Practice Short cuts

# What's new in the other general journals

BMJ 2006; 332 (Published 04 May 2006) Cite this as: BMJ 2006;332:1084
1. Alison Tonks (atonks{at}bmj.com), associate editor

## Implantable cardiac devices have substantial failure rate

Pacemakers and implantable cardioverter defibrillators save lives—but only when they work. Although millions of people have these devices, it's hard to say accurately how often they fail and how often the patient's life is at risk when they do. All we know is that malfunction is not a rare event.

Credit: JAMA

After appraising two new papers on the subject (pp 1901-6 and 1929-34), one commentator estimates that about seven in every 1000 pacemakers and about 21 in every 1000 cardioverter defibrillators malfunction badly enough to need replacing. Neither of these estimates included devices that broke down because of faulty leads. Nor did the estimates include people who died as a result of their device malfunctioning. Unreliable reporting means that both estimates are likely to be lower than the real rate, which, for pacemakers at least, seems to be going down. The trend for cardioverter defibrillators looks less predictable.

Credit: JAMA

Information on the safety of these devices is a complex mosaic, the author writes, and many of the pieces are still missing. What, for example, should doctors do when regulators issue a safety warning about a particular model? In a third paper (pp 1907-11), doctors from Canada electively replaced the “faulty” cardioverter defibrillator in about a fifth of their patients, although some replaced none and others replaced 45% (24/53). In 6% (31/533) of patients, the elective replacement caused serious surgical complications, most commonly bleeding and infections. Two patients died.

## Calls from a “prevention coach” help women remember cancer screening

Poor US women often miss out on recommended cancer screening services. In an attempt to reach them, researchers trained “prevention coaches” to telephone and motivate women to keep up to date with their cervical screening, mammography, and colorectal cancer screening. The coaches used a partially scripted interview technique to find out why women weren't attending and how any barriers to screening could be overcome. The coaches couldn't order screening tests but were able to organise transport to appointments, send reminders, and help with language barriers between women and their doctors.

The intervention seemed to work. In a randomised trial lasting 18 months, the proportion of women up to date with cancer screening increased from 71% to 78% for cervical cancer, from 58% to 68% for mammography, and from 39% to 63% for colorectal cancer (mostly fecal occult blood tests and colonoscopy). These gains were significantly bigger than those achieved by control women, who had usual care.

The 1413 women in this study were poor, and 60% spoke Spanish as a first language. All attended urban primary care clinics in New York city, and all were overdue at least one screening appointment when they were recruited. A mean of four telephone calls helped them access recommended screening services. But what happens when the calls stop?

## Relief poorly coordinated after Pakistan earthquake

Two disaster experts from the United States have called on the World Health Organization to develop an “A team” to coordinate the response to international disasters, such as the recent earthquake in Pakistan, which killed more than 74 000 people.

The relief effort was slow to get going, partly because of the geography and the scale of the disaster, but also because of poor coordination and a dearth of people with the leadership and organisational skills necessary to prioritise, set strategy, and make it all happen, write the experts. The relief effort continued to underperform as winter set in, and tents and shelters failed to materialise. Those that did often collapsed under the weight of the heavy snow typical of Pakistan's mountains. With half the available health facilities destroyed and a further 13% badly damaged, the health sector was quickly overwhelmed and failed to notice looming threats to public health caused by exposure, overcrowding, and poor sanitation.

WHO is already training people in the “science and practice of disaster management, and, especially, the art of coordination” to help to improve the outcome of future disasters. In the meantime, the basic needs of many thousands of people in Pakistan are still unmet, and many thousands of promised dollars have yet to arrive.

## Good clinical trials are excellent value for money

Clinical trials are an excellent investment, producing a return more than four times a matching investment in the stock market, or about 46% a year, research from the US National Institute of Neurological Disorders and Stroke has found.

The authors estimate that phase III clinical trials funded by the institute before January 2000 will produce net returns to society totalling about 470 000 extra healthy years of life worth $15bn (£8bn; €12bn) over 10 years (95% CI$0.67bn to $35bn). The estimates are based on published data on the use, costs, and effects of interventions tested in 25 clinical trials, which together cost$335m. The costs were repaid in health benefits in just over a year. The core data for the analysis came from eight interventions, including endarterectomy for carotid stenosis, anticoagulation for atrial fibrillation, interferon beta-1a for multiple sclerosis, and diazepam for repeat seizures.

The authors say that their figures almost certainly underestimate the real value of high class clinical trials. The benefits would be even greater if proved treatments were incorporated quickly into clinical practice. Good research may be expensive—one of the trials in this study cost $64m—but the payback seems to be more than enough to cover it. Credit: LANCET ## Smallpox virus persists for at least three weeks after vaccination The US authorities suspended smallpox vaccination in 1972 and then reintroduced it in 2002. Because the smallpox virus can persist after vaccination, a theoretical possibility of transmission exists between vaccinated and unvaccinated people. To try to quantify the risk, researchers vaccinated 97 members of the US air force and then monitored them for up to three weeks for persistence of viral DNA in their blood and in swabs from their site dressings, throats, and hands. Credit: JAMA After 6-8 days, researchers found vaccine DNA in at least one sample from 16% (12/74) of the men. After 22 days, they found viral DNA in 22% (11/51)—one had the virus in his blood, four had it in their throat swabs, and six on dressings. Overall, a quarter of the men had at least one positive sample during the study. The positive blood samples contained 1000 genome copies/ml, which, the authors say, corresponds to about 10 infectious viral particles. It's unclear how many are required to establish infection in a susceptible person, but the authors think that transmission is at least possible. The current guidelines recommend that vaccinated individuals should not give blood for at least three weeks. These authors want that period extended. ## Women should not take antioxidant vitamins to prevent pre-eclampsia, for now Pre-eclampsia is a leading cause of death and serious illness in pregnant women and their babies. Can a cheap and simple treatment such as antioxidant vitamins help prevent it? A definitive answer has yet to emerge despite a handful of clinical trials. The latest included 1877 healthy Australian women who took vitamins C (1000 mg a day) and E (400 IU a day) or an identical placebo throughout the second half of their first pregnancy. Compared with placebo, treatment had no effect on the incidence of pre-eclampsia (5% v 6%), death or other serious outcome for the infant (12% v 10%), or the risk of having a baby that was small for gestational age (10% v 9%). Among the many other comparisons, the authors report two that show that antioxidant vitamins may cause harm. Women taking them were more likely than controls to have to go into hospital because of high blood pressure (relative risk 1.54, 95% CI 1.00 to 2.39) and more likely to need antihypertensive drugs (1.67, 1.03 to 2.69) Various design problems mean the findings are not the final word, however. A linked editorial (pp 1841-3) says that the trial wasn't powerful enough to rule out moderate but clinically important benefits. The harms turned up among multiple comparisons, and may not be real. Bigger trials are already under way that should help clarify the situation. ## MSF expands its HIV programme in Malawi Malawi is an extremely poor country in which yearly per capita health expenditure is a little over$10 (£6; €8), and the citizens of Malawi can expect to live no longer than an average of 39 years. A 10th of the population is infected with HIV, and HIV related disease is the main cause of death in 20-49 year olds.

In 2002, the international medical charity Médecins Sans Frontières began increasing its operation to provide highly active antiretroviral therapy to the population of one district, and, by 2004, more than 220 people a month were starting treatment, with an easy to take combination of stavudine, lamivudine, and nevirapine. Treatment was free, and an observational study reports encouraging results.

Most people stuck with their treatment, and three quarters were still enrolled in the programme after a year. Although a fifth of the 1308 adults in the study died, the survivors had a median gain of CD4 cells of 165 x 106/l in the first year. Four fifths of a smaller sample had undetectable viral loads (< 400 copies/ml) at the end of follow-up. The authors say that these gains are comparable with those achieved in other cohorts from developed as well as developing countries. They show that large scale treatment with highly active antiretroviral therapy is possible even in the poorest of settings.

## Patients do better in hospitals that follow guidelines

Judging a hospital's quality by the outcomes enjoyed by its patients is complicated by variations in case mix that are hard to measure. Measuring processes, such as how often patients are treated according to guidelines, is easier, but only valid if better processes lead to better outcomes for patients. A study of 64 773 patients with acute coronary syndrome suggests that they do.

The authors found a clear and significant inverse relation between early mortality and adherence to national guidelines in 350 US hospitals. For each 10% increase in adherence to guidelines, patients were 10% less likely to die in hospital (adjusted odds ratio 0.9, 95% CI 0.84 to 0.97).

More than 75% of eligible patients received the recommended treatments, but this proportion varied from 63% in the worst hospitals (lowest quartile) to 82% in the best (highest quartile). Corresponding inhospital mortality was 6% and 4% (adjusted odds ratio 0.81, 0.68 to 0.97).

These data indicate, but do not prove, that hospitals that follow guidelines are better for patients. The authors say that they lend support to the current enthusiasm for using process measures based on guidelines to judge hospital performance and improve quality.

View Abstract