This week in the BMJ

Volume 326, Number 7398, Issue of 17 May 2003

Environmental tobacco smoke does not affect mortality
Mortality is high in young offenders
Standardised measures needed for data quality
Targeted programmes needed for hardcore smokers
Promising new technologies for heart failure are available

Environmental tobacco smoke does not affect mortality

Environmental tobacco smoke was not associated with mortalityfrom coronary heart disease or lung cancer at any level ofexposure in a large study of Californians followed for 40 years.Based on an analysis tracking 35 561 adults who had never smokedand their smoker spouses between the periods 1960 and 1998,Enstrom and Kabat (p 1057) suggest that the effects of environmentaltobacco smoke, particularly for coronary heart disease, areconsiderably smaller than generally believed. In contrast,active cigarette smoking was a strong, dose related risk factorfor coronary heart disease, lung cancer, and chronic obstructivepulmonary disease. Despite these findings, Davey Smith (p 1048)says debate about the role of environmental smoke, as wellas the risk of being married to a smoker, will continue. Additionalresearch on the link between second hand smoke and diseaseis needed.

NIKI BOROFSKY

Mortality is high in young offenders

Young people in Australia who have received custodial sentenceshave high mortality from drug overdose, suicide, and unintentionalinjury. In a retrospective cohort study, Coffey and colleagues (p 1064) compared death rates in young offenders with thosein young people in the general population of the same state.They found that rates in young offenders were nine times higherin males and 40 times higher in females. A quarter of all drugrelated deaths in 15-19 year old men were of young offenders.The authors say that social policies to address drug and mentalhealth problems and social disadvantage in young offendersare urgently needed.

SIPA/REX

Standardised measures needed for data quality

Quality measurement of electronic patient records in primarycare needs to be standardised. Thiru and colleagues (p 1070)conducted a systematic review of 52 publications on how dataquality in primary care is measured and reported. Most (92%)were concerned with patient identification data, and theirmeasurements varied. Reliability was typically measured throughrate comparisons, whereas validity was assessed through either sensitivity or predictive probabilities. The recording of diagnosticand prescription data tended to be better than that of othervariables. Until every practice is paperless and enters codeddata for every consultation, the authors recommend that setreference standards be used to determine completeness and accuracy.

ANTONIA REEVE/SPL

Targeted programmes needed for hardcore smokers

Hardcore smokers say they will keep on smoking. In a cross sectionalstudy of 7766 English smokers, Jarvis and colleagues (p 1061)found that 16% were hardcore, four times as many as previousresearch has suggested. Hardcore smokers were defined as thosewho had never attempted to stop smoking and did not wish orintend to give it up in the future, and those who had not gonea day without smoking cigarettes in the past five years. Hardcoresmoking was commoner in older and more dependent smokers, andin those from more socioeconomically deprived backgrounds.They tended to believe that their present health was unaffectedby smoking and was unlikely to be affected in the future. Despitehardcore smokers' recalcitrance, the authors say that targetedinterventions could help older and socially disadvantaged smokersto quit.

CRISTINA PEDRAZZINI/SPL

Promising new technologies for heart failure are available

The annual mortality from chronic heart failure remains as highas 40%. In this week's clinical review, Chow and colleagues (p 1073) discuss the effects of new pacing technologies onpatients with chronic heart failure. They report that the useof biventricular pacing improves symptoms and reduces the number of hospital admissions and lengths of stay. Implantable defibrillatorsreduce mortality, and forthcoming data from randomised controlledtrials suggest that a wider range of patients should receivethem. The authors outline the characteristics of patients whocould benefit from these technologies and the complicationsthat might arise.