Article Text

Download PDFPDF

Orbital cellulitis in Scotland: current incidence, aetiology, management and outcomes
  1. C Murphy1,
  2. I Livingstone2,
  3. B Foot3,
  4. H Murgatroyd1,
  5. C J MacEwen1
  1. 1Department of Ophthalmology, Ninewells Hospital, Dundee, UK
  2. 2Glasgow Centre for Ophthalmic Research, Greater Glasgow & Clyde, Glasgow, UK
  3. 3British Ophthalmic Surveillance Unit, The Royal College of Ophthalmologists, London, UK
  1. Correspondence to Dr Claire Murphy, Department of Ophthalmology, Ninewells Hospital, Dundee DD1 9SY, UK; c.murphy3{at}


Purpose Orbital cellulitis is a potentially blinding and life-threatening condition. There are little published data on the incidence of orbital cellulitis and little is known about the differences between children and adults affected. The purpose of this study was to identify the incidence, aetiology, management and outcome of orbital cellulitis in children and adults in Scotland.

Methods This study was a 1-year prospective observational study using the Scottish Ophthalmic Surveillance Unit reporting system among Scottish ophthalmologists.

Results The response rate from ophthalmologists was 66.4%. There were 15 children and 5 adults reported giving an incidence of 1.6 per 100 000 and 0.1 per 100 000 in children and adults, respectively. 47% of children had a preceding upper respiratory tract infection with 87% having radiological evidence of sinus disease. Within the adult group, there was preceding immunosuppression and trauma. Streptococcus (66%) and Haemophilus (46%) species were the most commonly isolated pathogens in children. Respiratory pathogens were less predictable in adults. All patients were treated with intravenous antibiotics. All children with orbital and subperiosteal abscesses had surgery; one adult with orbital abscess did not have surgery. There were two cases of series morbidity: one intracranial spread of infection and one evisceration.

Discussion The incidence of orbital cellulitis is higher in children than in adults. In children, it commonly follows upper respiratory infection and sinus disease; however, in adults, preceding illness and trauma are more common. Respiratory pathogens are common in affected children. Intravenous antibiotics and surgical treatment of abscesses remain the preferred management.

  • Orbit
  • Eye Lids
  • Infection
  • Epidemiology
  • Pathology

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Infective orbital cellulitis is an uncommon condition that involves the ocular adnexal structures behind the orbital septum. It is a potentially blinding and life-threatening condition that presents predominantly in children.1 Orbital cellulitis is a clinical diagnosis that is supported by radiological investigations. Management can be challenging in both adults and children, both requiring a multidisciplinary approach but often with differences depending on aetiology and comorbidity.

There are little published data regarding the incidence of orbital cellulitis, the current literature mainly consisting of single-centre retrospective case series.1–3 To our knowledge, to date there is only one paper that accurately describes the incidence of orbital cellulitis; in 2000, the incidence in California was 3.51 per 100 000 of the population.4 There is currently no reported incidence of orbital cellulitis from the UK.

The purpose of this observational study is to compare and contrast orbital cellulitis occurring in children and adults within Scotland over a 12-month period. The main aims are (i) to establish the difference in incidence of orbital cellulitis between Scottish children and adults, (ii) to identify differences in aetiology and microbiology of orbital cellulitis in children and adults and (iii) to compare the management and outcomes of orbital cellulitis in children and adults.

Materials and methods

Cases of orbital cellulitis were identified prospectively through the Scottish Ophthalmic Surveillance Unit (SOSU). The methodology is as commonly used by the sister organisation British Ophthalmic Surveillance Unit (BOSU).5 All Scottish ophthalmic consultants were asked to report any cases of newly diagnosed orbital cellulitis on a monthly reporting card. Reporting consultants were issued an incident questionnaire relating to demographics, clinical findings at presentation, investigations and management. A follow-up questionnaire was completed 3 months later regarding any readmissions and ocular or systemic morbidity. Only patients’ resident within Scotland with a new diagnosis of orbital cellulitis between November 2011 and October 2012 were included in the study. Using the 2011 Scottish census, the incidence per 100 000 of the population was calculated.6 Caldicott approval and ethics permission were granted.

In order to be included in the study, patients had to display lid swelling and erythema in addition to signs of orbital involvement, which included relative afferent pupillary defect, globe displacement, external ophthalmoplegia, chemosis and/or evidence of optic nerve compromise. Cases of preseptal cellulitis alone were not included in this study.

We validated our results by comparing the number of reported cases with the number of orbital cellulitis cases reported from Information and Statistics Department Scotland (ISD) across Scotland during the study period.



On average each month 66% of consultants returned their reporting cards. There were 20 reported cases of orbital cellulitis in the study period affecting 6 females and 14 males. In total, 15 of the 20 cases (75%) were in patients aged less than 15 years (paediatric group) and 5 cases (25%) occurred in patients over 17 years (adult group). Within the paediatric group, six children (40%) were less than 5 years, three children (20%) were between 5 and 10 years and six children (40%) were 11–15 years. In 2011, the Scottish population was 916 331 for those aged 0–15 and 4 379 072 for those aged 16 and over.6 The overall minimum incidence of orbital cellulitis in Scotland was 1.6/100 000 population in children 0–15 years and 0.1/100 000 in adults.

Data validation

Within the study period, ISD reported 364 cases in Scotland with International Classification of Diseases (ICD) 10 code H05.0 (unpublished data, supplied to us by ISD). ICD 10 code H05.0 is acute inflammation of the orbit that includes abscess, cellulitis, periostitis, osteomyelitis and tendonitis of orbit. Preseptal cellulitis, which was an exclusion criterion of our study, has no specific code and therefore will often be inaccurately coded as H05.0. Within one health board that was more fully investigated for data validation, during the study period there were 22 cases coded as H05.0. Review of case notes confirmed that only two (10%) of these were orbital cellulitis, the remainder being preseptal cellulitis involvement. Extrapolating this finding to the Scottish population—assuming 10% of the 364 cases coded as HO5.0 are true orbital cellulitis, then 20 reported cases seems reasonable (taking into account a 66% reporting rate).


In total, 7 out of 15 children (47%) had a preceding upper respiratory tract infection and sinus involvement was confirmed by CT scan in 13 of 15 (87%) children. In comparison, no adult had any antecedent respiratory infection; one adult was immunocompromised due to haematological malignancy and two adults had preceding trauma. One had preceding facial trauma and one had an intraorbital foreign body and endophthalmitis. Both trauma patients had evidence of sinus involvement on CT, although it is not known whether this was infective sinusitis or related to the trauma. One other adult had radiological evidence of asymptomatic sinus disease.

Pathogens were isolated in 70% of cases (table 1). Streptococci species and Haemophilus influenzae were the most frequently isolated pathogens in children and polymicrobial infection was common. In comparison, single organisms were isolated in adults.

Table 1

Microbiology of isolated pathogens


All patients received intravenous antibiotics on admission, and there were no obvious differences between the choices of empirical treatment between the groups. Dual therapy with a third-generation cephalosporin and flucloxacillin was the most common empirical therapy in nine cases followed by triple therapy with addition of metronidazole in four cases. One adult and child were treated empirically with co-amoxiclav monotherapy. Five patients received other dual or triple therapy with various antibiotic classes.

All patients underwent CT imaging. Twelve children (80%) and three adults (60%) had an abscess identified on CT scanning. Within the paediatric group, five (33%) children had orbital abscesses, five (33%) had subperiosteal abscesses (SPA) and two (13%) had both orbital and SPAs. The three adults had orbital abscesses only.

All children with abscesses and two adults with abscesses underwent surgical intervention. One adult with an orbital abscess had no surgery and recovered with antibiotics alone. Of those who were surgically drained, there was variation in the extent of drainage received; external drainage of abscess alone in seven patients (47%), sinus surgery alone in three patients (20%) and combined drainage of abscess and sinus surgery in four patients (27%).

Two children required multiple external drainage procedures one of whom initially had endoscopic drainage of abscess and later required external drainage.


All children completely recovered with no serious adverse outcomes. At 3 months, all but one child was discharged. One remained under review because of a tender scar secondary to external drainage surgery. The adult group had more serious morbidity; one patient had intracranial spread of infection requiring neurosurgical input and the patient with preceding intraorbital trauma and endophthalmitis progressed to evisceration. At 3 months, three adults were discharged. Of the two remaining under care, one had undergone evisceration and the other had leukaemia with diffuse orbital and lacrimal infiltrate.


Our data demonstrate the difference in incidence of orbital cellulitis in children and adults within Scotland; 1.6/100 000 in children and 0.1/100 000 in adults. The incidence of orbital cellulitis in California across all age groups in the year 2000 was 3.51 per 100 000.4 In this Californian study, the cases were obtained from a discharge database that relied on clinical coding; this may have led to an overestimation given a proportion of these cases may have been preseptal involvement only.

There are inaccuracies around the ICD-10 code for orbital cellulitis that makes cross-referencing data difficult. Preseptal cellulitis is not a defined clinical entity in ICD-10 classification; therefore, it may be grouped with code H05.0.7 ISD Scotland reported over 300 cases of code H05.0 Scotland wide during the study period. In our unit during the study period, only 10% of cases coded as orbital cellulitis were true cases, the remainder had preseptal involvement only. This emphasises the limitations of clinical coding. To review the case notes of every individual coded as H05.0 throughout Scotland during the study period would have been a large undertaking involving resources from multiple hospitals. Given the high number of false positive results from coding demonstrated within our unit, we felt this would not be a cost- or time-efficient use of resources. Given the monthly reporting rate was 66.4% similar to that of other BOSU studies, we believe we have likely captured the majority of relevant cases.5

An Australian paper reported 52 cases of orbital cellulitis within a 4-year study period of which 34 cases were less than 16 years of age.1 The peak incidence was in the 0–10 age group and a progressive decline in incidence occurred with increasing age.1 A study from the Middle East reported an approximately equal incidence within adults and children; a high incidence of preceding trauma accounted for the large number of adults affected.3 Similar to other studies, we identified a higher incidence of orbital cellulitis among males within our paediatric group with a roughly 2:1 male to female ratio although the reason for the male predilection remains unclear.1 ,2

Although the number of adults in our series is small, patterns can be seen in the differences in aetiology between children and adults. Orbital cellulitis most commonly occurs from direct extension from the paranasal sinuses in children.2 ,8 ,9 Children are more commonly affected than adults due to the incomplete immunological development and underdevelopment of the paranasal sinuses, septum and thin lamina paprecea.1 ,8 In our series, 47% of children had a preceding respiratory tract infection with 87% of children having radiological evidence of sinus disease.

In contrast, three adults in this series had significant prior or coexisting medical history: immunosuppression due to chronic lymphocytic leukaemia, orbital trauma and facial trauma. Various case reports reinforce the vulnerability of immunosuppressed patients to infection and opportunistic pathogens in the context of orbital cellulitis and rhinosinusitis.10–12 Trauma is an important aetiological factor in the development of orbital cellulitis, which may occur as a consequence of intraorbital foreign body.3 ,13 ,14 Orbital fractures can predispose to orbital cellulitis by allowing direct spread of pathogens from the sinuses to the orbit; in addition, any resultant ischaemia predisposes to anaerobic infections.15

Staphylococcus and Streptococcus species are the most commonly recorded pathogens in orbital cellulitis;1–3 ,8 ,16 however, our study had a very low yield of Staphylococcus. Streptococcus species were the most commonly isolated pathogen in children. Two children with Streptococcus pyogenes isolated from orbital aspirate had systemic evidence of Scarlet Fever. There are a few historic reports of subperiosteal abscess in patients with Scarlet Fever; however, a thorough PubMed search revealed no recent associations.17 ,18 The last recorded outbreak of Scarlet Fever in the UK was in Lancashire in 2009 also with no cases of orbital cellulitis.19 Prior to the introduction of the Haemophilus influenzae type B (Hib) vaccine, Hib was the most common organism causing orbital cellulitis.1 ,2 ,9 ,11 This particular species was extremely virulent and often spread to cause meningitis, cavernous sinus thrombosis and high mortatlity.9 A recent British national seroprevalence study has confirmed excellent Hib control in children up to 10 years of age and control of Hib disease in the England and Wales is at an all-time high.20 ,21 However interesting, the high number of H. influenzae isolated from children in this group is limited as we do not know the strains identified or the immunisation status of the children. Nonetheless, it emphasises the need for continued surveillance of colonisation and disease patterns.

In this observational study, all patients received intravenous antimicrobial therapy; as in other studies, third-generation cephalosporins and flucloxacillin were most commonly used.1–3 This is despite evidence that orbital cellulitis can be successfully treated in adults and children with empirical oral ciprofloxacin and clindamycin.22 Oral ciprofloxacin and clindamycin have similar bioavailability to the intravenous preparations and provide broad spectrum of activity against gram-positive and gram-negative bacteria.22 In patients who can tolerate oral antibiotics, the advantages include rapid delivery of the first dose, less disruption to delivery and prevention of complications of intravenous cannulation.22

In our series, 14 out of 15 patients with orbital or SPA underwent surgical intervention. Garcia and Harris have shown that under specific conditions, medical management alone can be a safe and effective way to deal with SPAs in patients less than 9 years of age.23 Evidence suggests increasing complexity of pathogens in SPA in those greater than 9 years old makes medical management more challenging and may demand early surgical intervention.24 Surgical management is generally advised in patients 9 years or over, the presence of frontal sinusitis, non-medial location of SPA, suspicion of anaerobic infection, recurrence of SPA after previous drainage, evidence of chronic sinusitis, acute optic nerve compromise and coexisting dental infection.23 ,24 When surgical intervention is required, there are variations in technique; external drainage of abscess was the most preferred option in our series with some surgeons performing sinus drainage alone. Evidence suggests that for larger abscesses combined sinus and SPA drainage is associated with improved outcomes.25

Two patients in this series had significant complications; one patient with intracranial spread of abscess and one patient requiring evisceration. Since the advent and widespread use of antibiotics, orbital cellulitis has had a low incidence of ocular and systemic complications.1–3 22–25


To our knowledge, we present the largest series of orbital cellulitis from the UK. Orbital cellulitis occurs more commonly in children than adults; sinusitis is the major preceding factor in children and adults are more likely to have trauma and comorbidities. Within Scotland, intravenous antibiotics are used invariably in adults and children and despite early empirical treatment most needed surgery. The current practice in Scotland is to surgically manage all paediatric SPAs and orbital abscesses, even in those less than 9 years of age. Streptococcus species are the most common pathogens in children and H. influenzae appears to be emerging as a frequent pathogen. Respiratory tract organisms are less predictable pathogens in adults, depending on preceding trauma and comorbidities.


We would like to thank the Ross Foundation for funding and the consultants who reported cases.



  • Contributors All authors met the ICMJE criteria for authorship.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.