You don't need to be signed in to read BMJ Blogs, but you can register here to receive updates about other BMJ products and services via our site.

Pulm- infection

Primary Care Corner with Geoffrey Modest MD: Lung ultrasound to diagnose heart failure

4 Aug, 15 | by EBM

By: Dr. Geoffrey Modest

A potentially very useful multicenter study was just published finding that lung ultrasound (LUS) was better than chest xray (CXR) in the diagnosis of acute decompensated heart failure (ADHF) in patients presenting to the emergency room (ER) with dyspnea (see CHEST 2015; 148(1): 202 – 210). This is an increasingly common issue, with evidently very different therapeutic approaches depending on the etiology.


 –1005 patients from 7 Italian ERs presented with acute dyspnea, and the ER MD was asked to categorize the dyspnea diagnosis as either from ADHF or noncardiogenic dyspnea, based on their initial clinical assessment (this was not prescribed but left up to the ER MD, to simulate a real-world situation) and after LUS. All patients also had CXR’s. After patient discharge, independent reviewers  determined the cause of the dyspnea.

–544 (51.4%) patients were in community hospitals, 461 (45.9%) academic medical centers. Median age 77, 46.2% women, 41.8% smokers, 40.7% with COPD, 63.5% hypertensive, 20.2% with congestive heart failure (CHF), 25.8% with ischemic cardiomyopathy/CAD, 27.9% diabetic. Most on lots of meds. So, pretty sick group overall


–final diagnosis of ADHF was in 463 patients (46%), with very high agreement between the 2 physician adjudicators of the diagnosis.  81.7% of the patients were admitted to the hospital.

–clinical workup and LUS alone were much more accurate than CXR (p<0.01):

–I included below 2 of their summary figures. The one to the left is the summary sensitivity, specificity, etc in the whole group (top) and in those who had BNP/NT-pro-BNP sent. The  figure on the right is the ROC curves, showing that clinical evaluation plus LUS were remarkably accurate: they achieved sensitivity, specificity and AUC values for diagnosis of ADHF of 97.0%, 97.4%, and 0.972!!! CXR had sensitivity as low as 69.5% (which is higher than in some studies…..) — so, absence of CXR findings does not even come close to excluding a diagnosis of ADHF!!

–the net reclassification improvement (NRI), an estimate of the % of subjects moving from one diagnosis to another, of the LUS approach vs standard clinical workup was 19.1% (p<0.01)

–the accuracy of BNP/NT-pro-BNP values (available on subgroup of 486 patients) was significantly lower than the LUS approach (p<0.01): the sensitivity/specificity was 85%/61.7% with AUC of 0.733, again being not nearly as sensitive as LUS.

–there was near perfect intraobserver agreement in reading LUS (0.97) among experts, and a really good agreement (0.92) for ER MDs with limited LUS training (<10 examinations performed)



 One of the strengths of this study is that it incorporated the clinical pretest probability (the “LUS-implemented” approach), which was much more accurate than LUS alone, which was better than CXR (though, to be fair, they did not look at a “CXR-implemented approach”, but the stand-alone CXR was still really inferior to the stand-alone LUS).  In this study LUS was done as a rapid, bedside test, with a positive test for ADHF defined as “bilateral presence of 2 or more zones showing the presence of at least 3 B-lines: vertical, hyperechoic reverberation artifacts extending from the pleural line to the bottom of the screen”, these artifacts are presumably from interstitial fluid in the lung. This was a stricter definition of a “positive” LUS than in prior studies, which could explain the greater accuracy here — prior studies have found more limited LUS accuracy (as low as 54%) in ADHF diagnosis. One potential bias in this study is that the same MD who made the initial clinical diagnosis also performed the LUS and was therefore not blinded to the results of the clinical workup in interpreting the LUS.

So, bottom line: this study comes on the tail of one earlier this year showing better accuracy of LUS than CXR for pneumonia (see here). Together, these studies reinforce the potential to use a noninvasive and nonradiation-associated evaluation for both CHF and pneumonia. Although this was an ER-based study, it is likely to be applicable to patients who come into primary care with decompensated heart failure, though it would be great to have a community-based study done — the majority of patients with ADHF I see at the health center are not sent to the ER or admitted.  I treat these patients based on clinical assessment and CXR, with very close followup, and they do fine clinically. But I imagine that the availability of LUS would allow for a more assured diagnosis.

Primary Care Corner with Geoffrey Modest MD: Lung ultrasound to diagnose pneumonia?

27 May, 15 | by EBM

By: Dr. Geoffrey Modest
An Italian prospective study was just published looking at the merits of ultrasound in diagnosing pneumonia (see Am J Emerg Med 2015 May; 33:620.). Prior studies have found that chest xray (CXR) is not so sensitive, 43.5% in one study when compared to chest CT (which is considered the gold standard). details:

–275 patients with respiratory complaints (60% dyspnea, 25% cough, 10% purulent sputum, 15% pleuritic chest pain) and underwent chest CT. Mean age 71, 53.3% women.
–Chest CT was positive for pneumonia (PNA) in 87 patients, with “almost perfect” concordance between the readings of 2 radiologists.
–lung ultrasound (LUS), done within 3 hours of the CT, and was positive in 81 patients. 72 of them (88.9%) also had a positive chest CT

–the 9 false positives were: 3 lung cancer nodules, 3 cases of impaired ventilation not due to PNA, 3 cases of fibrotic bands
–the 15 false negatives were: 3 cases of posterior consolidations (the LUS was only performed on anterior-lateral chest), 5 cases of deep infection without contact with the pleura
–overall sensitivity was 82.8% (73.2-90%) and specificity was 95.5% (91.5-97.9%)
–interobserver variability: “almost perfect”
–in subgroup of patients with pleuritic chest pain: sensitivity was 91.7% and specificity was 97.4%
–in subgroup wihtout pleuritic chest pain: sensitivity was 81.3% and specificity was 95% (other studies found LUS less reliable with deep-seated infections)

–CXR was done in addition to LUS in 190 patients:

–sensitivity of LUS was 81.4% and of CXR was 64.3% (difference significant with p=0.036); specificities were similar (94.2% vs 90%)

–so, overall, LUS rules in consolidation well (+ likelihood ratio of 18.2), and is moderately good in ruling out consolidation ( – LR=0.18)

so, pretty inmpressive. This study confirms the relatively poor showing of CXR in diagnosing PNA, and found that LUS was really quite good, and the usual ultrasound concern of intra-observer variability was minimal. Although this needs to be validated in other settings, this study does hold out the promise of a noninvasive and nonradiation-associated evaluation of pneumonia.

Primary Care Corner with Geoffrey Modest MD: Antibiotics for pneumonia

16 Apr, 15 | by EBM

By: Dr. Geoffrey Modest

A Dutch study looked at 2283 patients with community-acquired pneumonia (CAP) admitted to the hospital but not the ICU during three 4-month periods in 2011-2013, with different antibiotics as the preferred empirical treatment for each period, and with the order of these treatments randomized separately for the different hospitals (N Engl J Med 2015;372:1312-23).


–Case definition: at least 2 of the clinical criteria (below)

–Clinical criteria: cough, purulent sputum or change in sputum character, temp >38C or <36.1C, lung exam with rales and/or consolidation, WBC >10K or >15% bands, CRP >3x upper limit of normal, dyspnea/tachypnea/hypoxemia, new or increased infiltrate on xray

–Median age of patients 70. 2.5% had received PPSV23 vaccine and about 2% PCV13. 20% had cardiovascular disease, 40% COPD or asthma, 10% other pulmonary diseases, 16% cancer, 18% diabetes. 77% had radiologically confirmed CAP

–Intervention strategies:

–β-lactam: amoxicillin (in 31%), amoxacillin plus clavulanate (in 43%), or 3rd generation cephalosporin (ceftriaxone in 17%)

–β​-lactam/macrolide: penicillin, amoxacillin, amoxacillin plus clavulanate, or 3rd generation cephalosporin, in combination with azithromycin, erythromycin or clarithromycin

–fluoroquinolone: moxifloxacin or levofloxacin

–Primary outcome: all-cause mortality within 90 days of admission; secondary outcomes: time to starting oral treatment, length of hospital stay, minor or major complications in the hospital


–microbiology: 15.9% s pneumonia, 6.8% h influenza, 2.1% atypicals (mycoplasma pneumoniae, chlamydia pneumoniae, and legionella species), 2.5% viruses, 65% no pathogen identified.

–most frequent reason for deviation from the β-lactam strategy was perceived need to cover atypicals (8.1%)

–Comparing the strategies of β-lactam, β-lactam-macrolide combo, fluorquinolone, respectively, crude 90-day mortality was  9.0% (59 patients) vs 11.1% (82 patients) vs 8.8% (78  patients) – nonsignificant differences. Also nonsignificant when adjusted for pneumonia severity or if limited to group with xray-confirmed CAP.

–These results suggest noninferiority of the β-lactam strategy, even after controlling for providers’ deviating from the antibiotic strategy by clinical considerations.

–median length of stay 6 days for all strategies

–median time to starting oral therapy in the hospital was 3 days for fluoroquinolones and 4 days for others

— they did do urinary antigen testing for strep pneumoniae and legionella pneumophila. Those with legionella were preferentially given ciprofloxacin or other appropriate antibiotics

​–The pathogens found were similar among the strategy groups, but there was more bacterial resistance in the group on the β-lactam​ strategy. This was not associated with worse clinical outcomes.

Current US guidelines:

–US guidelines (IDSA/ATS –infect disease soc of America/amer thorac soc, 2007): for outpatient treatment

–previously healthy and no use of antimicrobials in past 3 months: macrolide (azithro, clarithro, erythro) OR doxycycline (prefer macrolide)

–presence of comorbidities (chronic heart, lung, liver, renal disease; diabetes; alcoholism; malignancy; asplenia; immunosuppression; or use of antimicrobials in the past 3 months): fluoroquinolone (moxifloxacin, gemifloxacin, or levofloxacin [750mg] OR β​-lactam (first line: high-dose amoxicillin, amoxicillin-clarulanate; alternatives: ceftriaxone, cefpodoxime, or cefuroxime) PLUS a macrolide.

–in regions with high rate (>25%) of infection with high-level (MIC>=16) macrolide-resistant s. pneumoniae, consider one of the alternative agents and don’t use a macrolide by itself. [note: the US overall had a resistance rate of 41.15% in 2009 (the UK by contrast, has 4.20%)

Overall, pretty low incidence of atypical CAP infections during this study. Although this was a hospital-based study, it seems that an outpatient strategy, with likely less sick patients, would yield similar results.  A major concern in this study is the remarkably low rate of vaccination (<5% had either of the pneumococcal vaccines). The vast majority of these patients, by age or comorbidities, would qualify for both PPSV23 and PCV13 vaccines, and this is especially important because of the worldwide spread of antibiotic-resistant s. pneumoniae. For example, a 2002 article found that 14-16% of pneumococci were multi-drug resistant (at least 3 classes, including β-lactam, macrolides, tetracyclines, sulfonamides, and chloramphenicol). In terms of atypicals), there was a meta-analysis comparing β-lactam antibiotics vs those active against atypical pathogens, finding that in 18 trial with 6749 patients with non-severe community acquired pneumonia, there was no advantage in using specific antibiotics against atypicals other than those with Legionella (in subgroup analysis), where there was a significantly lower failure rate in those on an antibiotic active against Legionella (RR 0.40, 0.19-0.85), a rare cause of pneumonia in this meta-analysis (only 75 patients in whole sample, or about 1%). In this meta-analysis, there was no difference in cure rates between antibiotics, including those considered ineffective for mycoplasma or chlamydia pneumoniae. (see BMJ. 2005;330(7489):456)​​.

So, what makes sense? First, i think we should be aggressive in vaccinating people against pneumococcal pneumonia, the most common of the bacterial pneumonias. Second, it is impressive that both mycoplasma and chlamydia seem to do okay with β-lactam antibiotics, in both this study and the BMJ meta-analysis. I continue to be very concerned about increasing bacterial resistance. Articles in the past have promoted levofloxacin as the go-to antibiotic for pneumonia. But, of course, that might well lead to unintended consequences. Specifically bacterial resistance both with lung pathogens and others (eg, the increasing resistance of H pylori to antibiotics, exacerbated by the fact that single agents are unable to cure the infection but quite good at causing resistance). Perhaps the safest strategy is to use a macrolide, when resistance levels are low, or doxycycline, as suggested by IDSA/ATS, or to use a β-lactam but make sure the CAP was not from legionella (a very unusual cause), which does do better with a fluorquinolone or macrolide. Based on the above studies, I do find the last strategy appealing, though would be great to have a prospective trial confirming it.

Primary Care Corner with Geoffrey Modest MD: Bronchiolitis guidelines for kids

28 Jan, 15 | by EBM

By: Dr. Geoffrey Modest

The American Academy of Pediatrics updated their 2006 guidelines on diagnosis and treatment of bronchiolitis In brief:

–For diagnosis: It’s all in the history and physical for both diagnosis and assessing disease severity — assess risk factors for severity (90%, chest PT, antibiotics (unless concomitant bacterial infection or strong suspicion). use nasogastric or IV fluids if infant cannot maintain hydration orally. Use pulse oximetry if clinically indicated.
–Give palivizumab in first year of life if hemodynamically significant heart disease or chronic lung disease of prematurity (eg 21% for at least the first 28 days of life). Give 5 monthly doses during RSV season.
–All people should disinfect hands before and after direct patient contact, use alcohol-based rubs for hand decontamination when taking care of kids with bronchiolitis, check on smoking exposure and counsel to avoid exposing the infant/child, encourage exclusive breast-feeding for first six months to decrease morbidity of respiratory infections.

So, the biggest change is to not use b-agonists/albuterol. Their recommendation is (it seems to me) a bit stronger than the data they reviewed — several meta-analyses found some improvement in symptom score (though this is noted by them to be observer dependent and may not be reliable). A few studies have looked at pulmonary function tests as an objective measure, and a couple of studies did not find any objective difference in infants in tidal breathing responses to albuterol​ — however, on looking further I found a recent Cochrane review on the subject of bronchodilators in infants with bronchiolitis (see DOI: 10.1002/14651858.CD001266.pub4.​10.1002/14651858.CD001266.pub4). I tried to look at the 3 specific PFT studies they cited from which they concluded there was no utility of albuterol , could only find 2, which had only 20 and 22 infants, one of which did find a statistically significant, but small and questionably clinically important improvement with albuterol, and one did not — the other study I could not find in PubMed or even the journal itself in the Harvard on-line library). In general, studies of albuterol have not found that they affect disease resolution, need for hospitalization, or length-of-stay in hospitals. And the guideline writers are concerned about cost and adverse effects (tachycardia, tremors). So, it seems that albuterol is unlikely to help most infants. But this does raise some questions. For example, we know that the correlation between PFTsand function in COPD and asthma is not great, so should we really consider this the objective gold standard in bronchiolitis (esp. when several meta-analyses have found some clinical improvement)? What about in kids with mild-to-moderate bronchiolitis? Effectively sending a very small infant away from the clinic and giving the parents nothing to treat them with may have some significant consequences: e.g, increased parental anxiety and increased likelihood of going to the emergency room. and, again, per the clinical guidelines: “several meta-analyses and systematic reviews [they cite 6] have shown that bronchodilators may improve clinical symptom scores”.​ Since the diagnosis of bronchiolitis is a clinical one, it is also likely that some of the kids have bronchospasm/reactive airway disease from a viral infection, which in fact may respond to b-agonists. All-in-all, in kids with mild to moderate bronchiolitis/wheezing, it seems that one could make a cogent argument to try albuterol and stop it if it is not working. So, I’m not so sure about this recommendation…..

Primary Care Corner with Geoffrey Modest MD: Antibiotics for URI

21 Mar, 14 | by EBM

BMJ online recently published an article showing utility of delaying antibiotics, or giving no antibiotics, in patients with upper resp infection (see doi: 10.1136/bmj.g1606 ). in brief,

–889 pts over age 3 (median age 30) in 25 practices in UK were given advice about symptom control (antipyretics, +/- vaporizer) and then either immediate antibiotics in 333 (37%) (the group felt to have more severe sx and felt more likely to have lower resp tract infection), and in 556 (63%) randomized to the following different strategies:

–having the patient subsequently recontact the clinic to request a prescription if needed — pts told that symptoms typically worst on days 2-4
–post-dating the prescription
–having the patient come back to the clinic just for a prescription if needed
–given a prescription but asking the patient to wait to fill it
–no antibiotic prescribed


 –mean symptom severity at 3-4 days no diff between getting no prescription vs any of the delayed methods.
–no diff in duration of sx rated moderately bad or worse (median 3-4 days), whether in the immediate script, delayed script, or no script groups
–nonsignif difference in patient satisfaction in the different groups, belief in antibiotics, or antibiotic use, though no benefit in symptom severity or duration of symptoms. no diff in patient revisits to the clinic.
–with the no prescription or delayed prescription policies, the median day that antibiotics were started was day 4, and <40% of patients actually took antibiotics (approx 25% of the no prescription group ultimately got antibiotics)

 the background here is that, per the NICE guidelines (natl institute for health and care excellence in the UK), it is suggested to give an antibiotic prescription advising the patient to wait for at least the expected natural history of the illness (the study found that delayed scripts only have a slightly higher fill rate than giving no antibiotics). the advantage of the delayed script over none is that the former decreases the likelihood of patients coming back for another visit. also didn’t seem to matter which method used for delayed scripts. and, in this study, only a minority of patients actually filled the prescriptions, even if they were given them with the advice to wait a few days to see how they were.  also, of note, no difference in disease response in the group given immediate antibiotics vs those in delayed or no antibiotics (not an RCT, however). my experience is that, at least in our clinic and perhaps through reiteration over the years of the lack of utility of antibiotics, most patients are happy to be reassured that “it is just a virus” and not insist on getting antibiotics. and, perhaps in some patients, it is a good idea to see them in a few days if they are not getting better/getting worse.


Primary Care Corner with Geoffrey Modest MD: Antibiotics and ibuprofen (still) don’t help acute bronchitis

16 Oct, 13 | by EBM

recent online article in BMJ finding lack of efficacy of antibiotics or ibuprofen on acute bronchitis with discolored sputum (see BMJ 2013;347:f5762 doi: 10.1136/bmj.f5762). 416 adults aged 18-70 in primary care centers in spain with sx of respiratory tract infection of less than 1 week, cough predominant with discolored sputum, and at least one other sx suggesting lower resp tract infection (dyspnea, wheezing, chest discomfort or pain). chest xray not required to r/o pneumonia — the dx of acute bronchitis was a clinical one. randomized to ibuprofen 600 TID, amox/clavulanic acid 500/125 TID, or placebo for 10 days.  patients reported sx in diary. overall, 40% of patients were smokers, 10% had diabetes, 8% with fever, >50% had increased CRP with 25% quite high CRP (>21). results


–median # days from the initial presentation of frequent cough slightly decreased with ibuprofen (9 days) vs those on abx (11 days) or placebo (11 days), not statistically different

–no diff in probability of cough resolution

–adverse effects common with abx (12%), vs ibuprofen (5%) or placebo (3%).


so, not a lot to do for patients with acute bronchitis. (note: this does not apply to patients with copd and exacerbation with change in color of their sputum). cochrane reviews do not show benefit from b-agonists or anti-tussives. it is pretty clear that providers do prescribe antibiotics a lot for acute bronchitis (>60% of the time), and even more so if there is discolored sputum (3.2 times as likely as in patients without discolored sputum). would have been better in this study if they evaluated utility of antibiotics or ibuprofen in patient subgroups (eg, smokers, or diabetics, or those with higher CRP), but overall this article provides further impetus to avoid prescribing antibiotics, with their adverse effects on the patients and on society overall from resistant bacteria.

EBM blog homepage

Evidence-Based Medicine blog

Analysis and discussion of developments in Evidence-Based Medicine Visit site

Creative Comms logo

Latest from Evidence-Based Medicine

Latest from EBM