Reference:  Shaikh N, et al. Identifying children likely to benefit from antibiotics for acute sinusitis: a randomized clinical trial. JAMA July 2023

Date: October 17, 2023

Dr. Alasdair Munro

Guest Skeptic: Dr. Alasdair Munro is a clinical research fellow specializing in pediatric infectious disease at the University of Southampton. He is currently involved with clinical trials of vaccines and antibiotics.

Case: A 4-year-old girl presents to your emergency department (ED) with fever and nasal drainage. Her vaccinations are all up to date. Symptoms have been present for the past 12 days. She initially had some cough and congestion which was diagnosed as a viral upper respiratory infection by her primary care doctor. Her symptoms have persisted and yesterday she developed fever (temperature of 38.3°C) and nasal drainage. On physical examination, she has nasal congestion with yellow-colored nasal discharge. The family says to you, “She’s been sick for almost 2 weeks and the color of her nasal drainage changed to yellow. Does this mean she has a bacterial infection that needs antibiotics?”

Background: Distinguishing between sinusitis and viral upper respiratory infections in children is challenging. The symptoms often overlap.

The latest clinical practice guidelines from the American Academy of Pediatrics (AAP) on the diagnosis and management of acute bacterial sinusitis in children was published in 2013 [1]. Based on those guidelines, a presumptive diagnosis of bacterial sinusitis can be made when a child with URI symptoms has:

  • persistent illness (nasal discharge, daytime cough) lasting more than 10 days without improvement
  • worsening course (new or worsening nasal discharge, daytime cough, or fever after initial improvement)
  • severe onset (fever ≥39°C, purulent nasal discharge for at least 3 consecutive days)

This recommendation only has an evidence quality of B.

We don’t routinely perform sinus aspiration on children, but it is thought that the most common pathogens involved in sinusitis include Streptococcus pneumoniae Hemophilus influenzae, or Moraxella catarrhalis. Untreated sinusitis is associated with complications such as pre septal cellulitis, orbital cellulitis. In bad cases, there can be intracranial involvement that includes cavernous venous thrombosis, osteomyelitis, meningitis, or intracranial abscess.

However, in the interest of antibiotic stewardship. We also do not want to be prescribing antibiotics for viral illnesses.

This issue was covered on SGEM #263. We should be thinking about implementing strategies to reduce the unnecessary prescribing of antibiotics in the emergency department.

Clinical Question: What are the potential benefits and harms of antibiotic treatment for children diagnosed with acute sinusitis and does it depend on bacterial pathogen colonization or color of nasal discharge?

  • Population: Children aged 2 to 11 years with persistent or worsening acute sinusitis (as per AAP practice guideline) and symptom score of => 9 on Pediatric Rhinosinusitis Scale (PRSS) [2]. This scale ranges from 0-40 with higher scores representing more severe symptoms. Persistent = nasal symptoms, cough or both for 11 – 30 days without improvement. Worsening = period of improvement followed by worsening nasal symptoms or daytime cough, or new onset fever on days 6 – 10.
    • Exclusion: severe presentation (presence of both colored nasal discharge and fever ≥39°C for 3 or more consecutive days, history of asthma, active wheezing, solely cough, history of allergic rhinitis, immotile cilia syndrome, cystic fibrosis, immunodeficiency, allergy to study medications, concurrent infections, systemic antibiotic use within the previous 15 days, prior sinus surgery, families did not have access to phone or were not English/Spanish speaking
  • Intervention: 10 days of Amoxicillin/Clavulanic acid
  • Comparison: Matching placebo
  • Outcome:
    • Primary Outcome: Symptom burden (PRSS) during first 10 days after diagnosis
    • Secondary Outcome: Treatment failure defined as (1) an increase in PRSS score of greater than 20% from enrollment at any time, (2) a decrease of less than 2 points in the PRSS score from enrollment to day 3, (3) a decrease in PRSS score of less than 20% from enrollment to day 4, (4) a decrease in PRSS score of less than 20% from enrollment on 2 consecutive occasions on days 5 to 11, or (5) a decrease in PRSS score by less than 50% from enrollment to the end-of-study follow-up visit. Acute otitis media (AOM). Receipt of another systemic antibiotic
    • Safety: Non-susceptible pathogen at follow up (S pneumonia or H influenzae). Clinically significant diarrhea (3 or more watery stools for 1 day or 2 watery stools on each of 2 consecutive days). Rash. Other resource utilization. Missed work days.
  • Type of Study: A multi-center, double-blind, placebo-controlled, randomized clinical trial.

Authors’ Conclusions: In children with acute sinusitis, antibiotic treatment had minimal benefit for those without nasopharyngeal bacterial pathogens on presentation, and its effects did not depend on the color of nasal discharge. Testing for specific bacteria on presentation may represent a strategy to reduce antibiotic use in this condition.

Quality Checklist for Randomized Clinical Trials:

  1. The study population included or focused on those in the emergency department. No
  2. The patients were adequately randomized. Yes
  3. The randomization process was concealed. Yes
  4. The patients were analyzed in the groups to which they were randomized. Yes
  5. The study patients were recruited consecutively (i.e. no selection bias). Unsure
  6. The patients in both groups were similar with respect to prognostic factors. Yes
  7. All participants (patients, clinicians, outcome assessors) were unaware of group allocation. Yes
  8. All groups were treated equally except for the intervention. Yes
  9. Follow-up was complete (i.e. at least 80% for both groups). Yes
  10. All patient-important outcomes were considered. Yes
  11. The treatment effect was large enough and precise enough to be clinically significant. Unsure
  12. Financial conflicts of interest. Multiple doctors receive grants or consult for pharmaceutical companies. Dr. Hoberman has patents for amoxicillin-clavulanate with the same company.

Results: There were 2,923 children screened and 515 were randomized. Around 60% were aged 2 to 5 years old. Most (~70%) were included because they had persistent symptoms. Two thirds had colored nasal discharge and around 70% had some pathogen (strep pneumo, haemophilus, moraxella) detected.

Key Results: Antibiotics may have a clinical benefit in children with acute sinusitis at the cost of increased diarrhea while the detection a pathogen had little impact and color of their nasal discharge did not matter.

Primary Outcome:

Between group difference in symptoms score of -1.69 (95% CI -2.07 to -1.31).

When they looked at the subgroups of children who had bacterial pathogens detected versus those who did not, they also noticed some improvement in the scores.

In those with pathogen detected, there was a difference of -1.95 (95% CI -2.4 to -1.51). In those without pathogen detected, there was still a difference but less at -0.88 (95% CI -1.63 to -0.12).

The effects of the antibiotics did not differ significantly between the groups that had clear vs. colored nasal discharge.

Median length of time to symptoms resolution was 7 days for the antibiotic group compared to 9 days for the placebo group (p=0.003)

Secondary Outcomes:

The antibiotic group was less likely to experience treatment failure, develop acute otitis media, or receive additional systemic antibiotics.

Treatment failure had a risk ratio of 0.69 (95%CI 0.54 to 0.88) and NNT 8.

Receiving another antibiotic had a risk ratio of 0.52 (95%CI 0.36 to 0.76) and NNT of 9.

Development of AOM had a risk ratio of 0 (95% CI of 0 to 0.48) and NNT of 32. This occurred in 3% of the placebo group.

They also did not see any statistically significant difference in resource use between the two groups.

Safety Outcomes:

There was not a difference in non-susceptible pathogens detected at follow up between the two groups. Risk ratio 1.16 (0.63 – 2.13).

The antibiotic group did have more diarrhea with a risk ratio 2.4 (1.26 – 4.59) but no significant difference in rash. Risk ratio 2.05 (0.19 – 22.37)

1) Nasal Discharge Color: Some people say that colored nasal discharge is associated with bacterial infections in comparison to clear discharge. A SRMA from 2014 in adults did not support this position [3].

It is interesting that the authors stratified the study by the color of nasal discharge, when this was removed from the PRSS scale due to the inability of parents to be able to distinguish the color of nasal discharge.

Correlation with pathogen detection was very poor. The authors found a Pearson correlation of 0.13 (95%CI 0.04-0.22). Reminder that a correlation of 1 is a perfect positive correlation.

Nasal discharge color is probably meaningless in determining whether there are bacteria.

2) Selection Bias: They screened 2,923 children and excluded 2,408 which is slightly over 80%. Most of these excluded did not meet diagnostic criteria for sinusitis (~46% of those excluded).

The proportion that declined to participate was also relatively high (19% of those excluded). Interestingly, this group that declined to participate had lower PRSS score (22 vs 24 in those who participated). They also excluded a few because they did not have enough symptoms based on the PRSS score. There were also 95 (4% of those excluded) for “Other reason” but we are never told what those reasons were.

3) Should We Perform Testing? A high percentage (70%) of the cohort included in this had some kind of identifiable pathogen detected (Strep pneumo, Haemophilus influenzae, Moraxella). These results are driven mainly by this population.

The authors did consider whether Moraxella may have been a colonizer rather than a true pathogen so ran separate analyses for just Strep pneumo and Haemophilus which you can see in Supplement 2. But there was a much bigger difference in PRSS score, -1.96 (CI -2.64 to -1.28), in when they only looked for the presence of those two organisms compared to all three (PRSS score -1.00 (CI -1.59 to -0.42).

Going back to the primary outcome, remember that whether children had a pathogen detected, both the antibiotic and placebo group did show improvement based on the PRSS score but this was a bigger difference in the group that had pathogens detected. This does make me go hmmm…

So, is it worth an uncomfortable test to start, delay, or stop antibiotics in some children?

4) Copper Standard: It’s time to talk about the diagnosis of sinusitis. The gold standard for diagnosing sinusitis is a bacterial culture from a sinus aspiration. Obviously, it does not make sense to do this for every single child that we might suspect sinusitis in. So, we’re left with the clinical practice guidelines that tell us to make a diagnosis based on persistent, worsening, or severe symptoms. We must acknowledge this is imperfect or a copper standard.

For example, many caretakers who have children in daycare or other childcare facilities will say that it seems like their toddler seems to get sick with upper respiratory symptoms (cough and rhinorrhea) over and over and over again to the point where it may feel like they are sick for months at a time. But does that mean they all have sinusitis? They would potentially meet the diagnostic criteria based on persistence of symptoms, but do they all really need antibiotics?

We think the authors did the best they could defining the study group that was included, but there is still the possibility that a proportion of the children included in this study were misdiagnosed. Especially when you consider that most were included (70%) for meeting criteria for persistent symptoms.

5) Clinical Significance vs Statistical Significance: We’ve talked about this issue before on the SGEM. Just because something is statistically significant, does not necessarily mean that it is clinically significant. In this study, we saw an overall improvement in the PRSS score between antibiotic group vs placebo of 1.69 in favor of the antibiotic group. But what does this mean?

The PRSS score evaluates for stuffy nose, runny nose, daytime cough, tiredness, irritability or fussiness, trouble breathing through the nose, cough in the night, and trouble sleeping at night. These factors are rated no, almost none, a little, some, a lot, and an extreme amount.

We would have trouble distinguishing between some of these like “almost none” or “a little” but that’s a one-point difference right there! Also, these criteria aren’t really weighted so where is the 1.69 points of improvement? We don’t know how this improvement was broken down by categories.

When it comes to possible harms, it’s important to point out that the side effect of diarrhea might be clinically important for families as it was worse in the antibiotic group with a number needed to harm of 16.

Comment on Authors’ Conclusion Compared to SGEM Conclusion: We agree with the authors’ conclusions.

SGEM Bottom Line: Antibiotics may offer some benefit for mild to moderate sinusitis in children but at the cost of a potential increase in diarrhea.

Case Resolution: You tell the family that their child likely has a viral infection or sinusitis and that the color of her nasal discharge does not necessarily tell you whether this is a bacterial infection. You engage in shared decision making and talk with the families about the potential harms and benefits of initiating antibiotic therapy. The family expresses that they would like a prescription for antibiotics, and you discharge them home.

Clinical Application: This is reassuring that most children with sinusitis will recover with or without antibiotics, even when the history has been prolonged and moderately severe. At present we generally avoid prescribing antibiotics for those with uncomplicated sinusitis in the first instance, and will continue to recommend this. For those with moderate to severe sinusitis, or with a more prolonged history, use this evidence to support shared decision making with parents and children and young people; in particular considering the modest benefit against the relatively common side effects.

What Do I Tell the Parent?  I’m sorry your child is still sick. It’s hard to know whether her symptoms are due to a virus or bacteria. Giving her fever medications and fluids can help. Another option is to give her some antibiotics. Antibiotics don’t work on viruses. Antibiotics may decrease how sick she is and how long she will be sick.  However, it can cause a runny bum (diarrhea) and there is a risk of allergic reaction. Which would you prefer, a running nose or a runny bum?

Remember to be skeptical of anything you learn, even if you heard it on the Skeptics Guide to Emergency Medicine.


  1. Wald ER, Applegate KE, Bordley C, et al. Clinical practice guideline for the diagnosis and management of acute bacterial sinusitis in children aged 1 to 18 years. Pediatrics. 2013;132(1):e262-280.
  2. Shaikh N, Wald ER, Jeong JH, et al. Development and modification of an outcome measure to follow symptoms of children with sinusitis. J Pediatr. 2019;207:103-108.e1.
  3. van den Broek MF, Gudden C, Kluijfhout WP, Stam-Slob MC, Aarts MC, Kaper NM, van der Heijden GJ. No evidence for distinguishing bacterial from viral acute rhinosinusitis using symptom duration and purulent rhinorrhea: a systematic review of the evidence base. Otolaryngol Head Neck Surg. 2014 Apr;150(4):533-7. doi: 10.1177/0194599814522595. Epub 2014 Feb 10. PMID: 24515968.