Comparative Effectiveness of Amoxicillin and Amoxicillin-Clavulanate Potassium in Acute Paranasal Sinus Infections in Children: A Double-Blind, Placebo-Controlled Trial

(1)

VOLUME 77. JUNE 1986 #{149}NUMBER 6

Comparative

Effectiveness

of Amoxicillin

and

Amoxicillin-Clavulanate

Potassium

in Acute

Paranasal

Sinus

Infections

in Children:

A

Double-Blind,

Placebo-Controlled

Trial

Ellen

R. Wald,

MD, Darleen

Chiponis,

CPNP,

MEd,

and

Jocyline

Ledesma-Medina,

MD

From the Departments of Pediatrics (Divisions of Infectious Disease and Ambulatory Care) and Radiology, University of Pittsburgh School of Medicine, Children’s Hospital of Pittsburgh, Pittsburgh

ABSTRACT. This study compared the relative effective-ness of two antimicrobial preparations, amoxicillin and

amoxicillin-clavulanate potassium (Augmentin), in the

treatment of acute maxillary sinusitis in children 2 to 16 years of age. Of 171 children with persistent (ten to 30 days’ duration) nasal discharge or daytime cough or both, 136 (80%) had abnormal maxillary sinus radiographs. These children were stratified by age and severity of symptoms and randomly assigned to receive either amox-icillin, amoxicillin-clavulanate potassium, or placebo. After the exclusion of 28 children with throat cultures positive for group A Streptococcus and 15 who did not complete their medication, the remaining 93 children were evaluated: 30 received amoxicillin, 28 received

amoxicillin-clavulanate potassium, and 35 received

pla-cebo. Clinical assessment was performed at three and ten days. On each occasion, children treated with an anti-biotic were more likely to be cured than children receiving placebo (P < .01 at three days, P < .05 at ten days). The overall cure rate was 67% for amoxicillin, 64% for amox-icillin-clavulanate potassium, and 43% for placebo. Pe-diatrics 1986;77:795-800; acute sinusitis, respiratory in-fection, antimicrobial treatment, amoxicillin, amoxicillin-clavukinate potassium.

Infections of the upper respiratory tract

consti-Received for publication July 29, 1985; accepted Aug 29, 1985. Reprint requests to (E.R.W.) Children’s Hospital of Pittsburgh,

125 DeSoto St, Pittsburgh, PA 15213.

tute the principal reason for seeking care in pedi-atric emergency facilities and are second only to well-baby visits in the offices of private pediatri-cians.’2 Few well-controlled, double-blind pro-spective studies have evaluated the effectiveness of any form of drug therapy in the management of upper respiratory tract infections in children. In studies that have evaluated antimicrobial therapy, illnesses have not been stratified according to Se-verity or persistence of symptoms.35

The natural history of acute maxillary sinus in-fections in children has not been investigated. In one study of children with clinical and radiographic evidence of maxillary sinusitis who underwent di-agnostic sinus puncture, bacteria in high colony count were recovered from 75% of the sinus aspi-rates.6 Subsequent treatment with amoxicillin or cefaclor was associated with a clinical cure rate of 80%. Nonetheless, it is possible that these infec-tions would have been brief and self-limited even without specific antimicrobial therapy.

(2)

Amoxicillin-clavulanate potassium potentially of-fers an expanded antibacterial spectrum, including 3-lactamase-producing bacterial species that have been recovered from children with acute sinusitis.6

SUBJECTS

AND

METHODS

Subject

Selection

and Initial

Clinical

Assessment

Subjects were selected as identified and available from patients receiving primary or secondary health services at the Children’s Hospital of Pittsburgh. Children were eligible for study if they had nasal discharge of any quality (ie, watery, mucoid, or purulent) or cough (not exclusively nocturnal) or both that parents estimated had been present for at least ten but no longer than 30 days and that did not seem to be improving.

Exclusion

Criteria

Patients were excluded if they had a history of penicillin allergy, had received antimicrobial ther-apy within the previous three days or were known to have asthma, allergic rhinitis, cystic fibrosis, sickle cell anemia, congenital heart disease, or an immunodeficiency syndrome. Patients also were excluded if they had otitis media or pneumonia or if they had severe headache or periorbital swell-ing-conditions that appeared to mandate antimi-crobial treatment. Similarly, patients whose throat or nasopharyngeal cultures were positive for group

A /3-hemolytic streptococci were also excluded,

usu-ally 24 h after entry, when the culture results were reported.

Informed consent for entry into the study was obtained from the parent of each subject. A single examiner performed a complete physical examina-tion on each child, including transillumination of the maxillary sinuses. Transillumination was per-formed in a darkened room with the examiner’s eyes light adjusted. A Welch-Allyn transilluminator was positioned at the midpoint of the inferior or-bital rim. Transmission of light through the hard palate, as viewed on inspection of the oral cavity, was judged to be either normal or absent.

Patients were assigned a clinical severity score on the basis of their presenting signs and symp-toms, including nasal or postnasal discharge, nasal congestion, cough, malodorous breath, periorbital swelling, facial tenderness, erythematous nasal mu-cosa, fever, and headache (Table 1). A score of less than 8 was defined as mild and those higher as severe.

Radiographs of the paranasal sinuses were ob-tamed for each subject, including anteroposterior and erect lateral views for all age groups. For

chil-TABLE 1. Clinical Severity Score*

Symptom or Sign Points

Abnormal nasal or postnasal

discharge

Minimal 1

Severe 2

Nasal congestion 1

Cough 2

Malodorous breath 1

Facial tenderness 3

Erythematous nasal mucosa 1 Fevert

<38.5#{176}C 1

38.5#{176}C 2

Headache (retro-orbital)

Severe 3

Mild 1

* Score: <8 = mild; 8 = severe.

t

Within 24 h of presentation, either observed or by

history and documented with thermometer.

dren less than 6 years of age, a mentooccipital view was performed, and for those older than 6, an occipitomental view was performed for evaluation of the maxillary sinuses. These films were inter-preted by a radiologist, who was not aware of the clinical symptoms. The maxillary sinus radiographs were considered to be abnormal if they showed partial or complete opacification, an air-fluid level, or mucosal swelling of at least 4 mm. Children with normal maxillary sinus radiographs were excluded from the present trial.

Surface

Cultures

Before initiation of treatment, a throat and/or nasopharyngeal culture was obtained from each subject to identify those children whose symptoms might represent group A streptococcal infection. Specimens were inoculated onto 5% sheep blood agar plates. A differential bacitracin disc (0.04 unit) was placed on the primary plate. Plates were incu-bated at 37#{176}Cin 5% CO2 for 18 to 24 hours and then inspected. Cultures were considered positive if they showed any f3-hemolytic colonies with strep-tococcal morphology that were inhibited by the bacitracin disc. All isolates were confirmed as group

A streptococci by specific serotyping using the

Streptex method (Wellcome Laboratories Research Center, England).

If group A streptococci were recovered, the pa-tient’s participation in the study was discontinued, and an appropriate dose of amoxicillin was pre-scribed. Because the concurrent sinusitis was likely to involve a pathogen in addition to group A

Strep-tococcus,6 amoxicillin was selected instead of

(3)

Assignment

to Treatment

Group and Compliance

Radiographic

Evaluation

Subjects were stratified by age (2 to 5 and 6 to

16 years) and clinical severity score (<8; at least 8) and randomly assigned to one of three treatment groups: amoxicillin, amoxicillin-clavulanate potas-sium, or a look-alike placebo. In each case that active drug (amoxicillin) was prescribed, the dose was 40 mg/kg/d for ten days in three divided doses. No antihistamine-decongestants, either systemic or local, were prescribed. Neither the investigators nor the parents were aware ofwhich treatment the child was receiving. Compliance was assessed by history taking and measurement of residual drug at the follow-up visit.

Follow-up

Patients were telephoned and evaluated using a standardized ten-point assessment questionnaire at

1, 2, 3, 5, and 7 days. The ten-point questionnaire was also administered on the day of entry and at

the follow-up visit. The symptoms included in the

questionnaire were fever, nasal discharge, nasal congestion, cough, headache, facial pain, facial swelling, reduced activity, impaired sleep, and im-paired appetite. If a particular symptom was

pres-ent initially, a score of 2 was given; if it was absent,

the score was 0. A maximum total score at entry

was 20. If a particular symptom became more

se-vere, less severe, or stayed the same during treat-ment, +1, -1, or 0, respectively, was added to the original score for each symptom. Any child whose

symptoms worsened was reexamined. If the

condi-tion of the subject was deemed worse by the exam-iner, erythromycin/sulfisoxazole was prescribed on

the basis of 50 and 150 mg/kg/d, respectively, in

four divided doses for ten days. This ensured that patients receiving an antibiotic would receive an alternative drug and those receiving the placebo would receive an antimicrobial drug.

Clinical

Outcome

At the return appointment, ten days after entry into the study or when treatment with erythromy-cin-sulfisoxazole was completed, an interval history was obtained, and a physical examination and re-peat radiographic examination were performed. Clinical outcome status was defined as follows: (1) cure, complete absence of symptoms or a ten-point questionnaire score of 2 or less; (2) improvement, subsidence of clinical symptoms with a ten-point questionnaire score of no more than half of the original score; (3) same; or (4) failure, worsening of symptoms while the patient was observing the

treatment protocol or persistence of symptoms

Un-changed at ten days. Clinical outcome was evalu-ated at three and at ten days.

When repeat radiographs were performed, each maxillary sinus was evaluated separately. Radio-graphic findings were classified as: (1) improved, if both maxillary sinuses showed increased aeration or if one sinus showed increased aeration and the other sinus was not worse; (2) unchanged, if the

radiographs were the same or if one sinus was

improved and the other sinus was worse; and (3)

worse, if both sinuses showed decreased aeration or

if one sinus showed decreased aeration and the other sinus was unchanged.

Statistical Methods

The standard 2 test was used to analyze the data on clinical effectiveness.

RESULTS

During the period from November 1982 through May 1984, 171 children with clinical symptoms

suggestive of sinusitis who fulfilled the eligibility

requirements of this study were identified. Of these 171, 136 (80%) were found to have abnormal max-illary sinus radiographs. For children younger than

6 years, maxillary sinus radiographs were abnormal in 88%; for those older then 6 years the radiographs were abnormal in 70% (P < .01). The remaining 35 subjects with normal sinus radiographs were ex-cluded from the present trial. Group A Streptococ-cus was recovered from the throat or nasopharyn-geal cultures of 28 of 136 (20.8%) children, and their participation in the study was discontinued. An additional 15 children were eliminated from the analysis of results because of noncompliance or drug toxicity (Figure). The remaining 93 children,

49 boys and 44 girls, ranged in age from 2 to 16 years. Sixty children were 6 years of age or younger. Fifty-seven children were white and 35 were black; there was one Asian child. As shown in Table 2, treatment groups were similar with regard to age, race, clinical severity, and presence of unilateral or bilateral involvement as determined radiographi-cally. The placebo group had 40% boys compared with 60% boys in the antibiotic treatment groups.

Radiographic Findings

Initial radiographs of 93 subjects and 186 maxil-lary sinuses showed 162 to be abnormal with the following interpretations: 78 with partial or corn-plete opacification, 83 with rnucosal thickening of at least 4 mm, and one with an air-fluid level. Twenty-four maxillary sinuses were normal; in each

case the contralateral sinus was abnormal. Of the

(4)

Amoic,IIin

-Clavulanols Potassium

28

Figure. Study design.

radiographic abnormalities in at least one addi-tional paranasal sinus-most commonly the eth-moid sinus. Follow-up radiographic evaluation was available for 59 patients who improved or were cured clinically after ten days of therapy: 37 were

improved or normal, 12 were unchanged, and ten were worse.

Transillumination

Transillumination of the maxillary sinuses was attempted on all children at entry into the study

and again at follow-up. In 24% ofpatients, predom-inantly in children less than 6 years of age, the procedure could not be accomplished because of inability of the subject to cooperate. In an addi-tional 15% of evaluations, the interpretation of transillumination was indeterminate-neither

“normal” nor “absent.” When the remaining results

F

SIGNS 8 SYMPTOMS

I

SINUSITIS

I 171

I

were compared to radiographic findings, there was

agreement in 41% and disagreement in 20%.

Exclusions

Streptococcus. Twenty-eight children were ex-cluded when group A streptococci were recovered from the throat or nasopharyngeal culture. The mean age of these children was 6/12 (with a range

of 2/12 to 10/12), and they had been symptomatic

for an average of 21.9 days (range ten to 30). Their

principal symptoms were purulent nasal discharge

(86%) and cough (86%); no patient had sore throat as a prominent complaint. In only three of 28 patients was fever documented at the initial visit. All children with cultures positive for streptococci

had been randomized to a treatment group prior to

the availability of culture results. Eighteen were

receiving antibiotic (ten, amoxicillin; eight, amoxi-cillin-clavulanate potassium), and ten were receiv-ing placebo. When culture results were reported, 14

of 18 children receiving antibiotic therapy but only

one of ten receiving placebo showed improvement

(P<.01).

Fifteen children were excluded from the study for

lack of compliance or drug toxicity: in three cases

the medication was not obtained by the parent, in four patients there was noncompliance (three, amoxicillin-clavulanate potassium; one, placebo),

in two patients a pruritic erythematous rash devel-oped which warranted discontinuation of the study

drug (one, amoxicillin; one, placebo), and in six

patients diarrhea developed, necessitating discon-tinuation of the study drug (five, amoxicillin-cla-vulanate potassium; one, placebo).

Clinical

Outcome

Thirty patients were treated with amoxicillin, 28

with amoxicillin-clavulanate potassium, and 35 with placebo. The clinical assessment of the

differ-ent treatment groups at three and ten days is shown

in Table 3. When the amoxicillin and amoxicillin-clavulanate potassium groups are considered sepa-rately or combined, children treated with antibiotic

were statistically significantly more likely to be

TABLE 2. Characteristics of Study Population According to Treatment Group*

Treatment Groups Amoxicillin Amoxicillin-Clavulanate

Potassium

Placebo

Mean age (yr)

Sex: boys/girls

Race: black/white/other

Radiographs: unilateral/bilateral involvement Clinical severity score (mean)

Clinical severity score 8

No.

56/12

18/12

12/18/0 4:26 6.53 3

30

5/12

18/10

8/20/0 5:23 6.57 3

28

6/12

14/21

11/23/1 5:30 6.51 3

(5)

TABLE 3. Clinical Outcome of 93 Patients

of Therapy*

With Acute Si nusitis According to Treatm ent Group on Days 3 and 10

Treatment Groups Clinical Status Total

Cure Improved Same Failure

----Day 3 Amoxicillin Amoxicillin-clavulanate potassium Placebo Day 10 Amoxicillin Amoxicillin-clavulanate potassium Placebo 13 (43) 13 (47) 4 (11) 20 (67) 18 (64) 15 (43)

13 (43) 2 (7)

9 (32) 4 (14)

14 (40) 11 (32)

5 (16) 0

3 (11) 0 6 (17) 0

2 (7) 2 (7) 6 (17) 5 (16) 7 (25) 14 (40) 30 28 35 30 28 35

* Numbers in parentheses indicate percentages.

cured at three days than children receiving placebo

(P < .01). At ten days, children treated with an antibiotic were more likely to be cured and less likely to have unsuccessful treatment than children receiving placebo (P < .05). The clinical cure rate at ten days was 67% for amoxicillin, 64% for amox-icillin-clavulanate potassium, and 43% for placebo.

In 26 children therapy failed: 14 had received

placebo; five, amoxicillin; and seven,

amoxicillin-clavulanate potassium. Of the 14 children in whom treatment with placebo failed, the conditions of 12 responded promptly (within 48 hours) and one

slowly to the initiation of treatment with erythro-mycin-sulfisoxazole. Treatment of the remaining child with placebo was deemed a failure on day 10 because his condition was unchanged. When he appeared for evaluation three days later, his con-dition was improving and, consequently, no

anti-microbial agent was prescribed. Of the seven

chil-dren in whom treatment with

amoxicillin-clavulan-ate potassium failed, the conditions of four

re-sponded promptly, two responded slowly, and the

remaining patient failed to respond to therapy with erythromycin-sulfisoxazole. A sinus aspirate

ob-tamed from this child grew 3-lactamase-negative

Haemophilus influenzae. Of five patients initially

treated with amoxicillin in whom therapy failed,

the conditions of four responded briskly to treat-ment with erythromycin-sulfisoxazole. The last pa-tient underwent a sinus puncture after which she recovered without further antibiotic therapy; the

sinus aspirate was sterile.

DISCUSSION

The cost of respiratory infections in children can

be calculated in terms of both morbidity and ab-sences from school. Treatment with

drugs-decon-gestants, antihistamines (applied topically or

ad-ministered orally), or antimicrobials-adds consid-erably to the economic burden of respiratory

ill-nesses. The effectiveness of antimicrobial therapy

has not previously been compared to that of placebo

in patients with protracted respiratory symptoms and signs.

The common bacterial isolates recovered from maxillary sinus aspirates are Streptococcus pneu-moniae, Branhamella catarrhalLs, and H influenzae,

accounting, respectively, for approximately 30%, 20%, and 20% of infections.7’8 In a study of the

treatment of acute sinusitis in children that com-pared amoxicillin to cefaclor after diagnostic sinus

puncture, the outcome for each antibiotic group

was similar, with cure rates of approximately 80%.

The conditions of most of the remaining children

were clinically improved. In three of the four

chil-then who experienced a clinical failure, a 13-lacta-mase-producing antibiotic-resistant bacterial spe-cies was recovered from the maxillary sinus aspi-rate. The expected advantage of cefaclor or

amoxi-cillin-clavulanate potassium would be potential ac-tivity against most, if not all,

$-lactamase-produc-ing bacterial species. The currently expected local prevalence of /3-lactamase-producing bacterial

spe-cies in maxillary sinus aspirates is approximately 11% (15% of H influenzae, 40% of B catarrhalis).7’8

However, therapy will not fail in all patients with

in vitro antibiotic-resistant bacterial isolates.7 This

may explain the similar outcome of patients treated with amoxicillin and amoxicillin-clavulanate

potas-sium. The potential advantage offered by the

anti-bacterial spectrum of amoxicillin-clavulanate

po-tassium may have been offset by the slightly in-creased side effect of the drug-diarrhea-which

led to discontinuation of the medication in five

patients. Unfortunately, three of four noncompliant

subjects were also in the amoxicillin-clavulanate

potassium group.

The results of this study suggest an important role for antimicrobial agents in the management of

selected patients with upper respiratory symptoms. Children treated with an antibiotic experienced earlier relief of symptoms and a higher cure rate

(6)

for streptococci who received antibiotic treatment compared to those receiving placebo. In addition, when initial treatment failed in children receiving

placebo, the conditions of 12 of 13 responded

promptly to the initiation of treatment with eryth-romycin-sulfisoxazole.

To our surprise, group A streptococci were

re-covered from the cultures of the throat or

naso-pharynx of 21% of patients entered into this study. Although the syndrome of “streptococcosis” is corn-mon in children less than 3 years of age,9

presen-tation of streptococcal infection in the older child, as other than well-localized pharyngitis, has been

infrequently described.’0 Although one could

spec-ulate that these children were simply carriers of

group A streptococci and not acutely infected, the carriage rate in asymptomatic children calculated in the same season (for another study) was 5#{149}5%h1

The conditions of most children with cultures pos-itive for streptococci responded to antibiotic

ther-apy promptly; in only one child of ten receiving

placebo was improvement seen in 24 hours. This observation provides support for the therapeutic

effectiveness of antirnicrobials in both streptococcal infection and sinusitis. In children with throat

cul-tures positive for streptococci, other bacterial spe-cies may be isolated from their maxillary sinus

aspirates.6 In these patients, a streptococcal

infec-tion rather than a viral upper respiratory infection may have initiated the mucosal inflammation, lead-ing to obstruction of the sinus ostia and purulent

sinusitis.

The sensitivity and specificity of sinus

radio-graphs in signaling infection of the paranasal

sin-uses has been a subject of considerable controversy.

A recent study has shown that maxillary sinus

radiographs are infrequently abnormal in a “nor-mal” population of children older than 1 year

with-out recent or acute signs or symptoms of respiratory

inflammation.’2 However, there is little doubt that in the presence of respiratory inflammation, of either an allergic or infectious nature, maxillary

sinus radiographs are frequently abnormal. The group of children studied here had a constellation of respiratory signs and symptoms present in ad-dition to abnormal radiographs. On the basis of the history alone, the occurrence of significantly ab-normal radiographs could be predicted for 80% of patients: 88% of patients less than 6 years and 70% of patients older than 6 years of age. Consequently, it is probably unnecessary to obtain sinus

radio-graphs in the majority of apparently uncomplicated

cases of suspected sinus infection, particularly in

children less than 6 years of age.

The physical examination has not been especially

useful in the diagnosis of acute sinusitis in children

unless periorbital swelling is present or purulent secretions can be seen coming from the middle meatus. The results of this study suggest that

trans-illumination of the maxillary sinuses is also not

helpful diagnostically in contrast to an experience with adult patients.13 Transillumination is often

difficult to perform in the younger age group, and interpretation may be misleading or indecisive.

Special attention to the careful delineation of res-piratory signs and symptoms that have persisted

for at least ten but less than 30 days without

parental impression of improvement identifies a patient likely to benefit from specific antimicrobial therapy. On the other hand, if spontaneous im-provement has already been noted, treatment with antimicrobial agents is not necessary, thereby avoiding additional expense and potential toxicity.

ACKNOWLEDGMENTS

This study was supported by a grant from Beecham Pharmaceuticals, Bristol, TN.

The authors thank Drs. J. L. Paradise, B. Dashefsky,

and R. Michaels for review of the manuscript and Helen

Schorner for secretarial assistance.

REFERENCES

1. Cypress BK: Patterns of ambulatory care in pediatrics: The National Ambulatory Medical Care Survey United States, January 1980-December 1981. National Health Survey,

series 13, No. 75, US Department of Health and Human Services, Hyattsville, MD, 1983

2. Came S: Care of children in general practice. Br Med J

1979;2:190-192

3. Ackerman BD: Treatment of undifferentiated respiratory infections in infants. Clin Pediatr 1968;7:391-395

4. Townsend EH Jr, Radenbaugh JF: Prevention of complica-tions of respiratory illnesses in pediatric practice. N Engi J Med 1962;226:683-689

5. Todd JK, Todd J, Damato J: Bacteriology and treatment of purulent nasopharyngitis: A double-bind, placebo-controlled evaluation. Pediatr Infect Dis 1984;3:226-232

6. Wald ER, Milmoe GJ, Bowen A’D, et al: Acute maxillary sinusitis in children. N Engi J Med 1981;304:749-754

7. Wald ER, Reilly JS, Casselbrant M, et a!: Treatment of

acute maxillary sinusitis in childhooth A comparative study of amoxicillin and cefaclor. J Pediatr 1984;104:297-302 8. Wald ER, Reilly JS, Casse!brant MC, Chiponis DM:

Treat-ment of acute sinusitis in children: Augmentin vs cefaclor.

Postgrad Med Custom Commun, Sept/Oct 1984, pp 133-136 9. Powers GF, Boisvert PD: Age as a factor in streptococcosis.

J Pediatr 1944;25:481-504

10. Breese BB: Streptococcal pharyngitis and scarlet fever. Am J Di.s Child 1978;132:612-616

11. Hofkosh D, Wald ER: Prevalence of non-group A beta-hemolytic streptococci in childhood pharyngitis. Presented at the meeting of the Ambulatory Pediatric Association, Washington, DC, 1985

12. Kovatch AL, Wald ER, Ledesma-Medina J, et al: Maxillary sinus radiographs in children with nonrespiratory corn-plaints. Pediatrics 1984:73:306-380

(7)

1986;77;795

Pediatrics

Ellen R. Wald, Darleen Chiponis and Jocyline Ledesma-Medina

Trial

Acute Paranasal Sinus Infections in Children: A Double-Blind, Placebo-Controlled