Literature Roundup: March 2011

ATS/IDSA guideline committee co-chairs Michael Niederman, MD and Donald Craven, MD respond.

Jonathon Truwit, MD, MBA shares his thoughts on the results of HARP, and the ongoing ARDSNet SAILS trial.

Donald LaVan, MD of the American Heart Association breaks down the Felker NEJM study.

William Meadow, MD, PhD confirms that predicting survival in the ICU is easy ... except when it's not.

Randomized Controlled Trials           (How do we make this list?)

Statins for acute lung injury? (HARP trial) Statins reduce healthy volunteers' inflammatory response to inhaled or injected lipopolysaccharide. Craig et al report the results of HARP, in which the UK investigators gave 80 mg of simvastatin or placebo to 60 acute lung injury patients for up to 14 days. There were no differences in mortality (30%), ventilator-free days or ICU/hospital stay. However, the one-third of the treated group who were left to analyze after 14 days had significantly lower SOFA organ dysfunction scores. They also had a non-statistically significant improvement in hemodynamics at day 14 (0 of 9 [simvastatin] vs. 3-4 of 10 [placebo] requiring vasopressors or inotropes, p=0.05-0.09), and significantly lower IL-8 in BAL fluid. No adverse events were noted. Larger trials are underway to explore this further. (n=60). AJRCCM 2011;183:620-626.  

Diuretics in acute decompensated CHF: How "hard" should we "hit" our patients with Lasix? And does that "hurt?" Felker et al asked that (sort of) in a NHLBI-funded trial. They gave furosemide to 308 patients hospitalized for ADHF in an IV dose either equivalent to their home dose, or at 2.5 times their home dose, and either at q12-hour intervals or by continuous infusion (i.e., patients assigned to one of four groups in a 2x2 table). No one method was superior on the primary endpoint (a composite global assessment of symptoms). Higher doses of Lasix resulted in more fluid output, without significant renal failure. Continuous drips were not superior to bolus infusion in safety or efficacy.(n=308). NEJM 2011;364:797-805.

Sirolimus was helpful for LAM, but with toxicity: McCormack et al randomized 89 women with lymphangioleiomyomatosis to receive 2 mg rapamycin or placebo daily for 12 months, followed by 12 months of drug-free observation. Women receiving sirolimus had stable lung function on spirometry, fewer symptoms, fewer "respiratory events," and better quality of life during the treatment period. This was tempered by GI, dermatologic, hematologic, and metabolic side effects, along with an episode of pericarditis, in the treated group. Symptoms and lung function decline resumed upon stopping rapamycin. Pfizer paid much of the tab, but reportedly didn't participate otherwise. (n=89). NEJM 2011; March 16 online

Dalteparin vs. unfractionated heparin for ICU DVT prophylaxis (PROTECT trial)Dalteparin is a low-molecular weight heparin that, unlike enoxaparin, is safe in people with renal failure. Among 3,746 ICU patients in 6 countries, followed until hospital discharge, there was no difference in the rate of proximal deep venous thrombosis (1' endpoint) between once-daily dalteparin and twice-daily UFH, as detected by twice-weekly ultrasounds. There were less pulmonary emboli in the dalteparin group (2' endpoint; 1.3% vs. 2.3%, p=0.01; number needed to treat with dalteparin to prevent one PE = 100). Unfortunately for physicians, each LWMH may have unique properties and efficacies in different patient populations, and this trial does not answer the question of dalteparin's performance compared to other LMWHs. (n=3,746). NEJM 2011; March 22 online.

Subgroup analysis: Patients were much more prone to clots while in the ICU than on the medical floor. Of the proximal DVTs, 182 of 205 occurred in the ICU rather than on the ward; of 67 PEs, 47 occurred in the ICU. Interestingly, only 13 of the 67 patients with a PE (19%) had a simultaneous proximal DVT.

CAGE MATCH: Tiotropium vs. Salmeterol ... scorecard: time-to-COPD exacerbation (POET-COPD trial):  Who's the best bronchodilator on the block? In 725 centers, in 25 countries, among COPD'ers with FEV1<=70% (moderate-or-worse) and an exacerbation in the past year, after using tiotropium or salmeterol (double-dummy) for 1 year, the winner is.... Tiotropium! Probably! With a time-to-first-exacerbation of 187 vs. 145 days -- but only in the quartile of study patients having the earliest exacerbations. That sounded confusing to us too. Only 36% of patients had an exacerbation during the year, so this is really an analysis of the first 1,844 to have exacerbations of the original 7,376 patients (if our math is right -- they don't provide this number). That said, tiotropium won hands down, with a small/modest benefit that was not apparently influenced by glucocorticoid use, drug discontinuation or other confounders. In absolute terms, tiotropium prevented 0.08 more exacerbations per patient per year (0.64 vs. 0.72) than did salmeterol, especially severe exacerbations. No difference in adverse events. Boehringer Ingelheim was heavily involved in the design and controlled the statistical analysis of the study. (n=7,376). NEJM 2011;364:1093-1103.

Inspiratory muscle training liberated the hard-to-wean: In a rare encouraging positive study in the chronically critically ill, 71% of vent-dependent patients (~6 weeks on MV) who did multiple sets of daily inspirations backwards through a PEEP valve weaned from mechanical ventilation (defined as >72 hrs off the vent), vs. 47% receiving a sham treatment. MIP also improved by 10 cmH2O in the treatment group, in University of Florida ICUs. (n=69). Critical Care 2011;15:R84. FREE FULL TEXT

Chantix worked in people with mild to moderate COPD: Tashkin et al randomized 504 patients at 27 centers with COPD and FEV1 > 49% to varenicline or placebo for 12 weeks, with 52 weeks follow up. Abstinence rate (carbon monoxide-confirmed) was 42% vs 9% in weeks 9 - 12 (1' endpoint), and 19% vs. 6% in weeks 9 - 52 (p<0.0001). And even better, no one killed themselves!  (n=504). CHEST 2011;139:591-599.

Xolair for inner-city kids with asthma (ICATA study):  Busse et al randomized 419 inner-city kids (age 6-20) with uncontrolled, skin-test-positive, mostly severe allergic asthma and IgE < 1,300 IU/mL to 60 weeks of omalizumab (Novartis) or placebo. The treated group had modest improvements: ~1 fewer day of symptoms per month and a 5% absolute reduction in hospitalizations (1.5% vs 6.3%); 18% fewer of them had exacerbations (30% vs 48%), especially during fall & spring. Novartis, other pharmas and the NIH funded this, and most of the investigators report receiving direct payments from Novartis. (n=419). NEJM 2011;364:1005-1015.

Chronic rhinosinusitis: oral steroid "priming" helped: Valdyanathan et al gave prednisolone 25 mg (equivalent to prednisone) or placebo daily for 2 weeks to sufferers of chronic rhinosinusitis in the UK, followed by 8 months of nasal fluticasone (drops, then spray) for all. The orally-treated group had a 2.1-point fall in their polyp grade and could smell better, at 6 months. Ann Int Med 2011;154:293-302. 

Apixaban better than aspirin for atrial fib: Warfarin is the best drug for stroke prevention in atrial fibrillation. In people with a-fib for whom warfarin was unsuitable (unstable INR, unlikely adherence, etc, but excluding those with prior bleeding on warfarin), twice daily oral direct Xa inhibitor apixaban had a hazard ratio of 0.45 over aspirin for the primary outcome of stroke (absolute risk for stroke of 1.6% vs 3.7% per year), with equivalent risk for major bleeding (~1.3% per year, <0.4% for intracranial hemorrhage). The trial was funded by industry, who are also investigating apixaban's efficacy in treating DVT/PE. (n=5,599). NEJM 2011;364:806-817.

Inhaled denufusol for early CF: Accurso et al report phase 3 results from industry-funded TIGER-1, testing t.i.d. inhaled denufusol (an ion channel regulator that alters sodium/chloride transport to improve mucociliary clearance) vs. placebo in early CF (FEV1>74% predicted). After 24 weeks, FEV1 was 45 mL better in the treated group, most of whom were younger than age 19. There was no difference in exacerbation rates. (n=352). AJRCCM 2011;183:627-634.

Corticosteroids in community-acquired pneumonia: Fernandez-Cerrano et al randomized 56 people with CAP in Spain (only 3 needed mechanical ventilation) to receive methylprednisolone bolus-and-taper or placebo, in addition to ceftriaxone and levofloxacin. The steroid-treated group had less hypoxemia (lower PaO2/FiO2). (n=56). Critical Care 2011;15:R96. FREE FULL TEXT
MORE ON THIS: In AJRCCM 2005, Confalonieri found less septic shock and need for mechanical ventilation in the hydrocortisone-treated arm of a randomized group of 46 much-sicker patients with CAP. [2005 article] FREE FULL TEXT

Hydrocortisone in multiple trauma (HYPOLYTE trial): Roquilly et ses amis francais randomized 150 critically ill victims of severe French trauma to hydrocortisone or placebo for 7 days. The treated group had an absolute 16-19% lower risk for hospital-acquired pneumonia, the primary endpoint (36% vs. 51-54% depending on which intention-to-treat analysis you prefer, p=0.007 to 0.01). (n=150). JAMA 2011;305:1201-1209.

Notable Journal Articles:

Budesonide (for asthma) doesn't cause pneumonia: After analyzing all AstraZeneca's clinical trials data on budesonide for asthma (26 placebo-controlled, n~14,000; 60 non-placebo-controlled, n~33,000), O'Byrne et al found no increased risk of pneumonia with use of budesonide by asthmatics, who contracted 1-2 detected pneumonias per 100 patient-years of treatment. AJRCCM 2011;183:589-595.

Critical Care

Empiric antibiotics for HAP/HCAP/VAP were appropriate - yet deadly? 
Authors' controversial advice: shelve the ATS/IDSA guidelines for now:  
In a well-intentioned performance improvement initiative, the IMPACT-HAP team prospectively observed 303 patients at risk for MDR pneumonia at four academic hospitals. Surprisingly, more of those who got appropriate antibiotics died (34%) than those treated off-ATS/IDSA guidelines (20%), with Kaplan-Meier-modeled 28-day survival of 65% vs. 79%, p=0.0044—woops. The authors were concerned enough to recommend against continued adherence to ATS/IDSA HCAP antibiotic guidelines, pending a randomized trial. The study was funded by Pfizer, who incidentally are getting out of the antibiotics business, to the dismay of those paid to worry about the ever-impending apocalypse of superbugs. Lancet Infect Dis 2011;11:181-189.
As you could guess, the authors of the ATS/IDSA guidelines had something to say about this paper and its conclusions. ** COMMENTARY by Drs. Michael Niederman & Donald Craven **

It's official: Someone's being too negative: Meadow et al followed 560 patients in one University of Chicago MICU throughout their ICU stay, cornering team members (attending, fellow, resident, nurse) each day and asking a simple question: Will this patient survive to discharge? The results were rich and fascinating. Of the 433 survivors, 77% had unanimous predictions of survival on all days. Of the 127 who died, 57% were unanimously predicted to eventually die on every day in the ICU. A single day of unanimous prognostication of death had an 84% positive predictive value, but when there was any discordance, PPV fell to 52-66%. Critical Care Med 2011;39:474-479.    **COMMENTS by Dr. William Meadow**

Ventilator-associated tracheobronchitis or pneumonia (VAT or VAP): For one year, Dallas et al prospectively followed 2,060 intubated MICU/SICU patients at Barnes-Jewish, and concluded that 83 (4%) developed VAP and 28 (1.4%) got VAT, which was defined as fever and 100,000 CFUs in tracheal secretions, without an infiltrate. By their reckoning, a third of VATs progressed to VAPs. Pathogens (mainly MDR bacteria) overlapped. No differences in mortality were seen. In light of evolving "never event" reimbursement issues, the associated editorial's title is priceless: "VAT: Public-Reporting Scam or Important Clinical Infection?" CHEST 2011;139:513-518.

Hypothermia looks unhelpful for out-of-hospital PEA/asystole: Dumas et al prospectively observed 1145 consecutive victims of  
cardiac arrest who survived to admission in France between 2000 and 2009. Among VT/VF patients, 65% got therapeutic hypothermia, while 60% of PEA/asystole patients did, with higher proportions receiving hypothermia later in the study period 
(86 and 73% respectively)
Most people in the PEA/asystole cohort had asystole (80%). 
Therapeutic hypothermia was strongly associated with improved neurologic outcome in VT/VF (44% in those cooled vs. 29% in those not, p<0.001), but seemed to make no difference in PEA/asystole (15 vs 17%).  Circulation 2011;123:877-886.

Weaning protocols for mechanical ventilation: Cochrane systematic review and meta-analysis: Pooling 11 randomized trials, n=1,971, Blackwood et al suggested there are some benefits to using a standardized weaning protocol. There was a reduction in weaning and ventilation time in the groups that had weaning protocols; we won't tell them to you though, because they would be misleading. They were reported in aggregates (as "geometric means," not recognizable clinical data like ventilator-days), due to the statistical sleight of hand imposed by high heterogeneity and variable outcomes reported across trials. If you're interested: BMJ 2011;342:c7237.  FREE FULL TEXT

No racial disparities seen in Cali ICUs: Sara Erickson et al examined data from 9,518 ICU patients in California hospitals, and concluded that race & ethnicity did not influence mortality or length of stay, after adjusting for disease severity and other factors. Crit Care Med 2011;39:429-435.

Chest tubes for transudative effusions? Kupfer et al looked retrospectively at 168 vented patients with transudative pleural effusions at Maimonides in Brooklyn. The half that got chest tubes (with an average of 1,200 mL drained) spent 3.8 days on the vent, vs 6.5 days for the group that got only thoracentesis, with no complications reported. That's interesting, but concluding and titling the paper "Chest Tube Drainage of Transudative Effusions Hastens Liberation from Mechanical Ventilation" without a randomized trial seems a bit, well, hasty. CHEST 2011;139:519-523.

REALITY CHECK:  literature review by Medscape this month concluded there is no good evidence for or against draining pleural effusions in vented patients ... because, astonishingly, none of the 19 studies reviewed (n=1,124) reported mortality, duration of ventilation, or ICU LOS. There's an editorial this month and a recent meta-analysis 
on this thorny subject, both 
in Critical Care, 
as well as the results of a multicenter survey on local practices in the CHEST 12/2010 issue. And if that's not enough for you, Seminars in Respiratory & Critical Care Medicine's most recent issue is all about pleural disease.

Hypertonic saline vs. mannitol for elevated ICP: meta-analysis

Hooman et al pooled 5 randomized trials with 112 patients, who had 184 episodes of elevated intracranial pressure. They concluded hypertonic saline is better at reducing ICP; but the relative-risk-for-ICP-control confidence interval was 1.0-1.3 and the mean ICP reduction included zero (-1.6 to 5.7 mm Hg). So, maybe. Crit Care Med 2011;39:554-559.

Hyperoxia after cardiac arrest maybe not so bad: The EMShockNet team previously claimed in JAMA that hyperoxia (paO2>300) after cardiac arrest could be deadly (63% in-hospital mortality vs 45% in normoxic patients). After looking at 12,108 database records in Australia/New Zealand, Bellomo et al beg to differ. By applying severity of illness/propensity models, they found no increased risk of death in those with high paO2 after cardiac arrest. Critical Care 2011;15:R90.
APC for septic shock: Sadaka et al add their two cents and a cohort study to the Xigris efficacy/safety debate. In their retrospective, propensity-matched analysis of 563 patients sourced from the Project IMPACT database, the 108 who received activated protein C had 35% mortality, vs 54% for the 108 who did not (p=0.005). There were unusually low rates of bleeding (2.8% GI bleeding in APC group, no brain hemorrhages), compared to previous trials of APC. Authors reported no competing interests. Critical Care 2011;15:R89.

Lung Cancer

TBNA usually types NSCLC correctly: In 186 patients who underwent TBNA at an experienced center, followed by endoscopic biopsy or surgery as a gold standard, the positive predictive value of FNA cytology for adenocarcinoma was 92% and for squamous cell carcinoma, 82%. J Thorac Oncol 2011;6:489-493. 

Quitting smoking heralds lung cancer? Almost half of 115 Philadelphia veterans diagnosed with lung cancer had recently quit smoking, at a median 2.7 years prior to their lung cancer diagnosis; only 11% recalled having any symptoms of lung cancer at the time of quitting.  

Infectious Disease, Tuberculosis

Blood test for invasive fungal infections: (1-->3) Beta-D-glucan, or BDG, is a component of the cell wall of most fungi (not Zygomycetes or Cryptococcus though). Karageorgopoulos et al pooled 16 studies, including 2,979 patients with definite or likely fungal infections, that reported the test characteristics of serum BDG. The pooled sensitivity of serum BDG for invasive fungal infection was 77% and its specificity was 85%; however, statistical heterogeneity was high among the studies and most used the EORTC-MSG clinical criteria (not culture or histopathology for example) as their diagnostic gold standard for comparison. Clin Infect Dis 2011;52:750-770. 
Fine-toothed comb blames inhaled steroids for TB:  Brassard et al analyzed a Canadian database cohort of >427,000 people and, finding 564 cases of tuberculosis, announced that users of inhaled corticosteroids had a rate ratio of ~1.26 to 1.97 over never-users, with current and high-dose ICS users on the higher end of that range. Somehow, they also found that oral corticosteroids reduced the risk association, though OCS are strongly believed to increase risk for TB. The whole thing seems a bit suspect, but is included here FYI. AJRCCM 2011;183:675-678.

Public Health, 
Cardiovascular Disease

No increased risk detected from quitting smoking just before surgery: 

In a literature review and meta-analysis (actually 3 meta-analyses), Myers et al examined 9 studies and found no evidence of increased complications (pulmonary or otherwise) associated with quitting smoking within 8 weeks before surgery of any kind. They include, and debunk, the 1989 cohort study that birthed the dogma that just-quit smokers are at increased risk for post-op pulmonary complications. That paper, which was not designed to answer the question at hand, in fact did not show a statistically significant difference in complication rates between continuing smokers and just-quit smokers (the salient comparison). There was wide heterogeneity between studies, hence the multiple meta-analyses. Their sensible conclusion: counsel everyone to quit smoking, regardless of surgery timing. Arch Int Med 2011:online FREE FULL TEXT

What's her name? As Andy Rooney might say, "With all the diagnoses doctors like to make these days, didjever wonder if anyone is actually healthy anymore?" When it comes to heart health, according to Bambs et al, the answer is yes: one person is. Circulation 2011;123:850-857.

Policy & Medical Education

Why academic care isn't the best care: Efficient private systems like Geisinger and Kaiser outperform academic institutions on  
accepted measures of quality
. Dhalla & Detsky say that's because academic docs aren't trying, because thanks to misguided incentives, their careers suffer if they do. They encourage hospitals and payers to take the lead to reward quality improvement, since universities are unlikely to. JAMA 2011;305:932-933.

The new iUnit, by Apple? An excess amount of the chaos and stress endemic to ICUs is due to poor systems engineering (think alarm fatigue); a rational, integrated approach to design & technology selection is needed to make ICUs more effective and safer for patients, argue Mathews & Pronovost. Academic health systems should fill the leadership vacuum, organize stakeholders and drive change by demanding more integrated systems from vendors and partnering with them to make them successful. JAMA 2011;305:934-935.

Where the wild things are: Arthur Kleinman, physician and ethnographer, challenges you to abandon comforting but simplistic value constructions and look unblinkingly at your divided medical soul, if you dare. Lancet 2011;377:804-805.

Review Articles:

Critical care for patients with stroke. Lukovits TG, CHEST 2011;139:694-700.

Heparin-induced thrombocytopenia in the ICU. Sakr Y, Critical Care 2011;15:211.

The effect of light on critical illness. Castro R, Critical Care 2011; 15:218

New treatments against gram-negative organisms. Bassetti M, Critical Care 2011;15:215.

Pulmonary Sarcoidosis: Concise Clinical Review. Baughman RP, AJRCCM 2011;183:573-581.

Lymphangioleiomyomatosis. Harari S, Eur Resp Review 2011;20:34-44.

DVT of the upper extremity: Kucher N, NEJM 2011;364:861-869.

Autonomy and beneficence, a historical perspective. Will JF, CHEST 2011;139:669-673.

Special Issues:

Pregnancy and respiratory disease: Radiation risks from imaging; pharmacy and lactation; critical illness; interventional chest procedures; asthma; smoking; cystic fibrosis; pneumonia; infiltrative lung diseases; pulmonary embolism; pulmonary hypertension, and sleep-disordered breathing in pregnancy. Coming in just under 200 pages. Clinics in Chest Medicine 2011;32:1-198.

 The upper airways and nasal passages: imaging, airflow, rhinosinusitis, congestion, allergies, the sense of smell  ... more than a dozen articles on "the other airway" is nothing to sneeze at. Proc Am Thorac Soc 2011; 8: 1-140.

The pleura and pleural diseases: Pleural imaging; CHF pleural effusions; hepatic hydrothorax; parapneumonic effusions and empyema; malignant pleural effusions and disease; chylothorax; thoracoscopy; the chest tube size debate; pneumothorax -- did they miss anything? In 130 pages, probably not much. Sem Resp Crit Care Med Dec 2010.

Bonus! More Free Full Text:

The glass is only half-full ... of cancer: People with lung cancer who scored high on optimism metrics on the MMPI personality test lived a remarkable 6 months longer than the pessimists, independent of treatment, stage, comorbidities, etc, in a multivariate analysis of 534 patients by Novotny et al. J Thorac Oncol 2010;5:326-332. FREE FULL TEXT

"Downstaging" NSCLC IIIa prior to resection may often be wishful thinking: Candela et al looked at 32 articles and considered the test characteristics for CT, PET, EUS/EBUS, and re-mediastinoscopy. They concluded that an unacceptably high 20-40% of patients considered "downstaged" to stage I or II after induction chemotherapy for NSCLC are due to false negatives of restaging tests, particularly CT. EUS/EBUS performed best as a restaging test, with a 15% false negative rate. Pathology at resection was the gold standard. J Thorac Oncol 2010;5:389-398. FREE FULL TEXT