161 Nabulsi AA et al. Association of hormone-replace- is estimated by the cardiac risk assessor computer ment therapy with various cardiovascular risk factors programme. When the CHD risk and stroke risk are in post-menopausal women. The Atherosclerosis added, this equates to 10-year CVD risk. Risk in Communities Investigators. N Engl J Med 1993; 328: 1069–1075. Family history: The estimated 10-year CHD risk 162 The writing group for the PEPI trial. Effect of oes- should be adjusted upwards by a factor of 1.5 for trogen or oestrogen/progestin regimens on heart dis- patients who have a first degree male relative ease risk factors in post-meopausal women. JAMA developing CHD or other atherosclerotic disease 1995; 273: 199–208. before the age of 55 years, or a female first degree 163 Hulley S et al. Randomised trial of estrogen plus pro- relative with a similar history before the age of 65 gestin for secondary prevention of coronary heart dis- years. ease in post-menopausal women. JAMA 1998; 280: 605–613. Ethnic minorities: The Framingham risk equation 164 Cappuccio FP, Cook DG, Atkinson RW, Strazzullo P. has not been validated in these populations. Prevalence, detection, and management of cardio- vascular risk factors in different ethnic groups in Joint British Societies CHD Risk Chart: When the south London. Heart 1997; 78: 555–563. 165 Relation of central obesity and insulin resistance use of the computer programme is impractical, the with diabetes prevalence and cardiovascular risk in Joint British Societies CHD risk chart provides a use- south Asians. Lancet 1991; 337: 382–386. ful alternative for the estimation of 10-year CHD % 166 Williams B. Westernised Asians and cardiovascular risk. CVD risk can easily be derived from CHD risk, disease: Nature or nuture? Lancet 1995; 345: 401– eg, a 10-year CHD risk of у15% is equivalent to a 402. 10-year CVD risk of у20%, or a 10-year CHD risk of 167 Balarajan R. Ethnic differences in mortality from у30% is equivalent to a 10-year CVD risk of у40%. Guidelines for management of hypertension British Hypertension Society 592 Appendix 2 Available from BMJ Publications • Blood Pressure Measurement – Recommendations Available from the British Hypertension Society of the British Hypertension Society. 3rd Edition. Information Service O’Brien E et al (ed) (£4.95). Blood Pressure Unit • BHS/BMJ CD-ROM – Recommendations for Blood St George’s Hospital Medical School Pressure Measurement. Cranmer Terrace The Joint British Societies ‘Cardiac Risk Assessor’ London SW17 0RE Computer programme and copies of the Joint British Tel: 0181 725 3412 Societies CHD risk assessment chart (Figure 2) can Fax: 0181 725 2959 be downloaded from the BHS web site. E-mail: bhsisȰsghms.ac.uk Website: www.hyp.ac.uk/bhsinfo and www.hyp.ac.uk/bhs • Patient information booklet: –Understanding High Blood Pressure. • Patient fact sheets: –Self-help Measures. –Antihypertensive Drugs –Blood Pressure Measurement –Reducing Dietary Salt –Blood Pressure and Kidney Disease • Diet sheet: –Healthy Eating Diet Sheet