Acute Periorbital Infections: Who Needs Emergent Imaging? Abstract

ARTICLES

Acute Periorbital Infections: Who Needs Emergent Imaging?

AUTHORS: Tiffany F. Rudloe, MD,a Marvin B. Harper, WHAT’S KNOWN ON THIS SUBJECT: Orbital infections may be MD,a,b Sanjay P. Prabhu, MBBS, MD,c Reza Rahbar, DMD, difficult to distinguish from periorbital cellulitis on the basis of MD,d Deborah VanderVeen, MD,e and Amir A. Kimia, MDa clinical examination. CT can be used to delineate orbital anatomy Divisions of a Emergency Medicine and bInfectious Disease, and determine if an abscess is present. Department of Medicine, cDivision of Neuroradiology, Department of Pediatric Radiology, and Departments WHAT THIS STUDY ADDS: We present here an algorithm to of d Otolaryngology and Communication Enhancement and eOphthalmology, Children’s Hospital Boston, Boston, stratify patients who are at risk for subperiosteal or orbital Massachusetts abscess and need emergent imaging, and to identify a population KEY WORDS of patients at low risk. periorbital cellulitis, orbital cellulitis, subperiosteal abscess, computed tomography, orbital infections ABBREVIATIONS CT—computed tomography EOM—extraocular movement abstract ED—emergency department WBC—white blood cell OBJECTIVES: Computed tomography (CT) is used often in the evalua- ANC—absolute neutrophil count tion of orbital infections to identify children who are most likely to URI—upper respiratory tract infection benefit from surgical intervention. Our objective was to identify predic- EMR—electronic medical record tors for intraorbital or intracranial abscess among children who CI—confidence interval IQR—interquartile range present with signs or symptoms of periorbital infection. These predictors could be used to better target patients for emergent CT. www.pediatrics.org/cgi/doi/10.1542/peds.2009-1709 METHODS: This was a retrospective cohort study of all patients admit- doi:10.1542/peds.2009-1709 ted to an urban pediatric tertiary care emergency department be- Accepted for publication Nov 12, 2009 tween 1995 and 2008. We included otherwise healthy patients with Address correspondence to Tiffany F. Rudloe, MD, Division of suspected acute clinical periorbital or orbital cellulitis without a his- Emergency Medicine, Children’s Hospital Boston, 300 Longwood tory of craniofacial surgery, trauma, or external source of infection. Ave, Boston, MA 02115. E-mail: Immunocompromised patients and patients with noninfectious causes [email protected] of periorbital swelling were excluded. Variables analyzed included PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275). age, duration of symptoms, highest recorded temperature, previous Copyright © 2010 by the American Academy of Pediatrics antibiotic therapy, physical examination findings, laboratory results, FINANCIAL DISCLOSURE: The authors have indicated they have and interpretation of imaging. CT scans of the orbit were reread by a no financial relationships relevant to this article to disclose. neuroradiologist. RESULTS: Nine hundred eighteen patients were included; 298 underwent a CT scan, and of those, 111 were shown to have an abscess. Although proptosis, pain with external ocular movement, and ophthalmoplegia were associated with presence of an abscess, 56 (50.5%) patients with abscess did not experience these symptoms. Other variables associated with the presence of an abscess in multivariate analysis were a peripheral blood neutrophil count greater than 10 000/␮L, absence of infectious conjunctivitis, periorbital edema, age greater than 3 years, and previous antibiotic therapy (P Ͻ .05 for all). Our recursive partitioning model identified all high-risk (44%) patients as well as a low-risk (0.4%–2%) group (Rsq ϭ 0.27). CONCLUSIONS: We confirmed that patients with proptosis and/or pain or limitation of extraocular movements are at high risk for intraorbital abscess, yet many do not have these predictors. Other features can identify patients who do not have such obvious predictors but do have significant risk of disease. A recursive partitioning model is presented. Pediatrics 2010;125:e719–e726

PEDIATRICS Volume 125, Number 4, April 2010 e719 Downloaded from www.aappublications.org/news by guest on September 28, 2021 Orbital cellulitis or abscess can occur as ver for Ͼ36 hours, and those with no not to be related to infection (patient a complication of acute sinusitis in chil- clinical improvement within 24 hours did not receive treatment with antibi- dren. This infectious process involves of first antibiotic administration.16 Oth- otic [eg, allergic reaction or nephrotic the structures posterior to the anatomic ers have suggested imaging only for syndrome]); had an underlying lesion boundary created by the orbital sep- patients with ophthalmoplegia, pain, such as a dacrocystocele; or were tum.1–3 Preseptal cellulitis is also com- or proptosis after their condition identified as having an external source mon in childhood, and may result from failed to respond to an initial antibiotic of infection (eg, infected laceration, sinusitis, percutaneous introduction of trial, reserving CT scanning for cases bug bites) or dentoalveolar abscess. In bacteria, or bacteremia.2,4–6 It is impor- in which surgical intervention is being cases of diagnostic uncertainty, we in- tant to distinguish between preseptal considered.3 cluded patients to whom antibiotic cellulitis, orbital cellulitis, and orbital ab- The aim of our study was to determine treatment was administered or pre- scess. Orbital and subperiosteal ab- clinical predictors of subperiosteal, scribed. For those patients with more scess place a patient at risk for blind- orbital, or intracranial complication in than 1 ED visit for periorbital symp- ness, intracranial infection, or death.1,7,8 cases of acute periorbital symptoms toms, subsequent visits were excluded External sources of infection, which caused by infection. Such predictors if they occurred within the first 24 could lead to periorbital swelling, are could be used to stratify patients into hours after the initial visit or Ͼ24 less likely to cross the anatomic bound- groups of those who are most likely or hours after the initial visit with no ary of the orbital septum and lead to least likely to have complications iden- changes in clinical variables. such complications.1,9,10 tified by CT scanning that would benefit Case Identification Evaluation of children with acute peri- from surgical intervention. Case identification was conducted in orbital swelling can be difficult. Clini- 2 phases. First, for the initial screen- cal findings alone may not be specific METHODS ing, we created a computer-assisted enough to distinguish preseptal from Study Design screening tool that is similar to key- orbital infections or those with compli- A retrospective cohort review of con- word search tools but uses regu- cations.11–13 To date, studies have iden- secutive patients seen at an urban ter- lar expressions.20 Regular expression tified proptosis and ophthalmoplegia tiary care pediatric emergency depart- matching provides a more comprehen- (limitation of extraocular movements) ment (ED) was conducted. The ED sive and inclusive search than key- as indicators of intraorbital inflamma- serves ϳ50 000 children per year. ED word searching because it includes tion but have not been able to distin- electronic medical records (EMRs) of possible misspelled and mistyped vari- guish between orbital cellulitis and patients who presented from October ations of the key word(s) of interest. abscess.13–15 Laboratory findings have 1995 through August 2008 with Second, a manual chart review was not proven helpful in making this periorbital-related complaints were performed by a single pediatric urgent determination.13 reviewed. The study protocol was ap- care medicine attending physician (Dr Concern for orbital involvement may proved by the Children’s Hospital Bos- Rudloe) using the inclusion and exclu- prompt the use of computed- ton internal review board committee. sion criteria above. tomography (CT) scanning of the orbit.9,13,16 CT imaging can detect exten- Study Population Data Extraction sion of infection into the orbit, identify All patients with available EMRs during Data regarding the ED or hospital an orbital abscess, or reveal sinus dis- the study period were evaluated for in- course and management were then re- ease.16–19 Although CT scanning is a clusion in this study. Otherwise healthy trieved from EMRs. Data abstracted in- useful tool, it exposes the pediatric pa- patients with acute periorbital edema cluded age, gender, height of fever, tient to ionizing radiation and, in some and/or erythema who were thought to duration of symptoms, previous upper cases, the risks associated with seda- have preseptal cellulitis, orbital cellu- respiratory tract infection (URI)/sinus- tion.14,16 Protocols in the current litera- litis, or intraorbital abscess were in- itis symptoms, previous antibiotic ture provide a wide range of indica- cluded in the study population. Pa- treatment, laboratory tests including tions for CT imaging that include tients were excluded if they had a complete blood count, absolute neu- ophthalmoplegia or proptosis but go history of immunodeficiency, previous trophil count (ANC), erythrocyte sedi- further to suggest imaging for those in ocular or craniofacial surgery, or ana- mentation rate, and measurement of whom visual acuity cannot be as- tomic anomaly (eg, craniosynostosis); C-reactive protein level. Physical exam- sessed, patients with persistence of fe- had orbital swelling that was thought ination findings (extraocular move-

e720 RUDLOE et al Downloaded from www.aappublications.org/news by guest on September 28, 2021 ARTICLES ment [EOM], degree of edema, propto- TABLE 1 Chandler Classification of Orbital Complications of Sinusitis, a Clinical Description sis), CT results, and procedures Chandler Stage Clinical Description and Definition performed were also recorded. EMRs Class were screened for return visits within I Inflammatory Eyelid edema and erythema edema Normal extraocular movement 1 week of the index visit. If a trainee Normal visual acuity was involved in the patient care, both II Orbital cellulitis Diffuse edema of orbital contents without discrete abscess formation the trainee and the attending physi- III Subperiosteal Collection of purulent exudatea beneath periosteum of lamina papyracea cian’s notes were reviewed. Findings abscess Displacement of globe downward/laterally IV Orbital abscess Purulent collection within orbita were considered present if they were Proptosis documented either by the resident or Chemosis attending physician. All cases were re- Ophthalmoplegia Decreased vision viewed to screen for a second ED visit V Cavernous sinus Bilateral eye findings or hospital admission within 1 week of thrombosis Prostration the index visit. Meningismus a The radiographic correlation of a subperiosteal or orbital abscess seen with CT is a contrast-enhancing mass in the Definitions extraconal/intraconal space, possibly with areas of cavitation, because purulence cannot be determined with CT scanning. We divided our clinical and laboratory data into binary results that represent Ͼ either presence or absence of a find- been administered for 24 hours be- those who require antibiotic therapy ing. Quantitative measures of propto- fore the ED visit. alone. sis were rarely recorded. Therefore, Outcome Measures CT Classification proptosis was considered simply as Imaging findings were categorized ac- present or absent. If there was no spe- Patients who underwent imaging cording to Chandler’s clinical classifica- cific finding of proptosis recorded in within 24 hours of their initial visit and tion (modified Huberts classification) for the chart, it was considered to be ab- were diagnosed with an abscess were complications of acute sinusitis (Table sent. Fever was defined as any temper- considered to have had the abscess on 1).21 This classification system was intro- ature of Ͼ38°C recorded during pa- their initial visit and classified as hav- duced before the advent of CT but is tient care (oral, rectal, or tympanic). ing a significant result. Patients who widely used today.10 Edema was considered significant if underwent imaging Ͼ24 hours from the documentation suggested exten- The Chandler classification system is their initial visit and were diagnosed sion beyond the eyelid. EOMs were cat- similar to others described based on with an abscess were excluded. The egorized as limited, intact, or unable to CT findings.22 Both ED imaging results authors could not determine if the ab- determine. Pain with EOM was coded and imaging results from outside hos- scess represented natural progress of as present if specifically recorded. Oth- pitals were included. All CT scans per- disease of an existing abscess from erwise, pain with EOM was coded as formed within 24 hours of the ED visit the index visit. Patients who did not not present/not addressed. For the were analyzed. All available CT images undergo imaging were classified as purposes of analysis the 3 results of were reviewed again by a single neuro- having a negative or “nonsignificant” proptosis, limitation of EOM, and pain radiologist (Dr Prabhu) who classified result. with EOM were also grouped into a sin- the readings by using Chandler’s cri- gle “clinical findings” category if any teria. Two-dimensional and coronal Statistical Analysis one was present. On the basis of the planes were reviewed. Discrepancies Data were analyzed by using SPSS for ability to communicate symptoms and between original readings and those Windows (SPSS Inc, Chicago, IL). Multi- primary analysis, age was treated as a of the study reviewer were resolved by variate analyses were used to examine binary variable, with the 2 groups be- consensus among investigators. A ␬ score relationships between patients with ing those aged 3 years and older and was calculated to assess agreement. significant CT findings and those with- those younger than 3 years. Duration A CT finding was classified as “signifi- out them. All patients were included in of symptoms was recorded and then cant” if the Chandler score was III or the analysis even if they did not have a analyzed as greater than or less than 1 higher (subperiosteal abscess, orbital CT scan. A recursive partitioning day of symptoms. Previous antibiotic abscess, or cavernous sinus thrombo- model was created by using JMP soft- treatment was considered present if sis), thus discriminating patients who ware (SAS Institute, Cary, NC) that cre- any oral or intravenous antibiotic had may need surgical intervention from ates a decision tree based on dichoto-

PEDIATRICS Volume 125, Number 4, April 2010 e721 Downloaded from www.aappublications.org/news by guest on September 28, 2021 mous dependent variables. Internal validation was performed by using K-fold validation.

RESULTS Study Group There were 670 056 ED visits from Oc- tober 1995 to August 2008 with electronic ED notes available for review. Fig 1 shows how the 918 patient visits that met study criteria were identiﬁed. Table 2 lists the patient demographics and clinical characteristics.

Radiologic Findings Two hundred ninety-eight patients had an orbital CT performed (32%). There were 295 CT scans obtained with contrast-enhanced images (Table 3). One hundred eleven patients, 12% of patients overall (95% confidence interval [CI]: 10.1–14.4) or 37% (95% CI: 32.0–43.2) of the 298 evaluable CT scans had significant CT findings. There was good agreement between radiology readings when original readings were compared with neuroradiologist rereads (K ϭ 0.9). The max- FIGURE 1 Case identification. PICC indicates peripherally inserted central catheter. illary and ethmoid sinuses were the most commonly involved, at 86% and 84%, respectively (Table 3). these, consults of both services were aureus Staphylococcus, 2 group A Of the 111 patients who had a signifi- common (148). streptococci, and 1 Streptococcus mil- cant CT finding (abscess), the median leri). Seven patients had sequelae of age was 7.3 years (interquartile range Patient Outcomes infection 3 months from the initial pre- [IQR]: 3.3–9.7); 20 patients were Four hundred sixty-nine patients (51%) younger than 3 years of age. The me- sentation: 2 patients suffered loss of were admitted to the hospital. The me- vision in the affected eye, 3 had persis- dian white blood cell (WBC) count dian hospital length of stay was 2.1 ␮ tence of ptosis, and 2 had persistence was 15 500/ L (IQR: 12.6–20 000), and days (IQR: 1.6–3.7). Seventy-two pa- ␮ of pain. None of the 39 remaining pa- median ANC was 12 210/ L (IQR: 9900– tients underwent a surgical proce- 16 640). Median duration of ocular tients required surgery after subse- dure. Sixty-two children had drainage quent review of the charts. symptoms was 2 days. Of the 111 pa- of a subperiosteal or orbital abscess, tients with abscess shown with CT, 56 and 10 had sinus surgery. Forty sinus Of the 450 patients discharged to home (50.5% [95% CI: 40.8–60.0]) of these or abscess specimens were obtained from the initial ED visit, 22 (4.8%) re- did not have ophthalmoplegia, proptosis, after antibiotic treatment. The organ- turned to the ED Ͼ24 hours but Ͻ1 or pain with extraocular movements. isms recovered are shown in Table 4. week after the initial visit. One patient Subspecialty consult requests were Three of 407 blood specimens that had a significant finding on orbit CT ordered for 240 patients. The most were sent for culture grew a pathogen scan (subperiosteal abscess) when common were otorhinolaryngology (Streptococcus pneumoniae), and 8 she returned on day 4 and was ex- (171) and ophthalmology (217). Among grew an unlikely pathogen (5 non- cluded per protocol.

e722 RUDLOE et al Downloaded from www.aappublications.org/news by guest on September 28, 2021 ARTICLES

TABLE 2 Patient Demographics TABLE 5 Multivariate Analysis Value 95% CI Variable Adjusted Odds ED visits during the study period, n 670 056 Ratio (95% CI) Included patients 918 ANC Ն 10 000 cells per ␮L 8.6 (4.4–16.9) Age, median (IQR), y 3.8 (1.7–7.4) Proptosis, limitation, or pain 6.6 (3.3–13.5) Female, % 46 with EOM Symptoms of URI or sinusitis, % 39 36–43 Absence of conjunctivitis 5.9 (1.5–23.9) Pretreated with antibiotics, % 34 32–43 Edema beyond the eyelid 3.6 (1.9–6.7) Topical antibiotics, % 6 Previous antibiotics 3.4 (1.8–6.3) Oral antibiotics, % 17 Age Ն 3 y 2.3 (1.1–5.0) Parenteral antibiotics, % 11 Presence of URI or sinusitis 0.9 (0.5–1.7) Duration of previous antibiotics, median, d 1 (by history) Duration of ocular symptoms, median (IQR), d 1 (1–2) Temperature Ն 38.0°C 0.6 (0.3–1.1) Previous conjunctivitis, % 19 16–21 Periorbital edema Ͼ eyelid, % 24 21–27 Pain with extraocular movement, % 12 10–13 Limited extraocular movement, % 13 12–17 tients with none of these findings, the Proptosis, % 4 3–5 Fever (Ն38.0°C), % 42 35–45 risk can be further stratified to low, Temperature, median (IQR), °C 37.3 (36.5–38.8) intermediate, and high risk on the ba- WBC count obtained, % 71 sis of the presence of significant ϫ ␮ WBC, mean (IQR), 1000/ L) 13.8 (9.9–17.7) Ͼ ␮ ANC, mean (IQR), ϫ1000/␮L) 8.1 (4.2–11.8) edema, an ANC of 10 000/ L, age, and previous antibiotic treatment, as described in Fig 2. TABLE 3 Radiologic Findings, Based on TABLE 4 Bacterial Pathogens From Sinus Chandler Stage of Orbital Findings Cultures DISCUSSION No. (%) of Patients Organism No. of Cultures In this large study in a pediatric ED set- Chandler stage Viridans streptococci 25 ting, we examined the factors that 0 12 (4) (Streptococcus intermedius) (17) 1 124 (42) Non-aureus Staphylococcus 13 identify candidates for early imaging 2 51 (17) Streptococcus pyogenes (group 10 of the orbit. In the case of patients who 3 100 (34) A Streptococci) presented with periorbital symptoms 4 6 (2) Staphylococcus aureus 7 5 4 (1) (MRSA) (1) that may indicate infection, the diag- Bilateral involvement 5 Streptococcus pneumoniae 5 noses of concern are intraorbital ab- Stage 1 4 Corynebacterium 3 scess, which may lead to blindness, Stage 2 1 Haemophilus influenzae 3 and intracranial spread of infection Sinus involvement Peptostreptococci 2 Maxillary 258 (86) Arcanobacterium pyogenes 1 or thromboses, which may lead to Ethmoid 252 (84) Eikenella corrodens 1 death.9,23,24 Prompt recognition of pa- Sphenoid 131 (44) Escherichia coli 1 tients who are at risk for complica- Frontal 89 (30) Group C streptococci 1 None 26 (9) Klebsiella pneumoniae 1 tions, with timely CT imaging and/or Unspecified 3 (1) Prevotella intermedia 1 specialty surgical consultation, is an Multiple 240 (80) Propionibacterium granulosum 1 early goal for these children. When Pseudomonas 1 The data should be interpreted with caution, because 75% identified, a discrete abscess collec- of our patients were younger than 7 years of age. Stomatococcus 1 Multiple organisms 23 tion that is sufficiently large to allow No growth 10 drainage may require surgical drain- Multivariate Analysis Organisms in parentheses are a subset of the group listed age.9,23 In a carefully monitored setting, directly above it. MRSA indicates methicillin-resistant Binary logistic regression was used Staphylococcus aureus. many patients with subperiosteal ab- to assess the association of each vari- scess respond well to medical man- able with positive CT findings. The vari- agement but must be identified early.25 ables associated with imaging findings all high-risk (44%) patients and a low- As a result, we have focused our atten- are presented in Table 5. risk (0.4%–2%) group (Rsq ϭ 0.27 tion on identifying orbital or subperi- [K-fold value of 0.24]). The model osteal abscesses that may require Recursive Partitioning shows that those patients with the clin- drainage rather than trying to discrim- A recursive partitioning model was de- ical findings of proptosis, ophthalmo- inate these infections as being presep- veloped and is shown in Fig 2. Our re- plegia, or pain with EOM fall directly tal versus postseptal. Other authors cursive partitioning model identified into the high-risk category. For pa- have taken a similar approach and

PEDIATRICS Volume 125, Number 4, April 2010 e723 Downloaded from www.aappublications.org/news by guest on September 28, 2021 FIGURE 2 Recursive partitioning model stratifying the risk for orbital abscess. A, Patients with proptosis, ophthalmoplegia, or pain with EOM were at high risk. B, Those with edema beyond the eyelid had a risk of 20%, whereas patients with no edema beyond the eyelid had a risk of 3.5%. C, The risk can be further stratified by adding ANC, age, previous antibiotics (shown), and previous conjunctivitis (not shown). have proposed refining the classifica- with postseptal disease compared also reported that subperiosteal ab- tion by simply considering those with with preseptal disease.14 However, scess can present with only swelling orbital complications of acute sinus- these findings could not discriminate and erythema. The proportion of pa- itis and dividing them into 3 main cat- between orbital cellulitis and orbital tients with an abscess who lack classic egories (orbital cellulitis, subperios- abscess.14 findings is concerning, because the lit- teal abscess, and orbital abscess) We present here 111 patients with erature suggests that delay in treat- rather than using the more traditional Chandler class III or greater, of which ment can result in blindness in up to preseptal or postseptal terminology almost half had either proptosis or 10% of affected patients.15 Two of our commonly used.22 This strategy leaves ophthalmoplegia. Our data confirm 111 patients had a course that re- preseptal cases in a separate cate- that clinical findings such as propto- sulted in blindness. As a result of the gory, because this is not a process that sis, ophthalmoplegia, or pain with ex- poor sensitivity of clinical findings, the directly affects the orbit. traocular movements are causes for clinician needs better tools to identify A previous study of 139 patients who concern not only for orbital inflamma- those at risk. Clinical findings also underwent CT scanning (38 had post- tion but also for abscess. It is impor- have poor specificity; 49% of our pa- septal disease and 101 had preseptal tant to draw attention to the 56 patients with findings had Chandler class only) showed that proptosis and oph- tients (50.5%) who did not have these I or II disease. thalmoplegia were found in a signifi- classic findings but did have CT dis- We also present an algorithm to fur- cantly higher proportion of patients ease stage III or higher. Rahbar et al15 ther stratify the risk for significant

e724 RUDLOE et al Downloaded from www.aappublications.org/news by guest on September 28, 2021 ARTICLES complications among patients who CT scans obtained was unchanged. The also possible that factors related to present without these classic symp- effect on number of images obtained how patients arrive at a pediatric ED toms or clinical findings. Our results in centers that do not have pediatric would introduce bias. If so, our screen- show that an ANC of Ͼ10 000 cells per subspecialty and/or pediatric emer- ing criteria may perform more poorly ␮L, moderate-to-severe periorbital gency medicine expertise remains to in other settings. In addition, there edema, absence of conjunctivitis as be seen. However, identifying patients may have been other reasons for ob- the presenting symptom, age Ͼ 3 at particularly high risk should pro- taining a CT scan that were not evident years, and previous antibiotic use are vide the opportunity to expedite the to us in this retrospective study. care of these patients with a goal of predictors of an orbital abscess. The It is also important to note that we decreasing associated morbidity. One suggestion that edema extending be- have not analyzed the role of MRI in the patient excluded from our study was a yond the eyelid margins is a strong management of these cases. Others 2-year-old girl who came back on day 4 predictor was previously described by have suggested that MRI be consid- with subperiosteal abscess. Had the Vu et al26 as part of a severity score. In ered for cases in which there is clinical suggested algorithm been applied to addition, the Vu et al score suggests or CT-based suspicion for intracranial her at her first visit, her risk would high fever (Ͼ39°C) as a marker of se- complications such as subdural empyema have been estimated at 14%. In our verity when it comes to predicting an ab- or cavernous sinus thrombosis.27,28 scess. Our data did not identify fever as a model, almost two thirds of the patients can be classified as being at low risk, predictor. It is interesting to note that CONCLUSIONS eliminating the need for emergent imag- 44% of the patients with an abscess did ing. We therefore recommend expedited We present a tool for physicians not have a documented temperature of imaging of those patients who fall into who are considering emergent CT im- Ͼ38°C recorded during their illness. the high-risk category, with strong con- aging versus antibiotic therapy and ex- By excluding patients with an evident sideration of imaging for those in the pectant management. In addition to external source for infection, we iden- intermediate-risk category. the known high-risk findings of oph- tified a high rate with sinus disease on This study has several limitations. We thalmoplegia and proptosis, or in the CT scan (94%) and an overall sobering considered those who did not undergo absence of these signs, we have shown risk of an abscess at 12%. The overall CT as not having Chandler III or worse that patients who have moderate-to- risk may differ depending on the prac- disease. As a result, we may have un- severe periorbital edema, or an ANC of tice setting (tertiary versus primary derestimated the percentage of chil- Ͼ10 000 cells per ␮L, also fall into a care). The incidence of surgical proce- dren with significant findings. A sub- high-risk category and their evaluation dures in our cohort (63.9%) was simi- group analysis of only those patients should be expedited. We also suggest lar to that in previous studies (66%).14 who had CT performed was evaluated, that those at low risk can be easily Once the algorithm was applied to our and the recursive partitioning yielded identified, and that CT imaging is of low patient population, the total number of similar results (data not shown). It is yield for these patients. REFERENCES

1. Lessner A, Stern GA. Preseptal and orbital 7. Reynolds DJ, Kodsi SR, Rubin SE, Rodgers IR. tors for admission of children with perior- cellulitis. Infect Dis Clin North Am. 1992;6(4): Intracranial infection associated with pre- bital cellulitis presenting to the pediatric 933–952 septal and orbital cellulitis in the pediatric emergency department. Pediatr Emerg 2. Givner LB. Periorbital versus orbital cellulitis. patient. J AAPOS. 2003;7(6):413–417 Care. 2008;24(5):279–283 Pediatr Infect Dis J. 2002;21(12):1157–1158 8. Hyto¨nen M, Atula T, Pitka¨ranta A. Complica- 13. Ho CF, Huang YC, Wang CJ, Chiu CH, Lin TY. 3. Uzca´tegui N, Warman R, Smith A, Howard tions of acute sinusitis in children. Acta Oto- Clinical analysis of computed tomography- CW. Clinical practice guidelines for the man- laryngol Suppl. 2000;543:154–157 staged orbital cellulitis in children. J Micro- agement of orbital cellulitis. J Pediatr Oph- 9. van Dissel JT, de Keizer RJ. Bacterial infec- biol Immunol Infect. 2007;40(6):518–524 thalmol Strabismus. 1998;35(2):73–79; quiz tions of the orbit. Orbit. 1998;17(4):227–235 14. Sobol SE, Marchand J, Tewﬁk TL, Manoukian 110–111 10. Younis RT, Lazar RH, Bustillo A, Anand VK. JJ, Schloss MD. Orbital complications of sinus- 4. Smith TF, O’Day D, Wright PF. Clinical implica- Orbital infection as a complication of sinusitis: itis in children. J Otolaryngol. 2002;31(3): tions of preseptal (periorbital) cellulitis in are diagnostic and treatment trends chang- 131–136 childhood. Pediatrics. 1978;62(6):1006–1009 ing? Ear Nose Throat J. 2002;81(11):771–775 15. Rahbar R, Robson CD, Petersen RA, et al. 5. Adnan SB. Instructive case: a swollen eye. 11. Starkey CR, Steele RW. Medical management Management of orbital subperiosteal ab- J Paediatr Child Health. 2000;36(2):179–181 of orbital cellulitis. Pediatr Infect Dis J. 2001; scess in children. Arch Otolaryngol Head 6. Wald ER. Periorbital and orbital infections. 20(10):1002–1005 Neck Surg. 2001;127(3):281–286 Pediatr Rev. 2004;25(9):312–320 12. Goldman RD, Dolansky G, Rogovik AL. Predic- 16. Howe L, Jones NS. Guidelines for the manage-

PEDIATRICS Volume 125, Number 4, April 2010 e725 Downloaded from www.aappublications.org/news by guest on September 28, 2021 ment of periorbital cellulitis/abscess. Clin Oto- ER. The pathogenesis of orbital complica- management of subperiosteal abscess of the laryngol Allied Sci. 2004;29(6):725–728 tions in acute sinusitis. Laryngoscope. 1970; orbit: analysis of outcomes 1988–1998. Oph- 17. Hirsch M, Lifshitz T. Computerized tomogra- 80(9):1414–1428 thalmology. 2000;107(8):1454–1456; discus- phy in the diagnosis and treatment of orbital 22. Velasco e Cruz AA, Demarco RC, Pereira sion 1457–1458 cellulitis. Pediatr Radiol. 1988;18(4):302–305 Valera FC, dos Santos AC, Anselmo-Lima WT, 26. Vu BL, Dick PT, Levin AV, Pirie J. Development da Silva Marquezini RM. Orbital complica- 18. Eustis HS, Mafee MF, Walton C, Mondonca J. of a clinical severity score for preseptal cel- tions of acute rhinosinusitis: a new classiﬁ- lulitis in children. Pediatr Emerg Care. 2003; MR imaging and CT of orbital infections and cation. Braz J Otorhinolaryngol. 2007;73(5): 19(5):302–307 complications in acute rhinosinusitis. Radiol 684–688 Clin North Am. 1998;36(6):1165–1183, xi 27. Kapur R, Sepahdari AR, Mafee MF, et al. MR 23. Ikeda K, Oshima T, Suzuki H, Kikuchi T, Suzuki imaging of orbital inﬂammatory syndrome, 19. Goldberg F, Berne AS, Oski FA. Differentia- M, Kobayashi T. Surgical treatment of sub- orbital cellulitis, and orbital lymphoid tion of orbital cellulitis from preseptal cel- periosteal abscess of the orbit: Sendai’s lesions: the role of diffusion-weighted imag- lulitis by computed tomography. Pediatrics. ten-year experience. Auris Nasus Larynx. ing. AJNR Am J Neuroradiol. 2009;30(1): 1978;62(6):1000–1005 2003;30(3):259–262 64–70 20. Friedl JEF. Mastering Regular Expressions. 24. Brook I, Friedman EM, Rodriguez WJ, Con- 28. Sepahdari AR, Aakalu VK, Kapur R, et al. MRI 3rd ed. Sebastopol, CA: O’Reilly Media Inc; troni G. Complications of sinusitis in chil- of orbital cellulitis and orbital abscess: the 2006. dren. Pediatrics. 1980;66(4):568–572 role of diffusion-weighted imaging. AJR 21. Chandler JR, Langenbrunner DJ, Stevens 25. Garcia GH, Harris GJ. Criteria for nonsurgical Am J Roentgenol. 2009;193(3):W244–W250

e726 RUDLOE et al Downloaded from www.aappublications.org/news by guest on September 28, 2021 Acute Periorbital Infections: Who Needs Emergent Imaging? Tiffany F. Rudloe, Marvin B. Harper, Sanjay P. Prabhu, Reza Rahbar, Deborah VanderVeen and Amir A. Kimia Pediatrics 2010;125;e719 DOI: 10.1542/peds.2009-1709 originally published online March 1, 2010;

Updated Information & including high resolution figures, can be found at: Services http://pediatrics.aappublications.org/content/125/4/e719 References This article cites 27 articles, 5 of which you can access for free at: http://pediatrics.aappublications.org/content/125/4/e719#BIBL Subspecialty Collections This article, along with others on similar topics, appears in the following collection(s): Dermatology http://www.aappublications.org/cgi/collection/dermatology_sub Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.aappublications.org/site/misc/Permissions.xhtml Reprints Information about ordering reprints can be found online: http://www.aappublications.org/site/misc/reprints.xhtml

Downloaded from www.aappublications.org/news by guest on September 28, 2021 Acute Periorbital Infections: Who Needs Emergent Imaging? Tiffany F. Rudloe, Marvin B. Harper, Sanjay P. Prabhu, Reza Rahbar, Deborah VanderVeen and Amir A. Kimia Pediatrics 2010;125;e719 DOI: 10.1542/peds.2009-1709 originally published online March 1, 2010;

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://pediatrics.aappublications.org/content/125/4/e719

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it has been published continuously since 1948. Pediatrics is owned, published, and trademarked by the American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 2010 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.

Downloaded from www.aappublications.org/news by guest on September 28, 2021