This week in the BMJ
Volume 328, Number 7450, Issue of 22 May 2004Infections increase the risk of atopic dermatitis
Infections increase the risk of atopic dermatitis
Infectious diseases in the first six months of life increase the risk of atopic dermatitis. Analysing 24 341 mother-child pairs, Benn and colleagues (p 1223) found that the risk of atopic dermatitis increased with each infectious disease, and decreased with exposure to siblings, day care, pet ownership, and farm residence. These findings challenge the hygiene hypothesis, which holds that infectious diseases protect against allergic diseases. Better hygiene decreases the production of regulatory T cells, resulting in the emergence of allergies, says Watts in a commentary (p 1226); preliminary attempts at vaccination are based on stimulating the production of these T cells.
|
![[To top]](/icons/back.gif){kind=icon}
Lesbian, gay, and bisexual health carers face undue anxiety
Lesbian, gay, and bisexual health carers use various strategies in clinical examinations to deal with sexuality. Interviewing 16 healthcare professionals, Riordan (p 1227) found that health carers may be more or less open about their sexuality; desexualisation and interaction strategies help them to manage sexual prejudice; and their sexual identity may facilitate clinical encounters with certain groups of patients. They are under added pressure and not trained to deal with ethical and medicolegal anxieties.
|
More editors and authors from poor countries are needed
International representation of editors and authors of tropical medicine journals is inadequate. Keiser and colleagues (p 1229) found that in 2003 only 5% of the members of the editorial and advisory boards of tropical medicine journals, and 14 % of the authors of articles published in 2000-2 in the six leading journals, were from countries with a low development index. International research collaborations were common, but the number of articles written only by scientists from underdeveloped countries was low.
Simple mortality data may be misleading...
League tables, ranking hospitals by death rates, may not adequately represent the quality of care provided. Outcomes after subarachnoid haemorrhage were significantly different between Nottingham and Newcastle (unfavoraurable outcome 15% v 35%) between 1992 and 1998, report Mitchell and colleagues (p 1234). These differences disappeared when the effect of age and presenting conditions was included. Using crude results to guide clinical governance and policy making would have been pernicious, say the authors.
... and affected by local facilities
Variations of in-hospital death rates may be explained by differences in provision of services in the area for dying people. Seagroatt and Goldacre (p 1235) adjusted the published hospital standardised mortality ratios to allow for geographical differences in the percentages of deaths occurring in hospital. In most cases the adjustment brought hospital results closer together, and it also changed the rankings. Simple death rates may be difficult to interpret as quality measures, because hospitals may also be a place for care of the dying.
School education programmes can reduce obesity
A simple educational programme delivered to children can help tackle obesity. James and colleagues (p 1237) randomised 644 children aged 7-11 to a one hour session of advice on healthy drinking, or to no intervention. After one year, children in the intervention group were drinking 0.6 fewer glasses of carbonate drinks over three days (p = 0.02), but those in the control group were drinking 0.2 glasses more (p = n.s.). The percentage of overweight and obese children increased in the control group by 7.5% and decreased in the intervention group by 0.2%, mean difference 7.7% (CI 2.2 to 13.1).
|
Doctors are not equipped to judge their peers
Instruments developed for physicians to rate their peers in practice are of poor quality, and their validity remains questionable. Reviewing more than 4500 papers, Evans and colleagues (p 1240) identified only three rating scales that had psychometric data about their development or their validity and reliability. None referred to a theoretical framework, and they all lacked construct and criterion validity. Instruments developed for peer assessment have not been developed in accordance with best practice, say the authors, and should be used with caution.
What's new
Keep updated
Services
Tools
Online poll
Resources
Print issues
