All you need to read in the other general journalsBMJ 2011; 343 doi: https://doi.org/10.1136/bmj.d4220 (Published 06 July 2011) Cite this as: BMJ 2011;343:d4220
Text messages help smokers to quit
“This is it! - QUIT DAY, throw away your fags. Today is the start of being QUIT forever, you can do it!” This and other motivational text messages helped adults from the UK quit smoking in a recent trial. The texts also told recipients where to go for help, how to beat cravings, and what to do after a lapse. After six months of texting, including five a day for the first five weeks, one in 10 recipients managed to quit completely (10.7%). Controls did half as well (4.9%, relative risk of quitting for intervention group 2.20 (95% CI 1.80 to 2.68)).⇓ Controls received fewer texts, which were unrelated to smoking, such as “Thanks for taking part! Without your input the study could not have gone ahead!”
Researchers verified abstinence with salivary concentrations of cotinine, and also checked that the texts didn’t cause any extra traffic accidents or thumb pain. Almost 600 adults took part. All wanted to quit, and most had tried before.
Two thirds of the world’s population now own a mobile phone, say the researchers, and scaling up an intervention like this could be relatively straight forward. A barrage of texts seems to work about as well as other behavioural approaches such as counselling. Cost effectiveness analyses are under way.
A 10% quit rate is low, says the accompanying editorial, but every little helps. Researchers should persevere with an approach that has the potential to reach all sections of society equally. Mobile phone ownership is largely independent of income and social class, unlike smoking.
CT screening for lung cancer looks promising; questions remain about harms
In 2002 researchers from the US launched a trial comparing low dose computed tomography (CT) with chest radiography for lung cancer screening among high risk adults. The results, announced last year and published last week, show a clear mortality advantage for computed tomography (relative reduction 20% (95% CI 6.8% to 26.7%) for lung cancer mortality and 6.7% (1.2% to 13.6%) for all cause mortality).⇓ US authorities are now considering whether to commission screening on a larger scale.
The authors and a linked editorial agree that it’s much too early, and decisions should wait for better information about overdiagnosis, other harms, and costs. Cost effectiveness data are on the way, but overdiagnosis—finding cancers that would never have caused symptoms or death—is hard to quantify without many years of follow-up. The key indication is an enduring difference in the number of cancers detected in screened and control populations. The difference between CT and radiography endured for 6.5 years in this trial, says the editorial, so overdiagnosis with CT looks likely.
These trial participants were aged 55–74 and had a history of heavy smoking. They were screened once a year for three years. Almost 40% of the 26 722 adults screened with CT had at least one positive test that required further investigation (39.1%). There were 18 146 positive tests in this group, and just 649 (3.6%) of the abnormalities turned out to be malignant. False positives were also common with radiography screening (94.5% of positive tests), although only 16% of adults in this group had a positive test.
Serious complications were rare during diagnostic work up for false positive results in both groups (0.06% for CT and 0.02% for radiography), although six adults died soon after an invasive procedure to investigate a suspicious but ultimately benign abnormality detected by CT.
Half the PCIs done for non-acute indications are inappropriate or uncertain
More than half a million US citizens have a percutaneous coronary intervention (PCI) each year. Most procedures done for acute indications such as heart attack are appropriate, say researchers, but around half of the PCIs done for non-acute indications may not be.
They matched data from an established PCI register to appropriateness criteria recently developed by six professional organisations. Among 500 154 PCIs, more than a quarter (28.9%, 144 737) were done for patients without an acute indication. Almost 12% (16 838) of these procedures were classified as inappropriate or unlikely to improve a patient’s health or prolong survival. A further 38% (54 988) were classified as uncertain. Most inappropriate procedures were done for patients with little or no angina, low risk ischaemia on non-invasive testing, or low risk coronary artery disease on angiography. Of these patients, 96% were treated inadequately with antianginal drugs.
More than 1000 US hospitals contributed to the registry. The proportion of non-acute PCIs that were inappropriate varied from none to 55% among hospitals. Differences in patient volumes or source of funding (private or public) explained very little of the variance.
These results suggest that hospitals, professionals, and their organisations need to find a way to target non-emergency PCI better, say the authors. Some hospitals need more help than others.
Top-up parenteral nutrition sooner or later for critically ill adults?
Parenteral feeding is usually reserved for critically ill adults who can’t get enough energy and nutrients from enteral feeding alone. The top-up can begin sooner or later after admission to an intensive care unit (ICU), and a head to head trial comparing the two approaches reports that later is probably better. Adults given early parenteral nutrition (from day two) stayed one day longer on ICU (4 v 3 days, P=0.02), two days longer in hospital (16 v 14, P=0.004), and had slightly but significantly more complications such as infections (605/2312 (26.2%) v 531/2328 (22.8%), P=0.008) than adults allowed no top-up parenteral nutrition until day eight.⇓ Early and late strategies were associated with similar death rates up to 90 days after randomisation.
All the adults in this trial were at risk of nutritional deficiency and began enteral nutrition on day two, combined with intravenous vitamins, minerals, and trace elements. Standardised protocols aimed to keep blood glucose concentrations within normal limits. Cardiac surgery was the most common reason for admission to ICU. A fifth of all participants had sepsis.
International guidelines currently disagree about the best time to start additional parenteral nutrition, and these findings should help reconcile their differences, or at least inspire further research, says a linked editorial. A delay of a week did these patients no harm and was associated with discernible benefits. We still don’t know why. Late initiators had a higher risk of hypoglycaemia (3.5% v 1.9%, P=0.001).
Water will do where soap is scarce
Hand washing with soap helps prevent childhood diarrhoea in rural areas of developing countries. But soap is expensive, hard to come by, and often kept some way from the communal water source. Hand washing with water alone is much more common in these communities, and may also protect children from diarrhoea, according to a study from rural Bangladesh. Young children in families where adults regularly washed both hands with water before preparing food had significantly less diarrhoea than children in families where adults did not regularly wash their hands before preparing food (odds ratio 0.67 (95% CI 0.51 to 0.89)).
Hand washing before preparing food or after a trip to the toilet looked more important than hand washing before feeding a child, before eating, or after cleaning a soiled child in these analyses. All were adjusted for characteristics that might confound the association between more washing and less diarrhoea, such as income and education. As expected, hand washing with soap, though unusual, was associated with the lowest risk of diarrhoea among children under 5 years old in these 347 households in 50 villages.
Local hygiene programmes continue to promote soap here and elsewhere, say the authors. They also promote an unrealistic schedule of hand washing that would leave low income parents with time for little else. A sharper focus on washing before preparing food and after going to the toilet might be a reasonable compromise, they write.
Longer telomeres, longer life (probably)
Chromosomes have nucleoprotein caps called telomeres that tend to shorten with age, a process that has been linked to increasing risk of cancer and death. Researchers recently confirmed the association in a cohort of 787 Italian adults, who had their telomeres measured in 1995 then again in 2005. Researchers tracked deaths and cancers in the cohort until 2010.
In fully adjusted analyses, adults with the shortest telomeres in 1995 were three times more likely to develop cancer (hazard ratio 3.02 (95% CI 1.84 to 4.97)) and eight times more likely to die from cancer (8.17 (2.86 to 23.29)) than those with the longest telomeres. Results were similar in analyses using an average of the two measurements, instead of baseline telomere length, and the link was independent of more than a dozen potential confounders including age, sex, social class, smoking, drinking, inflammatory markers, and chronic infection.
Participants were at least middle aged when recruited, and 137 developed cancer during 15 years of follow up. Enough to establish a general association, but not enough to explore different cancer types, say the authors. Mortality analyses included just 62 deaths, so the confidence intervals around these results are very wide.
Unhealthy lifestyles implicated in most sudden cardiac deaths among women
A new analysis from a well established cohort of US nurses suggests that around 80% of sudden cardiac deaths among women are attributable to poor diet, smoking, lack of exercise, and failure to maintain a normal body weight (population attributable risk 81% (95% CI 52% to 93%). The analysis included 321 sudden cardiac deaths among 81 722 women followed up for 26 years.
All four lifestyle factors were independently associated with risk of sudden cardiac death in the Nurses Health Study, and together they explained most of it. If the association is causal (still a big if), rates of sudden cardiac death would drop dramatically if all women adopted exemplary lifestyles, say the researchers.
For the nurses in this cohort that meant no smoking, a body mass index <25, a Mediterranean diet with a moderate intake of alcohol, and at least half an hour of exercise a day. Only 8% of the cohort managed all four components of a healthy lifestyle. Their risk of sudden cardiac death was 92% lower than the 3% of the cohort who managed none (relative risk 0.08 (95% CI 0.03 to 0.23)).
Cite this as: BMJ 2011;343:d4220