The objective of this study was to assess the cost-effectiveness of prophylaxis for venous thromboembolism (VTE) in acutely ill medical patients using 40 mg enoxaparin od compared with unfractionated heparin (UFH) and placebo. An established decision tree model based on epidemiological data, clinical trials, and a recent meta-analysis was used to evaluate costs and consequences of alternative means of thromboprophylaxis in medical patients. Primary outcome measures were episodes of VTE (deep vein thrombosis or pulmonary embolism) and major bleeding. Secondary effectiveness measures included estimated mortality. Inhospital drug costs and administration time, in conjunction with long-term VTE complications up to 15 years and major bleeding, were considered. Results were estimated for a hypothetical cohort of 100 patients. The expected cost per 100

No abstract

The central problem is often poor design and conduct of trialsEditor-We are concerned about some aspects of the recent articles on consent. Len Doyal claims that informed consent may not be necessary for three most vulnerable groups: young children, patients with learning difficulties, and unconscious or semiconscious patients. Yet young children (unlike all adult groups) have the protection of their parents' consent, and this should always be respected. 3 The other two groups show the limitations of applying Kantian respect for autonomy, designed for property owning 18th century gentlemen, to vulnerable dependent patients. There is an urgent need to agree new ways of making research decisions with and for these minority groups.As is usual in arguments against seeking informed consent, there is a tendency to concentrate on dramatic extremes: patients with severe mental impairment and patients receiving heroic cancer treatment. The development of principles from extremes is dangerous and should be discouraged. Difficulties with relatively small groups should not be used to excuse researchers from requesting the consent of the vast majority of the millions of people every year who help with research into mundane pharmaceutical trials of treatments of arthritis or everyday misery.A deeper problem with the articles is the assumption that consent is the central problem in research. We suggest that, more often, the central problem is the poor design and conduct of trials which alienate or distress people on whose practical support researchers depend. The solution here is not to tinker with consent but to clean up research. Health service users could help at every stage of clinical research: the selection of questions worth investigating; the design and conduct of trials, including the information materials; the interpretation and reporting of the evidence; dissemination; and working with practitioners to put findings into practice. Consumers for Ethics in Research has been working with health service users, researchers, and practitioners on these issues for the past eight years, partly through regular open meetings, during which many practical ideas have been advanced. Naomi Pfeffer Honorary treasurer Journals should require routine reporting of consent ratesEditor-We wish to contribute to the debate on informed consent. 1 Two of us (HC, SAMS) have recently conducted a review of randomised control trials published in the Archives of Diseases in Childhood from 1982 to 1996. We found that 112 (45%) of 249 trials did not report whether informed consent had been obtained. Of the trials that did note that informed consent had been obtained, 111 (81%) of 137 quoted consent rates of 100%. This proportion varied by study setting and paediatric subspecialty and was particularly high in trials in inpatients (90%) and trials in neonates (96%). Two of the trials that reported 100% consent rates included over 500 children.Some of the trials may have considered obtaining patients' consent to be part of the inclusion criteria for ...