MRI can be cost effective for breast cancer screening
Current guidelines recommend yearly mammography for women with BRCA1 or BRCA2 mutations, even though it isn't particularly sensitive. Adding magnetic resonance imaging (MRI) might help, but since MRI is 10 times more expensive could it ever be cost effective?
Researchers estimated that women screened by both methods every year would live two years longer with BRCA1 and about 18 months longer with BRCA2, compared with women screened by mammography alone. They also estimated that each extra quality adjusted life year (QALY) would cost between $45 000 (£24 000; €35 000) and $700 000, depending on the age range screened and the type of mutation. If we assume that society is willing to pay up to $100 000 for each extra QALY, then annual MRI would be cost effective for women with BRCA1 screened yearly between the ages of 35 and 54. The same programme would be too expensive for most women with BRCA2, who have a lower risk of breast cancer.
Figure 1.
Credit: JAMA
The researchers used an established mathematical model to produce these estimates, plugging in the best available data on the benefits and costs of various screening strategies. As usual, there were plenty of gaps. But MRI does at least look cost effective for some women some of the time. And it's likely to get cheaper. The final estimates were most sensitive to changes in women's life time risk of breast cancer.
JAMA 2006;295: 2374-84
Stop using pulmonary artery catheters
Intensivists have been debating the merits of pulmonary artery catheters for years, while continuing to use them for many patients with acute lung injury. It's probably time to stop, now that a large and convincing trial has found no evidence of benefit, says an editorial (pp 2273-4).
Patients who were managed with a pulmonary artery catheter were just as likely to die within 60 days as patients who were managed by central venous catheter (27.4% v 26.3%, P = 0.69), and they spent just as long being ventilated. Compared with central venous monitoring, the pulmonary artery catheters did not improve patients' renal function, blood pressure, or fluid balance. But they did cause more complications, usually arrhythmias.
This trial is convincing because of its size (n = 1000), and also because the researchers used a strict protocol to guide management in both groups of patients. It now seems likely that pulmonary artery catheters don't help, even when they are used properly.
New Engl J Med 2006;354: 2213-24
We need colonoscopy trials
The time has come for a proper randomised trial of screening for colorectal cancer with colonoscopy, says an editorial. More and more people are being screened this way, even though we can't say with any certainty whether or not it saves lives, or even mitigates the risk of death from colorectal cancer. Originally, recommendations for screening were based on positive results from trials of faecal occult blood testing. Colonoscopy has sneaked in relatively unchallenged since then, supported by only a few observational studies and some theoretical simulation.
The author, an academic public health expert from the United States, says that although researchers resolutely struggle on, evaluating colonoscopy with imagination but no proper prospective data, billions of dollars are spent screening an ageing US population using a tool that may or may not be cost effective. A decent randomised trial would cost very little in comparison.
The author made these comments after the publication of two studies that tried to find out firstly whether a 10 year screening interval is too long and, secondly, at what age screening should stop (pp 2366-73 and 2357-65). Both come up with answers (no; probably before 80 years) but neither is definitive because of weaknesses in their methods.
JAMA 2006;295: 2411-2
Liberal caesarean policy may do harm
High rates of caesarean section in hospitals throughout Latin America are associated with an increased risk of fetal death and serious complications for the mother, a recent survey has found. Higher rates of both elective and intrapartum caesarean section were associated with bad outcomes for both mother and baby, even after the researchers accounted for the baseline risk in the population (case mix) and for the sophistication of services on offer in each hospital. They conclude that a liberal policy towards caesarean section could be doing women and their babies more harm than good and that hospitals in Latin America should take a close and critical look at their obstetric practices.
Figure 2.
Credit: LANCET
The survey, which was done at the behest of the World Health Organization, included 120 hospitals in eight randomly selected countries. Most were moderately sophisticated secondary or tertiary care hospitals that operated within the public or social security sectors. The researchers surveyed more than 97 000 deliveries. The median caesarean section rate was 33%.
Lancet 2006. May 23, doi: 10.16/S0140-6736(06)68704-7