EU study shows wide range in antibiotic prescribing for febrile kids

A new study of European emergency departments has found that antibiotic prescribing for children who show up with a fever varies widely across the continent, and many of those children are discharged with a broad-spectrum antibiotic. The findings appeared yesterday in The Lancet Infectious Diseases.

The cross-sectional, observational analysis of more than 5,100 children admitted to 28 emergency departments in 11 countries across Europe over a 16-month period found that about a third of the children who presented with a fever but no other illnesses received an antibiotic on discharge, with the prescribing frequency ranging from 19% in Switzerland to 64% in Turkey.

More than 60% of those prescriptions, which were written mainly for respiratory tract infections, were for second-line antibiotics.

While the study did not evaluate whether the antibiotics were appropriately prescribed, the authors say the findings indicate a significant amount of overprescribing is occurring, particularly in children with respiratory infections, and that the drivers of this inappropriate prescribing need to be identified and addressed at the hospital and national level.

Respiratory infections drive prescribing

The study of hospitals participating in the Research in European Pediatric Emergency Medicine (REPEM) network looked at emergency department visits by children aged 1 month to 16 years who presented with a fever on one random sampling day per month from November 2014 to February 2016.

The researchers then determined the proportion of those children without comorbidities who received antibiotics on discharge by country, type of infection, and type of antibiotic. They then performed a detailed analysis of that population to assess the factors that influenced variability in prescribing.

Of the 5,177 children who were included in the analysis, 617 (12%) had comorbidities. Of the 4,560 children without comorbidities, 1,454 (32%) received an antibiotic, and 893 (61%) of those prescriptions were for second-line antibiotics, which included doxycycline, broad-spectrum penicillins, and second- and third-generation cephalosporins. The countries with the highest prescribing rates (Turkey, the United Kingdom, Hungary, Italy, and Romania), were also the most frequent prescribers of second-line antibiotics.

When the researchers looked at the type of infections for which antibiotics were prescribed in children without comorbidities, they found that respiratory tract infections (upper and lower) accounted for 83% (1,208 of 1,454) of all prescriptions. The variation in prescribing frequency across hospitals was even greater for these infections, ranging from 15% (Switzerland) to 67% (Turkey) for upper respiratory tract infections and 24% (Switzerland) to 87% (Turkey) for lower respiratory tract infections.

Analysis of patient characteristics found that older age (odds ratio [OR], 1.51; 95% confidence interval [CI], 1.08 to 2.13), longer duration of fever (OR, 1.45; 95% CI, 1.01 to 2.07), blood concentrations of C-reactive protein (OR, 2.31; 95% CI, 1.67 to 3.19), and chest x-ray results (OR,10.62; 95% CI, 5.65 to 19.94 for focal abnormalities; OR, 3.49; 95% CI, 1.59 to 7.64 for diffuse abnormalities) were all associated with the increased likelihood of receiving an antibiotic prescription.

After the researchers adjusted their analysis to account for patient characteristics, diagnostic assessment, and hospital characteristics (hospital type and national healthcare system), they found the rate of antibiotic prescribing for respiratory infections was still highly variable across hospitals and countries, which suggests others factors are involved.

"Even though specific determinants of antibiotic prescription could be identified in the whole population of patients, differences in patient mix, diagnostic assessment, or hospital characteristics could not explain all variability in antibiotic prescription," lead study author Josephine van de Maat, MD, and her co-authors write.

Diagnostic uncertainty

What could explain the variability? As van de Maat and her colleagues explain, even though most respiratory infections are viral in nature and don't require antibiotics, bacterial causes can't be ruled out. But since there is no gold standard for diagnosing bacterial respiratory tract infections, diagnostic uncertainty could influence the decision to prescribe an antibiotic.

And as Elise Launay, MD, and Christele Gras Le Guen, MD, PhD, of Centre Hospitalier Universitaire de Nantes in France, explain in an accompanying commentary, that diagnostic uncertainty could play a bigger role in hospitals in Turkey, where estimated mortality for bacterial pneumonia in children under 5 is significantly higher than it is in western European countries.

"When in doubt, clinicians generally might prescribe antibiotics, thinking that the treatment would be less harmful than missing a severe bacterial infection," they write. "Pneumonia is a good example of a doubtful situation that could explain why children with lower respiratory tract infections showed the most variation in antibiotic prescription frequency."

But, they note, that wouldn't explain why the prescribing rates for lower respiratory tract infections in hospitals were also high in countries like the United Kingdom and Italy, where mortality rates for pediatric bacterial pneumonia are lower.

Whatever the reasons, van de Maat and her colleagues suggest that identifying the source of the problem is critical for stewardship efforts across the continent.

"To ensure successful antibiotic stewardship initiatives in Europe aimed at reducing unnecessary prescription of antibiotics, variability of prescription across hospitals should be considered, drivers of suboptimal antibiotic prescription at the local level need to be identified, and European guidelines need to be devised," they write.

See also:

Feb 28 Lancet Infect Dis study

Feb 28 Lancet Infect Dis commentary

Newsletter Sign-up

Get CIDRAP news and other free newsletters.

Sign up now»


Unrestricted financial support provided by

Bentson Foundation 3MAccelerate DiagnosticsGilead 
Grant support for ASP provided by


  Become an underwriter»