ENDOPHTHALMITIS AFTER CATARACT SURGERY

An analysis of the methods used for reporting endophthalmitis after cataract surgery in a UK teaching hospital indicate that consulting the records of the Microbiology Department can be the most efficient approach, said Miss Aabgina Shafi, Department of Ophthalmology, Bradford Royal Infirmary, Bradford, UK.

“Endophthalmitis remains a devastating complication of cataract surgery. It can happen in outbreaks, making early recognition of an increasing trend in incidence of utmost importance,†Miss Shafi said at the XXX Congress of the ESCRS . Ms Shafi and her associates compared the different systems of data entry used at the Bradford Royal Infirmary to find the most failsafe method of identifying endophthalmitis cases occurring after cataract surgery. They analysed the hospital’s records from a five-year period, from January 2007 to January 2012.

Overall, from all the data sources available at their centre, they found 25 cases of endophthalmitis. Nine were bleb-related, and six occurred after phacoemulsification procedures, one of which was carried out in Pakistan. There were also four endogenous cases, three cases that occurred following intravitreal injection of triamcinolone acetonide, one which occurred following pars plana vitrectomy and one that was a result of trauma. “Out of 11,535 cataract surgeries performed over five years, there were five cases of endophthalmitis. That gives us an incidence of 0.04 per cent,†Miss Shafi said.

Data collecting software disappoints

Miss Shafi noted that the Medisoft software they used for the general recording of procedures gave particularly poor results. In fact, it only identified two cases, despite the fact that it logged nearly every cataract case performed, she said “I still hold that Medisoft would be the most practical method. It is available at the click of a finger. However, it is dependent on us actually inputting the data. Therefore the recommendation in our department is to log on every patient who has presumed endophthalmitis. We can always remove cases later if the diagnosis changes,†Miss Shafi added.

Critical incident reporting, which should have identified all cases according to the NHS policies, was even less effective and did not identify a single endophthalmitis case for the whole of the period under study. A review of the coding logs yielded a sample of 98 cases. Excluding the mis-codes reduced the sample to 68 cases and crossreferencing the sample with intervention codes yielded 21 cases, including some repeats, she noted.

The missed cases included two cases of endogenous endophthalmitis, one case that had received intravitreal triamcinolone acetonide, and one case that occurred following phacoemulsification. The information provided by the Microbiology Department identified 18 of 21 endophthalmitis cases. Those missed included one post-phacoemulsification cases, one trauma case and one case that occurred following phacotrabeculectomy. The pharmacy records of vancomycin usage identified 15 of 25 endophthalmitis cases. The missed cases included three that occurred following phacoemulsification and one that followed pars plana vitrectomy

“Consulting the microbiology records would be my preference because it gives us the least number of case notes to look at. Over the five years it really missed only one phaco-related endophthalmitis case and that was because during that year we sent some of our cases to an outside laboratory,†Miss Shafi added.