OBJECTIVE: The role of antibiotic therapy in managing acute bacterial sinusitis (ABS) in children is controversial. The purpose of this study was to determine the effectiveness of high-dose amoxicillin/potassium clavulanate in the treatment of children diagnosed with ABS.
METHODS: This was a randomized, double-blind, placebo-controlled study. Children 1 to 10 years of age with a clinical presentation compatible with ABS were eligible for participation. Patients were stratified according to age (<6 or ≥6 years) and clinical severity and randomly assigned to receive either amoxicillin (90 mg/kg) with potassium clavulanate (6.4 mg/kg) or placebo. A symptom survey was performed on days 0, 1, 2, 3, 5, 7, 10, 20, and 30. Patients were examined on day 14. Children's conditions were rated as cured, improved, or failed according to scoring rules.
RESULTS: Two thousand one hundred thirty-five children with respiratory complaints were screened for enrollment; 139 (6.5%) had ABS. Fifty-eight patients were enrolled, and 56 were randomly assigned. The mean age was 66 ± 30 months. Fifty (89%) patients presented with persistent symptoms, and 6 (11%) presented with nonpersistent symptoms. In 24 (43%) children, the illness was classified as mild, whereas in the remaining 32 (57%) children it was severe. Of the 28 children who received the antibiotic, 14 (50%) were cured, 4 (14%) were improved, 4 (14%) experienced treatment failure, and 6 (21%) withdrew. Of the 28 children who received placebo, 4 (14%) were cured, 5 (18%) improved, and 19 (68%) experienced treatment failure. Children receiving the antibiotic were more likely to be cured (50% vs 14%) and less likely to have treatment failure (14% vs 68%) than children receiving the placebo.
CONCLUSIONS: ABS is a common complication of viral upper respiratory infections. Amoxicillin/potassium clavulanate results in significantly more cures and fewer failures than placebo, according to parental report of time to resolution of clinical symptoms.
Acute bacterial sinusitis (ABS) is a common complication of upper respiratory infections in children. However, controversy remains regarding how best to manage this condition in children. There are only 2 randomized, placebo-controlled studies that have examined the efficacy of antibiotic therapy in children diagnosed with ABS.1,2 In 1986, Wald et al1 evaluated children between 2 and 16 years of age with persistent, nonimproving upper respiratory symptoms and abnormal radiographs of the paranasal sinuses. Subjects were randomly assigned to receive amoxicillin, amoxicillin/potassium clavulanate, or placebo. Children who received antibiotic therapy were cured more quickly and more often than those who received placebo.
Garbutt et al,2 using a study design similar to that used by Wald et al1 and comparing amoxicillin to placebo using clinical criteria without the performance of confirmatory images, failed to show a benefit of antimicrobial therapy in children diagnosed with ABS. Several methodologic issues, including the age of participants, the relatively low dose of antibiotic, and the stringency of entry criteria and evaluation may have contributed to different outcomes in these studies.3
The purpose of this study was to determine the effectiveness of amoxicillin/potassium clavulanate in the treatment of children diagnosed with ABS on the basis of stringent clinical criteria without performance of confirmatory radiographs.
Patients were recruited from January 2004 through June 2006 at 2 private pediatric practices (1 rural and 1 urban) and a hospital-based clinic in the Pittsburgh area.
Children between 1 and 10 years of age with 1 of 3 clinical presentations compatible with the diagnosis of ABS were eligible: children with persistent symptoms, children with acutely worsening symptoms, and children with severe symptoms. Persistent symptoms were defined as nasal discharge (of any quality) or daytime cough (which may be worse at night), or both, persisting for >10 days without evidence of improvement. Worsening symptoms were defined as nasal discharge or daytime cough that was worsening on or after the sixth day of symptoms, manifested by new-onset fever (temperature ≥100.5°F) or substantial increase in nasal discharge or cough after having experienced transient improvement of symptoms, and severe symptoms were defined as a temperature of at least 102°F and purulent (thick, colored, and opaque) nasal discharge present concurrently for least 3 consecutive days. Children with “persistent” or “severe” symptoms and significantly abnormal radiographs have been shown to have positive bacterial isolates from a maxillary sinus aspirate in 75% of cases.4 Onset with “worsening” symptoms is a classic presentation for a secondary bacterial infection of a viral upper respiratory infection (URI).
Recruitment and Randomization
A convenience sample of patients presenting with the chief complaint of respiratory symptoms was queried, if the research nurse was available, to determine the presence and characteristics of upper respiratory symptoms (nasal congestion, nasal drainage, cough, and fever). Eligibility was based on both the duration and status (improving, persisting, or worsening) of symptoms. If the pediatrician confirmed the diagnosis of probable ABS, the child and parent were asked to join the study; informed consent from the parents and assent from patients >7 years of age was obtained. This study was approved by the institutional review board of the University of Pittsburgh. Patients were assigned to a study group (persistent, nonpersistent), stratified according to age (<6, ≥6 years) and clinical severity (Table 1), and randomly assigned to receive either amoxicillin (90 mg/kg)/potassium clavulanate (6.4 mg/kg) in 2 divided doses or placebo on a 1:1 basis in blocks of 4. The allocation sequence was generated by the principal investigator. Participants, research nurses, and physicians were blinded to group assignment. The maximum dose of the study drug was 4.0 g twice daily. Administration of medications other than acetaminophen or ibuprofen for symptomatic relief was discouraged.
Patients were assigned a clinical severity score on the basis of their presenting signs and symptoms (Table 1).1 Scores <8 were defined as mild and scores ≥8 were defined as severe. This scoring system, which correlates highly with parental report of degree of illness, has been used since 1986.1
Children were excluded from the study if they (1) received antibiotics within 15 days of the onset of respiratory symptoms, (2) had symptoms for >30 days, (3) had another concurrent bacterial infection (eg, acute otitis media [AOM], pneumonia, or urinary tract infection), (4) were allergic to penicillin, (5) had symptoms that suggested a complication because of ABS necessitating hospitalization, intravenous antibiotics, or subspecialty evaluation, or (6) had been diagnosed with either immunodeficiency or anatomic abnormality of the upper respiratory tract.
A symptom score was obtained on day 0 at the time of enrollment and again on days 1, 2, 3, 5, 7, 10, 20, and 30 by telephone survey. The survey questions inquired about the presence of fever, nasal discharge, nasal congestion, cough, headache or irritability, facial pain, facial swelling, activity, sleep, and impaired appetite. If a particular symptom was present initially, a score of 2 was assigned; if it was absent, the score was 0. A maximum entry score was 20. If a particular symptom became more severe, less severe, or stayed the same during treatment, +1, −1, or 0, respectively, was added to the original score for each symptom. Patients were considered cured if their total symptom score was <2. Patients were considered to be improved if their symptom score decreased by at least 50%. Patients were considered to have experienced treatment failure if (1) they worsened at any time (symptom score increased by 4), (2) their scores were not reduced at 48 hours (entry score minus at least 2 points), (3) they were not improved by 72 hours (score ≤50% of the score at entry), or (4) their score was >5 at 14 days. These definitions, although arbitrary, were modeled on common clinical experiences and decision-making.
The survey also included 4 questions about common antibiotic adverse effects: rash, diarrhea, vomiting, and abdominal pain. Diarrhea was defined as a stool frequency of at least 2 per day above baseline of loose or watery stools. If any adverse effect was considered severe enough to discontinue the study drug, the patient was examined by the study nurse and local physician. All patients were followed for a total of 30 days. Each family received a calendar to assist with documentation and recall related to antibiotic adverse effects. The patient returned on day 14 (day 14–18 was acceptable) for an assessment. The development of AOM was considered a “failure.” The remaining volume of medication was measured as a check on compliance. Consumption of at least 80% of the medication was considered to be “compliant.”
Any patient who was deemed to have experienced treatment failure was seen by their pediatrician and the research nurse. The pediatrician determined if a new diagnosis was present (eg, AOM, pneumonia) or if the patient was still symptomatic because of ABS. When a child was classified as having treatment failure, the study drug was discontinued and cefpodoxime (10 mg/kg once daily) was prescribed for 10 days. This ensured that patients originally receiving the placebo were now receiving an antimicrobial and that children not responding to amoxicillin/clavulanate were now receiving an alternative antimicrobial. The principal investigator was available to discuss medication management at all times.
If patients developed respiratory symptoms after completing the study drug, they were classified as having either an uncomplicated URI or sinusitis. The major outcome measure for the study was the proportion of children in each group who were “cured” on day 14 of therapy. Secondary outcomes were (1) the proportion of children who experienced treatment failure and (2) the proportion of children who experienced adverse events. The number of children who experienced a recurrence of respiratory symptoms and the time to clinical improvement were calculated.
The sample size calculation for this trial (70 subjects in each group) was estimated as the number of patients per arm required to detect a 20% difference in the proportion of children who were cured at the 14 day visit, with at least 80% power at the 1-sided 5% significance level. Efficacy data were summarized for both the intent-to-treat and the per-protocol populations. The intent-to-treat population was defined as the population of all patients who were randomly assigned. The per-protocol population excluded withdrawn and noncompliant patients. Categorical data were summarized as proportions and percentages. Continuous data were summarized and reported as means, SDs, and ranges. The primary statistical analysis compared the proportions of patients who were cured on day 14 of therapy between the 2 study arms on an intent-to-treat basis. Primary and secondary end points were compared between study arms by using a χ2/Fisher's exact test (for proportions); these statistical tests were 2-sided, and a P value of < .05 was used to indicate statistical significance.
Two thousand one hundred thirty-five children with respiratory complaints were screened for enrollment (Fig 1). Of these, 1996 children were not eligible either because they did not meet criteria (1982 patients) or had AOM (14 patients). Of the 139 eligible patients, 77 declined, 3 patients' parents were unavailable for informed consent, and 1 family did not have a telephone. Fifty-eight patients were enrolled (signed informed consent). In 2 cases, consent was withdrawn before the drug was distributed, leaving 56 patients to be stratified according to age (<6 or ≥6 years) and severity (score of <8 or ≥8), and randomly assigned to receive the antimicrobial medication or placebo. Table 2 shows the comparability of groups receiving the antimicrobial medication and placebo.
Of the 28 children who received the antibiotic, 14 (50%) were cured, 4 (14%) were improved, 4 (14%) experienced treatment failure, and 6 (21%) withdrew from the study. In contrast, of the 28 children who received the placebo, 4 (14%) were cured, 5 (18%) improved, and 19 (68%) experienced treatment failure. Children receiving the antibiotic were more likely to be cured (50% vs 14%; P = .01) and less likely to experience treatment failure (14% vs 68%; P < .001) than children receiving the placebo (Table 3).
All 6 children who were withdrawn or lost to follow-up were assigned to receive the antibiotic (Table 4). Inclusion of these 6 children in the analysis of outcome is problematic. If all 6 children are considered to have treatment failure, the P value for the difference in proportion of children who experienced treatment failure for subjects receiving the antibiotic versus the placebo equals .032. If 4 children are considered to have treatment failure (excluding children A and B), the P value for the difference in proportion of children who experienced treatment failure for subjects receiving the antibiotic versus the placebo equals .007.
The overall rate of compliance (consuming ≥80% of the medication) was 85.7%, with 22 children receiving antibiotics and 26 subjects receiving the placebo. In the per protocol analysis (N = 48), 16 (80%) of the children receiving a full course of antibiotics were cured or improved compared with 8 (31%) of those receiving a full course of placebo (P = .001). The average time to improvement in symptoms among compliant children was similar in both groups (Table 3).
Nineteen (83%) of 23 children who experienced treatment failure received the placebo. The majority (17 of 19 [89.5%]) failed to respond within 72 hours of treatment. Treatment failed for 2 (11%) children on day 5. Four children who received the antibiotic were deemed to have clinically failed therapy. The response to treatment was sluggish in 3 children and was therefore interpreted as “failing” by our a priori criteria. The fourth patient had a prompt clinical response but developed recurrent symptoms on day 11 of treatment. Distinguishing between a new viral URI and recurrence of sinusitis was difficult; she met criteria for failure by having a score of >5 on day 14. Rescue therapy with cefpodoxime was successful in all children who experienced treatment failure.
A recurrence of respiratory symptoms after day 14 but before day 30 was observed in 4 of 18 (22%) children who experienced a clinical cure and 3 children who experienced treatment failure. These events were interpreted as new viral URIs, because they resolved spontaneously in <10 days. Of the 7 children who had a new URI, 4 had received the antibiotic and 3 had received the placebo. Only 1 child was treated with antibiotics for AOM. A recurrence of respiratory symptoms was observed in 4 of 9 children who experienced improvement.
Adverse events were reported in 12 (44%) children receiving antibiotics and in 4 (14%) children receiving the placebo (P = .014). The most common adverse event was diarrhea, which, in general, was self-limited.
ABS is an important cause of frequent morbidity and negatively impacts the quality of life of patients and families. The results of this study show that ABS is a relatively common complication of acute URIs in children (6.5%) and that among children fulfilling stringent criteria for a clinical diagnosis of ABS, even without the performance of confirmatory radiographs, those receiving antibiotic therapy experience a clinical cure more quickly and more often than children receiving placebo.
The diagnosis of sinusitis in children depends, to a great extent, on patient or parent report, as physical examination of the nasal cavity is not helpful in distinguishing between uncomplicated viral URI and ABS. The best measure or gold standard for the diagnosis of ABS is the performance of a sinus aspirate that yields bacteria in high density (≥104 colony forming units/mL).4 However, sinus aspiration is an invasive, time-consuming, and potentially painful procedure that should only be performed by a specialist.3 The recommendation to use clinical criteria to diagnose ABS is based on the impracticality of performing sinus aspirates or obtaining images of the paranasal sinuses in the office setting.3 Likewise, resolution of sinusitis must also depend on parental or patient report. The Food and Drug Administration recently promoted the use of patient-reported outcome measures in assessing medical devices and treatments.5
This is the third prospective, placebo-controlled clinical trial assessing the role of antibiotics in children with presumed ABS in the last 22 years in the United States.1,2 The differences between the results of this study and those reported by Garbutt et al2 require explanation. Our patients may have been more likely to have acute sinusitis rather than viral URI. Stringent entry requirements are critical to be sure that the number of children with uncomplicated viral URIs (who would not be expected to respond to antibiotics) who are admitted to the study is minimized. In the current study, 2135 children with respiratory complaints presenting to their primary care provider were screened. Only 139 (6.5%) children were eligible. This rate of ABS among children presenting with respiratory complaints is similar to that reported by others.6–8 In contrast, in the Garbutt et al2 study, the pool of patients from which study children were drawn was not described, and children deemed severely ill and most likely to benefit from antimicrobial therapy were excluded. In the present study and the one reported in 1986,1 strict a priori criteria were defined by which a child was judged to have experienced a clinical failure.3 Children who did not improve within 48 hours (at least 2 points below the entry score) or were not at half of their entry score by 72 hours were defined as clinical failures, and most of the children who experienced treatment failure (89%) did so by not improving within the first 72 hours. Although some children will eventually improve without therapy, the presumed benefit of and reason for treatment is a more rapid clinical cure. In contrast to the current study design, the Garbutt et al2 study evaluated the “natural history” of children with respiratory symptoms assigned to a treatment regimen. Adjustments to treatment were made by the primary care provider.
Concerns about the adverse effects of inappropriate use of antibiotics have led to the development of guidelines designed to improve both the diagnosis and treatment of patients with ABS. Three national organizations (Centers for Disease Control and Prevention, Sinus and Allergy Health Partnership [comprising the American Academy of Otolarygic Allergy, American Academy of Otolaryngology-Head and Neck Surgery, and the American Rhinologic Society], and the American Academy of Pediatrics) recently published 3 different guidelines.3,9,10 Each guideline recommends that the diagnosis and treatment of ABS should be based on clinical criteria (at least in some age groups) without the confirmation of imaging or other laboratory data. Unquestionably, the performance of radiographs or other images to confirm the diagnosis of ABS increases both the cost and complexity of clinical care. The similarity between the different guidelines suggests widespread consensus among “experts” to use clinical criteria to diagnosis ABS. This study provides evidence to support the use of stringent clinical criteria without radiographic imaging to diagnose ABS presenting with persistent symptoms, which is the most common presentation of ABS in children.
The benefit of antibiotic therapy must be weighed against risks. Although the rate of adverse events in children receiving amoxicillin/potassium clavulanate was substantial, only 3 children discontinued participation in the study as a result. Most episodes of diarrhea were mild and self-limited.
Recommendations for treatment of ABS include high-dose amoxicillin or high-dose amoxicillin/potassium clavulanate if risk factors exist for the likelihood of infection with antimicrobial-resistant organisms.3 The microbiology of ABS in children was established in the United States in 1981 and, unfortunately, more recent data on the microbiology of ABS in children are not available. However, the pathogenesis and microbiology of AOM and ABS are similar, and it is reasonable to regard recent data generated from cultures of middle ear fluid (obtained by tympanocentesis) from children with AOM as a surrogate for cultures of the paranasal sinuses.11 Several recent reports have highlighted a slight decrease in isolates of Streptococcus pneumoniae and an increase in isolates of Haemophilus influenzae.12,13 We chose amoxicillin/potassium clavulanate as the study drug to be most comprehensive in the ability to treat potential sinus pathogens (especially in view of increasing H influenzae isolates) and, therefore, ensure effectiveness in this proof-of-principle investigation.
The limitation of this study relates to sample size. A larger-than-expected difference in outcome permitted a demonstration of significant differences despite our difficulties with accrual. The major issue was parental refusal to participate, based on the long-time clinical practice in this community, which promised antimicrobial therapy for children whose respiratory symptoms had lasted >10 days. Nonetheless, the results of the Garbutt et al2 study provided sufficient clinical equipoise to permit the investigators to enroll patients in this study. Close follow-up and the provision of rescue therapy in the event of clinical failure were essential aspects of the study design and were stressed at the time of enrollment of subjects. This investigation, built on stringent entry criteria and stringent a priori definitions of clinical failure, clearly shows that sinusitis frequently complicates viral URIs in children, that sinusitis can be diagnosed clinically without confirmatory radiographs, and that antimicrobial therapy of children with clinically diagnosed ABS leads to significantly more cures and fewer failures than treatment with placebo. This message is easily applicable in clinical practice.
This study was funded by the Thrasher Research Fund.
- Accepted December 29, 2008.
- Address correspondence to Ellen R. Wald, MD, University of Wisconsin School of Medicine and Public Health, Department of Pediatrics, 600 Highland Ave, Box 4108, Madison, WI 53792. E-mail:
Financial Disclosure: The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject:
ABS is a common complication of viral URIs in children. Previous randomized, placebo-controlled studies that examined the efficacy of antibiotic therapy in children with ABS have produced discrepant results.
What This Study Adds:
ABS can be diagnosed by using stringent clinical criteria without the performance of imaging in children with persistent symptoms. Treatment with an appropriate antimicrobial agent results in significantly more cures and fewer failures than placebo.
- ↵Wald ER, Chiponis D, Ledesma-Medina J. Comparative effectiveness of amoxicillin and amoxicillin-clavulanate potassium in acute paranasal sinus infections in children: a double-blind placebo-controlled trial. Pediatrics.1986;77 (6):795– 780
- ↵Garbutt JM, Goldstein M, Gellman E, Whannon W, Littenberg B. A randomized, placebo-controlled trial of antimicrobial treatment for children with clinically diagnosed acute sinusitis. Pediatrics.2001;107 (4):619– 625
- ↵American Academy of Pediatrics. Subcommittee on Management of Sinusitis and Committee on Quality Improvement. Clinical practice guideline: management of sinusitis [published corrections appear in Pediatrics. 2001;108(5):A24; Pediatrics. 2002;109(5):40]. Pediatrics.2001;108 (3):798– 808
- ↵US Department of Health and Human Services FDA Center for Drug Evaluation and Research, US Department of Health and Human Services FDA Center for Biologics Evaluation and Research, US Department of Health and Human Services FDA Center for Devices and Radiological Health. Guidance for industry: patient-reported outcome measures: use in medical product development to support labeling claims: draft guidance. Health Qual Life Outcomes.2006;4 :79–98
- Copyright © 2009 by the American Academy of Pediatrics