- Email: [email protected]
Preseptal and orbital cellulitis in childhood
Preseptal and Orbital Cellulitis in Childhood A Changing Microbiologic Spectrum Sean P. Donahue, MD, PhD,1–3 Gary Schwartz, MD2,4 Objective: The authors sought to determine whether the microbiologic spectrum of preseptal and orbital cellulitis had changed over the past decade. Design: A retrospective chart review of all inpatient and outpatient children with an ICD-9 diagnosis of preseptal or orbital cellulitis seen at Vanderbilt University Medical Center since the introduction of the Haemophilus influenzae type-B (HiB) vaccine (1986 –1996). Main Outcome Measures: Blood and abscess cultures from children with preseptal and orbital cellulitis were tabulated. Results: During this period, 70 cases of preseptal cellulitis were seen. Blood cultures were obtained in 59 cases; only 6 were positive. Five cultures grew Streptococcus species. The one positive H. influenzae culture occurred in 1987 in a child who did not receive the HiB vaccine. There have been no new patients with preseptal cellulitis and H. influenzae bacteremia at Vanderbilt for 10 years. There were ten cases of orbital cellulitis, of which blood or abscess or both were cultured in eight. Six cases had positive cultures. Four cultures grew Streptococcus species. The other two grew H. influenzae and mixed H. influenzae/grampositive cocci. Conclusion: The incidence of hemophilus-associated bacteremia in patients with preseptal cellulitis has decreased dramatically over the past 10 years. Streptococcus species now are the predominant cause. Orbital cellulitis due to H. influenzae may still occur, but it is much less likely. A more conservative approach to the diagnosis and management of preseptal and orbital cellulitis may be warranted. Ophthalmology 1998;105:1902–1906 Orbital cellulitis in young children is a potentially sightthreatening emergency. Orbital cellulitis typically occurs because of either direct extension of preseptal cellulitis through the orbital septum, hematogenous seeding, or from direct extension from the paranasal sinuses. Aggressive organisms such as Haemophilus sp. can spread into the meninges and cause cavernous sinus thrombosis, meningitis, death, or severe neurologic impairment. This
Originally received: October 10, 1997. Revision accepted: May 19, 1998. Manuscript no. 97527. 1 Department of Ophthalmology, Vanderbilt University Medical Center, Nashville, Tennessee. 2 Department of Pediatrics, Vanderbilt University Medical Center, Nashville, Tennessee. 3 Department of Neurology, Vanderbilt University Medical Center, Nashville, Tennessee. 4 Department of Emergency Medicine, Vanderbilt University Medical Center, Nashville, Tennessee. Presented at the American Academy of Ophthalmology annual meeting, San Francisco, California, October 1997. Supported in part by a grant from Research to Prevent Blindness, Inc., New York, New York. Neither author has any proprietary interest in any product described in this article nor in any competing product. Address correspondence to Sean P. Donahue, MD, PhD, Department of Ophthalmology & Visual Sciences, 1215 21st Avenue, South, Nashville, TN 37232-8808.
1902
has led to the classic teaching that orbital cellulitis must be treated aggressively, especially when it occurs in young children. Preseptal cellulitis is not quite as urgent. However, rapid progression of preseptal cellulitis to orbital cellulitis can occur, and similar results are often seen, especially in young children. Before 1985, H. influenzae disease struck nearly 1 in 200 children younger than the age of 5 years.1 It caused nearly 70% of cases of childhood bacterial meningitis and left victims devastated.1 The H. influenzae type-B vaccine was first given to U.S. children in 1985, licensed for use in the United States in December 1987 for children older than 18 months of age and for children as young as 2 months of age in October 1990. It reduced substantially the incidence of hemophilus-associated meningitis, epiglottitis, buccal cellulitis, and otitis.2– 4 A recent New England Journal of Medicine article argues that there now are more than 7000 fewer cases of pediatric meningitis each year than in 1986.1 This massive decrease is arguably the most important advance in pediatric infectious disease since the elimination of polio. However, ophthalmology textbooks5 and current review articles6 still teach that Haemophilus sp. are a dangerous cause of preseptal and orbital cellulitis in young children. In light of the HiB vaccine-associated decrease in Haemophilus pathology, we thought that a reinvestigation of this classic teaching was warranted. This study is a continuation
Donahue and Schwartz 䡠 Bacteriology of Preseptal and Orbital Cellulitis Table 1. Preseptal and Orbital Cellulitis before and after HiB Year
Study
Relationship to HiB vaccine
Preseptal No.
No. H. influenzae
Orbital No.
No. H. influenzae
1978 1978 1994 1995 1998
Smith8 Gellady9 Harris15* Juan16 Donahue
Before Before Before and after Before and after After
37 45 N/A N/A 70
4/4 12/15 N/A N/A 1/6†
2 N/A 31 16 10
0/0 N/A 2/22 0/5 2/5†
N/A ⫽ not studied. * Subperiosteal abscess cultures only. † Children not immunized against H. influenzae.
of a previous report of 44 patients with preseptal cellulitis that showed H. influenzae is uncommon.7
Methods This study was a retrospective chart review. All inpatients and outpatients (younger than 18 years of age) seen at Vanderbilt University Medical Center who were given an ICD-9 diagnosis corresponding to preseptal cellulitis (373.13) or orbital cellulitis (376.01) were studied. We chose patients seen during the 11-year period from 1986 to 1996, as this followed the introduction of the HiB vaccine in 1985. We reviewed charts with specific attention paid to whether the cellulitis was orbital or performance. We excluded patients with preseptal cellulitis secondary to a chalazion. We also determined whether blood cultures had been obtained and the results of the cultures as well as the performance of lumbar puncture. We required documentation of ophthalmologic consultation to confirm the diagnosis of orbital cellulitis. For patients with orbital cellulitis, we specifically looked for radiologic evidence of subperiosteal or orbital abscesses and whether surgical drainage was performed. Finally, we determined whether the immunization against HiB had been specifically documented in the patient’s hospital chart (“immunizations up to date” was not adequate evidence of HiB immunization).
Results Preseptal Cellulitis There were 70 pediatric cases of preseptal cellulitis at Vanderbilt University Medical Center in the 11-year period from 1986 to 1996. Patients ranged in age from 2 months to 17 years. Although more than half of the patients had documentation that immunizations were up to date, only five had clear documentation in the chart of being immunized for Haemophilus type-B. Blood cultures were obtained in 59 (84%) of the 70 patients. These cultures were positive in only six cases (10%). Five blood cultures grew Streptococcus sp. (S. pneumonia, 2; group A Streptococcus, 3). One blood culture was positive for H. influenzae. This was in a 16month-old child who was seen in 1987 and had not been vaccinated with HiB.
Orbital Cellulitis Ten cases of orbital cellulitis documented by ophthalmologic consultation were seen during this 11-year period. All ten
patients had either blood or abscess cultures performed. Six patients had either orbital or subperiosteal abscesses as shown by either computed tomography or orbital magnetic resonance imaging. Of these, five underwent surgical drainage. All five had positive cultures from the abscess. Organisms found were Streptococcus sp., S. angiosis, S. pneumoniae, H. influenzae, and mixed H. influenzae/gram-positive cocci). The H. influenzae organisms were not typed in these individuals. Blood cultures were obtained in six of the ten individuals with documented orbital cellulitis. Only one patient had a positive blood culture. The blood culture grew alpha streptococcal species. Neither patient with a positive H. influenzae culture harbored extraorbital manifestations of disease other than an associated sinusitis.
Discussion Before the advent of the HiB vaccine, H. influenzae type-B was potentially the most feared organism in pediatric infectious disease. H. influenzae invasive disease affected nearly 1 in 200 children younger than the age of 5 years, caused 70% of bacterial meningitis, and produced devastating neurologic sequelae in approximately 3000 U.S. children each year.1 A study from this institution covering a 5-year period ending 20 years ago introduced Haemophilus sp. as the most common cause of preseptal and orbital cellulitis in children (Table 1).8 Thirty-nine patients with preseptal (37) or orbital (2) cellulitis were studied retrospectively. Although only four patients in that series had positive blood cultures, all grew H. influenzae. All four of these patients with positive blood cultures for Haemophilus were younger than 36 months of age, and these represented the only patients with positive blood cultures. A second study of 45 patients with preseptal cellulitis also supported the conclusions from the early Vanderbilt study.9 In this study, also performed in the 1970s, 15 patients had positive blood cultures and H. influenzae was the causative organism in 12 of the 15 patients. A University of Michigan study of 137 pediatric cases of preseptal and orbital cellulitis conducted before the HiB vaccine also found H. influenzae to be the most prevalent organism in each.10 H. influenzae type-B is now nearly nonexistent as a pathogen in childhood epiglottitis, meningitis, buccal cellulitis, and otitis.11–14 This corresponds closely to the introduction of the HiB vaccine in 1985 and its wide-
1903
Ophthalmology Volume 105, Number 10, October 1998 spread use since 1990. Whether the 95% decline in invasive Haemophilus disease of children14 applies to preseptal and orbital cellulitis, however, has been unclear. A study overlapping the introduction of the HiB vaccine and evaluating only patients with subperiosteal abscesses recently was performed by Harris.15 In this study, charts of 31 pediatric patients with orbital or subperiosteal abscesses seen between 1977 and 1992 were reviewed. Only one of these patients had H. influenzae cultured from the abscess. Similar results were reported from five children with culture-positive orbital cellulitis in Taiwan.16 The results of our study allow us to extend the conclusions regarding the prevalence of H. influenzae type-B infections to include preseptal cellulitis. Our patients (6 of 59) with preseptal cellulitis had a similar incidence of documented bacteremia; however, the bacteremia was caused by H. influenzae in only one of six cases. This case occurred more than 10 years ago in a patient who had not been immunized with HiB. Similarly, of our ten patients with orbital cellulitis, only two were documented to have H. influenzae as the causative organism. These two patients were 5 and 12 years of age and likely were not vaccinated because the vaccine was not available when they were infants. The age range of children with orbital cellulitis also has changed. In our series, only one patient was younger than 3 years of age, and he did not have an orbital or subperiosteal abscess. Similar conclusions were reached by Barone and Aiuto17 in a retrospective review of 134 patients with preseptal and orbital cellulitis published when this manuscript was in review. In their study, based in New York, H. influenzae type-B had not been identified as a causative agent in preseptal cellulitis since July 1987. Streptococcus sp. were predominant in cases of preseptal cellulitis. These authors found one patient with orbital cellulitis due to nontypeable H. influenzae. This additional report suggests that the change in bacteriology of preseptal and orbital cellulitis has occurred not only in the southeastern United States but instead is more typical of a stunning nationwide change in the bacteriology of invasive childhood infectious disease that has occurred since the introduction of the HiB vaccine. Past teaching encouraged aggressive management of young children with preseptal cellulitis,6 including hospitalization and intravenous antibiotics. With the advent of high-quality computed tomographic imaging, the use of intramuscular ceftriaxone,18,19 and the results of this study, it may be prudent to reassess the standard treatment for both preseptal and orbital cellulitis in children. Intravenous antibiotics and lumbar puncture definitely are not indicated in the absence of symptoms of bacteremia and meningitis.7,20 Children with orbital signs should have orbital imaging (computed tomography) to rule out subperiosteal abscesses or noninfectious causes of orbital cellulitis.21–23 Management of subperiosteal abscesses is controversial24 –27 and is not the subject of this article. Group A Streptococcus sp. are becoming more prevalent as a cause of bacteremia in children,28 and our study supports this finding. Empiric therapy should be directed toward these organisms. Inpatient
1904
management with cefuroxine (Zinacef; Glaxo Pharmaceuticals, Research Triangle Park, NC) 100 mg/kg per day intravenously in every 8-hour dosing would provide adequate parenteral coverage. Outpatient management with intramuscular ceftriaxone may be adequate for mild cases if close follow-up can be ensured. Older children with mild preseptal cellulitis can likely be treated with Augmentin (250 mg orally four times a day; Beecham Laboratories) and followed. It is unclear why the incidence of H. influenzae-associated orbital and periorbital disease has fallen so sharply when the majority of children still remain unimmunized against H. influenzae type-B. Our review likely underestimated the true incidence of children having received the HiB vaccine. However, the decrease in incidence of other HiB-associated diseases of childhood has fallen dramatically.1– 4,11–14 It is likely that our observed decrease represents a combination of individual vaccination along with a more widespread “herd” immunity, in which immunization has brought the levels of H. influenzae below a level at which it can be sustained as a pathogen. A similar phenomenon occurred during immunizations for other childhood diseases such as smallpox and polio. However, it is also possible that a natural decrease in Haemophilus-associated diseases was already occurring before the onset of the HiB vaccine, and this decline simply was accelerated by vaccination. The HiB vaccine is effective only against the invasive B serotype; less-invasive serotypes may cause infection but are not likely to progress to meningitis. The H. influenzae organisms cultured in this series were not serotyped, so it is possible that the true incidence of HiB disease is even less than reported here. There are several other limitations to this study. It is possible that the 11-year period without a positive blood culture for H. influenzae in the setting of preseptal cellulitis is not entirely because of the lack of H. influenzae bacteremia. Biases caused by earlier intervention, pretreatment as an outpatient with third-generation cephalosporins, and the failure to obtain blood cultures in patients with less-severe cellulitis could all prevent H. influenzae bacteremia from being detected. However, we think this is unlikely because most patients in the middleTennessee area diagnosed with preseptal cellulitis during this period received blood cultures, pediatricians here remain relatively cautious, and H. influenzae tends to present with a more severe cellulitis than do other bacteria. Finally, a chart review based simply on ICD-9 codes may be biased in that it would miss patients in whom more severe disease would develop, such as meningitis. To investigate this last potential bias, we reviewed all cases of ICD-9 coded H. influenzae childhood meningitis from 1986 to 1996 at Vanderbilt Medical Center and found no cases with preseptal or orbital involvement. In conclusion, use of the HiB vaccine has brought with it a substantial decrease in the incidence of preseptal and orbital cellulitis, mirroring the decrease in other Haemophilus-associated diseases of childhood. A more conservative approach to the treatment of preseptal and orbital cellulitis in childhood, possibly including outpa-
Donahue and Schwartz 䡠 Bacteriology of Preseptal and Orbital Cellulitis tient management with long-acting third-generation cephalosporins in mild cases, is now probably warranted.
References 1. Schuchat A, Robinson K, Wenger JD, et al. Bacterial meningitis in the United States in 1995. Active Surveillance Team. N Engl J Med 1997;337:970 – 6. 2. Broadhurst LE, Erickson RL, Kelley PW. Decreases in invasive Haemophilus influenzae diseases in U S Army children, 1984 through 1991. JAMA 1993;269:227–31. 3. Adams WG, Deaver KA, Cochi SL, et al. Decline of childhood Haemophilus influenzae type b (Hib) disease in the Hib vaccine era. JAMA 1993;269:221– 6. 4. Murphy T, White KE, Pastor P, et al. Declining incidence of Haemophilus influenzae type b disease since introduction of vaccination. JAMA 1993;269:246 – 8. 5. Cullom RD, Jr, Chang B, eds. The Wills eye manual: office and emergency room diagnosis and treatment of eye disease, 2nd ed. Philadelphia: Lippincott, 1994;159 –162. 6. Powell KR. Orbital and periorbital cellulitis. Pediatr Rev 1995;16:163–7. 7. Schwartz GR, Wright SW. Changing bacteriology of periorbital cellulitis. Ann Emerg Med 1996;28:617–20. 8. Smith TF, O’Day D, Wright PF. Clinical implications of preseptal (periorbital) cellulitis in childhood. Pediatrics 1978; 62:1006 –9. 9. Gellady AM, Shulman ST, Ayoub EM. Periorbital and orbital cellulitis in children. Pediatrics 1978;61:272–7. 10. Jackson K, Baker SR. Periorbital cellulitis. Head Neck Surg 1987;9:227–34. 11. Frantz TD, Rasgon BM. Acute epiglottitis: changing epidemiologic patterns. Otolaryngol Head Neck Surg 1993;109(3 Pt 1):457– 60. 12. Buchanan GA, Darville T. Impact of immunization against Haemophilus influenzae type b (HIB) on the incidence of HIB meningitis treated at Arkansas Children’s Hospital. South Med J 1994;87:38 – 40. 13. Hickerson SL, Kirby RS, Wheeler JG, Schutze GE. Epiglottitis: a 9-year case review. South Med J 1996;89:487– 90. 14. Progress toward elimination of Haemophilus influenzae type b disease among infants and children—United States, 1993–1994. MMWR Morb Mortal Wkly Rep 1995;44: 545–50.
15. Harris GJ. Subperiosteal abscess of the orbit. Age as a factor in the bacteriology and response to treatment. Ophthalmology 1994;101:585–95. 16. Juan CM, Chiu CH, Huang YC, Lin TY. Orbital cellulitis in children: clinical analysis of 16 cases. Chung Hua Min Kuo Hsiao Erh KoI Hsueh Hui Tsa Chih 1995;36:286 – 8. 17. Barone SR, Aiuto LT. Periorbital and orbital cellulitis in the Haemophilus influenzae vaccine era. J Pediatr Ophthalmol Strabismus 1997;34:293– 6. 18. Einhorn M, Fliss DM, Leiberman A, Dagan R. Otolaryngology and infectious disease team approach for outpatient management of serious pediatric infections requiring parenteral antibiotic therapy. Int J Pediatr Otorhinolaryngol 1992;24: 245–51. 19. Dagan R, Phillip M, Watemberg NM, Kassis I. Outpatient treatment of serious community-acquired pediatric infections using once daily intramuscular ceftriaxone. Pediatr Infect Dis J 1987;6:1080 – 4. 20. Ciarallo LR, Rowe PC. Lumbar puncture in children with periorbital and orbital cellulitis. J Pediatr 1993;122:355–9. 21. Pollard ZF. Acute rectus muscle palsy in children as a result of orbital myositis. J Pediatr 1996;128:230 –3. 22. Shields JA, Shields CL, Suvarnamani C, et al. Retinoblastoma manifesting as orbital cellulitis. Am J Ophthalmol 1991;112: 442–9. 23. Kelly SP, Lloyd IC, Anderson H, et al. Solitary extramedullary plasmacytoma of the maxillary antrum and orbit presenting as acute bacterial orbital cellulitis. Br J Ophthalmol 1991; 75:438 –9. 24. Rubin SE, Rubin LG, Zito J, et al. Medical management of orbital subperiosteal abscess in children. J Pediatr Ophthalmol Strabismus 1989;26:21–7. 25. Noel LP, Clarke WN, MacDonald N. Clinical management of orbital cellulitis in children. Can J Ophthalmol 1990;25:11– 6. 26. Catalano RA, Smoot CN. Subperiosteal orbital masses in children with orbital cellulitis: time for a reevaluation? J Pediatr Ophthalmol Strabismus 1990;27:141–2. 27. Souliere CR Jr, Antoine GA, Martin MP, et al. Selective non-surgical management of subperiosteal abscess of the orbit: computerized tomography and clinical course as indication for surgical drainage. Int J Pediatr Otorhinolaryngol 1990; 19:109 –19. 28. Doctor A, Harper MB, Fleisher GR. Group A beta-hemolytic streptococcal bacteremia: historical overview, changing incidence, and recent association with varicella. Pediatrics 1995; 96(3 Pt 1):428 –33.
Discussion by John P. Whitcher, MD, MPH As the authors have clearly documented in this retrospective study, the widely accepted dictum that Haemophilus sp. is the most common cause of preseptal and orbital cellulitis in young children must be reappraised. Since the introduction of Haemophilus influenzae type-B (HiB) vaccine in the United States in 1985, it is a fact that cases of nonophthalmic Haemophilus-associated infections in childhood such as epiglottitis, buccal cellulitis, and otitis media have been strikingly reduced. This study, however, is the first to document a corresponding
Francis I. Proctor Foundation, 95 Kirkham Street, San Francisco, CA 94143-0944.
reduction in ocular complications associated with this virulent pathogen. Of the 70 cases of preseptal cellulitis seen at Vanderbilt University Medical Center from 1986 to 1996, blood cultures obtained from 59 of the children were positive in 6 cases, only 1 of which was positive for H. influenzae, and that was in an individual who had not received the HiB vaccine. In the ten cases of orbital cellulitis during the same period, eight of which were cultured, six cultures were positive with only two growing H. influenzae or H. influenzae mixed with gram-positive cocci. In both groups of patients, all of the other positive cultures grew Streptococcus sp. This is a dramatic shift in the etiologic diagnosis of preseptal and orbital cellulitis in children in 1
1905
Ophthalmology Volume 105, Number 10, October 1998 decade that appears undoubtedly to be related to the introduction of the HiB vaccine. The study is somewhat limited, however, by the fact that it is retrospective and that prospective information regarding the immunization status of the children could, therefore, not be obtained. The causal relationship between the introduction of HiB vaccine and the reduction of Haemophilus sp. ocular infections is also weakened by the fact that the authors essentially are comparing their group of patients to historic control subjects. The unanswered question still remains of why the incidence of H. influenzae-associated orbital and preseptal disease has fallen so precipitously while the majority of children in
the United States still have not received the HiB vaccine. Other confounding factors may be playing at least a partial role in this dramatic epidemiologic shift. Undoubtedly, this is a very important article because it is the first to document the decline in Haemophilus-associated preseptal and orbital cellulitis in children. I would hope that this important finding would lead to the initiation of a large prospective study to investigate other confounding factors that may be responsible for this decline. Such factors might include fewer predisposing childhood illnesses, early aggressive treatment, and widespread use of new generations of broad-spectrum antibiotics in the pediatric population.
Historical Image Items from the E.B. Meyrowitz Catalog, ca. 1894.*
*Courtesy of the Museum of Ophthalmology, Foundation of the American Academy of Ophthalmology, San Francisco, California.
1906
Originally received: October 10, 1997. Revision accepted: May 19, 1998. Manuscript no. 97527. 1 Department of Ophthalmology, Vanderbilt University Medical Center, Nashville, Tennessee. 2 Department of Pediatrics, Vanderbilt University Medical Center, Nashville, Tennessee. 3 Department of Neurology, Vanderbilt University Medical Center, Nashville, Tennessee. 4 Department of Emergency Medicine, Vanderbilt University Medical Center, Nashville, Tennessee. Presented at the American Academy of Ophthalmology annual meeting, San Francisco, California, October 1997. Supported in part by a grant from Research to Prevent Blindness, Inc., New York, New York. Neither author has any proprietary interest in any product described in this article nor in any competing product. Address correspondence to Sean P. Donahue, MD, PhD, Department of Ophthalmology & Visual Sciences, 1215 21st Avenue, South, Nashville, TN 37232-8808.
1902
has led to the classic teaching that orbital cellulitis must be treated aggressively, especially when it occurs in young children. Preseptal cellulitis is not quite as urgent. However, rapid progression of preseptal cellulitis to orbital cellulitis can occur, and similar results are often seen, especially in young children. Before 1985, H. influenzae disease struck nearly 1 in 200 children younger than the age of 5 years.1 It caused nearly 70% of cases of childhood bacterial meningitis and left victims devastated.1 The H. influenzae type-B vaccine was first given to U.S. children in 1985, licensed for use in the United States in December 1987 for children older than 18 months of age and for children as young as 2 months of age in October 1990. It reduced substantially the incidence of hemophilus-associated meningitis, epiglottitis, buccal cellulitis, and otitis.2– 4 A recent New England Journal of Medicine article argues that there now are more than 7000 fewer cases of pediatric meningitis each year than in 1986.1 This massive decrease is arguably the most important advance in pediatric infectious disease since the elimination of polio. However, ophthalmology textbooks5 and current review articles6 still teach that Haemophilus sp. are a dangerous cause of preseptal and orbital cellulitis in young children. In light of the HiB vaccine-associated decrease in Haemophilus pathology, we thought that a reinvestigation of this classic teaching was warranted. This study is a continuation
Donahue and Schwartz 䡠 Bacteriology of Preseptal and Orbital Cellulitis Table 1. Preseptal and Orbital Cellulitis before and after HiB Year
Study
Relationship to HiB vaccine
Preseptal No.
No. H. influenzae
Orbital No.
No. H. influenzae
1978 1978 1994 1995 1998
Smith8 Gellady9 Harris15* Juan16 Donahue
Before Before Before and after Before and after After
37 45 N/A N/A 70
4/4 12/15 N/A N/A 1/6†
2 N/A 31 16 10
0/0 N/A 2/22 0/5 2/5†
N/A ⫽ not studied. * Subperiosteal abscess cultures only. † Children not immunized against H. influenzae.
of a previous report of 44 patients with preseptal cellulitis that showed H. influenzae is uncommon.7
Methods This study was a retrospective chart review. All inpatients and outpatients (younger than 18 years of age) seen at Vanderbilt University Medical Center who were given an ICD-9 diagnosis corresponding to preseptal cellulitis (373.13) or orbital cellulitis (376.01) were studied. We chose patients seen during the 11-year period from 1986 to 1996, as this followed the introduction of the HiB vaccine in 1985. We reviewed charts with specific attention paid to whether the cellulitis was orbital or performance. We excluded patients with preseptal cellulitis secondary to a chalazion. We also determined whether blood cultures had been obtained and the results of the cultures as well as the performance of lumbar puncture. We required documentation of ophthalmologic consultation to confirm the diagnosis of orbital cellulitis. For patients with orbital cellulitis, we specifically looked for radiologic evidence of subperiosteal or orbital abscesses and whether surgical drainage was performed. Finally, we determined whether the immunization against HiB had been specifically documented in the patient’s hospital chart (“immunizations up to date” was not adequate evidence of HiB immunization).
Results Preseptal Cellulitis There were 70 pediatric cases of preseptal cellulitis at Vanderbilt University Medical Center in the 11-year period from 1986 to 1996. Patients ranged in age from 2 months to 17 years. Although more than half of the patients had documentation that immunizations were up to date, only five had clear documentation in the chart of being immunized for Haemophilus type-B. Blood cultures were obtained in 59 (84%) of the 70 patients. These cultures were positive in only six cases (10%). Five blood cultures grew Streptococcus sp. (S. pneumonia, 2; group A Streptococcus, 3). One blood culture was positive for H. influenzae. This was in a 16month-old child who was seen in 1987 and had not been vaccinated with HiB.
Orbital Cellulitis Ten cases of orbital cellulitis documented by ophthalmologic consultation were seen during this 11-year period. All ten
patients had either blood or abscess cultures performed. Six patients had either orbital or subperiosteal abscesses as shown by either computed tomography or orbital magnetic resonance imaging. Of these, five underwent surgical drainage. All five had positive cultures from the abscess. Organisms found were Streptococcus sp., S. angiosis, S. pneumoniae, H. influenzae, and mixed H. influenzae/gram-positive cocci). The H. influenzae organisms were not typed in these individuals. Blood cultures were obtained in six of the ten individuals with documented orbital cellulitis. Only one patient had a positive blood culture. The blood culture grew alpha streptococcal species. Neither patient with a positive H. influenzae culture harbored extraorbital manifestations of disease other than an associated sinusitis.
Discussion Before the advent of the HiB vaccine, H. influenzae type-B was potentially the most feared organism in pediatric infectious disease. H. influenzae invasive disease affected nearly 1 in 200 children younger than the age of 5 years, caused 70% of bacterial meningitis, and produced devastating neurologic sequelae in approximately 3000 U.S. children each year.1 A study from this institution covering a 5-year period ending 20 years ago introduced Haemophilus sp. as the most common cause of preseptal and orbital cellulitis in children (Table 1).8 Thirty-nine patients with preseptal (37) or orbital (2) cellulitis were studied retrospectively. Although only four patients in that series had positive blood cultures, all grew H. influenzae. All four of these patients with positive blood cultures for Haemophilus were younger than 36 months of age, and these represented the only patients with positive blood cultures. A second study of 45 patients with preseptal cellulitis also supported the conclusions from the early Vanderbilt study.9 In this study, also performed in the 1970s, 15 patients had positive blood cultures and H. influenzae was the causative organism in 12 of the 15 patients. A University of Michigan study of 137 pediatric cases of preseptal and orbital cellulitis conducted before the HiB vaccine also found H. influenzae to be the most prevalent organism in each.10 H. influenzae type-B is now nearly nonexistent as a pathogen in childhood epiglottitis, meningitis, buccal cellulitis, and otitis.11–14 This corresponds closely to the introduction of the HiB vaccine in 1985 and its wide-
1903
Ophthalmology Volume 105, Number 10, October 1998 spread use since 1990. Whether the 95% decline in invasive Haemophilus disease of children14 applies to preseptal and orbital cellulitis, however, has been unclear. A study overlapping the introduction of the HiB vaccine and evaluating only patients with subperiosteal abscesses recently was performed by Harris.15 In this study, charts of 31 pediatric patients with orbital or subperiosteal abscesses seen between 1977 and 1992 were reviewed. Only one of these patients had H. influenzae cultured from the abscess. Similar results were reported from five children with culture-positive orbital cellulitis in Taiwan.16 The results of our study allow us to extend the conclusions regarding the prevalence of H. influenzae type-B infections to include preseptal cellulitis. Our patients (6 of 59) with preseptal cellulitis had a similar incidence of documented bacteremia; however, the bacteremia was caused by H. influenzae in only one of six cases. This case occurred more than 10 years ago in a patient who had not been immunized with HiB. Similarly, of our ten patients with orbital cellulitis, only two were documented to have H. influenzae as the causative organism. These two patients were 5 and 12 years of age and likely were not vaccinated because the vaccine was not available when they were infants. The age range of children with orbital cellulitis also has changed. In our series, only one patient was younger than 3 years of age, and he did not have an orbital or subperiosteal abscess. Similar conclusions were reached by Barone and Aiuto17 in a retrospective review of 134 patients with preseptal and orbital cellulitis published when this manuscript was in review. In their study, based in New York, H. influenzae type-B had not been identified as a causative agent in preseptal cellulitis since July 1987. Streptococcus sp. were predominant in cases of preseptal cellulitis. These authors found one patient with orbital cellulitis due to nontypeable H. influenzae. This additional report suggests that the change in bacteriology of preseptal and orbital cellulitis has occurred not only in the southeastern United States but instead is more typical of a stunning nationwide change in the bacteriology of invasive childhood infectious disease that has occurred since the introduction of the HiB vaccine. Past teaching encouraged aggressive management of young children with preseptal cellulitis,6 including hospitalization and intravenous antibiotics. With the advent of high-quality computed tomographic imaging, the use of intramuscular ceftriaxone,18,19 and the results of this study, it may be prudent to reassess the standard treatment for both preseptal and orbital cellulitis in children. Intravenous antibiotics and lumbar puncture definitely are not indicated in the absence of symptoms of bacteremia and meningitis.7,20 Children with orbital signs should have orbital imaging (computed tomography) to rule out subperiosteal abscesses or noninfectious causes of orbital cellulitis.21–23 Management of subperiosteal abscesses is controversial24 –27 and is not the subject of this article. Group A Streptococcus sp. are becoming more prevalent as a cause of bacteremia in children,28 and our study supports this finding. Empiric therapy should be directed toward these organisms. Inpatient
1904
management with cefuroxine (Zinacef; Glaxo Pharmaceuticals, Research Triangle Park, NC) 100 mg/kg per day intravenously in every 8-hour dosing would provide adequate parenteral coverage. Outpatient management with intramuscular ceftriaxone may be adequate for mild cases if close follow-up can be ensured. Older children with mild preseptal cellulitis can likely be treated with Augmentin (250 mg orally four times a day; Beecham Laboratories) and followed. It is unclear why the incidence of H. influenzae-associated orbital and periorbital disease has fallen so sharply when the majority of children still remain unimmunized against H. influenzae type-B. Our review likely underestimated the true incidence of children having received the HiB vaccine. However, the decrease in incidence of other HiB-associated diseases of childhood has fallen dramatically.1– 4,11–14 It is likely that our observed decrease represents a combination of individual vaccination along with a more widespread “herd” immunity, in which immunization has brought the levels of H. influenzae below a level at which it can be sustained as a pathogen. A similar phenomenon occurred during immunizations for other childhood diseases such as smallpox and polio. However, it is also possible that a natural decrease in Haemophilus-associated diseases was already occurring before the onset of the HiB vaccine, and this decline simply was accelerated by vaccination. The HiB vaccine is effective only against the invasive B serotype; less-invasive serotypes may cause infection but are not likely to progress to meningitis. The H. influenzae organisms cultured in this series were not serotyped, so it is possible that the true incidence of HiB disease is even less than reported here. There are several other limitations to this study. It is possible that the 11-year period without a positive blood culture for H. influenzae in the setting of preseptal cellulitis is not entirely because of the lack of H. influenzae bacteremia. Biases caused by earlier intervention, pretreatment as an outpatient with third-generation cephalosporins, and the failure to obtain blood cultures in patients with less-severe cellulitis could all prevent H. influenzae bacteremia from being detected. However, we think this is unlikely because most patients in the middleTennessee area diagnosed with preseptal cellulitis during this period received blood cultures, pediatricians here remain relatively cautious, and H. influenzae tends to present with a more severe cellulitis than do other bacteria. Finally, a chart review based simply on ICD-9 codes may be biased in that it would miss patients in whom more severe disease would develop, such as meningitis. To investigate this last potential bias, we reviewed all cases of ICD-9 coded H. influenzae childhood meningitis from 1986 to 1996 at Vanderbilt Medical Center and found no cases with preseptal or orbital involvement. In conclusion, use of the HiB vaccine has brought with it a substantial decrease in the incidence of preseptal and orbital cellulitis, mirroring the decrease in other Haemophilus-associated diseases of childhood. A more conservative approach to the treatment of preseptal and orbital cellulitis in childhood, possibly including outpa-
Donahue and Schwartz 䡠 Bacteriology of Preseptal and Orbital Cellulitis tient management with long-acting third-generation cephalosporins in mild cases, is now probably warranted.
References 1. Schuchat A, Robinson K, Wenger JD, et al. Bacterial meningitis in the United States in 1995. Active Surveillance Team. N Engl J Med 1997;337:970 – 6. 2. Broadhurst LE, Erickson RL, Kelley PW. Decreases in invasive Haemophilus influenzae diseases in U S Army children, 1984 through 1991. JAMA 1993;269:227–31. 3. Adams WG, Deaver KA, Cochi SL, et al. Decline of childhood Haemophilus influenzae type b (Hib) disease in the Hib vaccine era. JAMA 1993;269:221– 6. 4. Murphy T, White KE, Pastor P, et al. Declining incidence of Haemophilus influenzae type b disease since introduction of vaccination. JAMA 1993;269:246 – 8. 5. Cullom RD, Jr, Chang B, eds. The Wills eye manual: office and emergency room diagnosis and treatment of eye disease, 2nd ed. Philadelphia: Lippincott, 1994;159 –162. 6. Powell KR. Orbital and periorbital cellulitis. Pediatr Rev 1995;16:163–7. 7. Schwartz GR, Wright SW. Changing bacteriology of periorbital cellulitis. Ann Emerg Med 1996;28:617–20. 8. Smith TF, O’Day D, Wright PF. Clinical implications of preseptal (periorbital) cellulitis in childhood. Pediatrics 1978; 62:1006 –9. 9. Gellady AM, Shulman ST, Ayoub EM. Periorbital and orbital cellulitis in children. Pediatrics 1978;61:272–7. 10. Jackson K, Baker SR. Periorbital cellulitis. Head Neck Surg 1987;9:227–34. 11. Frantz TD, Rasgon BM. Acute epiglottitis: changing epidemiologic patterns. Otolaryngol Head Neck Surg 1993;109(3 Pt 1):457– 60. 12. Buchanan GA, Darville T. Impact of immunization against Haemophilus influenzae type b (HIB) on the incidence of HIB meningitis treated at Arkansas Children’s Hospital. South Med J 1994;87:38 – 40. 13. Hickerson SL, Kirby RS, Wheeler JG, Schutze GE. Epiglottitis: a 9-year case review. South Med J 1996;89:487– 90. 14. Progress toward elimination of Haemophilus influenzae type b disease among infants and children—United States, 1993–1994. MMWR Morb Mortal Wkly Rep 1995;44: 545–50.
15. Harris GJ. Subperiosteal abscess of the orbit. Age as a factor in the bacteriology and response to treatment. Ophthalmology 1994;101:585–95. 16. Juan CM, Chiu CH, Huang YC, Lin TY. Orbital cellulitis in children: clinical analysis of 16 cases. Chung Hua Min Kuo Hsiao Erh KoI Hsueh Hui Tsa Chih 1995;36:286 – 8. 17. Barone SR, Aiuto LT. Periorbital and orbital cellulitis in the Haemophilus influenzae vaccine era. J Pediatr Ophthalmol Strabismus 1997;34:293– 6. 18. Einhorn M, Fliss DM, Leiberman A, Dagan R. Otolaryngology and infectious disease team approach for outpatient management of serious pediatric infections requiring parenteral antibiotic therapy. Int J Pediatr Otorhinolaryngol 1992;24: 245–51. 19. Dagan R, Phillip M, Watemberg NM, Kassis I. Outpatient treatment of serious community-acquired pediatric infections using once daily intramuscular ceftriaxone. Pediatr Infect Dis J 1987;6:1080 – 4. 20. Ciarallo LR, Rowe PC. Lumbar puncture in children with periorbital and orbital cellulitis. J Pediatr 1993;122:355–9. 21. Pollard ZF. Acute rectus muscle palsy in children as a result of orbital myositis. J Pediatr 1996;128:230 –3. 22. Shields JA, Shields CL, Suvarnamani C, et al. Retinoblastoma manifesting as orbital cellulitis. Am J Ophthalmol 1991;112: 442–9. 23. Kelly SP, Lloyd IC, Anderson H, et al. Solitary extramedullary plasmacytoma of the maxillary antrum and orbit presenting as acute bacterial orbital cellulitis. Br J Ophthalmol 1991; 75:438 –9. 24. Rubin SE, Rubin LG, Zito J, et al. Medical management of orbital subperiosteal abscess in children. J Pediatr Ophthalmol Strabismus 1989;26:21–7. 25. Noel LP, Clarke WN, MacDonald N. Clinical management of orbital cellulitis in children. Can J Ophthalmol 1990;25:11– 6. 26. Catalano RA, Smoot CN. Subperiosteal orbital masses in children with orbital cellulitis: time for a reevaluation? J Pediatr Ophthalmol Strabismus 1990;27:141–2. 27. Souliere CR Jr, Antoine GA, Martin MP, et al. Selective non-surgical management of subperiosteal abscess of the orbit: computerized tomography and clinical course as indication for surgical drainage. Int J Pediatr Otorhinolaryngol 1990; 19:109 –19. 28. Doctor A, Harper MB, Fleisher GR. Group A beta-hemolytic streptococcal bacteremia: historical overview, changing incidence, and recent association with varicella. Pediatrics 1995; 96(3 Pt 1):428 –33.
Discussion by John P. Whitcher, MD, MPH As the authors have clearly documented in this retrospective study, the widely accepted dictum that Haemophilus sp. is the most common cause of preseptal and orbital cellulitis in young children must be reappraised. Since the introduction of Haemophilus influenzae type-B (HiB) vaccine in the United States in 1985, it is a fact that cases of nonophthalmic Haemophilus-associated infections in childhood such as epiglottitis, buccal cellulitis, and otitis media have been strikingly reduced. This study, however, is the first to document a corresponding
Francis I. Proctor Foundation, 95 Kirkham Street, San Francisco, CA 94143-0944.
reduction in ocular complications associated with this virulent pathogen. Of the 70 cases of preseptal cellulitis seen at Vanderbilt University Medical Center from 1986 to 1996, blood cultures obtained from 59 of the children were positive in 6 cases, only 1 of which was positive for H. influenzae, and that was in an individual who had not received the HiB vaccine. In the ten cases of orbital cellulitis during the same period, eight of which were cultured, six cultures were positive with only two growing H. influenzae or H. influenzae mixed with gram-positive cocci. In both groups of patients, all of the other positive cultures grew Streptococcus sp. This is a dramatic shift in the etiologic diagnosis of preseptal and orbital cellulitis in children in 1
1905
Ophthalmology Volume 105, Number 10, October 1998 decade that appears undoubtedly to be related to the introduction of the HiB vaccine. The study is somewhat limited, however, by the fact that it is retrospective and that prospective information regarding the immunization status of the children could, therefore, not be obtained. The causal relationship between the introduction of HiB vaccine and the reduction of Haemophilus sp. ocular infections is also weakened by the fact that the authors essentially are comparing their group of patients to historic control subjects. The unanswered question still remains of why the incidence of H. influenzae-associated orbital and preseptal disease has fallen so precipitously while the majority of children in
the United States still have not received the HiB vaccine. Other confounding factors may be playing at least a partial role in this dramatic epidemiologic shift. Undoubtedly, this is a very important article because it is the first to document the decline in Haemophilus-associated preseptal and orbital cellulitis in children. I would hope that this important finding would lead to the initiation of a large prospective study to investigate other confounding factors that may be responsible for this decline. Such factors might include fewer predisposing childhood illnesses, early aggressive treatment, and widespread use of new generations of broad-spectrum antibiotics in the pediatric population.
Historical Image Items from the E.B. Meyrowitz Catalog, ca. 1894.*
*Courtesy of the Museum of Ophthalmology, Foundation of the American Academy of Ophthalmology, San Francisco, California.
1906