What's new in the other general journals
http://www.100md.com
《英国医生杂志》
Data from cardiac patients supports ban on specialist hospitals in the US
In 2003 the US government put a temporary moratorium on the development of specialist hospitals that are partly owned by the doctors using them. Officials were responding to concerns that specialist hospitals cherry picked healthy patients having low risk but lucrative procedures. A retrospective analysis of Medicare data from 69 011 coronary patients supports this view. Those who had their percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG) in specialist cardiac hospitals were healthier (fewer comorbidities such as diabetes), better off (from neighbourhoods with a higher mean income), and more likely to be classed as "elective" than patients who had their procedure in general hospitals. Predictably, crude death rates were lower in specialist hospitals (2.1% v 3.1% for PCI and 5.0% v 5.8% for CABG; P < 0.001 for each comparison), but the differences disappeared once the data had been adjusted for patients' characteristics and for the higher patient volumes in specialist hospitals (adjusted odds ratio for death 1.05, 95% CI 0.83 to 1.32 for PCI; 0.91, 0.74 to 1.11 for CABG).
A report commissioned by Congress has also reported recently (pp 1405-7) that specialist hospitals owned by doctors tend to treat healthier patients who have more "profitable" diseases than do general hospitals. The report, which was compiled by the Medicare Payment Advisory Commission (MedPAC), recommended that Congress extend the ban on specialist hospitals until January 2007.
New England Journal of Medicine 2005;352: 1454-62
Suicide screening is unlikely to harm American high school students
Preventing suicide in young people is moving up the political and social agenda in the United States, and screening high school students is an increasingly important part of the strategy. There's some evidence that screening for mental health problems can work, but experts still worry that discussing suicide with teenagers might encourage them to try it.
Credit: ANNALS OF INTERNAL MEDICINE
The first randomised trial looking specifically for harmful effects from screening provides reassurance.
Researchers screened 2342 students from six high schools in New York state, using a variety of mental health questionnaires, some of which had been doctored to remove direct questions about suicide. The students, who had a mean age of nearly 15, were randomised by class to be screened with or without the suicide questions.
The researchers found no increase in distress, depression, or suicidal thoughts among students who had been asked about suicide. In fact, the questions were beneficial in a subgroup of high risk students with depression, a history of suicide attempts, or substance use.
JAMA 2005;293: 1635-43
Low dose dopamine does not protect patients with vulnerable kidney function
The latest meta-analysis evaluating renal dose dopamine reports that it doesn't reduce mortality or delay acute renal failure in patients with vulnerable renal function (relative risk for mortality 0.96, 95% CI 0.78 to 1.19; RR for renal replacement therapy 0.93, 0.76 to 1.15) (figure). Two previous meta-analyses found the same. So why are intensive care units across the developed world still using renal dose dopamine? Possibly because in this meta-analysis, renal dose dopamine caused an immediate, visible, and reassuring increase in urine output (mean increase 24%, 14% to 35%) and a smaller but still significant decrease in serum concentrations of creatinine (4% decrease relative to placebo or no treatment, 1% to 7%). The improvements didn't last, however. Urine output and creatinine clearance were back where they started by day 2.
These authors looked hard for all the relevant data, including data that were never published. They found 61 randomised or quasi-randomised trials in 3359 patients from a variety of clinical settings. Results of trials were fairly consistent, and the overall picture was resoundingly neutral—no real benefits, but no real harms either.
Annals of Internal Medicine 2005;142: 510-24
New treatment for alcohol addiction reduces heavy drinking
Naltrexone has been licensed as a treatment for alcohol addiction since 1994 in the US, but patients don't like it because they have to take a pill every day. A new long acting version could be the answer; it certainly looks promising in a recent placebo controlled trial in a selected group of 624 men and women.
Participants were treated for six months with monthly intramuscular injections of placebo, low dose naltrexone, or high dose naltrexone. They also had a monthly session of supportive therapy. At the end of the treatment period, self reported heavy drinking was reduced by 25% in the high dose group (P = 0.03) and by 17% in the low dose group (P = 0.07) compared with placebo (figure). Men did better than women in this trial, and so did the minority of participants (53/571) who had stopped drinking before the start of the trial.
Credit: JAMA
The company that developed injectable naltrexone funded, designed, conducted, and analysed this study, and the authors have modest hopes that their new product will help people who are traditionally hard to help. The news is not all good, however: long acting naltrexone did not reduce "risky" drinking (more than two drinks a day for men and more than one drink a day for women), nor did it help patients quit drinking altogether.
JAMA 2005;293: 1617-25
Anti-obesity drugs can work—for a while
A systematic search for trials of anti-obesity drugs found three recent meta-analyses and 79 separate trials of nine different drugs. Orlistat and sibutramine were the best evaluated; sertraline and zonisamide were the worst, with only one trial each. Overall, these drugs when combined with a diet produce a modest weight loss (up to 5 kg) in 12 months.
Estimates of weight loss relative to placebo were 4.45 kg for sibutramine, 2.89 kg for orlistat, 3.6 kg for phentermine, and 2.77 kg for bupropion, and 6.5% of pretreatment weight for topiramate. Results for fluoxetine and diethylpropion were less convincing, but the authors conclude that they probably help some people in the short term. Sertraline probably doesn't—in the only trial, people taking sertraline to maintain weight loss gained more than 17 kg on average over 54 weeks.
Questions remain about the long term benefits and safety of all these drugs. There are very few data evaluating their impact on heart disease or diabetes, for example. It's also unclear how they perform relative to each other. These authors found only one head to head trial, and even that was inadequately reported.
Annals of Internal Medicine 2005;142: 532-46
MRSA: a serious problem in the community
A community surveillance project covering Atlanta, Baltimore, and Minnesota identified 1647 cases of community acquired MRSA (methicillin resistant Staphylococcus aureus) infection between 2001 and 2003. That's an annual incidence of 18-26 cases per 100 000 population, or between 8% and 20% of all MRSA infections identified over that period. Over three quarters of patients had skin and soft tissue infections such as abscesses or cellulitis, 23% overall were admitted to hospital, and 73% overall had infections that were resistant to their prescribed antibiotic. Mortality in this survey was low; 95% of admitted patients went back home after a median stay of four days.
MRSA is now a common and serious problem in the community, say the authors, and doctors out there should be aware of it. They advise taking samples for culture from all patients with a suspected staphylococcus infection, followed by susceptibility testing of any cultured staphylococci. It's less clear what doctors should do next for the majority of people with infections who are well enough to stay at home. Prescribe antibiotics, presumably, but which ones? A linked editorial (pp 1485-7) says cheap oral drugs such as trimethoprim-sulfamethoxazole, doxycycline, and clindamycin could work in theory and should be tested quickly in clinical trials to see if they work in practice.
New England Journal of Medicine 2005;352: 1436-44
Platelets recover quickly after NSAIDs
The non-steroidal anti-inflammatory drug ibuprofen interferes with platelet function, but not always, and not for long, according to an experimental study in 11 healthy volunteers. The volunteers (seven women and four men aged between 32 and 64) took ibuprofen (600 mg 8 hourly) for a week. Forty minutes after the last dose, only seven of them had prolonged platelet aggregation times and even in these people the platelet function returned to normal within 24 hours (figure). The other four volunteers had normal platelet function throughout the experiment; interestingly, three of the four were taking oral contraceptives. The authors used a platelet function analyser that timed platelets aggregating into a plug big enough to close a small hole in a membrane.
Credit: ANNALS OF MEDICINE INTERNAL
Although very preliminary, these findings could help surgeons and primary care doctors develop more rational advice to patients about when to stop taking NSAIDs before elective surgery. Typically, patients stop taking their NSAIDs about a week before surgery, but there's no consensus and until now no data to inform the decision.
Annals of Internal Medicine 2005;142: 506-9
Many infant deaths in Belgium follow a clear end of life decision
More than half (57%) of neonatal and infant deaths in Flanders, Belgium were preceded by an acknowledged end of life decision, a recent survey found. A third of the 253 deaths studied were preceded by a decision to withhold or withdraw treatment, but in 16% the decision was based on the principle of double effect: giving drugs to relieve pain in doses likely to hasten death. Active euthanasia remains illegal in Belgium but 7% of deaths were preceded by a lethal injection, and 68% of the doctors questioned said they would be willing to end a child's life this way to relieve suffering. Lethal drugs were most commonly used in the early neonatal period (15/117 (13%) of early neonatal deaths v 2/77 (3%) of later deaths, P = 0.018), usually in babies who were very premature or severely brain damaged by intracranial haemorrhage or had serious or multiple congenital abnormalities.
The anonymous survey was sent to 292 doctors who had completed death certificates for the infants under 1 year old who had died between August 1999 and July 2000; 253 (87%) of them responded.
In 2003 the US government put a temporary moratorium on the development of specialist hospitals that are partly owned by the doctors using them. Officials were responding to concerns that specialist hospitals cherry picked healthy patients having low risk but lucrative procedures. A retrospective analysis of Medicare data from 69 011 coronary patients supports this view. Those who had their percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG) in specialist cardiac hospitals were healthier (fewer comorbidities such as diabetes), better off (from neighbourhoods with a higher mean income), and more likely to be classed as "elective" than patients who had their procedure in general hospitals. Predictably, crude death rates were lower in specialist hospitals (2.1% v 3.1% for PCI and 5.0% v 5.8% for CABG; P < 0.001 for each comparison), but the differences disappeared once the data had been adjusted for patients' characteristics and for the higher patient volumes in specialist hospitals (adjusted odds ratio for death 1.05, 95% CI 0.83 to 1.32 for PCI; 0.91, 0.74 to 1.11 for CABG).
A report commissioned by Congress has also reported recently (pp 1405-7) that specialist hospitals owned by doctors tend to treat healthier patients who have more "profitable" diseases than do general hospitals. The report, which was compiled by the Medicare Payment Advisory Commission (MedPAC), recommended that Congress extend the ban on specialist hospitals until January 2007.
New England Journal of Medicine 2005;352: 1454-62
Suicide screening is unlikely to harm American high school students
Preventing suicide in young people is moving up the political and social agenda in the United States, and screening high school students is an increasingly important part of the strategy. There's some evidence that screening for mental health problems can work, but experts still worry that discussing suicide with teenagers might encourage them to try it.
Credit: ANNALS OF INTERNAL MEDICINE
The first randomised trial looking specifically for harmful effects from screening provides reassurance.
Researchers screened 2342 students from six high schools in New York state, using a variety of mental health questionnaires, some of which had been doctored to remove direct questions about suicide. The students, who had a mean age of nearly 15, were randomised by class to be screened with or without the suicide questions.
The researchers found no increase in distress, depression, or suicidal thoughts among students who had been asked about suicide. In fact, the questions were beneficial in a subgroup of high risk students with depression, a history of suicide attempts, or substance use.
JAMA 2005;293: 1635-43
Low dose dopamine does not protect patients with vulnerable kidney function
The latest meta-analysis evaluating renal dose dopamine reports that it doesn't reduce mortality or delay acute renal failure in patients with vulnerable renal function (relative risk for mortality 0.96, 95% CI 0.78 to 1.19; RR for renal replacement therapy 0.93, 0.76 to 1.15) (figure). Two previous meta-analyses found the same. So why are intensive care units across the developed world still using renal dose dopamine? Possibly because in this meta-analysis, renal dose dopamine caused an immediate, visible, and reassuring increase in urine output (mean increase 24%, 14% to 35%) and a smaller but still significant decrease in serum concentrations of creatinine (4% decrease relative to placebo or no treatment, 1% to 7%). The improvements didn't last, however. Urine output and creatinine clearance were back where they started by day 2.
These authors looked hard for all the relevant data, including data that were never published. They found 61 randomised or quasi-randomised trials in 3359 patients from a variety of clinical settings. Results of trials were fairly consistent, and the overall picture was resoundingly neutral—no real benefits, but no real harms either.
Annals of Internal Medicine 2005;142: 510-24
New treatment for alcohol addiction reduces heavy drinking
Naltrexone has been licensed as a treatment for alcohol addiction since 1994 in the US, but patients don't like it because they have to take a pill every day. A new long acting version could be the answer; it certainly looks promising in a recent placebo controlled trial in a selected group of 624 men and women.
Participants were treated for six months with monthly intramuscular injections of placebo, low dose naltrexone, or high dose naltrexone. They also had a monthly session of supportive therapy. At the end of the treatment period, self reported heavy drinking was reduced by 25% in the high dose group (P = 0.03) and by 17% in the low dose group (P = 0.07) compared with placebo (figure). Men did better than women in this trial, and so did the minority of participants (53/571) who had stopped drinking before the start of the trial.
Credit: JAMA
The company that developed injectable naltrexone funded, designed, conducted, and analysed this study, and the authors have modest hopes that their new product will help people who are traditionally hard to help. The news is not all good, however: long acting naltrexone did not reduce "risky" drinking (more than two drinks a day for men and more than one drink a day for women), nor did it help patients quit drinking altogether.
JAMA 2005;293: 1617-25
Anti-obesity drugs can work—for a while
A systematic search for trials of anti-obesity drugs found three recent meta-analyses and 79 separate trials of nine different drugs. Orlistat and sibutramine were the best evaluated; sertraline and zonisamide were the worst, with only one trial each. Overall, these drugs when combined with a diet produce a modest weight loss (up to 5 kg) in 12 months.
Estimates of weight loss relative to placebo were 4.45 kg for sibutramine, 2.89 kg for orlistat, 3.6 kg for phentermine, and 2.77 kg for bupropion, and 6.5% of pretreatment weight for topiramate. Results for fluoxetine and diethylpropion were less convincing, but the authors conclude that they probably help some people in the short term. Sertraline probably doesn't—in the only trial, people taking sertraline to maintain weight loss gained more than 17 kg on average over 54 weeks.
Questions remain about the long term benefits and safety of all these drugs. There are very few data evaluating their impact on heart disease or diabetes, for example. It's also unclear how they perform relative to each other. These authors found only one head to head trial, and even that was inadequately reported.
Annals of Internal Medicine 2005;142: 532-46
MRSA: a serious problem in the community
A community surveillance project covering Atlanta, Baltimore, and Minnesota identified 1647 cases of community acquired MRSA (methicillin resistant Staphylococcus aureus) infection between 2001 and 2003. That's an annual incidence of 18-26 cases per 100 000 population, or between 8% and 20% of all MRSA infections identified over that period. Over three quarters of patients had skin and soft tissue infections such as abscesses or cellulitis, 23% overall were admitted to hospital, and 73% overall had infections that were resistant to their prescribed antibiotic. Mortality in this survey was low; 95% of admitted patients went back home after a median stay of four days.
MRSA is now a common and serious problem in the community, say the authors, and doctors out there should be aware of it. They advise taking samples for culture from all patients with a suspected staphylococcus infection, followed by susceptibility testing of any cultured staphylococci. It's less clear what doctors should do next for the majority of people with infections who are well enough to stay at home. Prescribe antibiotics, presumably, but which ones? A linked editorial (pp 1485-7) says cheap oral drugs such as trimethoprim-sulfamethoxazole, doxycycline, and clindamycin could work in theory and should be tested quickly in clinical trials to see if they work in practice.
New England Journal of Medicine 2005;352: 1436-44
Platelets recover quickly after NSAIDs
The non-steroidal anti-inflammatory drug ibuprofen interferes with platelet function, but not always, and not for long, according to an experimental study in 11 healthy volunteers. The volunteers (seven women and four men aged between 32 and 64) took ibuprofen (600 mg 8 hourly) for a week. Forty minutes after the last dose, only seven of them had prolonged platelet aggregation times and even in these people the platelet function returned to normal within 24 hours (figure). The other four volunteers had normal platelet function throughout the experiment; interestingly, three of the four were taking oral contraceptives. The authors used a platelet function analyser that timed platelets aggregating into a plug big enough to close a small hole in a membrane.
Credit: ANNALS OF MEDICINE INTERNAL
Although very preliminary, these findings could help surgeons and primary care doctors develop more rational advice to patients about when to stop taking NSAIDs before elective surgery. Typically, patients stop taking their NSAIDs about a week before surgery, but there's no consensus and until now no data to inform the decision.
Annals of Internal Medicine 2005;142: 506-9
Many infant deaths in Belgium follow a clear end of life decision
More than half (57%) of neonatal and infant deaths in Flanders, Belgium were preceded by an acknowledged end of life decision, a recent survey found. A third of the 253 deaths studied were preceded by a decision to withhold or withdraw treatment, but in 16% the decision was based on the principle of double effect: giving drugs to relieve pain in doses likely to hasten death. Active euthanasia remains illegal in Belgium but 7% of deaths were preceded by a lethal injection, and 68% of the doctors questioned said they would be willing to end a child's life this way to relieve suffering. Lethal drugs were most commonly used in the early neonatal period (15/117 (13%) of early neonatal deaths v 2/77 (3%) of later deaths, P = 0.018), usually in babies who were very premature or severely brain damaged by intracranial haemorrhage or had serious or multiple congenital abnormalities.
The anonymous survey was sent to 292 doctors who had completed death certificates for the infants under 1 year old who had died between August 1999 and July 2000; 253 (87%) of them responded.