Medical versus surgical management of pediatric orbital subperiosteal abscesses


Send correspondence to Joshua R. Bedwell, MD, Division of Otolaryngology, Children's National Medical Center, 111 Michigan Ave NW, Washington, DC 20010. E-mail:


Which children with orbital subperiosteal abscess are able to be managed medically versus surgically?


Orbital subperiosteal abscess (SPA) describes a process in which a collection of pus forms between the bony orbit and the periorbita. SPA most commonly arises as a complication of acute sinusitis, where it exists along a spectrum of orbital involvement from limited (preseptal cellulitis) to severe (cavernous sinus thrombosis). The surgical management of SPA in children has evolved from drainage via an external approach to an endoscopic approach. Controversy remains, however, in deciding which SPAs require drainage, and which may be managed conservatively with systemic antibiotics.


The advent of high-resolution computed tomography (CT) scanning allowed for definitive diagnosis of orbital SPAs. Prior to the 1980s, evidence of SPA on CT was an absolute indication for immediate surgical drainage. In 1990, Souliere reported on a series of 10 patients with CT-documented SPA and sinusitis.[1] In this retrospective series, five patients were successfully managed with systemic antibiotics and decongestants, while five patients ultimately required surgery. Due to the small series, no significant differences between the groups could be found, but the series did illustrate that SPA need not be an absolute indication for surgery. The authors concluded that in patients without worsening ophthalmologic findings or evidence of intracranial spread or systemic toxicity, a 48-hour trial of antibiotics, along with serial examinations (every 2 hours) was a reasonable approach.

Other groups have presented retrospective series aiming to tease out the differences between those that are successively treated with antibiotics alone versus those who require surgery. Rahbar examined a series of 19 patients with SPA due to sinusitis.[2] Of these, five patients improved with medical therapy (intravenous antibiotics and topical oxymetazoline) and 14 patients required surgical management. Patients were monitored with twice daily ophthalmologic examinations. Patients who failed conservative management went to surgery an average of 2 days after the initiation of therapy. In a multivariate analysis including age, sex, gaze restriction, number of extraocular muscles involved, orbital fat involvement, rim enhancement, proptosis, and abscess volume, only the degree of proptosis was found to be a significant predictor for surgery. Only one of the medically managed patients had proptosis in this series. The abscess volume on CT was found to be significantly larger in the surgical group, but this significance did not hold up in multivariate analysis.

Oxford reported on a series of 43 patients with SPA.[3] Forty-two percent of the patients were able to be managed medically. Their analysis found a number of significant differences between the medical and surgical groups. A higher proportion of patients in the surgical group had findings of chemosis, severe gaze restriction, and proptosis. In contrast to Rahbar's series, however, 55% of the patients managed medically had proptosis greater than 2 mm, demonstrating that the finding of proptosis in and of itself should not lead to surgery. None of the patients in the medical group had elevated intraocular pressure (IOP), while 50% of those requiring surgery had IOP >20 mmHg. Patients requiring surgery had larger abscesses than those who did not, with the width of the abscess demonstrating the largest difference between the groups. The widest abscess in the medically managed group was 4 mm. The authors recommended a trial of medical management in patients with normal vision, normal IOP, no gaze restriction, proptosis <5 mm, and an abscess width <4 mm. Notably, age was not found to be a significant factor in either of these two series.

Ryan analyzed a series of 68 patients with SPA, of which 69% were managed medically and 31% were managed surgically.[4] Patients were managed with intravenous antibiotics, topical decongestants, and nasal saline irrigation. Indications for surgery were large abscess, worsening ophthalmologic examination findings, clinical deterioration, or failure to improve after 24 to 48 hours. Age and temperature on admission were found to be significant differences between the two groups. Average age in the surgical group was 8.3 years, versus 6.2 in the medical group. The authors analyzed the effect of abscess size by defining “large” abscesses as >10 mm and “small” abscesses as <10 mm. Ninety-two percent of patients with large abscesses required surgery, compared to 19% with small collections. Among those with small abscesses, there was no difference in age between those who were adequately treated medically versus those requiring surgery. The authors concluded that conservative management was more likely to be successful in younger patients (<6 years) with abscesses <10 mm wide.

Garcia and Harris published the only prospective study of children with SPA.[5] Based on evidence that older children with SPA may have a different bacteriologic profile, the authors defined a set of criteria to select patients for medical management. Patients qualified for conservative management if they were less than 9 years old, had a medially located SPA that was not “large” (size criteria were not defined), no concern for anaerobic infection or dental origin, no frontal sinusitis, no chronic sinusitis, and no optic nerve or retinal compromise. Treatment consisted of intravenous antibiotics for at least 4 days, followed by 3 weeks of an oral antibiotic. Patients were monitored with serial ophthalmologic examinations every 6 hours for at least 48 hours. Treatment failure prompted surgical management, and was defined by: 1) development of visual loss or afferent pupillary defect; 2) continued fevers for longer than 36 hours after initiation of treatment; 3) clinical deterioration after 48 hours; or 4) no clinical improvement after 72 hours of treatment. Of the 29 patients that met the inclusion criteria, 27 (93%) were successfully treated without surgery. With these results in mind, it is worth noting that 14 of the 47 medically treated patients (30%) in the series reported by Ryan et al. were age 10 or older.[4]


Taken as a whole (Table 1), the literature suggests that orbital SPA in children is not an absolute indication for immediate surgical intervention. Patients presenting with advanced ophthalmologic findings (impaired visual acuity, elevated IOP, ophthalmoplegia, proptosis ≥5 mm) or with large abscesses (width >10 mm) are best treated surgically. Patients with a less serious presentation may improve with conservative management (intravenous antibiotics, nasal saline lavage, topical decongestants). There is evidence that older patients are less likely to be adequately treated with medications alone; however, age should not be a contraindication to a trial of conservative management. All patients must be monitored closely with serial ophthalmologic examinations, and any deterioration should lead to timely drainage. Failure to improve after 48 hours likely reflects treatment failure and should prompt consideration for surgical intervention.

Table 1. Review of Relevant Studies.
AuthorNumber of PatientsMedicalSurgicalDifferences Between GroupsRecommendations for Medical Management
  1. a

    Only proptosis significant in multivariate analysis. IOP=intraocular pressure.

Souliere (1990)1055n/a48-hr trial of antibiotics
     Q2hr opthalmologic exam
Garcia (2000)29272n/aAge <9 yrs
     Q6hr opthalmologic exam
Rahbar(2001)19514ProptosisTwice daily opthalmologic exam
    Abscess volumea 
Oxford (2006)431825ChemosisNormal vision
    ProptosisNo gaze restriction
    Severe gaze restrictionIOP <20 mmHg
    IOP >20 mmHgProptosis <5 mm Width <4 mm
Ryan (2009)684721AgeAge<6 yrs
    Temperature on admissionWidth<10 mm
    Size (10 mm cutoff) 


All studies are level 4 (case series), with all but one being retrospective. Further research in the area with a large prospective randomized trial would be ideal, but given the relatively low incidence this would be very difficult to achieve.