POSTOPERATIVE INTRAOCULAR INFECTION                                      Return to Contents page

Denis Stark                                                                                                 30/07/99

Introduction

Bacterial intraocular infection is fortunately a rare but devastating complication of cataract surgery.

It is, by common usage, referred to as Endophthalmitis. Duke Elder[1] defines Endophthalmitis as "vitreous cavity suppuration confined to the posterior segment" whereas when the organisms diffuse throughout the globe the condition should be "Panophthalmitis". Current authors do not differentiate.

As we approach the second millennium a certain confidence regarding surgical outcome exists. Surgeons discuss the possibility of bilateral simultaneous cataract surgery. Do we control this unpredictable complication sufficiently for this?
 
 

Incidence:

Historical

Duke Elder[2] reports prior to 1945 a combination of series of that era showed an incidence of 11 per 1000 operations (1.1%). Marked improvement occurred in the following 20 years. Review of a series of reports in1966, each of more than 1000 reported operations, showed an average incidence of 2 per 1000 operations (0.2%).

As a word of warning Forster[3] advised that Christy and Lall in a series of 77093 Cataract operations in Pakistan, 1973, with an incidence of 0.5% infection, had many "1000 consecutive procedures" without infection. This means that smaller reported numbers without infection are meaningless.
 
 

Current

With all the advances of cataract surgery today our results have advanced little. Infection rates have improved little in 30 years whether we look at the results from Sweden[4] (0.26%), Canada [5](0.22-0.3%), USA[6] (0.12%-0.17%), USA[7] (0.093%), Denmark[8] (0.18%). One survey from USA, Aaberg 1998[9] showed a marginal increase in infection rate from their previous survey 10 years previously. Australasia's figures have to be viewed with caution because of the survey technique[10](0.11%). There do not appear to be other reliable figures for Australia.

Table 1 Incidence of Postoperative Endophthalmitis
 
 

* 1980-84 series had an incidence of 0.073%- current series 1990-1994

Bacteriology

The bactericidal properties of tears have long been noted. Ridley[12] is credited with as identifying the responsible enzyme- a lysozyme. The concentration of this is effected by dry eyes, epiphora, and infections. The presence of this enzyme together with the lower temperature, the mechanical action of blinking and the sluicing effect of lacrimal flow all are thought to contribute to the low rate of postoperative infections.

Staphylococci are considered more prevalent in hotter climates. No recent studies exist but several earlier studies are reported by Duke Elder[13]. 62% Queensland, (Gibson 1951), 95% Egypt (Kamel 1949) and 34% England Smith 1954. No recent studies from Queensland have been found.

More recently conjunctival cultures performed preoperatively have confirmed a high prevalence of organisms.

Mistelberger et al[14] studied 700 consecutive patients- 76% of patients had a positive culture and coagulase negative Staphylococcus was present in 75% of those. A further study of the conjunctival flora showed relatively no change in organisms even in the association of lacrimal duct obstruction[15] Gram +ve in 62% (Staph Epidermidis in 27%) and Gram -ve in 20%.

Table 2 Mistelberger Operative cultures.
 
 

There were no cases of Endophthalmitis in his 700 patients in spite of the bacterial counts. Nor did surgical technique, preoperative antibiotics or lacrimal lavage alter significantly the bacterial count.

Tervo[16] similarly demonstrated bacteria on the lid margins, conjunctivae, lacrimal lake and anterior chamber:

Table 3 Tervo Operative Cultures
 
 

The bacteria cultured from cases of postoperative infection tend to follow the expected distribution[17]. Coagulase -ve Staph predominates (60%) of those cultured. Other organisms included Staphylococcus Aureus, Streptococci and Gram -ve Bacteria). In this series organisms with a low virulence (Coagulase -ve Staph, Proprionibacterium Acnes) had a better visual prognosis. This scatter of organisms is the rule in most studies*[18].

Table 4 Organisms Cultured
 
 

Factors perhaps contributing to Infection

No study admits inadequate surgical practices or inadequate sterile technique is implicated as a factor in the incidence of Endophthalmitis. There is no clear answer for prevention.

Bacteria are introduced into the eye during surgery. If the innoculum is sufficient and the immune system is overwhelmed then an infection will occur.

Peri-operative Prophylaxis

Morlet's Australasian review has limited ability to evaluate the incidence of the disorder but he has displayed well the myriad practices that are currently advocated by this region's ophthalmologists. His final conclusion supported the older ophthalmologist or those who performed more than 300 cataract procedures per year. A cynical appraisal may suggest that this results from poor recall in the first group and a poorer recall of patients in the latter. At least, as he suggests, it confirms that a common sense approach is the way to go.
 
 

Surgical

A number of specific surgical occurrences (Jeddi (France)[19], Ozer-Arasli(Germany)[20] , Bainbridge[21]) are thought to increase the risk of this occurring.

The following is a composite list:

Prolonged surgery

Rupture of posterior capsule

Vitrectomy

Retention of lens fragments

Wound dehiscence

Polypropylene Haptics (Bainbridge)

Left eye (Jeddi)
 
 

But many surveys did not find full statistical support for these findings. ( Mistlberger[22], Montan[23]). The former considering surgical technique, preoperative antibiotics and preoperative lacrimal system irrigation had no significant statistical effect on bacterial contamination.

The latter largely agreed but felt wound abnormality was significant and the use of a heparinised-coated lens was of value.

No survey compared sutured wounds with non sutured wounds.

The Bainbridge study with polypropylene loops was small and the incidence of infection very high (3.3%) but this has previously been reported.

No consistent variation in incidence has been seen between Phako extraction and Extracapsular extraction.

The use of Poviodine-Iodine was felt to be significant by Bohigian but again was not supported statistically.

The use of intra-operative antibiotics either in the irrigating solution[24] , the operating field, subconjunctival or intracameral have also not been supported by studies of bacteria counts pre and postoperatively.
 
 

Patient

Three factors immerge in the many papers offered. The immune state appears repeatedly as an important factor. Age is more frequently debated as a causative factor. Diabetes is mentioned at times but not consistently as a factor.

Outcome of Infection

The visual prognosis of affected patients is still poor in many cases.

The bacteria involved influences this. Surgical management has improved since the days of Duke Elder when many of these eyes were lost. Today eyes are not so frequently lost but visual prognosis remains poor.

Today, at time of presentation, the visual acuity of the affected eye is the single most predictive factor[25] of the final visual function.

The Endophthalmitis Vitrectomy Study showed as would be expected better results with the less virulent organisms-

Table 5 Visual Prognosis v. Organism
 
 

Somani[26] reported visual acuity of 20/50 or better in 46% of the Coagulase -ve Staph group but only 10% of those affected by the more virulent organism. 18% of their patients ended with no perception of light. The culture negative group had a better prognosis.

Management of Infection

This is beyond the scope of this document. But early diagnosis and prompt management is imperative. Appropriate postoperative review is necessary to facilitate this early diagnosis. Surgical evacuation, culture and antibiotic cover is mandatory.
 
 

Conclusions

No surgical team should be complacent.

The bacteria usually involved are those organisms usually inhabiting the conjunctival sac.

Recurrent infection by less common bacteria could point to an external source.

All normal surgical prophylactic procedures should be followed.

Orthopaedic teams have an immense respect for sterile procedure. Ophthalmic teams perhaps could review the standard of care of one of the best Orthopaedic units in their area.

Routine surveillance should be performed of practices by all personnel and possible sources of contamination considered- e.g.

Medications- reuse of medications vials, bottles, drop or ointments.

Hands- routine washing of hands between patients for all personnel- surgical, operative staff, scouts, anaesthetists, technicians, postoperative

Theatre wear - Should include masks, gloves, gowns, shoe cover Appropriate cover of theatre dress for external visits.

Remember Forster's warning, "1000 infection free cases does not give room for complacency." There is no "one way" to avoid infection but opportunities of contamination should be minimised?

References