The ability to combat bacterial infection with antibiotics transformed human history in the mid-20th century. Over the ensuing 70 years, antibiotics have been administered in medicine not only to treat infected patients but also in an effort to prevent infection. Recently, there has been a growing appreciation of the detrimental consequences associated with widespread antibiotic use. These include the selection of multidrug-resistant bacteria, adverse effects such as Clostridium difficile disease and devastating allergic reactions, and rising health care costs. There is early evidence that antibiotics given to infants and children, because antibiotics affect the human microbiome, may be contributing to the spate of modern health epidemics such as asthma, food allergies, and obesity.1
Approximately one-half of prophylactic antibiotics are administered in association with surgery.2 In the United States, several national guidelines have addressed the use of prophylactic antibiotics for various types of surgery (understood to be given intravenously to achieve the desired serum levels at the time surgery begins, and to keep the stomach empty during induction of general anesthesia).3,4 However, the role of antibiotic prophylaxis for orbital surgery has not been thoroughly vetted, and many experienced orbital surgeons have reported years of success without using antibiotics.5 In cases in which antibiotic prophylaxis is indicated, such as for clean surgery in which an implant is placed or for clean-contaminated surgery, most guidelines recommend starting intravenous (IV) prophylactic antibiotics within 60 minutes prior to surgical incision (120 minutes for vancomycin or fluoroquinolones).4 Guidelines published by the World Health Organization in 2016 and Centers for Disease Control and Prevention (CDC) in 2017 state that prophylactic antibiotics, when they are indicated, should be stopped at the end of surgery.3,6 In short surgical cases typical of orbital surgery, this means that a single preoperative antibiotic dose is given. Modified recommendations have been made for cardiac surgery and other specific types of surgery.7
The risk of surgical site infection (SSI) after orbital surgery is low, with recent studies reporting SSI rates of 0.3% to 1.1%.5,7 Furthermore, great disparity in antibiotic prescribing patterns among oculoplastic surgeons has been demonstrated.8 Therefore, the current study was undertaken to determine the risk of infection associated with orbital surgery and the efficacy of IV prophylactic antibiotics in preventing such complications.
A deidentified registry of orbital surgery cases was designed by members of the Orbital Society, an international association of senior orbital surgeons. The study received a “not human studies” determination from the Massachusetts Eye and Ear Infirmary Institutional Review Board and was conducted in accordance with the Declaration of Helsinki and the Health Information Portability and Accountability Act. All orbital surgery cases were included, with each contributing surgeon pledging to include all consecutive orbital surgery cases throughout the study period, Data collection began October 1, 2013, and concluded March 1, 2015. Surgeons were not asked to alter their ongoing practice of administering or omitting perioperative antibiotics. Several aspects of orbital surgery cases pertinent to the usage of antibiotics and development of infections were recorded, including indication for surgery, administration of antibiotics, antibiotic regimen, paranasal sinus entry, presence of osteotomy, corticosteroid use, the placement of an alloplastic orbital implant (e.g., polyethylene, titanium, or other synthetic spheres, sheets, plates, or screws), and the development and course of any SSIs. Cases with pre-existing infection at the time of surgery or incomplete information were excluded from the analysis. The primary outcome measure was the presence or absence of SSI within 2 weeks following surgery. Surgical site infection was determined clinically and included cases with cellulitis only. (This is in contrast to the CDC SSI definition, which excludes cases with cellulitis alone,9 but consistent with the SSI definition by the European Centre for Disease Control and Prevention, which allows diagnosis of SSI by a surgeon or attending physician and does not specifically exclude cases of cellulitis.)10
Following accrual, data were normalized, including the designation of 1 of 6 indications for surgery: tumor, thyroid-related orbitopathy, unspecified inflammatory disease, fracture, prior infection (e.g., enucleation for endophthalmitis), and “other.” Antibiotic regimens were categorized by the class of antibiotic and duration of antibiotics (no antibiotics, 1 perioperative dose, or multiple doses of antibiotics).
Rates of infection in the groups segregated by antibiotic exposure, as well as several subgroups defined by the aforementioned surgical characteristics, were calculated using scripts written in Matlab (The MathWorks, Natick, MA). Fisher exact test was used for discrete variables and Student’s t test for continuous variables. Multivariate analysis was performed with logistic regressions using binomial distribution with a logit link function. p values less than 0.05 were considered significant.
Of the 1,250 consecutive orbital surgery cases collected, 1,225 met inclusion criteria (3 cases were excluded due to incomplete antibiotic data, 1 case for incomplete sinus entry data, 2 cases for incomplete implant data, and 19 cases for incomplete steroid data). Of the 1,225 cases, 603 received no antibiotics (group A), and 605 received a single dose of IV perioperative antibiotic within 60 minutes prior to incision (group B). A third group of 17 patients received a single dose of IV antibiotics perioperatively, and also received 7 days of postoperative oral antibiotics (group C) (Table 1). The mean age of patients in group A was 49.3 years and in group B was 46.9 years (p = 0.30). In group A, 54.7% were female while in group B, 51.7% were female (Table 2). The most common indication for surgery was neoplasm (42.9%), followed by thyroid-related orbitopathy (23.0%) and trauma (7.5%). Types of surgery and numbers of SSI are presented in Table 3.
Surgical Site Infections.
Five (0.41%) patients developed SSIs; 3 in group A (0.5%) and 2 in group B (0.33%). The difference in SSI rates between group A and group B was not significant (odds ratio 0.69, 95% confidence interval, 0.11–4.0; p = 0.66). The 5 SSIs included 2 cases of orbital cellulitis and 3 cases of preseptal cellulitis (Table 4).
Both orbital infections occurred in group A are were due to Staphylococcus aureus (1 was methicillin-resistant). The difference between the rate of orbital infection in group A (0.33%) and group B (0.0%) was not significant (p = 0.25).
One of the orbital infections followed resection of a sphenoid wing meningioma, and the other after evisceration for a ruptured globe after penetrating keratoplasty. In neither case were the paranasal sinuses entered. Both infections were successfully treated with antibiotics. The meningioma patient received IV vancomycin and amoxicillin/clavulanate for 3 days, followed by amoxicillin/clavulanate by mouth. The evisceration patient received oral trimethoprim/sulfamethoxazole and removal of the implant. Wound cultures grew methicillin-susceptible S. aureus in the meningioma case and methicillin-resistant S. aureus in the evisceration case.
Among the 3 patients who developed preseptal cellulitis, 1 was in group A, while 2 were in group B. The difference in the rate of development of postoperative preseptal cellulitis between group A (0.17%) and group B (0.33%) was not significant (p = 1.0). None of the 5 patients with SSIs had underlying diabetes or immunosuppression.
Two hundred ninety-six cases involved entry into one or more of the paranasal sinuses (e.g., orbital decompression, orbital fracture repair). Three of these patients developed preseptal cellulitis and none developed orbital cellulitis.
Corticosteroids were administered intravenously either preoperatively or intraoperatively, or orally postoperatively, in 697 patients. Postoperative infections developed in 3 of the 697 cases. Of the cases without corticosteroid administration, 2 SSIs occurred.
An implant was placed during surgery in 340 cases, of whom 2 patients developed preseptal and 1 orbital cellulitis. One of these cases was an evisceration with implant, one was a nonporous sphere inserted secondarily, and one was a titanium microanchor. Among patients not receiving an orbital implant, one developed preseptal cellulitis and another developed orbital cellulitis. The SSI rate in cases with implants was not significantly higher than in nonimplant cases (p = 0.27).
To control for associations between antibiotic use and factors such as sinus entry, corticosteroid administration, and implant placement, multivariate regression analysis was performed.
Logistic regression demonstrated that prophylactic antibiotic administration (p = 0.26), sinus entry (p = 0.067), corticosteroid administration (p = 0.86), and implant placement (p = 0.12) were not significantly associated with SSI (whether preseptal or orbital). These factors were not associated with preseptal cellulitis or orbital cellulitis (Table 5).
Adverse Events Related to Antibiotics.
Two cases of antibiotic-related adverse events were documented, both in group C. One patient developed diarrhea following clindamycin administration, while the other developed acute renal injury related to ceftriaxone. The number needed to harm was 8.5 (Table 6).
Three-quarters of a century into the antibiotics era, enthusiasm for widespread prescribing has begun to wane, largely due to increasing appreciation for the adverse effects of medical and industrial antibiotics. Increasingly, antibiotic-resistant and virulent bacterial strains are not only more difficult to eradicate in individual patients, but can also produce more aggressive and destructive disease.11 Although rare, devastating immunological reactions, such as Stevens-Johnson syndrome and toxic epidermal necrolysis syndrome, are generally known to physicians. Less widely appreciated are the detrimental effects of antibiotics on the human microbiota. Alterations in these symbiotic bacteria have been linked to childhood obesity12 as well as asthma,13 food allergies,14 irritable bowel syndrome,15 and other disorders, collectively termed “modern epidemics.”1
Prophylaxis against SSI predates the modern era of antibiotics. As early as the 1930s, patients undergoing colorectal surgery were given the then newly discovered sulfanilamide.16 Experimental animal studies in the ensuing decades demonstrated the suppression of skin infections when IV antibiotics were given within 3 hours of wound contamination.17,18 By the 1980s, the CDC had classified surgical wounds according to the likelihood and degree of wound contamination at the time of surgery and published its first guidelines for prevention of SSI.19 These guidelines, though repeatedly updated, cannot consider all possible variations in anatomical location, surgical environs, procedures, and indications.20 The orbit, for example, possesses a luxuriant vascular network that may protect against infection from microbial contamination.
The current study is the largest to evaluate perioperative prophylaxis in orbital surgery. The SSI rate was very low (0.41%), and similar to that reported in other studies.5,7 No antibiotic prophylaxis was given to 603 patients, yet their SSI rate was not different from 608 patients who received prophylaxis. Moreover, this study also found successful treatment of SSI, should signs or symptoms indeed arise. No patient in this series went on to experience serious sequelae due to SSI.
Practices for prescribing antibiotic prophylaxis vary widely among orbital surgeons, as demonstrated by earlier reports8 and the present study. Many prominent orbital surgeons have not used any prophylactic antibiotics of any sort for decades, yet report virtually no significant SSIs. In cohorts studied thus far, neither IV prophylaxis (present study) nor oral postoperative antibiotics reduced the odds of SSI in patients undergoing orbital surgery.5,7 No prospective study of SSI following any type of surgery has demonstrated efficacy of postoperative prophylactic antibiotics. In this study, half of the patients received antibiotic prophylaxis, and multivariate analysis demonstrated no significant association between SSI rate and antibiotic prophylaxis, sinus entry, implants, or IV corticosteroids.
The adverse effects of antibiotics are greatly underestimated by oculoplastic surgeons,8 and some of these adverse effects were observed in this study. Extended (1 week) oral prophylaxis was associated with a 12% incidence of complications (diarrhea, renal injury). Prolonged antibiotic therapy is known to be associated with multiple potential risks to the individual patient, including severe allergic reactions and potential effects on the microbiome.11–15 Prolonged surgical antibiotic prophylaxis is never indicated; both CDC and World Health Organization guidelines specify stopping antibiotic prophylaxis—when any surgical prophylaxis is indicated—at the end of surgery at the latest.6,9 Prolonging antibiotic prophylaxis past the duration recommended has been associated with significant adverse events, including a 6.7 odds ratio of developing C. difficile infection.21
Problems with antibiotics can be categorized into individual and community concerns. Allergic reactions and gastrointestinal events constitute the well-known personal adverse considerations; antibiotic resistance, progressive bacterial virulence, and healthcare expenditures constitute the community concerns. Up to 20% of people demonstrate signs of allergic reaction when treated with topical sulfa medications. More important, however, are the severe, life-threatening reactions to a single dose of systemic antibiotics: Stevens-Johnson syndrome and toxic epidermal necrolysis syndrome. In this series, the individual adverse events occurred among the 17 patients who received postoperative antibiotics in addition to the single preoperative IV dose. A separate analysis of this small cohort yielded a number needed to harm of only 8.5, a staggering number given that no actual infection was being treated.
In addition to individual adverse events, the community is also hurt by indiscriminate deployment of antibiotics. Antibiotic resistance spreads exponentially among microbes not only because of genetic inheritance but also from direct gene-sharing through plasmids. This leaves the community at greater risk of aggressive and untreatable infection. Indeed, the European Antibiotics Awareness campaign states that responsible stewardship is the best hope for combating ever-increasingly resistant and virulent bacteria. Finally, the added expense of unnecessary medications contributes to already problematic health care costs. This is one of the reasons that the CDC and the European Antibiotics Awareness campaign have evolved guidelines for appropriate administration of antibiotics.
There are important limitations to this study. Based on previously published data and the results of this series, a study powered to 80% would require approximately 28,000 consecutive cases (https://www.stat.ubc.ca/~rollin/stats/ssize/b2.html). This large number of cases required underscores the rarity of the event sought and the extremely small difference in rates of SSI between the 2 groups in this study.
Second, because the study was not randomized, surgeons may have been more likely to prophylax what they perceived as higher-risk patients while withholding prophylaxis in a perceived lower risk population. For example, some surgeons administer antibiotics to patients with poor personal hygiene. It is possible that such an arrangement could coincidentally reduce the risk of infection among “high-risk” patients to the statistically identical risk of infection among untreated, “normal-risk” patients. To control for this possibility, all entries in the database included an indication for surgery, which allowed analysis by risk category. Only 17 of the 1,208 patients in groups A and B had experienced infections preoperatively (that were treated and cleared by the time of surgery). Exclusion of these cases from analysis resulted in no change in the rates of infection in either group A or group B (0.50% and 0.33%, respectively). The limited number of cases performed for infectious indications precludes specific conclusions for cases performed for infectious indications. However, the main conclusions of the study are broadly applicable to orbital surgeries for noninfectious indications, which comprised most of the orbital surgery cases in this cohort.
A third important limitation was a failure to record patient comorbidities, a potential source of increased infection risk. It is possible that patients with diabetes, distant infections, or immunosuppression, for example, are at higher than normal risk of infection from orbital surgery, and would, therefore, benefit from antibiotic prophylaxis. Indeed, such patients in this study may have been more likely to receive prophylaxis. Post hoc analysis, however, showed no comorbidities among the 5 patients who developed infection, regardless of whether they received antibiotics. Still, comorbidities should be taken into consideration when planning individual patient care.
Fourth, CDC SSI definitions specify a 30-day follow up period for orbital surgery, whereas the patients in this study were surveilled for infection specifically in the first 2 postoperative weeks. Based on the CDC definitions, a study period of 30 days would have been useful for comparison with other studies. Fortunately, all participating surgeons in this study routinely follow their surgical patients closely for at least 1 full month, and no additional SSIs were identified in this cohort during the 3rd and 4th postoperative weeks. CDC guidelines are summarized nicely in a recent retrospective study of SSI in cases of enucleation and evisceration.7
Fifth, the present study may have overestimated the SSI rate, because cases of cellulitis alone were included. Only 1 case met CDC criteria for SSI (implant removed) and counting only this case would yield a 0.08% SSI rate.
There are relevant variables that this study did not evaluate. Among them are environmental factors such as smoking, geographical location, local climate, operating room conditions, surgical technique including instrument maintenance, sterile technique, electrocautery, wound closure, implant type, and postoperative wound care. Some of these variables would be more easily evaluated in a single-center study.
In conclusion, given the observations in this study and others, patients undergoing orbital surgery do not appear to benefit from routinely administered prophylactic antibiotics. This agrees with recommendations for other types of clean surgeries in the head and neck.4 Surgeons must consider comorbidities and other infection risks not addressed in this study when treating individual patients. They should educate patients on the early signs and symptoms of postoperative SSI, but be aware that there remains no evidence to support the routine use of antibiotic prophylaxis in orbital surgery.
1. Blaser MJ. Missing Microbes. 2014.New York: Holt and Co.
2. Radda TM, Grasl MM, Gnad HD. Perioperative prevention of infection in ophthalmic surgery. Antibiot Chemother (1971) 1985;33:184–197.
3. Berríos-Torres SI, Umscheid CA, Bratzler DW, et al.; Healthcare Infection Control Practices Advisory Committee. Centers for Disease Control and Prevention guideline for the prevention of surgical site infection, 2017. JAMA Surg 2017;152:784–791.
4. Bratzler DW, Dellinger EP, Olsen KM, et al.; American Society of Health-System Pharmacists; Infectious Disease Society of America; Surgical Infection Society; Society for Healthcare Epidemiology of America. Clinical practice guidelines for antimicrobial prophylaxis in surgery. Am J Health Syst Pharm 2013;70:195–283.
5. Fay A, Nallasamy N, Nallassamy N, et al.; New England Oculoplastics Society Study Group. Prophylactic postoperative antibiotics for enucleation and evisceration. Ophthalmic Plast Reconstr Surg 2013;29:281–285.
6. World Health Organization. Global guidelines for the prevention of surgical site infection. http://www.who.int/gpsc/ssi-guidelines/en/
. Accessed November 2017
7. Pariseau B, Fox B, Dutton J. Prophylactic antibiotics for enucleation and evisceration: a retrospective study and systematic literature review. Ophthalmic Plast Reconstr Surg 2018;34:49–54.
8. Fay A, Nallasamy N, Bernardini F, et al. Multinational comparison of prophylactic antibiotic use for eyelid surgery. JAMA Ophthalmol. 2015;133:778–784.
9. Centers for Disease Control and Prevention (CDC). Surgical Site Infection (SSI) Event. January 2017 update. http://www.cdc.gov/nhsn/pdfs/pscmanual/9pscssicurrent.pdf
. Accessed November 2017
10. European Centre for Disease Prevention and Control (ECDC). Surveillance of surgical site infections in European hospitals – HAISSI protocol. Protocol version 1.02. 2012. Stockholm: ECDC, http://ecdc.europa.eu/sites/portal/files/media/en/publications/Publications/120215_TED_SSI_protocol.pdf
11. Beceiro A, Tomás M, Bou G. Antimicrobial resistance and virulence: a successful or deleterious association in the bacterial world?. Clin Microbiol Rev 2013; 26:185–230.
12. Azad MB, Bridgman SL, Becker AB, et al. Infant antibiotic exposure and the development of childhood overweight and central adiposity. Int J Obes (Lond) 2014;38:1290–1298.
13. Russell SL, Gold MJ, Reynolds LA, et al. Perinatal antibiotic-induced shifts in gut microbiota have differential effects on inflammatory lung diseases. J Allergy Clin Immunol 2015;135:100–109.
14. Stefka AT, Feehley T, Tripathi P, et al. Commensal bacteria protect against food allergen sensitization. Proc Natl Acad Sci U S A 2014;111:13145–13150.
15. Theodorou V, Ait Belgnaoui A, Agostini S, et al. Effect of commensals and probiotics on visceral sensitivity and pain in irritable bowel syndrome. Gut Microbes 2014;5:430–436.
16. Garlock JH, Seley GP. The use of sulfanilamide in surgery of the colon and rectum: preliminary report. Surgery 1939;5:787–790.
17. Miles AA, Miles EM, Burke J. The value and duration of defence reactions of the skin to the primary lodgement of bacteria. Br J Exp Pathol 1957;38:79–96.
18. Burke JF. The effective period of preventive antibiotic action in experimental incisions and dermal lesions. Surgery 1961;50:161–168.
19. Garner JS. CDC guideline for prevention of surgical wound infections, 1985. Supersedes guideline for prevention of surgical wound infections published in 1982. (Originally published in November 1985). Revised. Infect Control 1986;7:193–200.
20. Mangram AJ, Horan TC, Pearson ML, et al. Guideline for prevention of surgical site infection, 1999. Hospital infection control practices advisory committee. Infect Control Hosp Epidemiol 1999;20:250–78; quiz 279–80.
21. Balch A, Wendelboe AM, Vesely SK, et al. Antibiotic prophylaxis for surgical site infections as a risk factor for infection with Clostridium difficile
. PLoS One 2017;12:e0179117.