Indian Journal of Ophthalmology

: 2000  |  Volume : 48  |  Issue : 2  |  Page : 123--8

Spectrum of aetiological agents of postoperative endophthalmitis and antibiotic susceptibility of bacterial isolates.

AR Anand, KL Therese, HN Madhavan 
 Vision Research Foundation, Chennai, India

Correspondence Address:
A R Anand
Vision Research Foundation, Chennai


PURPOSE: To determine the spectrum of infectious agents of postoperative endophthalmitis, the relationship with the time of onset of symptoms after surgery and the antibiotic susceptibilities of the aerobic bacterial isolates. METHODS: A retrospective review of microbiological records from January 1995 to December 1998 yielded 173 isolates from intraocular specimen of 170 patients with culture-proven postoperative endophthalmitis. Antibiotic susceptibility of these isolates was determined for various ocular antibiotics using the Kirby-Bauer disk-diffusion test. Based on the time of onset of illness, clinical presentation was classified into acute, delayed and chronic. RESULTS: Among 170 cases, 71 (41.7%) were attributable to gram-negative, 64 (37.6%) to gram-positive bacteria, and 37 (21.8%) to fungi. Gram-negative bacteria included P. aeruginosa (29;17.1%), other Pseudomonas spp (15;8.8%), non-fermenters (18;10.6%) and others (10;5.8%). Among these, 40 of 72 (55.5%) were sensitive to gentamicin, 47 of 72 (65.2%) to cefotaxime, 47 of 69 (68.1%) to amikacin, 52 of 71 (73.2%) to ciprofloxacin, and 25 of 40 (62.5%) to ceftazidime. The gram-positive bacteria included S. epidermidis (22;12.9%), S. aureus (13;7.6%), P. acnes (10;5.9%), Enterococcus spp (4;2.3%), Streptococcus spp (7;4.1%) and others (8;4.8%). Among these, 41 of 53 (77.3%) were sensitive to gentamicin, 47 of 53 (88.6%) to cefotaxime, 46 of 52 (88.4%) to ciprofloxacin, 38 of 41 (92.6%) to cefazolin and 27 of 37 (72.9%) to ceftazidime. All gram-positive bacteria were sensitive to vancomycin. CONCLUSION: In this large series of postoperative endophthalmitis, gram-negative bacilli followed by fungi accounted for the largest number of cases. A high degree of resistance of gram-negative bacilli to gentamicin, cefotaxime, amikacin and ceftazidime was recorded.

How to cite this article:
Anand A R, Therese K L, Madhavan H N. Spectrum of aetiological agents of postoperative endophthalmitis and antibiotic susceptibility of bacterial isolates. Indian J Ophthalmol 2000;48:123-8

How to cite this URL:
Anand A R, Therese K L, Madhavan H N. Spectrum of aetiological agents of postoperative endophthalmitis and antibiotic susceptibility of bacterial isolates. Indian J Ophthalmol [serial online] 2000 [cited 2021 Sep 20 ];48:123-8
Available from:

Full Text

Postoperative endophthalmitis, though rare, is one of the most devastating complications of intraocular surgery. It is also the most common form of endophthalmitis, accounting for approximately 70% of infective endophthalmitis.1 There have been a few large series of reports on the types of infective agents causing postoperative endophthalmitis in Western literature.[1-5] However, with the exception of one large series of postoperative endophthalmitis,[6] only isolated and anecdotal reports so far have been available from India.[7, 8]

The present study is based on a review of a large series of culture-proven postoperative endophthalmitis from among patients referred to our centre during 4-year period, January 1995 to December 1998. The objectives of the study were to determine the frequency of different aetiological agents of postoperative endophthalmitis, the correlation of the time of onset of symptoms after surgery with the organism isolated, and the pattern of susceptibilities of the aerobic bacterial isolates to the commonly used antibiotics.

 Materials and Methods

 Patients and collection of specimens

A retrospective review of clinical and microbiological records at our hospital revealed 173 isolates from intraocular specimen of 170 cases of consecutive culture-proven postoperative endophthalmitis out of a total of 382 clinically diagnosed, between January 1995 and December 1998. These patients were grouped into 3 distinct categories based on the time of onset of symptoms after surgery as described by Megevand and Pournares.[9] This was: acute, up to 4 days; delayed, 5 days to 1 month; and chronic; 1 month onwards. Using a syringe with needle, 150-200 l of aqueous humor was collected as a diagnostic sample and 500-1000 l of undiluted vitreous fluid was collected either as a diagnostic sample or at the beginning of the therapeutic vitrectomy procedure. After the air in the syringe was expelled to prevent inactivation of anaerobic bacteria, the needle was fixed on to a sterile rubber bung and placed in a sterile large test tube container, which was immediately transported to the laboratory.

 Microbiological investigations

The specimens were processed within 15-30 minutes of their collection. Cultures and smears for the detection of bacterial and fungal agents were carried out on the intraocular specimens.

 Culture studies

Inoculation of media for culture of bacteria and fungi was done first. For inoculation on to solid media, 1-2 drops (20-30 l) of the specimen was expressed through the needle on to the agar plate. A loop was not used to spread the material to avoid the possibility of contamination. Another 2-3 drops (30-40 l) were inoculated on to liquid media. The culture media used were: blood agar (BA - incubated aerobically at 37C); Brucella blood agar (BBA - incubated at 37C in an anaerobic work station), (Don Whitley, India); and chocolate agar (CA - incubated at 37C in an atmosphere of 10% CO2) (Forma Scientific, USA). The liquid media used were: Brain heart infusion broth (BHIB) and Robertson cooked meat medium (RCM). The specimen inoculated on to Sabouraud's dextrose agar (SDA) was incubated at 25C in a cooling incubator (Remi, India). All the above media were prepared by reconstituting commercially available dehydrated media (HiMedia, India). If cultures on BA, CA and BHIB showed no growth at the end of 48 hours, they were placed in the cooling incubator for another 7 days for growth of fungus. The cultures for anaerobic bacteria were incubated for up to 2 weeks. The isolated bacteria and fungi were identified by standard methods.[10-12]

 Direct smear studies

Three to four smears from the remaining AH/VF specimens were made using the cytospin machine (Shandon, UK). One KOH- Calcofluor white preparation was made for detection of fungus, a gram stain for bacteria and a Giemsa stain for cytology. For study purposes, the results of smear were used for correlation with that of the cultures to satisfy the criteria mentioned below.

 Antibiotic sensitivity studies

The antibiotic susceptibility testing of the aerobic isolates was done by the standard Kirby-Bauer disk-diffusion technique.[13] The antibiotic disks used were: gentamicin (10 g), amikacin (30g), vancomycin (30g), ciprofloxacin (5g), cefazolin (30g), cefotaxime (30g) and ccftazidime (30g), (HiMedia) India. Quality control was carried out using standard strains of Escherichia coli (ATCC 25922), Pseudomonas aeruginosa (ATCC 27853), Staphylococcus aureus (ATCC 25923) and Enterococcus faecalis (ATCC 29212 ).

The criteria used to consider the isolated bacterium or fungus as the causative agent were: growth from VF/AH on two or more of the inoculated media, or growth on a single medium correlating with direct smear findings, or repeated isolation of the same organism from two or more intraocular specimens of the patient.


Based on the clinical presentation, 170 patients were distributed among the acute (n=68), delayed (n=52) and chronic (n=50) endophthalmitis groups. All the bleb-associated endophthalmitis cases (n=9) were chronic. A total of 173 infective agents were isolated from the 170 patients. This was from a total of 382 clinically diagnosed endophthalmitis patients; thus the overall culture positivity was 44.5%. The isolated organisms were considered causative agents if the criteria mentioned earlier were met. Among these 170 patients, 167 had single isolations which included 132 (77.6%) bacterial, 35 (20.6%) fungal agents; 3 (1.8%) patients had multiple bacterial or fungal isolations. The distribution of the isolates among the 170 patients is shown in the Figure. In all, bacteria were implicated in 135 (79.4%) patients, among which 71 (41.7%) had gram-negative bacilli and 64 (37.6%), gram-positive bacterial infections; fungi were isolated in 37 (21.8%) patients.

Pseudomonas aeruginosa was the single most frequent aetiological agent, accounting for 29 (17.1%) of all cases. The other gram-negative bacilli encountered were 18 (10.6%) other non-fermenting bacilli, 15 (8.8 %) Pseudomonas spp other than P.aeruginosa, 5 (2.9%) from the Enterobacteriaceae group of bacilli, and 5 (2.9 %) others that were classified as miscellaneous. P. aeruginosa and other Pseudomonas spp were mainly associated with acute endophthalmitis, in 27 of 29 patients (93.10%) and 12 of 15 patients (80%) respectively. Of 18 cases from which non-fermenting bacilli were isolated, 16 (88.8%) were of delayed or chronic endophthalmitis.

Among gram-positive bacteria, S. epidermidis was the most common, accounting for 22 (12.9%) cases followed by S. aureus 13 (7.6%), P. acnes 10 (5.9%), Streptococcus spp 7 (4.1%), E. faecalis 4 (2.4 %), Micrococcus spp 4 (2.4%), Corynebacterium spp 2 (1.2%), and Bacillus spp 2 (1.2%). S epidermidis was equally distributed among the clinical types: acute, 31.8%, delayed, 31.8%; and chronic, 36.4%. Of 13 cases in which S. aureus was implicated, 10 (76.9%) were delayed or chronic. Streptococcus spp and E. faecalis accounted for 5 of the 9 bleb-associated cases (55.6%).

Fungi were isolated from the intraocular specimens of 37 patients and the most common fungus was Aspergillus spp, accounting for 19 (11.2 %) cases, followed by Acremonium fatciforme and Candida spp in 2 patients each (1.2%) and Fusarium spp and Paecilomyces spp in 1 patient each (0.6 %). Eleven isolated fungi could not be identified due to their failure to sporulate. Of these 37 patients, 32 (86.5%) had either delayed or chronic endophthalmitis.

Polymicrobial infection was seen in 3 patients, they were combination of Candida albicans and E. faecalis; Bacillus megaterium and Acremonium falciforme; and Pseudomonas aeruginosa and Flavobacterium spp.

The distribution of bacterial and fungal isolates in relation to the acute, delayed and chronic clinical presentations are shown in [Table:1].

The antibiotic susceptibility tests were not done for all the isolates against all the listed antibiotics for various reasons. The results of the susceptibility tests of the gram-positive aerobic bacterial and gram-negative bacterial isolates to the commonly used intravitreal antibiotics such as gentamicin, amikacin, vancomycin, cefazolin, cefotaxime and ceftazidime, as well as to ciprofloxacin which is often used topically were analysed. The results are presented in [Table:2]. Among the gram-positive bacteria tested, 41 of 53 (77.3%) were sensitive to gentamicin, 47 of 53 (88.6%) to cefotaxime, 46 of 52 (88.4%) to ciprofloxacin, 38 of 41 (92.6%) to cefazolin, and 27 of 37 (72.9%) to ceftazidime. Among the gram-negative bacteria, 40 of 72 (55.5%) were sensitive to gentamicin, 47 of 72 (65.2%) to cefotaxime 47 of 69 (68.1%) to amikacin, 52 of 71 (73.2%) to ciprofloxacin, and 25 of 40 (62.5%) to ceftazidime. Only 77 isolates were tested against ceftazidime. All the gram-positive bacteria (100%) were sensitive to vancomycin while all the gram-negative bacteria, were resistant to it.


The aetiology of postoperative endophthalmitis and its true incidence in India are still unclear. Other than one recently published series on postoperative endophthalmitis from Hyderabad,[6] case reports[7, 8, 14, 15] have been isolated and sporadic. Outbreaks have never been reported. Our knowledge of the incidence and aetiology of postoperative endophthalmitis is mainly based on American[2-4] and to some extent, European studies.[5] The present study may represent to a large extent the spectrum of infective agents causing postoperative endophthalmitis in India, since the patients in this series had undergone surgery in different parts of the country and were referred for management after developing endophthalmitis.

The analysis of the results of our study on the intraocular isolates reveals a distinct spectrum of organisms causing postoperative endophthalmitis. Earlier studies from western literature 1964 and 1974 revealed a high percentage of S. aureus and Pseudomonas infections and very few anaerobic infections.[16, 17] However, recent reports overwhelmingly suggest that S. epidermidis is the single most common causative agent[2, 18] and P. acnes is the most common cause in delayed onset endophthalmitis.[19] Fungal aetiology has been very rare.[3, 4]

In our study, gram-negative bacilli were isolated from 41.7% of the patients and to the best of our knowledge, such a high association of gram-negative bacteria with postoperative endophthalmitis has never been reported before. Previously reported rates of infection due to gram-negative bacteria range from 3%-26%[2-4, 6, 20-22] of the culture-positive cases. P. aeruginosa, though reported as the most frequent gram-negative bacillus causing postoperative endophthalmitis,[2, 23] had until now never been reported as the single most common organism (as found in this study). Majority of the isolates (93.1%) of this bacterium in our study were from patients with acute onset endophthalmitis. Other Pseudomonas spp not known to be as virulent as P. aeruginosa were also isolated in a significant number of (8.8%) patients.

Other non-fermenting gram-negative bacilli such as Acinetobacter spp, Alcaligenes spp, Flavobacterium spp and Achromobacter spp, formed 10.6% of our isolates. These neglected bacteria were considered non-pathogenic and of little clinical significance. Anecdotal reports of postoperative endophthalmitis caused by these bacteria have been found in literature,[24-26] but their incidence (10.6%) in this study was probably the highest and they were mostly associated with either delayed or chronic postoperative endophthaimitis.

The incidence (38.8%) of gram-positive bacteria was low compared to the previously reported figures, which ranged between 57-90 %[2-4, 6, 30] of the isolates. Though S. epidermidis was the most common isolate in our study, its isolation rate was low (12.9%) compared to previous studies (33-63%).[2-4, 6, 20, 27] As in the previous reports,[27] the clinical presentations of S. epidermidis postoperative endophthalmitis included acute, delayed and chronic in almost equal numbers. The combined incidence of S. aureus and Streptococcus spp. (11.7%) was comparable with other reports (10-30%).[2, 3, 15, 17] Surprisingly, S. aureus was associated mainly with delayed and chronic clinical presentations. Streptococcus spp. have been known to account for over 50% of late-onset bleb-induced endophthalmitis cases.[28, 29] Our results were also comparable with the above studies, with Streptococcus spp and Enterococcus spp accounting for 5 of 9 isolates from bleb-associated endophthalmitis. Though P. acnes did form a considerable percentage (11.9%) of chronic postoperative endophthalmitis, it was less than those due to fungus.

We believe that the present study has shown the highest incidence (21.8%) of fungal postoperative endophthalmitis as compared to other reports (3-17 %).[3, 4, 6,20] Aspergillus spp. was the most common isolate with relatively low incidence of Candida spp. These results were in contrast to western literature.[4, 19, 30] Since a majority of the fungi caused delayed or chronic postoperative endophthalmitis, fungal aetiology should first be excluded in all such clinical presentations before deciding on the mode of therapy.

Polymicrobial infection is relatively less common in postoperative endophthalmitis (1.6-9,3%),[1, 2, 4] compared to post-traumatic endophthalmitis.[31, 32] In our study polymicrobial infection accounted for 3 (1.8%) cases, but comparable to a few reports.[1-4]

The sources of infections in all these patients could only be conjectural, as they were referrals to our hospital. The infective agents could have been introduced into the eye at the time of intraocular surgery, the chief sources being infected eyelids or lacrimal sac, airborne microbial agents, contaminated irrigating fluids, instruments, implants and bacterial flora of the conjunctiva. It has been postulated that the patient's own conjunctival flora could be the most common source of infection,[33, 34] particularly in infections due to S. epidermidis, S.aureus, P.acnes and Corynebacterium spp. But the majority of our isolates were gram-negative bacilli (non-Enterobacteriaceae group) and fungi, which do not commonly form part of normal conjunctival flora. Hence, the source could be elsewhere. A failure to achieve perfect sterility of the surgical field may, in all probability, have played a role in development of these infections.

The treatment of choice in cases of established bacterial endophthalmitis is intravitreal injections of antibiotics which include vancomycin for all gram-positive bacteria and an aminoglycoside-gentamicin/amikacin or ceftazidime which show broad spectrum activity against gram-negative bacteria. Regarding antibiotic sensitivity patterns, vancomycin resistant gram-positive bacteria were not encountered in this study, but a high degree of resistance to amikacin, gentamicin, cefotaxime and ceftazidime varying from 31.9% to 44.5% was detected among gram-negative bacilli. This is in sharp contrast to the previously reported series of high sensitivity of intraocular isolates of gram-negative bacilli to these antibiotics.[2, 23] While determining choice antibiotics for intravitreal injections, the high degree of in-vitro resistance of gram-negative bacilli to gentamicin, amikacin, cefotaxime and ceftazidime should be considered. The resistance of gram-positive bacteria to ceftazidime and ciprofloxacin was, however comparatively lower than that found in the endophthalmitis vitrectomy study (EVS).[2]

In summary, our study emphasizes the importance of gram-negative bacilli and fungi in causing postoperative endophthalmitis, which in fact represents a failure to maintain absolute sterility during surgery. This is in agreement with the recently published prospective study on the microbial spectrum and susceptibilities in postoperative endophthalmitis, where a relatively high prevalence of gram-negative bacteria and fungi causing postoperative endophthalmitis was noted.[6] In addition, our study also stresses the high degree of inherent resistance to aminoglycosides and ceftazidime among gram-negative bacteria. As suggested earlier,[35] a national reporting system to identify factors for future prospective studies may be an important first step, though it is possible that surgeons may be resistant or reticent to participate in such a venture. Our findings in this study appear very alarming, in light of which it is all the more essential for such a system to be introduced. This will help us form a clear picture of the incidence and aetiology of postoperative endophthalmitis in India.


1Forster R.K, Abbott RL, Gelender H. Management of infectious endophthalmitis. Opthalmology 1980;87:313-15.
2Han DP, Wisniewski SR, Wilson LA, Barza M, Vine AK, Doft BH et al. Spectrum and susceptibilities of microbiologic isolates in the endophthalmitis vitrectomy study. Am J Ophthalmol 1996;122:1-17.
3Puliafito CA, Baker AS, Haaf J, Foster CS. Infectious endophthalmitis. Ophthalmology 1982;89:921-29.
4Driebe WT Jr, Mandelbaum S, Forster RK, Schwartz LK, Culbertson WW. Pseudophakic endophthalmitis: diagnosis and management. Ophthalmology 1986;93:442-48.
5Fisch A, Salvanet A, Prazuck T, Forestier F, Gerband L, Coscas G, et al. Epidemiology of infective endophthalmitis in France. Lancet 1991;338:1373-76.
6Kunimoto DY, Das T, Sharma S, Jalali S, Majji A, Gopinathan U, et al. Microbiologic spectrum and susceptibility of isolates. Part I. Postoperative endophthalmitis. Am J Ophthalmol 1999;128:240-42.
7Sharma T, Gopal L, Therese L. Endophthalmitis in eyes following vitrectomy. Ophthalmic Surg & Lasers 1998;29:857-59.
8Mahajan V M. Post-operative ocular infections: An analysis of laboratory data in 750 cases. Ann Opthalmol 1984;16:847-48.
9Megevand GS, Pournares CJ. Current approach to postoperative endophthalmitiy. Br J Ophthalmol 1997;81:1006-15.
10Allen SD. Anaerobic bacteria. In: Lennete H, editor. Manual of Clinical Microbiology. 4th edition. Washington D.C., American Society for Microbiology 1985. PP 413-72.
11Baron EJ, Peterson LR, Finegold SM, editors. Bailey and Scott's Diagnostic Microbiology. 9th edition. St. Louis: Mosby Co. 1994. PP 79-136.
12Collee JG, Miles RS, Watt B. Tests for identification of bacteria. In: Collee JG, Fraser AG, Marmion BP, Simmons A. Mackie and McCartney's. Practical Medical Microbiology. 14th Ed. Edinburgh: Churchill Livingstone, 1996. pp 131-50.
13Bauer AW, Kirby WMM, Sherris JC. Antibiotic susceptibility testing by a standardised single disc method. Am J Clin Pathol 1966;45:493-96.
14Reddy PR. Incidence of post-operative endophthalmitis in India. Indian J Ophthalmol 1987;35:60-62
15Sharma T, Gopal L, Parikh S, Badrinath SS, Madhavan HN, Mukesh BN. Endophthalmitis caused by anaerobic bacteria. Indian J Ophthalmol 1995;44:191-94.
16Allen HF, Mangiaracine AB. Bacterial endophthalmitis after cataract extraction II. Incidence in 36,000 consecutive operations with special reference to preoperative topical antibiotics. Arch Ophthalmol 1974;91:3-7.
17Allen H F, Mangiaracine A B. Bacterial endophthalmitis after cataract extraction. Arch Opthalmol 1964;72:454-62.
18Rowsey JJ, Newsome DL, Sexton DJ, Harms WK. Endophthalmitis - Current approaches. Ophthalmology 1982;89:1055-66.
19Fox GM, Joondeph BC, Flynn HW Jr, Pflugfelder SC, Roussel TJ. Delayed onset pseudophakic endophthalmitis Am J Ophthalmol 1991;111:163-73.
20Olson JC, Flynn HW Jr, Forster RK, Culbertson WW. Results in treatment of postoperative endophthalmitis. Ophthalmology 1983;90:692-99.
21Kattan HM, Flynn HW Jr, Pflugfelder SC, Robertson C, Forster RK. Nosocomial endophthalmitis survey: Current incidence of infection after intraocular surgery. Ophthalmology 1991;98:227-38.
22Heaven CJ, Mam PJ, Boase DL. Endophthalmitis following extracapsular cataract surgery: A review of 32 cases. Br J Ophthalmol 1992;76:419-23.
23Irvine WD, Flynn HW Jr, Miller D, Pflugfelder SC. Endophthalmitis caused by Gram negative organisms. Arch Ophthalmol 1992;110:1450-54.
24Ficker L, Meredith TA, Wilson LA, Kaplan HJ, Kozarsky AM. Chronic bacterial endophthalmitis. Am J Opthalmol 1987;103:745-48.
25Peyman GA, Raichand M, Bennett TO. Management of endophthalmitis with pars plana vitrectomy Br J Ophthalmol 1980;64:472-75.
26Le Francois M, Baum JL Flavobacterium endophthalmitis following keratoplasty: Use of a tissue culture medium stored cornea. Arch Ophthalmol 1976;94:1907-9.
27Ormerod LD, Ho DD, Becker LE, Cruise RJ, Grohar HI, Paton BG, et al. Endophthalmitis caused by the coagulase-negative Staphylococci. Opthalmology 1993;100:715-23.
28Mandelbaum S, Forster RK, Gelender H, Culbertson WW. Late onset endophthalmitis associated with filtering blebs. Ophthalmology 1985;92:964-72.
29Forster RK. Endophthalmitis - Diagnostic cultures and visual results. Arch Ophthalmol 1974;92:387-92.
30Pflugfelder SC, Flynn HW, Zwickey TA, Forster RK, Tsiligianni A, Culbertson WW et al. Exogenous fungal endophthalmitis. Ophthalmology 1988;95:19-30.
31Boldt HC, Pulido JS, Blodi CF, Weingeist TA. Rural endophthalmitis. Ophthalmology 1989;96:1722-26.
32Alfaro DV, Roth D, Ligett PE. Post traumatic endophthalmitis: Causative organisms, treatment and prevention. Retina 1994;14:206-11.
33Speaker MG, Milch FA, Shah MK. The role of external bacterial flora in the pathogenesis of acute postoperative endophthalmitis. Ophthalmology 1991;98:639-49.
34Bannerman TL, Roden DL, McAllister SK, Miller JM, Wilson LA, for the EVS group. Source of Coagulase negative Staphylococci in the Endophthalmitis Vitrectomy Study. Arch Ophthalmol 1997;115:357-61.
35Hughes DS, Hill RJ. Infectious endophthalmitis after cataract surgery. Br J Ophthalmol 1994;78:227-32.