Wednesday, August 14, 2013

Clinical Practice Guideline: Sinusitis in Kids

Ellen R. Wald, MD, Laurie E. Scudder, DNP, NPAug 05, 2013Medscape: The guideline describes 3 clinical scenarios that should trigger a presumptive diagnosis of acute bacterial sinusitis. The first 2 -- persistent illness of 10 days without improvement or a child with a worsening course of symptoms after initial improvement -- are likely very familiar situations to pediatric primary care providers. However, the third scenario describes a child with a severe initial presentation that includes fever ≥39° C (102.2° F) and purulent nasal discharge for 3 days. Can you speak a little more to this situation? What are the key features that distinguish this situation from the much more common situation of a child with a viral illness with fever and respiratory symptoms?
Dr. Wald: It is the least common scenario and, I believe, the hardest to distinguish from an uncomplicated upper respiratory infection (URI). The key issue in differentiating an uncomplicated viral URI from an episode of acute bacterial sinusitis primarily rests with the duration and the quality of the different respiratory symptoms. In an uncomplicated cold, fever is not usually present at all. When it is present, most often it is observed at the very onset of the illness, often in concert with other constitutional symptoms such as headache and myalgia. Most commonly, the fever and other symptoms subside completely in 24-36 hours, and then the respiratory symptoms become more prominent.
Most times in an uncomplicated viral URI, the nasal discharge begins as clear and watery. After several days, it becomes thicker and more mucoid, and then, finally, it may become purulent (thick, colored, and opaque). Then the situation reverses with that purulent discharge becoming more mucoid again, and then either it becomes clear or it simply dries. That transition from clear to purulent and then back again occurs in uncomplicated colds without any antibiotic treatment at all. So what distinguishes this so-called severe presentation is that the fever comes on and stays longer than usual, and the purulent nasal discharge appears early in concert with the fever. It is the earlier appearance of the purulent discharge and the persistence of the fever and their concurrence for at least 3-4 days that should help the clinician distinguish this child from the child who has an uncomplicated cold.
Medscape: The guideline is very definitive in stating that children with a presumptive diagnosis of acute bacterial sinusitis should not undergo imaging studies. Equally, imaging should not be used to eliminate the diagnosis. Can you define the narrow circumstances under which imaging should be considered?
Dr. Wald: Let me just focus for a moment on why we try to de-emphasize the use of imaging as a diagnostic tool. It is important to recognize that, historically, images were used to confirm a diagnosis of acute bacterial sinusitis. There were particular findings on x-rays that made us think that we had confirmed the diagnosis of acute sinusitis. However, during the last 20 years, there has been increasing evidence that images of any kind, whether they're plain x-rays, CT, or MRI, obtained when children have an uncomplicated cold are also abnormal much of the time. And the abnormalities that you see in these children are exactly the same as those that in the past we attributed to acute sinusitis. That makes sense because the mucosa of the respiratory tree is a continuous membrane. So the membranes that line the nose are continuous with the membranes that line the eustachian tube, the paranasal sinuses, the throat, the airway, and the lower airway. Therefore, when there is an inflammation of the nasal membranes as part of a cold, that process often involves the sinus mucosa as well. Images cannot distinguish whether those mucosal changes found in the sinuses during an uncomplicated cold are due to viruses or bacteria or allergic inflammation.
So while it is true that images interpreted as normal obtained in a child with respiratory symptoms indicate that those symptoms are not attributable to sinusitis, it is not recommended that the images be obtained either for confirmation or denial of the diagnosis.
The only time that images are recommended is when complications of sinusitis are thought to be present. Those complications might involve either the orbit or the central nervous system. In those cases, the image will confirm the presence of the complication and will define its extent and, therefore, will be very helpful in determining the potential need for surgical intervention.
Medscape: The recommendations clearly state that children presenting with 2 of the described clinical scenarios -- a severe onset of symptoms or a clinical course that worsens after an initial improvement -- should receive antibiotic therapy. In a change from the 2001 guideline, the recommendations this time allow for continued observation for children with the third described clinical picture -- 10 days of symptoms without improvement. Can you speak to some of the evidence underlying these varying recommendations for management?
Dr. Wald: I think we should be clear on the benefits of antibiotics. Children with acute sinusitis who are treated with the antibiotics get better more quickly and more often than those who are treated with placebo. However, it is also true that some children with sinusitis get better spontaneously.
The most common clinical presentation for acute sinusitis, which we call onset with persistent symptoms, includes respiratory symptoms such as nasal discharge or cough or both, that are present and not improving at the 10-day mark. In contrast, most uncomplicated viral URIs last 5-7 days, or sometimes even fewer. Although the patients may not be free of symptoms on the tenth day, their symptoms have usually peaked in severity and are almost always improving. Those children are clearly on the mend. We therefore use 10 days as a guideline, and we strongly suspect that a patient is having an episode of sinusitis if their respiratory symptoms have not begun to improve by that 10-day mark. Now, could you wait for 11 or 12 days? Sure. There is probably no harm in waiting to see if the respiratory symptoms will turn around by themselves in the next day or 2. The parent and the clinician can make this decision together. However, my experience is that most parents will not want to wait any longer because they have been educated by the medical community over the last decade to realize that we usually wait for about 10 days before we endorse treatment for respiratory symptoms.
In the other 2 presentations, onset with worsening symptoms and onset with severe symptoms, there is a lot less likelihood that that episode of sinusitis will resolve spontaneously, and therefore treatment is recommended to begin immediately. 

Treating Sinusitis: Which Antibiotic Is Best?

Medscape: First-line therapy for these children continues to be amoxicillin with or without clavulanate. In contrast to 2012 recommendations from the Infectious Diseases Society of America (IDSA), which recommends amoxicillin-clavulanate as initial therapy for all patients with this condition, AAP suggests that some children may be treated with amoxicillin at either the standard dose (45 mg/kg/day in 2 divided doses) or a high dose (80-90 mg/kg/day in 2 divided doses). The guideline emphasizes the importance of clinicians being familiar with local resistance patterns as they make the decision regarding the most appropriate antimicrobial for a given child. With the fairly limited numbers of situations in which amoxicillin should be considered, combined with varying patterns of resistance, why the decision to recommend amoxicillin as first-line treatment?
Dr. Wald: To clarify, the key action statement in the AAP guideline says amoxicillin with or without clavulanic acid. Knowing your regional resistance patterns is really the key here because there is almost no information currently. While we say you should know it, the clinician has almost no way toknow it. In fact, there is an overall paucity of data, which makes this recommendation a difficult one.
To plan therapy for acute sinusitis, clearly you need to know the bacteriology. And that has become very difficult in the last few years. The last time that a study of the pathogenesis of acute sinusitis based on results from sinus aspiration was conducted was 30 years ago in the mid-1980s. There have been no studies of sinus aspirates in the United States since then. What we have leaned on instead, because the pathogenesis of acute otitis media is very similar to that of acute sinusitis, is information derived from studies of tympanocentesis performed in a child who has acute otitis media.
But even in this situation, most centers that were doing tympanocentesis in the '90s and in the early 2000s are no longer doing them, and so there is a near absence of data.
We know that the common agents that cause acute sinusitis are Streptococcus pneumoniae,Haemophilus influenzae, and Moraxella catarrhalis. As most of your readers know, some S pneumoniae may be resistant to amoxicillin as a result of an alteration of penicillin-binding protein. That particular kind of resistance can be overcome in large part by increasing the dose of amoxicillin from 45 mg/kg to 80-90 mg/kg.
That was the origin of high-dose amoxicillin or high-dose amoxicillin/potassium clavulanate recommendations intended to cover the possibility of resistant S pneumoniae.
Some H influenzae and all M catarrhalis are also resistant to amoxicillin, though that mechanism is the result of beta lactamase production. Beta lactamase resistance can be overcome by using a drug that is inherently resistant to the actions of the beta lactamase, and that is the case with some second- and third-generation cephalosporins. The alternative way to get around the issue of resistance is to combine amoxicillin with an irreversible beta lactamase inhibitor, like potassium clavulanate.
Now the desire to continue to use amoxicillin as first-line therapy, whether it is for acute otitis media or acute sinusitis, relates to the fact that amoxicillin is a safe drug. It has a narrow spectrum, is inexpensive, and it works a lot of the time. If you think that the most likely cause of a child's acute sinusitis is S pneumoniae and that H influenzae is less likely, then recognizing that most S pneumoniaestrains are susceptible to penicillin, you can use amoxicillin. If you are concerned about resistant S pneumoniae, you would use high-dose rather than regular-dose amoxicillin.
We believe that because of the effect of the new pneumococcal conjugate vaccine (PCV13), there are probably fewer cases of sinusitis caused by S pneumoniae and more caused by H influenzae. On the basis of, unfortunately, very few data, we are worried that the prevalence of beta lactamase-producingH influenzae is actually increasing. If this is the situation -- and I think it is -- then amoxicillin/potassium clavulanate is really a more comprehensive drug.
Concerns about use of amoxicillin/potassium clavulanate when amoxicillin would suffice relate primarily to cost and adverse reactions. But the differences are really tiny. Amoxicillin/potassium clavulanate is a little more expensive than amoxicillin, but not by much because both are available as generics. It is also a little bit more likely to cause diarrhea, though that is less of a concern with the new formulation, and the increase in incidence when compared with amoxicillin is not enormous.
So if a clinician does not have any information about the relative prevalence of S pneumoniaeH influenzae, and beta lactamase-producing strains, amoxicillin/potassium clavulanate is the most comprehensive option. It is probably safe to use it in regular doses rather than high doses.
Again, I think we have reason to think that the likelihood of resistant S pneumoniae currently is extremely low. Could this change in 2 or 3 years? Absolutely. But right now, it is likely that the prevalence of S pneumoniae is low, the prevalence of H influenzae is high, and the prevalence of beta lactamase-producing H influenzae may also be high.
I sat on the IDSA committee when we formulated those guidelines 2 or 3 years ago. That was just when PCV13 was released. At that time, there was concern about resistant H influenzae and resistant S pneumoniae. We made recommendations on that basis. Now, 2-3 years later, we have a little bit more information.
The only person still performing tympanocentesis in the United States for the purposes of determining etiologic agents is Dr. Michael Pichichero in Rochester, New York. His most recent data, published in 2013, documented the prevalence of S pneumoniae in middle ear disease -- which I am using as a surrogate for sinus infections -- to be ~15%, a low number.[2] The prevalence of H influenzae was high, and at least 50% of those strains of H influenzae were beta lactamase producing.
If that is reflective of what is going on across the United States, then amoxicillin/potassium clavulanate is the best choice. It is really a tough question.
Medscape: The guideline notes the paucity of data regarding duration of therapy. Most clinicians prescribe for 10 days and sometimes 14 days. Is that a reasonable practice for most situations?
Dr. Wald: As you know, there has just been no study of this question. We are currently engaged in a study comparing 5 days to 15 days of amoxicillin/potassium clavulanate for acute sinusitis. There are no data in children on short courses of therapy. So until those data are available, the recommendation is a 10-day course. That practice is based on the fact that most kids start to get better after 72 hours, and then if you treat them another week you probably have given them a good course of therapy. It makes the likelihood of relapse pretty low. The other option that I have sort of popularized is the notion of treating a patient until they are free of symptoms and then for an additional 7 days. That strategy protects children who respond more sluggishly to antibiotics. Although they are somewhat better at 72 hours, their symptoms do not completely resolve until they have received 7 or 8 days of treatment. In those children, I've always been concerned that 10 days is just too short of an interval after they became symptom-free. I like to give them another few days of treatment in the hopes that it will fully eradicate the infection, and there will not be a recurrence of symptoms.
Medscape: The guideline also thoroughly discusses the issue of when to change antibiotics. This seems more straightforward in children with any worsening of symptoms. However, the guidance regarding children who fail to improve after 72 hours of antibiotics is more nuanced. Can you discuss some of the factors affecting this decision?
Dr. Wald: The guideline made this distinction between getting worse, which is an easy decision (antibiotics must be changed), vs not improving (the child hasn't gotten worse but also hasn't gotten better). In that situation, the guideline gives the practitioner and the parent the option of either giving the antibiotic another day or 2 to start working or changing the antibiotic. My personal belief is that if a patient has not improved in 3 days, the antibiotic should be changed.
The expected course for children being treated for a presumed bacterial infection is that they will get better in that timeframe. Kids don't get better instantaneously; it takes at least 24-48 hours to improve. But if a child is not improving by 72 hours and the clinician is still confident of the diagnosis, then that says to me that I have not selected the right drug.
What are the options? You are right -- it becomes a lot more complicated. The guideline did not emphasize this point, but I would feel confident in emphasizing it now. As an alternative to amoxicillin/clavulanic acid, the guideline suggests use of a combination of clindamycin and cefixime. The rationale there was that clindamycin might cover a resistant S pneumoniae, and cefixime would cover a beta lactamase-producing H influenzae.
Clearly, though, 2 drugs are a clumsy choice. The other option is linezolid plus cefixime. Linezolid will cover all penicillin- or amoxicillin-resistant S pneumoniae, so it was offered as a rescue drug if a child was not responding to treatment with amoxicillin/potassium clavulanate because of an infection caused by resistant S pneumoniae. Once again, the cefixime is in that combination to cover for beta lactamase-producing H influenzae.
The guideline also suggests levofloxacin, which is a quinolone with high activity against both S pneumoniae and H influenzae. We know that we haven't been enthusiastic about using quinolones in kids, though not because we are really worried about the arthropathy. We are more worried about the fact that the organisms will become resistant to the quinolones if we use them broadly. Respiratory infections are common, and so the temptation to use quinolones broadly should be resisted. What the guideline did not emphasize, but which I would feel comfortable recommending today, are second- or third-generation cephalosporins. Cefuroxime, cefdinir, and cefpodoxime are 3 cephalosporins that could be used in a patient who failed to respond to amoxicillin/potassium clavulanate.
We did not emphasize cephalosporins as an option when this guideline was drafted because, at that time, there was still a lot of concern about a high prevalence of S pneumoniae, particularly penicillin-resistant S pneumoniae. In data from 2007 and 2008, cefuroxime, cefdinir, and cefpodoxime only covered 60%-70% of resistant S pneumoniae. So we did not want to offer them as a failsafe or rescue drug for the child who had failed amoxicillin/potassium clavulanate.
However, now I believe that the prevalence of S pneumoniae, including resistant strains, is low. Accordingly, we can use cefuroxime or cefdinir or cefpodoxime in a child who has not responded to amoxicillin/potassium clavulanate.

Adjuvant Treatment

Medscape: There were several adjuvant therapies noted in the guideline, though none are recommended due to a lack of evidence. What do you suggest as comfort measures for children either not receiving antibiotics or for those in the initial couple of days of treatment who have not yet responded?
Dr. Wald: There are 2 comfort measures that I would endorse. No surprise, the first is ibuprofen or acetaminophen for discomfort or fever. I would not hesitate to use these agents in appropriate doses for a couple of days at the beginning of treatment before the antibiotics have had a chance to work.
The second is the use of saline nasal spray. As you mentioned, we really can't prove that this works, but if the child is comfortable using the nasal spray -- and definitely some kids are not -- nasal spray helps to loosen the mucous crust that may have dried in the nose and usually causes discomfort and adds to the child's sense of nasal obstruction and congestion. The advantage, of course, of a saline spray is that it can be used as many times as the child wants. And it is very helpful in keeping the nose clean, which theoretically would facilitate the drainage of secretions from the sinus cavity into the nose.
Humidifiers and vaporizers probably have the same impact as nasal spray. If you can keep the room humidified, then the material in the nose stays a little bit more liquid, and then it's probably easier to sneeze out or, in kids who know to blow their nose, to blow out.
Medscape: Are there other key recommendations from this guideline you would like to highlight for clinicians?
Dr. Wald: Whether antibiotics are initiated or not, there does need to be a plan for follow-up. Parents need to be carefully instructed that they should call any time if their child worsens. That is just always good advice to give to parents. Sometimes they feel so reassured by having seen the doctor that even though the child gets worse, they may hesitate to give you a call.
And there should be a check-in at 72 hours just to make sure that the child is improving. If they are not, then antibiotics really should be started if they were not started at the outset, or they should be changed.
The only other point I would make is that most children with a runny nose or cough or both don't need antibiotics. Only about 6% or 7% of all viral URIs are complicated by sinusitis.
I am hopeful that these updated guidelines will help pediatricians and other primary care providers who care for children to accurately diagnose children with a common cold (which antibiotics are surely not going to help) and those with a bacterial infection that will get better faster with antibiotics.

No comments: