Intended for healthcare professionals

General Practice

Case-control study of stroke and the quality of hypertension control in north west England

BMJ 1997; 314 doi: (Published 25 January 1997) Cite this as: BMJ 1997;314:272
  1. Xianglin Du, research fellow in clinical epidemiologya,
  2. Kennedy Cruickshank, senior lecturer in clinical epidemiologya (clinep{at},
  3. Roseanne McNamee, senior lecturer in medical statisticsa,
  4. Mohamad Saraee, research fellowb,
  5. Joan Sourbutts, research nurse, clinical epidemiology unita,
  6. Alison Summers, consultant in public health medicinec,
  7. Nick Roberts, consultant physiciand,
  8. Elizabeth Walton, public health intelligence officerc,
  9. Stephen Holmes, principal in general practice and clinical tutore
  1. a School of Epidemiology and Health Sciences, University of Manchester Medical School, Manchester M13 9PT
  2. b Department of Computation, University of Manchester, Institute of Sciences and Technology, Manchester M60 1QD
  3. c Department of Public Health, East Lancashire Health Authority, Nelson, Lancashire BB9 5SZ
  4. d Department of Medicine for the Elderly, Queen's Park Hospital, Blackburn, Lancashire BB2 3HH
  5. e The Surgery, Earby, Lancashire BB8 6QT
  1. Correspondence to: Dr Du or Dr Cruickshank
  • Accepted 21 November 1996


Objective: To examine the risk of stroke in relation to quality of hypertension control in routine general practice across an entire health district.

Design: Population based matched case-control study.

Setting: East Lancashire Health District with a participating population of 388 821 aged (80.

Subjects: Cases were patients under 80 with their first stroke identified from a population based stroke register between 1 July 1994 and 30 June 1995. For each case two controls matched with the case for age and sex were selected from the same practice register. Hypertension was defined as systolic blood pressure !160 mm Hg or diastolic blood pressure !95 mm Hg, or both, on at least two occasions within any three month period or any history of treatment with antihypertensive drugs.

Main outcome measures: Prevalence of hypertension and quality of control of hypertension (assessed by using the mean blood pressure recorded before stroke) and odds ratios of stroke (derived from conditional logistic regression).

Results: Records of 267 cases and 534 controls were examined; 61% and 42% of these subjects respectively were hypertensive. Compared with non-hypertensive subjects hypertensive patients receiving treatment whose average pre-event systolic blood pressure was controlled to <140 mm Hg had an adjusted odds ratio for stroke of 1.3 (95% confidence interval 0.6 to 2.7). Those fairly well controlled (140-149 mm Hg), moderately controlled (150-159 mm Hg), or poorly controlled (!160 mm Hg) or untreated had progressively raised odds ratios of 1.6, 2.2, 3.2, and 3.5 respectively. Results for diastolic pressure were similar; both were independent of initial pressures before treatment. Around 21% of strokes were thus attributable to inadequate control with treatment, or 46 first events yearly per 100 000 population aged 40-79.

Conclusions: Risk of stroke was clearly related to quality of control of blood pressure with treatment. In routine practice consistent control of blood pressure to below 150/90 mm Hg seems to be required for optimal stroke prevention.

Key messages

Key messages A case-control study based on the community stroke register and practice records showed a prevalence of hypertension of 61% for stroke patients and 42% in controls

Quality of control of blood pressure was clearly related to the risk of stroke, independent of baseline blood pressure

Detection and treatment rates of hypertension were high but control of blood pressure to below 150/90 mm Hg in treated hypertensive patients was only 33% in cases and 42% in controls

When achieving optimal control of hypertension (to <150/90 mm Hg) in the most at risk and treatable age range (40-79 years) 86 hypertensive patients currently not well controlled need to be treated over five years to prevent one stroke


    • Accepted 21 November 1996
    View Full Text