This site is intended for health professionals only

At the heart of general practice since 1960

Tips for achieving maximum points for hypertension

Are hypertension drugs really saving lives?

Dr Malcolm Kendrick argues that the results of clinical trials of hypertensives are wildly out of kilter with the benefits claimed

Post-hoc statistical interpretation is rife in hypertension studies. Terms like 'risk reduction, number needed to treat (NNT), odds ratios, lives saved, and disease prevention' are used with little explanation.

In truth these terms carry very little meaning. Risk reduction means almost nothing without knowing if this is relative or absolute. Equally, odds ratios are almost impossible to interpret unless you know what the baseline risk may have been. For example, the risk of lung cancer for a non-smoker is extremely small. Yet, if passive smoking increases the risk from one in 1,000 to one in 500 this can be presented as increasing the risk by 100 per cent.

If enough individuals are studied, this information can also be trumpeted as highly statistically significant with very narrow confidence intervals, and may take the form: (Absolute mortality RR: 2.0 95% CI 1.9-2.1 P<0.0001). which="" may="" look="" impressive,="" but="" the="" absolute="" risk="" is="" rather="">

Lives saved?

Perhaps the most misleading terminology of all is the current trend to talk about 'lives saved' by treatment. In reality, it is only possible to state that a life has been saved if a disease has been fully cured and will never return.

The mere fact that more people are alive in the treatment arm than the placebo arm at the end of a clinical trial doesn't mean that lives have been saved, nor clinical events prevented. All that can be said is that death has been delayed by some amount.

In truth, cardiology should follow the lead of oncology in reporting trial results. In this discipline the terminology generally used is 'increase in median survival'.

Which means, by how much longer do 50 per cent of patients live if they are given the treatment?

This may seem somewhat austere. All doctors prefer to believe they are saving lives, rather than increasing median survival, but unless we are truly 'curing' a disease, or preventing it from happening ­ forever ­ we cannot define treatment as lifesaving.

This principle is critical to any meaningful calculation of the absolute benefits of antihypertensive treatment. When, for example, someone states that 20,000 lives a year could be saved by improved blood pressure control, the claim should really be that the deaths were delayed by 'some period of time'. And the period of time gained needs to be worked out.

Small rewards

Using the 'time period gained' principle, when we look at the hypertension trials we need to strip aside probabilities, confidence intervals, odds ratios, risk reduction and lives saved and look at the actual figures. All we need to know are three basic pieces of information:

·How long was treatment given for?

·How many more people were alive at the end of the trial in the treatment arm?

·What is the likely increased survival time, on average, of those people kept alive?

Using this information we can calculate the number of years of treatment required to gain one extra year of life. It should be said that in the analysis that follows I have excluded non-fatal events (eg non-fatal stroke, or MI). I have also ignored the issue of side-effects, or events that may have been caused by the treatment. In short I have simply reviewed life-years gained using active treatment against placebo in randomised controlled clinical trials in hypertension.

My analysis of lives saved

There have been 20 such trials1-20. Many other hypertension trials have been done, but they have looked at the use of one antihypertensive against another ­ no major differences seen.

It's true that the 20 trials had different protocols, and lasted for different time periods, and looked at different degrees of hypertension, in different age groups. However, in this 'meta-analysis' they have simply been aggregated.

There were 42,972 people in all trials which lasted, on average, just under 3.5 years, representing 66,870 years of drug treatment. At the end of all trials 1,303 patients in the treatment arms had died, compared with 1,453 in the placebo arm. In short, 66,870 years of drug treatment resulted in 150 more people being kept alive. Which is a total of 445.8 years of treatment for each extra life.

My analysis of life expectancy increase

The next question to answer is what was the average increase in life expectancy of those 150 people?

Figures show 1,303 patients dying in the treatment arm and 1,453 in the placebo arm at the end of 3.5 years. By extending the treatment arm to the point where 1,453 deaths would also have occurred ­ assuming a reasonably linear progression ­ it's clear that it would take about six months to 'catch up' with placebo. Which means that the average increase in life expectancy gained is around six months. Therefore 66,870 years of antihypertensive treatment resulted in 75 extra years of life, which works out at 892 years of treatment for each year.

Taking this to the individual level, a patient taking antihypertensive treatment for 30 years can expect to gain 30/892 extra years. Which is a total of 12.2 days.

And that is the absolute benefit of antihypertensive treatment on mortality.


11 Veterans Administration

Co-operative Study Group on Antihypertensive Agents: Effects of treatment on morbidity in hypertension: results in patients with diastolic blood pressures averaging 115 through 129mmHg. JAMA 1967;202:116-22

12 Veterans Administration

Co-operative Study Group on Antihypertensive Agents II. Effects of treatment in

patients with diastolic blood pressure averaging 90

through 114 mmHg. JAMA.


13 Carter AB. Hypotensive therapy in stroke survivors. Lancet 1970;1:485-9

14 Barraclough M, Joy MD et al. Control of moderately raised blood pressure: report of a co-operative randomised controlled trial. BMJ. 1973;3:434-6

15 Hypertension-Stroke Co-operative Study Group. Effect of antihypertensive treatment on stroke recurrence.

JAMA. 1974:229:409-18

16 Smith WM. Treatment of mild hypertension: results of a 10-year interventional trial.

Circ Res. 1977;40(5 suppl 1)

17 Perry Jr HM, Goldman AI et al. Evaluation of drug treatment in mild hypertension: VA-NHLBI feasibility study. Ann N Y Acad Sci. 1978;304:267-88

18 Hegeland A. Treatment of mild hypertension: a five-year controlled trial: the Oslo Study.

Am J Med. 1980;69:725-32

19 The Australian therapeutic trial in mild hypertension: report by the management committee.

Lancet 1980;1:1261-7

10 Kuramoto K et al. Prospective study on the treatment of mild hypertension in the aged.

Jpn Heart J. 1981:22:75-85

11 Amery A et al. Mortality and Morbidity from the European Working Party on High Blood Pressure in the Elderly trial.

Lancet 1985;1:1349-54

12 Medical Research Council Working Party. MRC trial of treatment of mild hypertension: principal results. BMJ 1985;291:97-104

13 Coope J, Warrender TS. Randomised trial of treatment of hypertension in elderly patients in primary care.

BMJ 1986;293:1145-51

14 SHEP Co-operative Research Group. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension: final results of the Systolic Hypertension in the Elderly Program (SHEP) JAMA. 1991;265:3255-64

15 Dalhof B, Lindholm LH et al. Morbidity and mortality in the Swedish Trial in Old Patients with Hypertension (STOP-Hypertension). Lancet. 1991;338:1281-85

16 Medical Research Council Working Party. Medical Research Council trial of treatment of hypertension in older adults: principal results.

BMJ. 1992;304:405-12

17 The Dutch TIA trial study group. Trial of secondary prevention with atenolol after transient ischaemic attack or non-disabling ischaemic stroke. Stroke. 1993;24:543-8

18 PATS collaborating group. Post-Stroke Antihypertensive Treatment Study: a preliminary report. Clin Med J (Engl) 1995;108:710-17

19 Eriksson S et al, for the TEST study group. Atenolol in secondary prevention after stroke.

Cerebrovasc Dis. 1995;5:21-5

20 Staessen JA, Thijs L et al. Update on the Systolic Hypertension in Europe (SYST-EUR) Trial. Hypertension 1999;33:1476-7

Malcolm Kendrick is a GP in Macclesfield, Cheshire, and the most active UK member of the International Network of Cholesterol Skeptics ­ Dr Kendrick developed the educational website for the European Society of Cardiology

Rate this article 

Click to rate

  • 1 star out of 5
  • 2 stars out of 5
  • 3 stars out of 5
  • 4 stars out of 5
  • 5 stars out of 5

0 out of 5 stars

Have your say