Management of mild hypertension: Too little emphasis on metabolic factorsBMJ 1994; 308 doi: https://doi.org/10.1136/bmj.308.6926.470 (Published 12 February 1994) Cite this as: BMJ 1994;308:470
- P H Winocour
EDITOR, - I was disappointed that the World Health Organisation-International Society of Hypertension placed such little emphasis on the impact of metabolic factors on the management of essential hypertension.1 Insulin insensitivity accompanying hypertension is associated with established or subsequent dyslipoproteinaemia (low high density lipoprotein cholesterol concentration, hypercholesterolaemia, and hypertrigly-ceridaemia) and glucose intolerance in at least 30% of patients.2,3 These disturbances will be compounded by treatment with preparations that combine a thiazide with a β blocker, leading to an 11-fold increased incidence of diabetes over the following 10 years.4
Although the published guidelines recommend treatment with a diuretic in combination with a β blocker for mild hypertension,1 they contain no suggestion that screening for metabolic dysfunction should be routine in the work up of hypertensive patients as an aid to the selection of appropriate pharmacological treatment. Nor is the need for continued surveillance for the development of diabetes mentioned. This seems paradoxical since the paper acknowledges that “serious cardiovascular disease is also determined by . . . diabetes, dyslipidaemia, central obesity.”
I suspect that I am not alone in frequently stopping treatment with a preparation combining a thiazide with a β blocker in patients recently diagnosed as diabetic who are overweight or elderly and have antecedent hypertension; this often leads to improved metabolic control.
The introduction of the paper states that guidelines should provide critical and balanced information on benefits and limitations of the various therapeutic interventions. These guidelines seem lopsided to me.
Target systolic blood pressure too low
- J Coope
EDITOR, - The summary of the World Health Organisation-International Society of Hypertension guidelines on the management of mild hypertension contains a recommendation for target systolic blood pressure in elderly patients that is not supported by evidence from trials and might be dangerous if adopted in clinical practice.1 The guidelines advise that a goal of 140 mm Hg should be aimed at. Although the rider “if this is tolerated” is added, this can be tested only by attempts to lower the pressure to this level. The five randomised trials in elderly patients adopted different goals and achieved different reductions in the treatment groups (table).
Variations in the blood pressure achieved reflect to some extent different baseline blood pressures. In no instance, however, was 140 mm Hg adopted as a goal, and this would be difficult to achieve in most elderly patients with the pharmacological treatments available at present.
The substantial reductions in stroke produced by lowering blood pressure in these trials have been the result of the regimens that were tested, and these should form the basis of clinical recommendations.
- A Zanchetti
EDITOR, - Peter H Winocour's concern about metabolic disturbances in hypertension is well placed, and I believe that one of the novel aspects of the guidelines is the emphasis on the need for an overall assessment of cardiovascular risk in patients with hypertension. The paper published in the BMJ is a condensed version of the full guidelines; this summary does not include the section on evaluation, which mentions measurement of total and high density lipoprotein cholesterol and blood glucose concentrations among minimum investigations to be performed in hypertensive patients. Insistence on the role of additional risk factors, however, clearly implies the need to assess metabolic variables.
We also omitted a detailed discussion on the benefits and limitations of each of the main classes of antihypertensive agents, but the full text recognises that “diuretics may cause a variety of unwanted metabolic effects (principally potassium depletion and reduced glucose tolerance)” and that β blockers “have limitations in patients with dyslipidaemia or reduced glucose tolerance.” Although these contraindications are not specifically mentioned in the summary paper, they are implicit in the statement: “The appropriate choice of a particular class of antihypertensive drugs for a patient may also be determined by the person's other characteristics because differences in the risk profile and in side effects are extensive in different patients”.
John Coope also raises an important point - namely, the blood pressure that is the goal of treatment. Unfortunately, no firm guidelines can yet be provided by prospective trials as no trials, even those that Coope mentions, have randomised hypertensive patients to different levels of achieved blood pressure. The hypertension optimal treatment trial, which is being carried out at the moment, aims at providing an answer. Until this trial is completed the blood pressure chosen as the goal of treatment remains arbitrary. Coope seems to prefer to aim at the average blood pressure reached in trials (which was as low as 144/68 mm Hg in one of the trials he mentions). On the other hand, the guidelines committee chose to follow the indications provided by observational epidemiological studies and suggests that the goal of treatment should be the maximum reduction in blood pressure that can be tolerated.
The guidelines committee believes that “guidelines should not be seen as rigid constraints on a practising doctor's decisions.” Particularly when no firm evidence is available, as in the case of the blood pressure that should be the goal of treatment, it is certainly possible for Coope to take a more conservative approach.
Which guidelines to follow?
- G Feder
EDITOR, - The fact that the BMJ has published two sets of guidelines on managing hypertension - from the British Hypertension Society1 and the World Health Organisation-International Society of Hypertension2 - in one year raises questions about the development of guidelines. What are we to make of the differences between the two sets of guidelines? The overall treatment strategy and many of the details are the same, but there are slight differences in the recommended threshold for treatment and important differences in the goals of treatment and advice on exercise and or stopping drug treatment.
Both guidelines were developed by informal consensus, without explicit criteria for choosing the evidence for specific recommendations. There is surprisingly little overlap in the references cited The opaque relation between the guidelines and evidence from research makes it difficult for bemused clinicians to assess the validity of recommendations that differ between the two sets of guidelines. This may require a longer reference document or tables of evidence, but these do not exist for either guideline. The longer version of the WHO-International Society of Hypertension's guidelines is no more explicit about how evidence was chosen or weighted.
What is the responsibility of a medical journal in assessing clinical guidelines submitted for publication? Is peer review sufficient? Perhaps journals should use one of the tools for appraising guidelines developed by the American guidelines industry or one of the shorter versions being developed in Britain. With the current profusion of clinical guidelines, the BMJ has a role in encouraging higher standards of development.
*** Medical journals certainly have a responsibility to assess rigorously the clinical guidelines that are submitted to them. The International Committee of Medical Editors has this issue on the agenda for its meeting in Oslo in June, and we at the BMJ are trying to raise the quality of our assessment of guidelines. The most crucial issue is how the group producing the guidelines has assembled and assessed evidence in reaching its conclusions. Another important question is whether all the detail of how the evidence was assembled and assessed should be included in the published guidelines. - ED, BMJ