All you need to read in the other general journalsBMJ 2011; 343 doi: http://dx.doi.org/10.1136/bmj.d4746 (Published 27 July 2011) Cite this as: BMJ 2011;343:d4746
Antidepressants don’t work for older people with depression and dementia
Doctors should think carefully before prescribing antidepressants to depressed adults with Alzheimer’s disease, say researchers from the UK, after a landmark trial reported that the two most widely prescribed agents worked no better than placebo. Sertraline and mirtazapine did not reduce depression scores for older people referred to psychiatric services with dementia and depression, and they had no lasting impact on more than a dozen secondary outcome measures, including quality of life for patients or carers. Both drugs caused significantly more side effects than the placebo⇓.
All three groups of patients improved during the first 13 weeks of the trial. The researchers are confident that improvements had little to do with either drug treatment but more to do with the usual care given to all participants by specialist psychiatric services for older people. Usual care depended on local protocols, but it would have included a variety of supportive and problem solving strategies in the first instance.
The 326 participants had a mean age of 80, moderately severe Alzheimer’s disease, and clinically important depression, often for more than six months. The researchers and a linked comment (doi:10.1016/S0140-6736(11)61031-3) agree that an antidepressant now looks like the wrong first line choice for the many patients with similar profiles. Adults at risk of suicide were excluded from this placebo controlled trial.
Rural isolation is linked to higher mortality from COPD
In a cohort of 26 591 men admitted to US hospitals with an exacerbation of chronic obstructive pulmonary disease (COPD), those living in isolated rural areas were significantly more likely to die than those living in more urban areas (5.0% v 3.8%; adjusted odds ratio 1.42, 95% CI 1.07 to 1.89).
All were admitted to hospitals managed by the Veterans Affairs Administration, which provides patient care to veterans and their dependents, between 2006 and 2008. The link between living in an isolated small town and mortality from COPD was independent of case mix and some hospital characteristics, including volume of patients with COPD managed during the study. Distance between home and hospital didn’t explain the findings either. In fact, those who travelled further for treatment had lower mortality in this study.
The authors found no significant increase in mortality for men living in country towns that were not isolated.
We don’t yet know why living in an isolated small town might disadvantage men with COPD, because the authors could explore only those factors that are routinely recorded and stored by Veterans Affairs hospitals. They suggest that future studies should take a closer look at specialist services that might be hard to access from isolated small towns, including specialist pulmonary physicians.
Risks and benefits are easier to understand as percentages
Patients need accurate information about risks and benefits before they make important decisions about drug treatments. They also need to understand that information, and researchers recently compared five different ways of presenting it, in a randomised trial⇓.
US adults seemed to understand percentages best. Participants given numerical information in this format were more likely to pass a comprehension test than those given the same information as frequencies (so many events out of a hundred or a thousand). The difference was small, but consistent. Combining percentages with frequencies did not improve test scores any further.
Researchers drafted tables giving the absolute benefits and harms associated with a drug to treat heartburn and a statin, so respondents had to grapple with rare events (muscle breakdown), as well as big improvements in a common symptom (heartburn). Natural frequencies, the format favoured by many expert groups, did not improve comprehension in this trial.
It may be time for a rethink, say the researchers. Percentages are simpler than frequencies and avoid the confusion caused by “denominator neglect.” Participants given variable frequencies were told that the chance of muscle breakdown after taking a statin was four in 10 000, and the chance of liver inflammation was one in 100. Only around 40% correctly understood that muscle breakdown was less common. The rest saw only that four is bigger than one so assumed muscle breakdown was the bigger risk.
Internal carotid measurements improve prediction of cardiovascular disease
Abnormal thickening of the carotid wall is a risk factor for cardiovascular events, although experts disagree about whether it adds anything useful to prediction based on clinical factors such as age, sex, smoking, and cholesterol concentrations. Researchers from the US recently reported that ultrasound measurements of the carotid wall can improve prediction, compared with the Framingham score alone, but that measuring the wall of the internal carotid artery adds more than measuring the wall of the common carotid. Looking for plaque in the internal carotid also added predictive power, although both improvements were fairly modest.
The researchers used data from the well established cohort of Framingham offspring—nearly 3000 US adults questioned and examined during the late 1990s when they were all free of cardiovascular disease. There were 296 cardiovascular events during just over seven years of follow-up.
As expected, all measures of carotid wall thickness helped predict events. But only the maximum thickness of the intima media layer of the internal carotid added clinically useful power to the Framingham risk score, correctly reclassifying 7.6% of adults as high, low, or intermediate risk. The presence of plaque (intima media thickness >1.5 mm) correctly reclassified 7.3% of participants, also a significant improvement.
US guidance currently emphasises measurement of the common carotid, say the researchers. That may have to change.
No need to hunt for micrometastases in women with early breast cancer
Routine examination of sentinel lymph nodes can miss isolated occult metastases in women with early breast cancer. Should pathologists look harder, using more sophisticated immunological tests? A cohort study from the US suggests it would make little difference to overall survival. Using immunochemical techniques, pathologists found occult metastases in 10.5% of 3326 women whose sentinel lymph nodes looked free of cancer on routine tests. These women were no more likely to die during the next five years than women with no occult metastases (95.1% v 95.7% survival; adjusted hazard ratio 0.88, 95% CI 0.45 to 1.71). Immunochemical test results did not inform treatment, and most women had both radiotherapy and adjuvant chemotherapy after their lumpectomy (2498/3247; 76.9%)⇓.
Pathologists also looked for occult metastases in bone marrow aspirates from 3413 women in the same cohort. Just 104 (3.0%) aspirates were positive, and while crude analyses hinted at reduced survival for these women, the association disappeared when researchers adjusted for well known prognostic factors such as age, tumour type, and the presence or absence of oestrogen receptors (1.83, 0.79 to 4.26).
All participants had early breast cancer—mostly stage 1 (83.3%) invasive ductal carcinomas (80.1%) that were oestrogen receptor positive (81.2%). Hunting for micrometastases in either sentinel lymph nodes or bone marrow isn’t justified for these women, say the researchers.
Another reason to keeping moving in old age
Two observational studies add to growing evidence that regular exercise can help maintain cognitive function in later life. Both find a significant association between physical activity and later cognitive function in cohorts of older US adults.
The first study reports secondary analyses from a large randomised trial of antioxidant vitamins that was launched in 1995. The authors used data from 2809 older female health professionals with a high risk of cardiovascular events. The most active women had the slowest decline in cognitive function over five years. The second reports a small substudy (n=197) of an established cohort of men and women with a mean age of 75 years. The authors used doubly labelled water to capture all activity including fidgeting, standing, and other low intensity movements. This technique measures the elimination of hydrogen-2 and oxygen-18 after a bolus dose of both isotopes and provides a proxy measure of total energy expenditure. Higher active energy expenditure was associated with a lower incidence of cognitive impairment over two to five years.
These studies may be limited by their observational designs, but they do support the growing consensus (informed by at least two trials), that exercise is good for mind as well as body, says a linked comment (p 1258). Doctors should be advising adults in midlife and beyond to keep moving for as long as possible.
Cranberry capsules or antibiotics for women with recurrent urinary infections?
Cranberries have a reputation for preventing recurrent urinary tract infections, thanks to at least two successful placebo controlled trials. Cranberries looked less successful in a recent head to head trial against standard prophylaxis with trimethoprim-sulfamethoxazole: women taking the “natural” treatment for a year developed significantly more urinary tract infections, significantly faster than controls (4 v 1.8 infections, P=0.02; 4 v 8 months to first infection, P=0.03).
Antibiotics clearly worked better for this selection of 221 young women, who reported six infections on average during the year before recruitment. But the authors and a linked comment (p 1279) agree that women shouldn’t give up on cranberries just yet. Prophylaxis with trimethoprim-sulfamethoxazole caused resistance to develop quickly in isolates of Escherichia coli grown from samples of both faeces and urine. After a month of antibiotics, 86.3% of faecal isolates and 90.5% of urinary isolates were resistant to this antibiotic. Resistance rates were 23.7% and 28.1% after a month of cranberry capsules. Women should still be given a choice, say the authors.
Future head to head trials should consider using bigger doses of cranberries or cranberry extract, says the comment. The bioavailability of the active ingredient (type A proanthocyanidins) is notoriously poor, and the doses of cranberry used in this trial were probably no match for an antibiotic with a bioavailability of 90%.
Cite this as: BMJ 2011;343:d4746