Skip navigation

NNTs for stroke prevention

Bandolier 17 carried a table of NNTs (numbers needed to treat) for cardiac interventions. People seemed to like this gathering together of information, though it isn't easy work. This month we have gathered together some information on NNTs for stroke prevention calculated from randomised trials of primary and secondary prevention.

NNT table

For NNTs to be comprehensive and comprehensible you have to know the population studied, their disease and its severity, the treatment, the comparator, and the duration of treatment as well as the outcome. So the table Bandolier offers has this information.

Randomised trials with outcome of total stroke (excluding TIA)

Study Population Treatments Period Relative risk NNT
Primary prevention (compared with placebo)
MRC [1] 17,354 individuals 36-64 years diastolic 90-109 mmHg benzoflurazide propranolol placebo 5.5 years 850 to prevent one stroke at one year
SHEP [3] 4736 individuals 60 years or older systolic 160-219 mmHg and diastolic <90 mmHg clorthalidone atenolol placebo 4.5 years 0.65 (0.51 - 0.83) 43 (27 - 95)
STOP [4] 1627 individuals 70-84 years systolic 180-230 mmHg and diastolic ≥90 mmHg or diastolic 105-120 mmHg atenolol hydrochlorothiazide plus amiloride metoprolol pindolol placebo 4 years 0.55 (0.30 - 0.97) 34 (20 - 123)
MRC [6] 4396 individuals 65-74 years systolic 160-209 mmHg diastolic <115 mmHg diuretic ß-blocker placebo 5.8 years 0.76 (0.59 - 0.98) 70 (36 - 997)
WOSCOPS [8] 6595 men 45-64 years cholesterol over 6.5 mmol/L privastatin placebo 4.9 years 0.90 (0.61 - 1.34) 641 (135 - no benefit)
Secondary prevention
CATS [2] 1072 patients 1 week to 4 months after stroke ticlopidine placebo 2 years 0.61 (0.44 - 0.84) 15 (9 - 41)
SALT [5] 1360 patients 50-79 years TIA or minor stroke in previous 3 months low dose aspirin placebo 2.7 years 0.84 (0.65 - 1.08) 38 (16 - 85)
4S [7] 4444 patients 35 - 70 years angina or MI cholesterol 5.5-8.0 mmol/L simvastatin placebo 5.4 years 0.64 (0.47 - 0.88) 65 (38 - 224)
ESPS2 [9] 18 years TIA or stroke in previous 3 months aspirin
aspirin plus dipyridamole placebo
2 years 0.82 (0.69 - 0.97) 0.84 (0.71 - 1.00) 0.63 (0.52 - 0.76) 1000) 18 (13 - 29)


The outcome we chose to extract from papers was total strokes - that is all non-fatal and fatal strokes. Some papers offered transient ischaemic attack (TIA) as well, but we chose not to include this. Only comparisons with placebo are offered, without any discrimination for particular drug treatment, either because it was not given in the original paper or because there was no difference except for one study [9].

The numbers of events which occurred over the study period were taken as the numerator. The denominator was the number of patients originally randomised. Because some of these studies included or concentrated on elderly people, after some years the numbers of people still in the study had fallen. So this form of analysis may give slightly different estimates from those in the original papers which used different methods. The table provides relative risk and NNTs with 95% confidence intervals. The order of papers is chronological.


Primary prevention

Two studies of antihypertensive treatment in hypertensive people over 60 years [3,4] have NNTs of about 40 to prevent one stroke over 4 years compared with placebo. This means that 40 people have to be treated for four years to prevent a stroke in one of them, who would have had a stroke if they had been given a placebo. The confidence intervals were wide, though.

The MRC study of treatment of hypertension in older adults [6] concluded that there was a significant reduction in strokes (101/2183 treated patients compared with 134/2213 with placebo). Differences became really apparent after 4 years, and the NNT was 70.

These results for single trials have to be compared with the overall NNT of 43 (31 - 69) over five years found in a meta-analysis of hypertension treatment in the elderly ( Bandolier 15 , [10]).

Secondary prevention

One of the secondary prevention studies [7] was designed to assess the effectiveness of cholesterol lowering. It did, however, show a benefit in reducing strokes, with a NNT of 65.

Three studies examining treatments in patients who had already had a stroke or TIA had lower (better) NNTs. Low-dose aspirin (75 mg) over a mean of 2.7 years had an NNT of 38 [5]. Ticlopidine over two years had an NNT of 15 [2]. Very low dose aspirin (25 mg) or dipyridamole alone had an NNT of about 40 over two years, but the effects were additive with a two-year NNT of 18 for the combined treatment [9].


People have to make their own minds up about the value of these treatments. They also have to remember that the NNT refers to a time period, and that it is probably not quite legitimate to "normalise" the NNTs by multiplying by the time period to get NNTs/year.

Nevertheless, knowing that there are treatments which prevent another stroke occurring, in a patient who has already had one, with NNTs of below 20 for two years, is encouraging.


  1. MRC trial of treatment of mild hypertension: principal results. British Medical Journal 1985 291: 97-104.
  2. M Gent et al. The Canadian American ticlopidine study (CATS) in thromboembolic stroke. Lancet 1989 i: 1215-20.
  3. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension. Journal of the American Medical Association 1991 265: 3255-64.
  4. B Dahlöf et al. Morbidity and mortality in the Swedish trial in old patients with hypertension (STOP-hypertension). Lancet 1991 338:1281-5.
  5. Swedish aspirin low-dose trial (SALT) of 75 mg aspirin as secondary prophylaxis after cerebrovascular ischaemic events. Lancet 1991 338:1345-9.
  6. MRC trial of treatment of hypertension in older adults: principal results. British Medical Journal 1992 304:405-12.
  7. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian simvastatin survival study (4S). Lancet 1994 344: 1383-9.
  8. J Shepherd et al. Prevention of coronary heart disease with privastatin in men with hypercholesterolaemia. New England Journal of Medicine 1995 333:1301-7.
  9. HC Diener et al. European stroke prevention study 2. Dipyridamole and acetylsalicylic acid in the secondary prevention of stroke. Journal of Neurological Sciences 1996 143: 1-13.
  10. CD Mulrow et al. Hypertension in the elderly. Implications and generalizability of randomized trials. Journal of the American Medical Association 1994 272: 1932-8.

previous or next story in this issue