What's new in the other general journals
http://www.100md.com
《英国医生杂志》
Intrapleural streptokinase is no better than placebo for patients with pleural infection
Intrapleural streptokinase is a well established treatment for infected pleural effusion, but a large randomised trial in 427 patients suggests that it doesn't work. Compared with placebo, intrapleural streptokinase did not reduce deaths or the need for surgery (64/206, 31% of the streptokinase group, v 60/221, 27% of the placebo group; relative risk 1.14; 95% CI 0.85 to 1.54). Streptokinase didn't reduce hospital stay either (figure), but it did increase the risk of serious adverse effects, including chest pain, allergy, and fever (14/208, 7% with streptokinase v 6/222, 3% with placebo; relative risk, 2.49; 0.98 to 6.36).
Credit: NEJM
The patients in this trial were recruited from hospitals all over the United Kingdom, where national guidelines recommend intrapleural streptokinase as part of a package of measures that includes antibiotics and chest drainage. The recommendations were based on promising results from smaller trials, along with the biological plausibility of a treatment that clears fibrin adhesions, allowing an effusion to drain more easily. In this long awaited trial, the theoretical benefits did not translate into better outcomes for patients, and the authors counsel against routine intrapleural streptokinase for patients with pleural infection.
New England Journal of Medicine 2005;352: 865-74
Lifestyle intervention increased the risk of type 2 diabetes in middle aged smokers
We already know that losing weight and exercising more can help prevent type 2 diabetes in people with impaired glucose tolerance. The picture is less straightforward in people without it, according to a new analysis of data from a cardiovascular prevention trial. A package of interventions including lifestyle advice to eat less, exercise more, and stop smoking had no overall effect on the risk of type 2 diabetes in nearly 12 000 middle aged men with normal glucose tolerance. During six years of follow-up, 11.5% of men who were given the advice and 10.8% of men who weren't developed diabetes (hazard ratio 1.08, 95% CI 0.96 to 1.2). Worse, the intervention increased the risk of diabetes among the 63% of men who smoked at the start of the trial (1.26, 1.1 to 1.45). The authors say the increase could be linked to weight gain after the smokers quit, or a shift in body fat distribution. Among non-smokers, the lifestyle intervention reduced the incidence of type 2 diabetes by 18% (0.82, 0.68 to 0.98).
Annals of Internal Medicine 2005;142: 313-22
Protecting the heart's microcirculation: good in theory, disappointing in practice
During a percutaneous coronary intervention in someone with a myocardial infarction, it makes good sense to protect the distal circulation from clots and other debris that can work loose and threaten reperfusion. Small preliminary trials of this technique looked promising, but a larger more definitive randomised trial has failed to deliver any of the expected benefits, such as better reperfusion, smaller infarcts, and longer survival.
The trial included 501 adults with proved myocardial infarction who presented within six hours of onset of symptoms and needed an angioplasty. The intervention was done with or without a balloon and suction system to protect the distal microcirculation from debris. Half an hour later, the groups did not differ in the proportion of patients with resolved ST segment elevation (63.3% (152/240) with the system v 61.9% (148/239) without; absolute difference 1.4%, 95% CI -7.7% to 10.5%). The groups did not differ in infarct size (determined 5-14 days after treatment), mortality at six months (8/243 (3.4%) v 8/233 (3.3%), or the proportion of patients who had had a serious ischaemic cardiac event at six months (24/243 (10%) v 26/233 (11%)).
JAMA 2005;293: 1063-72
Pneumococcal vaccine reduces incidence of infection and prevalence of antibiotic resistance
Population surveillance data from Atlanta, Georgia, show a substantial fall in the incidence of invasive pneumococcal infections since the introduction in 2000 of a conjugate pneumococcal vaccine for children. By 2003 between 58% and 67% of young children in the metropolitan region had received at least three of the four recommended doses, leading to an 82% decrease in the incidence of disease among children under 2 years (figure), and a 71% decrease in children aged between 2 and 4 years. Introduction of the vaccine led to significant herd immunity, with decreases of 54%, 25%, and 39% in adults aged 20-39, 40-64, and 65 and over, most of whom had not been immunised. In all age groups combined, the mean annual incidence fell from 30.2 per 100 000 population before the vaccine (1994-9) to 13.1 per 100 000 afterwards (2002).
Credit: LANCET
Resistance to macrolide antibiotics such as erythromycin, which increased rapidly during the 1990s, fell by 69% after the vaccine was licensed, from 7.7 per 100 000 in 2000 to 2.9 per 100 000 in 2002. Again, the most dramatic reduction occurred among children under 2 years (figure) The authors say that decreased antibiotic resistance is due to the vaccine targeting resistant pneumococcal strains, not to more responsible prescribing; doctors in Atlanta and throughout the United States continue to prescribe large quantities of macrolides, including the newer agents azithromycin and clarithromycin. If they continue to ignore warnings about inappropriate use of these agents, the downward trend in resistance could soon be reversed.
Lancet 2005;365: 855-63
Patients' preferences need not compromise randomised trials
Patients' preferences for a particular treatment (or their doctors' preferences) can hamper recruitment to randomised trials. But leaving out patients with a preference, or letting them have the treatment they want, probably doesn't seriously bias the results, according to a systematic review. Reviewers studied 32 randomised trials that took account of patients' or doctors' preferences by including at least one preference arm.
Overall, a substantial minority of participants (12-74%) refused to be randomised. There were few consistent differences between those who refused and those who didn't, although a minority of studies reported that refusers tended to be educated (four trials), and employed (three trials), and therefore empowered. In general, differences in results between randomised and preference groups were small, particularly in large trials.
The reviewers say that patients' preferences are likely to become more important as patients become more active participants in research. Their findings should reassure other researchers that giving patients what they want needn't seriously compromise the validity of research.
JAMA 2005;293: 1089-99
Strict control of blood pressure delays the need for dialysis in patients with chronic renal disease
Aiming for a low target blood pressure can delay end stage renal failure and may even reduce mortality. These findings come from an extended follow-up of an old randomised trial testing the effects of a low target blood pressure (mean arterial blood pressure < 92 mm Hg) and low protein diet on 840 adults with mostly non-diabetic renal disease. The original results were inconclusive, but six years after the end of the trial, fewer patients in the low blood pressure groups had progressed to dialysis or a renal transplant (hazard ratio 0.68, 95% CI 0.57 to 0.82) (figure). A composite outcome combining end stage renal failure and death from any cause was also less likely in patients with a target mean blood pressure less than 92 mm Hg (0.77, 0.65 to 0.91).
Credit: ANNALS OF INTERNAL MEDICINE
During the original trial, treatment lasted an average of 2.2 years, but it's unclear what happened to patients' treatment or their blood pressure when the trial finished. It's possible that two years of intensive blood pressure control gave lasting benefits, but it's also possible that the benefits were due to continued intensive treatment after the end of the trial.
Annals of Internal Medicine 2005;142: 342-51
Immunisation against hepatitis B protects for up to 15 years
Nearly 25 years ago, the inhabitants of 15 Alaskan villages were immunised against hepatitis B. Fifteen years later, serological testing showed that most of them were still immune (661/783, 84%). Immunity fell most slowly among those immunised as young adults, and fell most quickly among those immunised before the age of 4. Male sex and a good antibody response to immunisation were also linked to persisting protection. The researchers tracking this cohort found only 16 cases of breakthrough infection over the 15 years, a yearly incidence of 0.84 (95% CI 0.48 to 1.37) per 1000 people. Breakthrough infection was significantly more likely among people with a weak initial response to immunisation (rate of infections per 1000 persons per year 0.61 in responders v 4.22 in non responders, P < 0.01). All the breakthrough infections in this study were asymptomatic.
These encouraging results are let down slightly by the 47% loss to follow up, but they still confirm that immunisation against hepatitis B works, and that it keeps working for many years in people that mount a good initial response. It's still unclear, however, whether immunised children need a booster later on. A linked editorial (pp 384-5) says these children may be better protected than it would seem because useful immunological memory persists even when antibody levels fall.
Annals of Internal Medicine 2005;142: 333-41
Lower serum concentration of homocysteine may reduce hip fractures in older people with stroke
To test the theory that serum concentrations of homocysteine are somehow linked to risk of osteoporotic fracture, Japanese researchers recruited 628 older men and women with a high risk of hip fracture and gave them a homocysteine lowering treatment or a placebo. After two years, those who took folate (5 mg) and vitamin B-12 (1500 μg) had lower serum concentrations of homocysteine than the others. They were also 80% (95% CI 92% to 50%) less likely to have a hip fracture than those who took placebo (figure). The participants were all over 65 and had a history of ischaemic stroke and longstanding hemiplegia. They also had higher than usual circulating concentrations of homocysteine at the start of the trial.
Credit: JAMA
The study's authors don't know why their treatment worked. The vitamin B-12 and folate reduced mean serum concentrations of homocysteine but didn't prevent falls or increase patients' bone mineral density. It's even possible that the benefits were due to higher serum concentrations of vitamin B-12, rather than lower concentrations of homocysteine.
JAMA 2005;293: 1082-8
Review supports more optimistic view of phase I trials in adults with cancer
Phase I trials testing new treatments for patients with cancer are controversial because reviews suggest that only 4-6% of participants respond to the new treatment, while about 0.5% die from side effects. A new study revises the overall response rate up to 10.6%, but confirms that nearly 1 in 200 participants dies from toxicity.
This study's authors reviewed 460 phase I trials including 11 935 adults with cancer. The trials were heterogeneous and tested a wide range of anticancer treatments: cytoxic drugs alone or in combination, immunomodulators, vaccines, new compounds, and combinations of treatments that included drugs already approved by the US Food and Drug Administration. Response rates varied between 0% and 26% depending on the type of trial. Those that included at least one approved anticancer agent had a high response rate (17.8%); The response rate in classic phase I trials of a single new agent was much lower (4.4%).
The authors say the overall picture is not as gloomy as critics have suggested, but it's certainly more complex. They hope their study will encourage patients, oncologists, investigators, and institutional review boards not to make simplistic judgments about the risks and benefits of this diverse group of trials.
Intrapleural streptokinase is a well established treatment for infected pleural effusion, but a large randomised trial in 427 patients suggests that it doesn't work. Compared with placebo, intrapleural streptokinase did not reduce deaths or the need for surgery (64/206, 31% of the streptokinase group, v 60/221, 27% of the placebo group; relative risk 1.14; 95% CI 0.85 to 1.54). Streptokinase didn't reduce hospital stay either (figure), but it did increase the risk of serious adverse effects, including chest pain, allergy, and fever (14/208, 7% with streptokinase v 6/222, 3% with placebo; relative risk, 2.49; 0.98 to 6.36).
Credit: NEJM
The patients in this trial were recruited from hospitals all over the United Kingdom, where national guidelines recommend intrapleural streptokinase as part of a package of measures that includes antibiotics and chest drainage. The recommendations were based on promising results from smaller trials, along with the biological plausibility of a treatment that clears fibrin adhesions, allowing an effusion to drain more easily. In this long awaited trial, the theoretical benefits did not translate into better outcomes for patients, and the authors counsel against routine intrapleural streptokinase for patients with pleural infection.
New England Journal of Medicine 2005;352: 865-74
Lifestyle intervention increased the risk of type 2 diabetes in middle aged smokers
We already know that losing weight and exercising more can help prevent type 2 diabetes in people with impaired glucose tolerance. The picture is less straightforward in people without it, according to a new analysis of data from a cardiovascular prevention trial. A package of interventions including lifestyle advice to eat less, exercise more, and stop smoking had no overall effect on the risk of type 2 diabetes in nearly 12 000 middle aged men with normal glucose tolerance. During six years of follow-up, 11.5% of men who were given the advice and 10.8% of men who weren't developed diabetes (hazard ratio 1.08, 95% CI 0.96 to 1.2). Worse, the intervention increased the risk of diabetes among the 63% of men who smoked at the start of the trial (1.26, 1.1 to 1.45). The authors say the increase could be linked to weight gain after the smokers quit, or a shift in body fat distribution. Among non-smokers, the lifestyle intervention reduced the incidence of type 2 diabetes by 18% (0.82, 0.68 to 0.98).
Annals of Internal Medicine 2005;142: 313-22
Protecting the heart's microcirculation: good in theory, disappointing in practice
During a percutaneous coronary intervention in someone with a myocardial infarction, it makes good sense to protect the distal circulation from clots and other debris that can work loose and threaten reperfusion. Small preliminary trials of this technique looked promising, but a larger more definitive randomised trial has failed to deliver any of the expected benefits, such as better reperfusion, smaller infarcts, and longer survival.
The trial included 501 adults with proved myocardial infarction who presented within six hours of onset of symptoms and needed an angioplasty. The intervention was done with or without a balloon and suction system to protect the distal microcirculation from debris. Half an hour later, the groups did not differ in the proportion of patients with resolved ST segment elevation (63.3% (152/240) with the system v 61.9% (148/239) without; absolute difference 1.4%, 95% CI -7.7% to 10.5%). The groups did not differ in infarct size (determined 5-14 days after treatment), mortality at six months (8/243 (3.4%) v 8/233 (3.3%), or the proportion of patients who had had a serious ischaemic cardiac event at six months (24/243 (10%) v 26/233 (11%)).
JAMA 2005;293: 1063-72
Pneumococcal vaccine reduces incidence of infection and prevalence of antibiotic resistance
Population surveillance data from Atlanta, Georgia, show a substantial fall in the incidence of invasive pneumococcal infections since the introduction in 2000 of a conjugate pneumococcal vaccine for children. By 2003 between 58% and 67% of young children in the metropolitan region had received at least three of the four recommended doses, leading to an 82% decrease in the incidence of disease among children under 2 years (figure), and a 71% decrease in children aged between 2 and 4 years. Introduction of the vaccine led to significant herd immunity, with decreases of 54%, 25%, and 39% in adults aged 20-39, 40-64, and 65 and over, most of whom had not been immunised. In all age groups combined, the mean annual incidence fell from 30.2 per 100 000 population before the vaccine (1994-9) to 13.1 per 100 000 afterwards (2002).
Credit: LANCET
Resistance to macrolide antibiotics such as erythromycin, which increased rapidly during the 1990s, fell by 69% after the vaccine was licensed, from 7.7 per 100 000 in 2000 to 2.9 per 100 000 in 2002. Again, the most dramatic reduction occurred among children under 2 years (figure) The authors say that decreased antibiotic resistance is due to the vaccine targeting resistant pneumococcal strains, not to more responsible prescribing; doctors in Atlanta and throughout the United States continue to prescribe large quantities of macrolides, including the newer agents azithromycin and clarithromycin. If they continue to ignore warnings about inappropriate use of these agents, the downward trend in resistance could soon be reversed.
Lancet 2005;365: 855-63
Patients' preferences need not compromise randomised trials
Patients' preferences for a particular treatment (or their doctors' preferences) can hamper recruitment to randomised trials. But leaving out patients with a preference, or letting them have the treatment they want, probably doesn't seriously bias the results, according to a systematic review. Reviewers studied 32 randomised trials that took account of patients' or doctors' preferences by including at least one preference arm.
Overall, a substantial minority of participants (12-74%) refused to be randomised. There were few consistent differences between those who refused and those who didn't, although a minority of studies reported that refusers tended to be educated (four trials), and employed (three trials), and therefore empowered. In general, differences in results between randomised and preference groups were small, particularly in large trials.
The reviewers say that patients' preferences are likely to become more important as patients become more active participants in research. Their findings should reassure other researchers that giving patients what they want needn't seriously compromise the validity of research.
JAMA 2005;293: 1089-99
Strict control of blood pressure delays the need for dialysis in patients with chronic renal disease
Aiming for a low target blood pressure can delay end stage renal failure and may even reduce mortality. These findings come from an extended follow-up of an old randomised trial testing the effects of a low target blood pressure (mean arterial blood pressure < 92 mm Hg) and low protein diet on 840 adults with mostly non-diabetic renal disease. The original results were inconclusive, but six years after the end of the trial, fewer patients in the low blood pressure groups had progressed to dialysis or a renal transplant (hazard ratio 0.68, 95% CI 0.57 to 0.82) (figure). A composite outcome combining end stage renal failure and death from any cause was also less likely in patients with a target mean blood pressure less than 92 mm Hg (0.77, 0.65 to 0.91).
Credit: ANNALS OF INTERNAL MEDICINE
During the original trial, treatment lasted an average of 2.2 years, but it's unclear what happened to patients' treatment or their blood pressure when the trial finished. It's possible that two years of intensive blood pressure control gave lasting benefits, but it's also possible that the benefits were due to continued intensive treatment after the end of the trial.
Annals of Internal Medicine 2005;142: 342-51
Immunisation against hepatitis B protects for up to 15 years
Nearly 25 years ago, the inhabitants of 15 Alaskan villages were immunised against hepatitis B. Fifteen years later, serological testing showed that most of them were still immune (661/783, 84%). Immunity fell most slowly among those immunised as young adults, and fell most quickly among those immunised before the age of 4. Male sex and a good antibody response to immunisation were also linked to persisting protection. The researchers tracking this cohort found only 16 cases of breakthrough infection over the 15 years, a yearly incidence of 0.84 (95% CI 0.48 to 1.37) per 1000 people. Breakthrough infection was significantly more likely among people with a weak initial response to immunisation (rate of infections per 1000 persons per year 0.61 in responders v 4.22 in non responders, P < 0.01). All the breakthrough infections in this study were asymptomatic.
These encouraging results are let down slightly by the 47% loss to follow up, but they still confirm that immunisation against hepatitis B works, and that it keeps working for many years in people that mount a good initial response. It's still unclear, however, whether immunised children need a booster later on. A linked editorial (pp 384-5) says these children may be better protected than it would seem because useful immunological memory persists even when antibody levels fall.
Annals of Internal Medicine 2005;142: 333-41
Lower serum concentration of homocysteine may reduce hip fractures in older people with stroke
To test the theory that serum concentrations of homocysteine are somehow linked to risk of osteoporotic fracture, Japanese researchers recruited 628 older men and women with a high risk of hip fracture and gave them a homocysteine lowering treatment or a placebo. After two years, those who took folate (5 mg) and vitamin B-12 (1500 μg) had lower serum concentrations of homocysteine than the others. They were also 80% (95% CI 92% to 50%) less likely to have a hip fracture than those who took placebo (figure). The participants were all over 65 and had a history of ischaemic stroke and longstanding hemiplegia. They also had higher than usual circulating concentrations of homocysteine at the start of the trial.
Credit: JAMA
The study's authors don't know why their treatment worked. The vitamin B-12 and folate reduced mean serum concentrations of homocysteine but didn't prevent falls or increase patients' bone mineral density. It's even possible that the benefits were due to higher serum concentrations of vitamin B-12, rather than lower concentrations of homocysteine.
JAMA 2005;293: 1082-8
Review supports more optimistic view of phase I trials in adults with cancer
Phase I trials testing new treatments for patients with cancer are controversial because reviews suggest that only 4-6% of participants respond to the new treatment, while about 0.5% die from side effects. A new study revises the overall response rate up to 10.6%, but confirms that nearly 1 in 200 participants dies from toxicity.
This study's authors reviewed 460 phase I trials including 11 935 adults with cancer. The trials were heterogeneous and tested a wide range of anticancer treatments: cytoxic drugs alone or in combination, immunomodulators, vaccines, new compounds, and combinations of treatments that included drugs already approved by the US Food and Drug Administration. Response rates varied between 0% and 26% depending on the type of trial. Those that included at least one approved anticancer agent had a high response rate (17.8%); The response rate in classic phase I trials of a single new agent was much lower (4.4%).
The authors say the overall picture is not as gloomy as critics have suggested, but it's certainly more complex. They hope their study will encourage patients, oncologists, investigators, and institutional review boards not to make simplistic judgments about the risks and benefits of this diverse group of trials.