This prospective study investigated the common belief that pyrexia is frequently absent in elderly patients with infection. Oral temperature was closely monitored using both a mercury and an electronic thermometer in 150 ill elderly patients (mean age 81 years) of whom 80% were new admissions to this Unit. A scoring system was devised, based on investigation results and excluding temperature, to assess objectively the likelihood of infection. Seventy-one patients (47%) had 'definite' infection: 95% were pyrexial. A further eight of the nine patients with probable infection were pyrexial. There were no significant differences in mean temperature or other indices of infection between those who died of their infection and those who survived. Ten per cent of all pyrexias were detected only on the electronic thermometer, not on mercury measurement. In 12% of pyrexial patients, the pyrexia first appeared more than 12 h after temperature measurement started. With effective monitoring, pyrexia is detectable in the vast majority of infected elderly patients.

No abstract

The haemodynamic response to postural stress (60 degrees foot-down tilt) was measured by impedance cardiography in six elderly cardiovascular-normal patients and 39 with symptomatic postural hypotension (systolic blood pressure drop greater than or equal to 20 mmHg or more). In the normal elderly the mean increase in heart rate, fall in blood pressure and cardiac output, and rise in peripheral resistance was less than that described in younger subjects. The changes were at their maximum in 1 min, and there was little further change over the next 5 min. In those with postural hypotension, orthostatic reduction (or failure to rise) of the peripheral resistance was the mechanism in 83% of cases, whatever the cause, and the time course of the haemodynamic changes was the same in the majority as in the normals. Serial tests in patients whose postural hypotension was controlled (by cessation of causal drugs, often multiple, by fludrocortisone, or by dihydroergotamine) showed return to normal.

A postal questionnaire sent to all consultant geriatricians in Great Britain and Northern Ireland determined that less than one consultant in five offered influenza vaccine to patients in continuing-care wards. The main reasons given were that vaccine was inappropriate or unnecessary. This information prompted a prospective study of viral illness during the winter months of 1986-87 in eight continuing-care wards with a population of 196 patients. There were 70 episodes of influenza-like illness (ILI), but only 17 viruses were isolated, the commonest being rhinovirus (seven patients). As most cases of ILI in this population were caused by viruses other than influenza, the reluctance of most geriatricians to give influenza vaccine to continuing-care patients appears justified.

It has been suggested that patients with essential hypertension may fall into three groups8: those with high renin levels, who respond best to beta-blockers; those with low renin levels, who respond best to diuretics; and those with normal renin, who respond equally to beta-blockers or diuretics. We could not identify any such subgroups in these few patients or any biochemical of physical marker that could have predicted a response to a diuretic or beta-blocker. Certainly atenolol had a greater hypotensive effect than bendrofluazide in some patients, but an equal number did better with bendrofluazide than with atenolol. This difference in response bore no relation to the hyporeninaemic effect of atenolol.The decision to include patients with severe hypertension in a trial incorporating a placebo period was not taken lightly. Most of the patients in the trial were referred to us because they were not being controlled satisfactorily on existing regimens and it was not clear whether this was due to the regimens themselves or to lack of patient compliance. We regarded it as essential to establish the true level of untreated blood pressure and the degree of patient compliance by having a closely supervised placebo period and incorporating a riboflavin marker into one of the tablets. We found that the pressure levels recorded during our placebo period differed little from those achieved when the patients were on their previous "treatment" regimens.The agent of first choice for treating hypertension is likely to depend on many factors. So far as atenolol and bendrofluazide are concerned there was no significant difference in their effect on systolic blood pressure, although atenolol was more effective than bendrofluazide on diastolic blood pressure (P <0-05). The biochemical effects produced by the two agents may, however, be important in deciding which should be regarded as first-choice treatment. The acute and long-term effects of bendrofluazide (hypokalaemia, hyperuricaemia, and a tendency towards hyperglycaemia) are well known but are clinically not important. The long-term effects of atenolol are not yet known but it appears to have several, possibly advantageous biochemical effects-for example, reduction in plasma renin, a slight increase in serum potassium, and a small reduction in urate. It remains to be seen whether atenolol confers the same benefit in respect of myocardial infarction as has been shown with practolol9 and alprenolol.10 11Drugs and matching placebos were kindly supplied by the Boots Company Limited, Ciba, and ICI Limited. Renin measurements were made by Dr David Craven in the professorial department of obstetrics and gynaecology of the City Hospital, Nottingham.