Fewer than 15 cases of cellulitis after extraocular muscle surgery have been reported in detail. 1,9 Four of these cases were in the pre-antibiotic era.1"* The best incidence data are from Columbia Presbyterian Hospital in New York City; 11 cases of cellulitis, 1 abscess, and 1 wound infection were reported from an experience with 12 263 muscle surgeries for the years 1950 to 1968.10 While this is the single large series published, the only information given about the cases is that all cultures grew Staphylococcus aureus.10 In a recent survey of the Costenbader Alumni Society, the incidence was estimated at between 1 case in 1000 and 1 case in 1900 surgeries.11 Because serious infections are rare, it would be unusual for one surgeon to accumulate a substantial experience with this complication. Thus, cases were collected from members of the American Association for Pediatrie Ophthalmology and Strabismus (AAPO&S) who have assisted us by forming the Periocular Infection Study Group.
Members of AAPO&S were sent a questionnaire which stated "I have seen (fill in blank) patients with orbital cellulitis after strabismus surgery in (fill in blank) years of practice. Case details are available on (fill in blank) patients." The blanks were filled in by the members who also indicated if they were interested in participating in a group study. A standardized work sheet for extracting specific data from the medical record was sent to participants (Table 1). To be included, cellulitis cases had to have patient records providing evidence of infection deep to the conjunctiva12 (Table 2) as well as information on the patient's presentation, findings, treatment, and course. Incomplete records led to the exclusion of approximately 30 cases of mild eyelid infection, several S. aureus cellulitis cases, one case of bilateral Group A ß-hemolytic Streptococcus orbital cellulitis, and one adult case where nasal secretions contacted the eye at the end of the surgery.
Questions on Data Collection Sheet
Subconjunctival abscess cases were included if the patients had increasing pain and swelling and their physicians found a localized collection of pus which contained bacteria. All data were collected retrospectively except for the status of four patients 6 weeks after the infection, which was recorded concurrently.
Of the 419 questionnaires sent, 308 (73%) were returned. Two hundred thirty-eight respondents, or slightly more than half the AAPO&S membership, recalled no case of periocular infection after strabismus surgery in 2 to 40 years of practice (mean, 13 years). For the 289 respondents who listed both the number of cases and their years in practice, a total of 128 infections occurred over 3,983 years in practice or approximately 1 case in 32 years of practice.
From the 34 AAPO&S members who were interested in participating in the case series, 68 cases were submitted. Twenty-five cases of cellulitis and three cases of subconjunctival abscess met the inclusion criteria and were analyzed.
Two cases of cellulitis occurred after separate surgeries in the same patient. Twenty-two cases occurred in the years 1982 to 1993; the other three cases occurred in 1961, 1974, and 1979.
Twenty-one of the 25 patients were less than 13 years old with 10 being in the 2- to 5-year range. Nine patients (36%) had one or more factors felt to predispose them to infection (Table 3). When listed as predisposing factors, the degree of eye rubbing, the degree of unreliability of the family, and the lack of hygiene exceeded the usual experience of the surgeon in his or her practice. Horizontal deviations and thus horizontal muscle surgical procedures predominated (Table 4). Reflecting the more recent occurrence of most of the cases, only 1 of the 19 surgeries where suture type was indicated was done with chromic suture instead of polyglactin 910 (Vicryl) or polyglycolic acid (Dexon). Polyester (Mersilene) was used in one other case. While two-thirds of the surgeries were bilateral, all the infections were unilateral (Table 4). No sutures were adjusted after surgery.
Inclusion Criteria for Cellulitis Cases
Possible Predisposing Factors
Ocular and Systemic Symptoms In Cellulitis Cases
The chemical preparation of the skin included povidoneiodine solution or soap in 13 cases. Aqueous methiolate in combination with povidone-iodine solution, and hexachlorophene were used on the skin in one patient each. Five patients received povidone-iodine solution in the conjunctival cul-de-sac. There was no information on preparation for 10 cases.
Six patients (24%) received intraoperative antibiotics; one received subconjunctival gentamicin after nasal secretions came close to the field while the others received topical antibiotic or steroid-antibiotic combinations. After surgery, a topical steroid-antibiotic combination was prescribed for 19 patients (76%); one patient stated he did not use the medication.
FIGURE: Onset of symptoms in cellulitis patients.
Three patients received subconjunctival steroids at the end of surgery. One patient undergoing a re-operation received 5-fluorouracü subconjunctivally to try to prevent scarring. After surgery, as mentioned above, 19 patients were prescribed a topical steroid as part of a steroidantibiotic combination.
The time at which the patient and/or the family became aware of a problem ranged from 1 to 5 days after surgery (Figure). Prior to the onset of clinical infection, 14 patients (56%) had a routine postoperative visit in which they were felt to have normal postoperative findings.
Swelling and pain were the most common presenting complaints (Table 5).
Diagnosis and Treatment
The tissues involved with the infection varied. Ten patiente (40%) had no clinical signs such as decreased muscle function, proptosis, or pupillary changes which would indicate deeper orbital involvement although two of these patients had orbital edema on computed tomography (CT). Fifteen patients (60%) had one or more orbital signs.
Ten patients had CT scans and two had magnetic resonance imaging (MRI) scans. Most demonstrated edema and proptosis consistent with the clinical findings. Unexpected results included two cases of maxillary sinusitis and one case of pansinusitis as well as the two cases with orbital findings not suspected clinically. No patient had a subperiosteal abscess.
Conjunctiva! cultures were obtained in 22 cases (Table 6). S. aureus was the most common organism. Blood cultures were obtained in 11 cases yielding S. aureus in one case and Streptococcus pneumoniae in another. These two cases had the same organism cultured from the conjunctiva. The one subconjunctival aspirate yielded seven colonies of S. epidermidis.
Oral antibiotic therapy was the initial treatment for 11 cases (44%); only in three cases was the original antibiotic felt to be successful in eliminating the infection. Changing to a different oral antibiotic, adding a second oral antibiotic and topical steroid, and adding oral prednisone for a suture granuloma led to resolution of the condition in three other cases. Worsening of clinical findings in four cases and lack of improvement in one case led to a change in therapy to intravenous (IV) antibiotics.
IV antibiotics were employed in a total of 20 cases (80%) and intramuscular (IM) antibiotic in one case; no parenteral agents were changed because of lack of response (Table 7). Clinical improvement occurred in the first 24 hours on parenteral antibiotics in 10 patients (50% of patients treated). Two patients experienced worsening of edema and one patient developed poor abduction 1 day after IV therapy was initiated; these three patients improved the following day; on the same antibiotics. The amount of time that parenteral therapy was administered varied from 2 to 10 days with an average of 5 days. Oral antibiotics were prescribed for 5 to 21 days after discontinuing IV therapy in 14 cases. Topical therapy was listed for seven patients.
Culture Results in Cellulitis Cases
Eighteen patients (72%) had the postoperative ocular alignment result expected by their surgeons. The unexpected results were no change in the ocular deviation in one patient, undercorrection in two patients, and overcorrection in four patients. In one patient there was mild persistent motility limitation past the period of infection. No patient had unusual conjunctival scarring. Four patients had a subsequent strabismus surgery; one developed orbital cellulitis a second time.
Three patients developed localized subconjunctival abscesses postoperatively rather than diffuse lid or orbit infection (Table 8). No predisposing factors were apparent.
Patients 1 and 2 were started on oral antibiotics in the first week after surgery to treat excess edema and injection over a recessed muscle. After initial improvement, a discrete abscess formed over the new muscle insertion. One patient underwent incision and drainage and a course of IV cefamandole. The other abscess drained spontaneously and finally resolved after 6 weeks of oral cefaclor.
Patient 3 presented 1 week after surgery with increased pain and swelling. An examination 3 days previously had revealed no signs of infection. An antibiotic/steroid ointment was being used. A localized abscess which had formed over the new medial rectus muscle insertion was drained. A corresponding subretinal exúdate was also seen without overlying vitreous cells. No scierai perforation was identified. Because of concern that the infection could have entered the eye, cryotherapy was applied to the sclera in the area of the abscess. There was immediate improvement in pain after the procedures. IV cefotelam was used with complete resolution of the infection.
Antibiotics Used in Cellulitis Cases
Sutures were not removed in these patients. All had the expected alignment result without unusual conjunctival scarring or motility defects.
Periocular infections after strabismus surgery are rare but potentially very serious. Assuming an incidence of one case per 1100 surgeries,10 most strabismus surgeons will not accumulate much experience with managing these infections. Thus, the experience of the AAPO&S members interested in participating was collected and presented here. Because not all AAPO&S members chose to participate in the case series, there are missing data. We do not know if additional severe cases exist but have not been contributed due to lack of records or perhaps to reticence to report a complication. Lack of records prevented inclusion of approximately 30 mild infections that responded to oral antibiotics. Since no member of the study group uses oral antibiotics routinely after strabismus surgery, the experience may be different from the group within AAPO&S that does.
This series, similar to previously published cases,4,5,8-10 had a preponderance of S. aureus among the positive culture results (56%). Since the incidence of positive conjunctival cultures for S. aureus in normal individuals is only 8% to 12%13,14 and one patient was bacteremic with this organism, it seems reasonable to consider it the pathogen in most cases. Other gram-positive organisms such as Streptococcus can cause this infection as seen here and in the case of Wilson and Paul.6 The case of Proteus abscess is the only gram-negative organism that has been reported. It is interesting that Haemophilus influenzae was not cultured considering the larger number of preschoolers involved, the frequency of H. influenzae in spontaneous orbital cellulitis,15 and the inclusion of cases that occurred prior to the general use of vaccination in infancy for invasive H. influenzae B.
The onset of the cellulitis was concentrated in the first 4 postoperative days. This is similar to the information from other reports.3,7,9,16 In view of the large number of normal initial postoperative visits in this series and the range of onset times, a specific postoperative day cannot be recommended for a follow-up examination. Scheduling a return visit for between 1 and 5 days after surgery may still miss the infection. Concern about families' ability to recognize an infection and inclination to contact their physician has been raised by Parks.8 Two cases reported here were initially managed by primary care physicians who did not appreciate the seriousness of the condition. Recurrence of symptoms can be significant as seen in two abscess patients.
It is probably artificial to distinguish between periorbital and orbital infections after surgery in the orbit other than to indicate the extent and seriousness of the infection. It is surprising that the infection can be limited to the lids even on CT scanning. However, even with apparent limitation to the lids, the infection can disseminate in young children as shown in the present series by the bacteremia with Streptococcus pneumoniae that occurred in an 8-month-old infant. Bacteremia is common in spontaneous periorbital cellulitis.12
No patient in this series had endophthalmitis. The patient with the S. aureus abscess had a subretinal infiltrate in the area of the subconjunctival abscess but no visible scierai perforation externally or internally. While an association between orbital cellulitis after strabismus surgery and endophthalmitis has been observed,4·8 the question of unsuspected scierai perforation when this association occurs has not been answered Ui the literature. Endophthalmitis occurring from an abscess around a muscle suture in the absence of sclera! perforation has been observed by Dr Robert Reinecke. He stressed the importance of removing the suture to promote healing and to prevent this complication (personal communication, 1993).
There are multiple possible sources of the infections. Contamination at the time of surgery could have occurred even though no patient had bilateral infections. Some time delay while the bacteria multiplied to a level to cause clinical disease would be expected. Self-contamination after surgery by a young, non-hygienic patient or by the family seems a distinct possibility particularly in the patient who became infected after separate surgeries. Colonization with staphylococci after a course of systemic antibiotic such as amoxicillin is a theoretical possibility in children.
Steroid use intraoperatively or postoperatively could promote infection. Only one patient in our series received subconjunctival steroid and 5-fluorouracil without antibiotic because the prescribed steroid-antibiotic médication was not used. There is no evidence that the steroidantibiotic combinations commonly used after strabismus surgery17 promote infection. In one controlled study, even postoperative symptoms and signs did not differ between groups using artificial tears versus steroid-antibiotic drops.18
Finally, in view of the occurrence of spontaneous orbital and lid cellulitis in children, coincidental infection is a possibility. One member of the study group has had two patients who spontaneously developed preseptal cellulitis after their strabismus surgery had been canceled because of an upper respiratory infection (URI). Another member had a patient with bronchopulmonary dysplasia whose H. inffuenzae otitis and bronchitis were discovered when the surgery was completed and she was waking up. She developed preseptal cellulitis in the unoperated eye the following day despite IV antibiotics.
While sinusitis is a frequent cause of orbital and Ud cellulitis in children, the evidence from the three patients with sinusitis included here is not conclusive. Cultures of the eye were not taken in the 3-year-old, and cultures from the 12-year-old and 31-year-old showed no growth. No nasal cultures were taken. During the surgery in the 12-year-old, there was massive nasal discharge but it did not enter the eyes.
Could these cases have been prevented? Avoiding surgery m patients with bacterial URI is prudent, but it is difficult to make an accurate diagnosis of sinusitis in children without a CT scan. While persistence of upper respiratory symptoms past 10 days, daytime and nighttune cough, and nasal discharge can occur in sinusitis, no symptoms are specific.19'20 URIs are frequent in children and general anesthesia is often permitted shortly thereafter if only rhinorrhea remains. Whether a resolving viral URI could predispose a patient to periocular infection after surgery is unknown. Concern regarding manual spread of nasal infection to the eye is valid for young children, but avoiding surgery in a patient with any viral rhinorrhea probably would unnecessarily postpone many surgeries, as the frequency of postoperative infection is low. Postoperative instructions on hygiene of the eyes and reasons to contact the physician are appropriate but may not be followed. If the patient develops a URI or infection elsewhere after surgery, it would probably be safer to prescribe antibiotics to cover a bacteria infection of the involved area than to assume a viral origin.
Reduction of the bacterial load which could be carried into the wound during the incision can reduce the likelihood of a postoperative infection.21,22 In this area, there are specific data on the marked effectiveness of 5% povidone-iodine solution and/or neomycin drops to reduce the colony counts of conjunctiva! cultures.23,24 Povidone-iodine solution also reduced the incidence of endophthalxnitis after intraocular surgery in a recent study.25
Conjunctivitis after strabismus surgery probably occurs much more often than cellulitis, but the true incidence is not known. In our cases, there was not a clear progression from lid sticking and discharge to cellulitis in any patient. If some cases of cellulitis are due to organisms introduced into the eye after surgery, this progression may have occurred. A topical antibiotic might interrupt this progression although one study on steroid-antibiotic drops versus tear solution showed no difference.18 A very small trial of no treatment versus topical antibiotic was abandoned when three of eight patients with no antibiotic developed severe mucopurulent conjunctivitis.26
With the infrequency of periocular infection, the relative benefit of systemic prophylactic antibiotics for all patients versus the potential for adverse drug reactions seems small. Since the incidence of infection among subpopulations such as preschoolers or the non-hygienic is unknown, it is not certain that specific patient subpopulations with higher infection rates could be chosen for prophylaxis. No prospective studies have been done on systemic antibiotic prophylaxis for ocular or orbital surgery.27 In a consecutive study, the use of oral antibiotics after open lacrimal surgery reduced the infection rate from 7.9% to 1.6%.28 In otologie or minor nasopharyngeal surgery, systemic antibiotic prophylaxis does not reduce the infection rate.22 The literature em prophylactic antibiotic use in cardiac, gynecologic, orthopedic, and general surgery indicates that it is most effective when initiated sufficiently before the incision is made, and continued intraoperatively, to maintain therapeutic drug levels at the operative site.22-29 Cephalosporins, semisynthetic penicillins such as nafcillin, and vancomycin have been effective; aminoglycosides and metronidazole have been less well studied.22,29 Routine, prolongation of prophylactic antibiotic administration beyond 1 day has been shown to provide no additional protection.29 Single-dose systemic antibiotic prophylaxis has been as effective as multidose regimens in gastric, biliary, and transurethral operations, hysterectomies, and caesarean sections.30 Initiation of antibiotics after wound closure is considered to be without value.21,22,29
In evaluating patients with postoperative infections, conjunctiva! cultures offer useful information for continuing management. Blood cultures are appropriate in view of the occurrence of bacteremia in two of our cases and in many cases of spontaneous orbital cellulitis.12'15 This would be particularly appropriate for patients with toxic findings such as fever and lethargy and in children under 5 years of age. Because meningitis has occurred with advanced cases of orbital cellulitis, cerebrospinal fluid examination and culture should be kept in mind for very ill patients,2-12 particularly infants.
CT scanning is appropriate for cases of suspected sinusitis or subperiosteal abscess. The presence of sinusitis could influence the choice of antibiotic as Streptococcus pneumoniae, H. influenzae, and Moraxelia catarrhalis are common pathogens in acute sinusitis. Anaerobes, a-hemolytic streptococci and S. aureus can also occur with chronic sinusitis.19 Sinusitis must be treated medically for 2 to 3 weeks for complete eradication and follow-up CT scanning is often recommended.31 In patients with proptosis, CT scanning can identify a subperiosteal abscess which would need to be surgically drained if there is not a rapid response to antibiotics. An initial response to antibiotics followed by a relapse would be an indication for rescanning to look for an orbital abscess.32
Surgical drainage of a subconjunctival abscess will provide a useful culture and speed recovery. It also may prevent scierai necrosis.33
Intravenous therapy and hospitalization are warranted for cases with orbital signs and systemic illness, and for cases with known scierai perforation in view of the potential for endophthalmitis34 and central nervous system spread.2,12 Oral antibiotic therapy may be effective in some cases but in this series was ineffective for S. aureus, the most common organism.
1. Haase CG. Tenotomia muscoli recti externi: Phlegmonä Entzündung dee Orbitalizellgewebes mit Ausgang in Atrophia nervi optici. Arch Augenheilk. 1880 ;9: 442-445.
2. Holt EE. Orbital cellulitis, the inflammation spreading to the temporal region, then to the neck, obstructing deglutition, extending to the brain, and producing death. With remarks upon the same and brief reporte of five other cases. Trans Am Opht halmol Soc. 1892;6:295-306.
3. May CH. A caee of orbital ceUulitis following tenotomy for squint. Ann Ophthalmol Otolaryngol. 1894;3:261-264.
4. OTiriea CS, Leinfelder PJ. Unilateral exophthalmos. Etiologic and diagnostic studies in 82 consecutive cases. Am J Ophthalmol. 1935;18:123-132.
5. von Noorden GK. Orbital cellulitis following extraocular muscle surgery. Am J Ophthalmol. 1972;74:627-629.
6. Wilson ME, Paul TO. Orbital cellulitis following strabismus surgery. Ophthalmic Surg. 1987;18:92-94.
7. Casu L, Pitzorno E. Cellulite orbitaire au decours d'une chirurgie pour strabisme. J Fr Ophtalmol. 1988;11:771-772.
8. Parks MM. Routine antibiotic coverage in eye muscle surgery. Binocular Vision Quarterly. 1989;4: 152-153.
9. Weakley DR. Orbital cellulitis complicating strabismus surgery: a case report and review of the literature, Ann Ophthalmol. 1991 123:454-45 7.
10. Locatcher-Khorazo D, Seegal BC, Gutiérrez EH. Postoperative infections of the eye. In: Locatcher-Khorazo D, Seegal BC, eds. Microbiology of the Eye. St Louis, Mo: CV Mosby; 1972: 80-82.
11. Ing MR. Infection following strabismus eurgery. Ophthalmic Surg. 1991;22:41-43.
12. Weise A, Friendly D, Eglin K, Chang M, Gold B. Bacterial periorbital and orbital cellulitis in childhood. Ophthalmology. 1983 ;90: 195-203.
13. Smith CH. Bacteriology of the healthy conjunctiva. Br J Ophthalmol. 1954;38:719-726.
14. Perkins RE, Kundsin RB, Pratt MV, Abrahamsen I, Leibowitz HM. Bacteriology of normal and infected conjunctiva. J Clin Microbiol, 1975;1: 147- 149.
15. Watten EC, Wallar PH, Hues DA, Michaels RH. Acute orbital ceüiúiüs.ArchOphihalmol. 1976;94:785-788.
16. Folk ER. Antibiotics and timing of follow up visits in routine postoperative care: a survey of 25 strabismus surgeons. Binocular Vision Quarterly. 1990;5:7.
17. von Noorden GK Binocular Vision and Ocular Motility, 3rd ed. St Louis, Mo: CV Mosby; 1985:476.
18. Wortham EV, Anandakrishnan I, Kraft SP, Smith D, Morin JD. Are antibiotic-steroid drops necessary following strabismus surgery? A prospective, randomized, masked trial. J Pediatr Ophthalmol Strabismus. 1990;27: 205-207.
19. Ott NL, O'Connell EJ, Hoffinan AD, Beatty CW, Sachs MI. Childhood sinusitis. Mayo Clin PTOC. 1991;66:1238-1247.
20. Williams JW Jr, Simel DL. Does this patient have sinusitis? Diagnosing acute sinusitis by history and physical examination. JAMA, 1993;270: 1242-1246.
21. Burke JF. The effective period of preventive antibiotic action in experimental incisions and dermal lesions. Surgery 19euro1;50:161-180.
22. Burnakis TG. Surgical antimicrobial prophylaxis: principles and guidelines. Pharmacotherapy 1984;4:248-271.
23. Isenberg SI, Apt L, Yoshimori R, Khwarg S. Chemical preparation of the eye in ophthalmic surgery. Arch Ophthalmol. 1985;103:1340-1342.
24. Apt L, Isenberg SJ, Yoshimori R, Spierer A. Outpatient topical use of povidone-iodine in preparing the eye for surgery. Ophthalmology 1989;96:289-292.
25. Speaker MG, Menikoff IA. Prophylaxis of endophthalmitis with topical povidone-iodine. Ophthalmology 1991 #8: 1769- 1775.
26. Kearns PP, Cullen JF. Fucithalmic, chloramphenicol or no treatment after squint surgery in children. A single blind randomised study. Acta Ophthalmol. 1992;70:132-134.
27. Hurley LD, Weadall CT, Shore JW. Prophylactic use of antibiotics in oculoplastk surgery, lnt Ophthalmol Clin. 1992;32:165-178.
28. Wollend MJ, ROBC GE. Soft tissue infections after open lacrimal surgery. Ophthalmology 1994;101:608-611.
29. Kaiser AB. Antimicrobial prophylaxis in surgery, ff Engt J Afed. 1986;315:1129-1138.
30. DiPiro JT, Cheung RPF, Bowden TA Jr, Mansberger JA. Single dose systemic antibiotic prophylaxis of surgical wound infections. Am J Surg. 1986;152:552-559.
31. Giebink GS. Criteria for evaluation of antimicrobial agents and current therapies for acute sinusitis in children. Clin Infect Dis. 1992;14(suppl 2):S212-S215.
32. Bullock JD, Fleishman JA. Orbital cellulitis following dental extraction. Trans Am Ophthalmol Soc. 1984;82:111-133.
33. McNeer KW. Three complications of strabismus surgery. Ann Ophthalmol. 1975;7:441-446.
34. Simon YW, Lininger LL, Scheraga JL. Recognized scierai perforation during eye muscle surgery: incidence and sequelae. J Pediatr Ophthalmol Strabismus. 1992-29:273-275.
Questions on Data Collection Sheet
Inclusion Criteria for Cellulitis Cases
Possible Predisposing Factors
Ocular and Systemic Symptoms In Cellulitis Cases
Culture Results in Cellulitis Cases
Antibiotics Used in Cellulitis Cases