All the best in PulmCCM (Roundup #6)

All the best in the pulmonary and critical care medicine literature from our ongoing journal survey. Browse all the PulmCCM Roundups to stay up to date.

Thrombolytics for Pulmonary Embolism: New Metaanalysis

Most patients with massive pulmonary embolism (PE with shock) should receive thrombolytics, but it's unclear from randomized trial data which patients with submassive pulmonary embolism (without shock, but with right ventricular dysfunction or other high risk markers) should receive thrombolytics. (See our review of the PEITHO study, and TOPCOAT, although there have been many others.) Authors of a new meta-analysis in JAMA of 16 trials (n=2,115, of both massive and submassive PE patients) make the case that thrombolytics' benefit has been underestimated by individual trials in the past. Thrombolytics were associated with odds ratios of about 0.50 for death, for both massive and submassive PE. However, thrombolytic therapy tripled the odds of major bleeding in massive and submassive PE patients. Among all comers, the number needed to treat with thrombolytics to save a life (59) was comparable to the number needed to treat to cause a devastating intracranial hemorrhage (78); authors estimated similar equipoise in the submassive PE group, considered alone. Contrary to earlier-era trials, bleeding risk was considerably lower in patients younger than 65, making age a potential risk stratifier for consideration of thrombolytics, especially for submassive pulmonary embolism.

Blood Transfusions Down 30% in U.S.

Physicians' transfusion practices may be coming in line with published evidence ... finally. The New York Times reports that blood transfusions in the U.S. fell from 15 million units in 2012 to 11 million for 2013, a decline by almost one-third. Changes to professional society guidelines are probably responsible. Thoracic surgery consumes a large proportion of the blood supply; a 2011 update to their guidelines discouraged liberal blood transfusions above hemoglobin of 7 g/dL. More recently, the Choosing Wisely campaign advised that non-bleeding medical patients not receive blood transfusions until their hemoglobin falls below 7 g/dL (for the critically ill) or 7-8 g/dL (for the non-critically-ill). The result: there's actually a surplus in the blood supply today, hospitals are demanding discounts from the going rate of $230 per unit, and the Red Cross has had to lay off some workers and reduce others' working hours. Maybe my previous half-serious suggestion of a special slogan aimed at physicians -- "Save a Life: Don't Give Blood" -- should go back on the shelf.

Does Ultrasound During Thoracentesis Really Prevent Pneumothorax?

A clinical review and metaanalysis in JAMA concludes that Light's criteria are highly sensitive for ruling out an exudative effusion when all three criteria are negative; high LDH (>200 U/L) or high cholesterol were most specific for exudates.

More provocatively, after reviewing 37 studies on thoracentesis and adverse event rates, the authors found that the use of ultrasound was not associated with a decrease in pneumothorax (odds ratios of 0.37 and 0.55 for skin marking and any use of ultrasound, respectively -- but with wide confidence intervals, thus not statistically significant). The implication that procedural ultrasound may be unnecessary will draw fire from the ultrasound-advocacy crowd, who like to cite observational studies such as this one, this one and this one, and mention the atrociously high rates of iatrogenic pneumothorax in the pre-ultrasound era to argue that introducing ultrasound into training programs for thoracentesis does reduce the rate of pneumothorax.

The Agency for Healthcare Research and Quality (AHRQ) includes the use of ultrasound during thoracentesis (marking or realtime visualization of pleural fluid) on its list of best practices. The observed rate of pneumothorax in the series reviewed in the metaanalysis was 6%, but expert centers (using ultrasound) report a pneumothorax rate of 1-2%.

Influenza Vaccine Protects Pregnant Women (and their Newborns)

Influenza vaccination is a standard recommendation for pregnant women, but that isn't based on strong randomized trial data. That's mainly because it's culturally difficult to perform large placebo-controlled randomized trials on pregnant women in the U.S. and Europe. Those barriers are apparently lower in Africa, creating research opportunities that you can call global health advocacy or philanthropic neo-colonialism, depending on your cup of political tea.

At any rate, researchers (funded by Bill & Melinda Gates) randomized thousands of pregnant women in South Africa to receive either influenza vaccine or placebo. The vaccine worked, reducing rtPCR-confirmed influenza from 3.6% to 1.8% in the HIV-uninfected women and their newborns (50% efficacy), and from 17% to 7% in the HIV-infected women (58% efficacy). No word yet on the consequences of preventable influenza in the placebo-vaccinated mothers and their babies.

Smoking Cessation and Electronic Cigarettes

Quit while you're ahead. Hospitalizations represent a "quittable moment" for smokers, but follow-up is rarely sustained after discharge. In a randomized trial (n=397), providing 3 months of free smoking cessation drugs and regular automated interactive phone calls after hospital discharge resulted in biochemically-confirmed abstinence rates of 26% at 6 months, vs 15% for those given advice while hospitalized only. Nicotine --> Cocaine? Biochemical and epidemiologic research in mice and men shows that regular use of nicotine makes cocaine use more pleasurable and addictive, according to a new article in NEJM. Regular smoking primes neurochemical pathways that enhance the effects of cocaine, which may be why most cocaine users smoke. The authors then make the dubious leap that e-cigarettes could become a gateway drug leading to cocaine use among minors. E-cigs and kids. Speaking of e-cigarettes, another widely reported study showed that according to survey data, adolescents who used e-cigarettes are also likely to smoke real cigarettes, and that e-cigarettes weren't clearly linked to quitting smoking. Also, more adolescents who have never smoked are experimenting with e-cigarettes. All in all, not surprising nor worrying, since limited use of e-cigarettes is probably not harmful, and if someone has already started smoking real cigarettes, who cares if they then try e-cigarettes? Although these studies have been misconstrued and hyped in the lay press as "e-cigarettes are a gateway drug to smoking", no study has yet answered the burning question: Among never-smoking teens, does experimentation with electronic cigarette use lead to the regular use of real cigarettes? Since this is a relatively simple question to ask and answer in these survey designs ("Did you try e-cigarettes or real cigarettes first?") my guess is, probably not, but stay tuned. Closer to science: another randomized trial supports varenicline (Chantix)'s efficacy for smoking cessation. This one showed that using nicotine patches along with Chantix increased continuous smoking abstinence for 6 months, as compared to varenicline plus placebo patches -- although both groups did well (49% vs 33%). Numerous studies have previously confirmed varenicline's utility for smoking cessation.