Search

The practice patterns of individual clinicians are fundamental determinants of the quality, ethical standards, and cost-effectiveness of health services. The uncertainties inherent in medical practice are the direct result of biological variability and an enormous range of interchanges between a host of factors. Until recently the uncertainties have been ignored or obscured, but their implications are now being exposed through changes in health service organisation. The implications for the quality and costs of health care are such that it has become mandatory for clinicians, purchasers, and providers to recognise and understand the nature, importance, and influence of uncertainty in determining patterns of action. Paradoxically, it is becoming recognised that power traditionally vested within the health professions, especially doctors, seems to rest as much on uncertainty as on technical expertise. 1 Uncertainty has been identified as a major factor common to the key controversies being debated in relation to health-care policy determination.2 More constructive approaches to uncertainty in health care must start with clinicians. For this to happen doctors first need to have a clearer understanding of the causes and impact of uncertainty on their practice and to develop new strategies for responding to it. Clinical decision making is complex, involving intuitive as well as rational thinking. It requires consideration extending beyond the illness or disease in order to take account of value judgments and social, psychological, and moral factors. Although uncertainty is inevitable, clinicians frequently downplay its importance. To some extent this is a defensive reaction taken consciously or unconsciously in response to challenges within an increasingly adversarial environment. Much of the criticism and challenge to traditional attitudes and practice comes from individuals and groups who previously have not been obvious participants in decision making within health services. Unfortunately, the ways in which clinicians respond to uncertainty and questioning of their practice may have adverse effects on complex decision-making processes,3 leading, for example, to unnecessary overinvestigation of patients in the desire to reduce diagnostic uncertainty to a minimum. Katz4 has gone so far as to say that the doctor's defences against ignorance and uncertainty are a greater difficulty in the physician-patient interaction than the patient's ignorance.

OBJECTIVE--To investigate long term changes in total cholesterol, high density lipoprotein cholesterol, and low density lipoprotein cholesterol concentrations and in measures of other risk factors for coronary heart disease and to assess their importance for the development of coronary heart disease in Scottish men. DESIGN--Longitudinal study entailing follow up in 1988-9 of men investigated during a study in 1976. SETTING--Edinburgh, Scotland. SUBJECTS--107 men from Edinburgh who had taken part in a comparative study of risk factors for heart disease with Swedish men in 1976 when aged 40. INTERVENTION--The men were invited to attend a follow up clinic in 1988-9 for measurement of cholesterol concentrations and other risk factor measurements. Eighty three attended and 24 refused to or could not attend. MAIN OUTCOME MEASURES--Changes in total cholesterol, high density lipoprotein cholesterol, and low density lipoprotein cholesterol concentrations, body weight, weight to height index, prevalence of smoking, and alcohol intake; number of coronary artery disease events. RESULTS--Mean serum total cholesterol concentration increased over the 12 years mainly due to an increase in the low density lipoprotein cholesterol fraction (from 3.53 (SD 0.09) to 4.56 (0.11) mmol/l) despite a reduction in high density lipoprotein cholesterol concentration. Body weight and weight to height index increased. Fewer men smoked more than 15 cigarettes/day in 1988-9 than in 1976. Blood pressure remained stable and fasting triglyceride concentrations did not change. The frequency of corneal arcus doubled. Alcohol consumption decreased significantly. Eleven men developed clinical coronary heart disease. High low density lipoprotein and low high density lipoprotein cholesterol concentrations in 1976, but not total cholesterol concentration, significantly predicted coronary heart disease (p = 0.05). Almost all of the men who developed coronary heart disease were smokers (91% v 53%, p less than 0.05). CONCLUSION--Over 12 years the lipid profile deteriorated significantly in this healthy cohort of young men. Smoking, a low high density lipoprotein concentration and a raised low density lipoprotein concentration were all associated with coronary heart disease in middle aged Scottish men, whereas there was no association for total cholesterol concentration. The findings have implications for screening programmes.

The insulin response to a standard oral glucose tolerance test (OGTT) and other anthropometric and biochemical risk factors for coronary heart disease (CHD) were measured in a random sample of 107 Edinburgh men, who were initially studied in 1976 when they were 40 and who were reexamined in 1988-89. Fasting glucose and glucose response to OGTT were higher in 1988-89 than in 1976. In contrast, insulin levels did not differ between the initial and follow-up study either before or after the glucose load. Body mass indices increased, except triceps skinfold. Changing patterns in both fasting and OGTT insulin or glucose levels in individuals were related to changes in bodyweight or in subscapular skinfolds. Modifications in serum total and HDL cholesterol were related to changes in fasting insulin and insulin area, respectively, but not to glucose data. Eleven men developed clinical CHD. Neither glucose nor insulin measures obtained in 1976 differed between those with and without CHD. Weight-height index and abdominal skin-folds were higher in those with CHD. HDL cholesterol was significantly lower (P less than 0.05). Abdominal skin-fold but not body mass index remained significant when adjusted for HDL cholesterol. This small study confirms the importance of central obesity and low HDL cholesterol but failed to identify insulin as a risk factor for CHD in this Scottish population.

107 healthy men in Edinburgh and 82 in Stockholm, all aged 40 and selected at random, took part in a study aimed at identifying factors which might explain why, in men of this age, the mortality-rates from ischaemic heart disease are three times greater in Edinburgh men. Clinical measurements and electrocardiographic procedures were standardised, and steps taken to eliminate interlaboratory differences. Edinburgh men were shorter and fatter; had higher systolic and diastolic blood-pressures; smoked more cigarettes (but the number of cigarette smokers in the two groups was similar); drank more alcohol; had more E.C.G. abnormalities and lower exercise tolerance. They had higher levels of total serum-triglycerides and triglycerides in very low (V.L.D.L.), low (L.D.L.), and high (H.D.L.) density lipoproteins. Triglyceride clearance (assessed by an intravenous fat-tolerance test) was similar. Serum-cholesterol and L.D.L.-cholesterol concentrations were similar in both cities but cholesterol in V.L.D.L. was higher, and cholesterol in H.D.L. was lower, in Edinburgh men. The relative linoleic-acid content of plasma-triglycerides and of plasma-cholesterol esters were lower in Edinburgh men, and their subcutaneous adipose tissue glycerides had a lower polyunsaturated/ saturated fatty-acid ratio and a strikingly lower relative linoleic acid content. Edinburgh men had a greater total plasma-insulin response to a standardised glucose-tolerance test. The glucose response was the same in both cities. Plasma free fatty-acids and free glycerol were reduced by a lesser extent in Edinburgh men in response to a glucose load. Although the high incidence of ischaemic heart-disease in Scotland may be due to many factors, some of which are well established, this study points out metabolic factors which may provide new information on the pathogenesis of ischaemic heart-disease.

The effect of linoleic acid on the induction of fatal ventricular fibrillation by intravenous CaCl2 (10%), was studied in rats fed for a month from weaning on a diet with either a high or low content of linoleic acid. Studies were performed in the basal state and after pretreatment with noradrenaline, which increased the sensitivity to CaCl2 equally in animals from both diet groups. Despite considerable differences in the linoleic acid levels in the plasma and myocardium, the two groups did not differ in the incidence of fatal ventricular fibrillation. Our conclusions concerning the effect of linoleic acid on cardiac arrhythmias, and sudden death in particular, are compared with those from other studies.

Free fatty acid concentrations in plasma measured after in vivo heparinisation are often overestimated because of ex vivo lipolysis of variable degrees. A new method has been developed using immediate extraction of blood which obviates this and shows that the true rise in plasma free fatty acid concentration after heparin in ambulant ward patients and in patients with acute myocardial infarction is less than previously reported. The small rise in plasma free fatty acid concentration after heparin is unlikely to have adverse metabolic effects in most patients during acute myocardial infarction.

The mortality from coronary heart disease (CHD) (ICD 410) in men aged 40 years is 3 times higher in Edinburgh than in Stockholm. This assessment is based on two community surveys of CHD incidence (1,2) and WHO mortality data (3). A pilot survey of CHD risk factors (4) suggested that there are significantly higher serum triglycerides, blood pressure, insulin production and cigarette smoking in Edinburgh compared with Stockholm but that serum cholesterol and weight are similar.

A trial of intravenous followed by oral propranolol, started within 4 h of onset of suspected myocardial infarction and continued over 27 h, was carried out in 735 patients; 364 received propranolol, 371 were controls. Ventricular fibrillation during the first 48 h after entry to the trial occurred in 2 treated patients and in 14 controls (p=0·006). Rates of hospital mortality, complications other than ventricular fibrillation, and progression from threatened to completed infarction did not differ between treated and control patients. Ventricular fibrillation was not apparently prevented by prior betablocker treatment, which was not a reason for exclusion from the trial. This intravenous/oral propranolol regimen seems to prevent ventricular fibrillation due to evolving myocardial infarction.