This Week in BMJ
BMJ No 7061 Volume 313
Saturday 5 October 1996
Measures of patient satisfaction may vary with
question wording
Summarising the first national survey of what users of the NHS in Scotland think about the service, the management executive suggested that around 90% of hospital patients were satisfied with the amount of information they received and with their involvement in decisions. Cohen et al (841), however, report a survey in one Scottish region which disputes this. Their survey used comparable questions to those in the national survey but differed quite radically in other respects. The survey results diverged in relation to patients being given time to ask questions about their treatment and get involved in decisions. An interesting contrast emerged between negative statements about hospital care and a corresponding list of positive statements. Patients seemed less likely to say that something "bad" happened than to say that something "good" failed to happen.
Thallium-201 imaging predicts cardiac events after
heart attack and thrombolysis
Thrombolysis has reduced mortality, but patients remain susceptible to farther cardiac events. Exercise electrocardiography poorly predicts future course, and alternative methods have rarely been tested in this setting. In their study of risk stratification techniques Basu et al (p 844) performed exercise electrocardiography and exercise and rest thallium-201 imaging in 100 patients with acute myocardial infarction treated with thrombolysis. Patients were classified according to the presence or absence of ischaemia on exercise electrocardiography and thallium-201 imaging. Over a mean follow up period of 21 months, 37 patients had events, with thallium-201 imaging predicting the event in 33 patients. Patients with no ischaemia on thallium-201 imaging had a low (12%) risk of events. The authors suggest that stress and rest thallium-201 imaging may be used to identify those patients who would benefit most from invasive investigations.
Findings linking heart disease and diabetes in one
ethnic group cannot be generalised
Heart disease and diabetes are intimately related. But assumptions about disease relationships may not apply to all ethnic groups. People of Black African descent worldwide appear to enjoy a degree of protection from heart disease, even though the prevalence of diabetes is high. A 20 year follow-up study of 77 African Caribbean and 150 European people with diabetes in London showed that heart disease deaths were significantly fewer in African Caribbeans than in Europeans (p 848). The African Caribbean protection from heart disease persists in people with diabetes; establishing the reasons for this would improve understanding of heart disease and may help reduce risks for all populations.
Women do no worse than men after a cardiac
event
Women have been reported to be more likely than men to die after admission to hospital after an acute cardiac event, despite lower incidence rates. Little is known, however, about the relation between sex and death rates before admission to hospital and how it may affect rates after admission. On p 853 Sonke et al studied case fatality in men and women before and after admission to hospital after an acute cardiac event by using data from a community based coronary heart disease register. Despite less favourable baseline characteristics, women have a lower case fatality before admission.This effect was balanced by a higher case fatality in those admitted to hospital. Adjustment for differences in baseline risk largely explained the differences after admission and farther improved the outcome in women before admission. These results show that the paradoxical higher mortality in women admitted to hospital after an acute cardiac infarction can be explained by differences in baseline characteristics and mortality before admission.
Cluster analysis is not for the unwary
Growing public awareness about environmental issues has led to an increased demand for an understanding of the basic principles underlying investigations into clusters and other spatial patterns of disease. On p 863 Olsen et al clarify the main statistical problems in investigating alleged clusters and exemplify them using the identification in the early 1980s of an excess of cases of leukaemia near the Sellafield nuclear site in north west England. They deal with the problems of identifying clusters and spatial gradients during disease surveillance and explain the pitfalls as well as the advantages of using disease maps. Although cluster analysis has proved inefficient in identifying causes of disease, it is important in addressing public concerns.
Lay people fear becoming addicted to
antidepressants
The national Defeat Depression Campaign was set up to meet concerns that up to 50% of people in the community with clinical depression did not go to their doctors for help and that depression was not recognised in many of those who did consult their doctor. On p 858 Priest et al report the results of a MORI survey for the campaign about the lay public's attitudes to depression. Most people interviewed (78%) thought that antidepressants were addictive, and only 16% thought that depressed people should be offered antidepressants. By contrast 91% thought they should be offered counselling. Patients being treated for depression often cannot be persuaded to continue treatment for the recommended six months. To ensure compliance doctors should emphasise early on that antidepressants are not addictive.