[Summary text]
Postoperative endophthalmitis is probably the most feared complication after cataract surgery.
Randomized controlled clinical trials and observational studies were considered.
Studies performed on patients undergoing surgery for age-related cataract were included.
The primary outcome was postoperative endophthalmitis both culture proven and non-proven cases.
Retrospective, observational study.
Compares two time periods with different use of intracameral antibiotic
(vancomycin)
Country and clinic:Ophthalmic Department at Warrington Hospital, England
Patients undergoing cataract surgery by phacoemulsification
Demographics of study groups: no data on age, sex distribution or preoperative
visual acuities in the study groups was provided
Patient assignment procedure: patients operated before December 31, 2000 did
not received intracameral vancomycin. Patients operated after January 1rst 2001
received intracameral vancomycin.
Group 1: no intracameral antibiotic
Group 2: intracameral vancomycin (1 mg in 0.1 ml of saline)
All patients received 5% povidone-iodine preoperatively and prednisolone and
fusidic acid postoperatively
Endophthalmitis rate was 13/3904 in Group 1 and 1/12702 in Group 2
The study did not receive any financial support. No conflict of interests reported
Prospective, observational study
Compares two different time periods with different use of intracameral antibiotic
(cefuroxime)
Country and clinic: Ophthalmology Department at Dyputren Hospital, France
Patients undergoing cataract surgery by phacoemulsification
Demographics of study groups: no data on age, sex distribution or preoperative
visual acuities in the study groups was provided
Patient assignment procedure: Patients operated before May 31 2006 did not
receive cefuroxime. All patients operated after June 1rst 2006 received
cefuroxime.
Group 1: no intracameral antibiotic
Group 2: intracameral cefuroxime (1 mg in 0.1 ml of saline)
All patients received 5% povidone-iodine preoperatively. Dexamethasone,
tobramycin, diclofenac and tropicamide was used postoperatively in all patients
Endophthalmitis rate was 35/2826 in Group 1 and 1/2289 in Group 2
No information on financial support or conflict of interests provided
RCT
Compares endophthalmitis rate in patients receiving intracameral cefuroxime
versus no intracameral cefuroxime
2x2 factorial design with intracameral cefuroxime and topical levofloxacin in the 4
treatment groups
Country: multicenter study across Europe (Austria, Belgium, Germany, Italy,
Poland, Portugal, Spain, Turkey, UK)
Patients undergoing cataract surgery by phacoemulsification
Demographics of study group: 58% women. Median age of women 75 yrs.
Median age of men 73 yrs
Randomization procedure: computer generated randomization
No of patients excluded after allocation: not reported
No of patients lost to follow-up: 324 (2%)
Blinding of patients: yes
Blinding of outcome assessment: yes
Group 1: intracameral cefuroxime 1 mg in 0.1 ml saline
Group 2: no intracameral cefuroxime
All patients received povidone-iodine 5% drops in the conjunctival sac and onto
the cornea preoperatively. All patients used levofloxacin 0.5% eyedrops 4 times
daily for 6 days postoperatively
Endophthalmitis rate was 5/8108 in Group 1 and 24/8103 in Group 2 (all
endophthalmitis cases (culture and PCR proven + non-proven/intent to treat)
The study was funded by ESCRS with Santen GmbH supplying levofloxacin and
placebo eye drops free of charge along with an unrestricted educational grant.
Observational study
Compares use of intracameral cefuroxime versus no cefuroxime from 2005
through 2010
Country and clinics: multicenter study comprising 98% of all cataract surgeries
performed in Sweden
Patients undergoing cataract surgery
Demographics of study population: 393.298 patients between 10 and 84 yrs,
71.483 patients older than 85 yrs. 61.9% women
Patient assignment procedure: surgeons choise whether patients received
cefuroxime or not
Group 1: intracameral antibiotic
Group 2: no intracameral cefuroxime
Endophthalmitis rate was 121/455054 in Group 1 (121/455054 patients receiving
intracameral cefuroxime and 2/6897 receiving moxifloxacin) and 11/2804 in
Group 2
The study received financial support from the Swedish Association of Local
Authorities and Regions
Observational study
Compares two time periods with different intracameral antibiotic regimes
Country and clinic: Department of Ophthalmology Hospital General de
L´Hospitalet, Barcelona, Spain
Patients undergoing cataract surgery by phacoemulsification
Demographics of Group 1: patient age (mean (SD)) was 73.8 (8.8) in 2002 and 74.2 (8.5) in 2003. 60.4% women
Demographics of Group 2: patient age (mean (SD)) was 74.4 (8.3) in 2004, 73.8
(8.4) in 2005, 73.2 (8.4) in 2006, 73.6 (8.3) in 2007. 59.3% women
Patient assignment procedure: patients operated before January 2004 did not
receive intracameral cefazolin, patients operated after January 2004 received
intracameral cefazolin.
Group 1: no intracameral antibiotics
Group 2: intracameral cefazolin (2.5 mg in 0.1 ml Ringer lactate)
All patients received 5% povidone-iodine preoperatively. Postoperatively
dexamethasone-tobramycin, oxofloxacin and diclofenac was used in all patients.
Endophthalmitis rate was 25/5 930 in Group 1 and 6/12 649 in Group 2.
Outbreak of endophthalmitis in Group 1 included 3 cases caused by
contamination of air-condition.
No conflict of interests reported.
Prospective, observational study
Compares two time periods with different use of intracameral cefuroxime
Country and clinic: University Hospital Fundación Alcorcón, Madrid, Spain
Patients undergoing cataract surgery by phacoemulsification
Demographics of study population: no data on age, sex distribution or
preoperative visual acuities in the two groups was provided
Patient assignment procedure: patients operated before October 2005 did not
receive intracameral cefuroxime. Patients operated after October 2005 received
intracameral cefuroxime
Group 1: no intracameral antibiotic
Group 2: intracameral cefuroxime (1 mg in 0.1 ml saline)
All patients received 5% povidone-iodine preoperatively. Ofloxacin, prednisolone
and ketorolac was used postoperatively in all patients
Endophthalmitis rate was 39/6595 in Group 1 and 3/7057 in Group 2.
No information on financing of the study of conflict of interests provided
Prospective, observational study
Compares use of intracameral use of cefuroxime versus no cefuroxime from
January 2002 through December 2004
Country and clinics: multicenter, nationwide study in Sweden
Patients undergoing cataract surgery by phacoemulsification
Demographics of study groups: no data on age, sex distribution or preoperative
visual acuities in the two groups was provided
Patient assignment procedure: surgeons choise whether patients received
cefuroxime (99%) or not (1%)
Group 1: no intracameral antibiotic
Group 2. intracameral cefuroxime (1 mg)
Patients received preoperative chlorhexidine rinsing, <5% received postoperative
regime of topical antibiotics
Endophthalmitis rate was 8/2315 in Group 1 and 100/223156 in Group 2
The study received financial support from public and private funds. No conflict of
interests reported.
Prospective, observational study
Compares two time periods with different intracameral antibiotic regimes
Country and clinic: 38 public and private hospital in Barcelona, Spain
Patients undergoing cataract surgery by phacoemulsification
Demographics of Group 1: age (mean (SD)) was 69.8 (7.6), 58.0% women
Demographics of Group 2: age (mean (SD)) was 66.2 (7.8), 58.0% women
Patient assignment procedure: patients operated before December 2002 did not
receive intracameral cefazolin. Patients operated after January 2003 received
intracameral cefazolin
Group 1: no intracameral antibiotics
Group 2: intracameral cefazolin (1 mg in 0.1 ml)
All patients recived 5% povidone-iodine conjunctival wash preoperatively.
Tobramycin, dexamethasone and diclofenac was used postoperatively in all
patients.
Endophthalmitis rate was 76/11696 in Group 1 versus 7/13305 in Group 2
No financial support was received. No conflict of interests reported
Observational study
Compares different use of intracameral antibiotic in different time periods
Country and clinic: Kaiser Permanente, Diable Service Area, California
Patients undergoing cataract surgery by phacoemulsification
Demographics on study groups was not reported
Patient assignment procedure: before September 2007 no patient received
intracameral antibiotics. Between September 2007 and December 2009 patients
received intracameral antibiotics depending on surgeon preference. After
January 2010 all patients received intracameral antibiotics.
Group 1: no intracameral antibiotic
Group 2: intracameral antibiotic (cefuroxime (1 mg in 0.1 ml): n=10644,
moxifloxacin (0.1% in 0.1 ml): n=1890, vancomycin (1 mg in 0.1 ml): n=75)
All patients received 5% povidone-iodine preoperatively. Postoperative topical
antibiotic use changed over time.
Endophthalmitis rate was 15/3655 in Group 1 and 4/12609 in Group 2
No conflict of interests reported
Retrospective, observational study
Compares two time periods with different intracameral antibiotic regimes
Country and clinic: Tan Tock Seng Hospital, Singapore
Patients undergoing cataract surgery by phacoemulsification and ECCE (12-15%
of cases)
Demographics of the study groups was not reported
Patient assignment procedure: patients operated before June 2006 received
subconjunctival cefazolin. Patients operated after July 2006 received
intracameral cefazolin
Group 1: no intracameral antibiotics but with subconjunctival injection of cefazolin
(1 mg in 0.1 ml)
Group 2: intracameral cefazolin (1.0 mg in 0.1 ml)
All patients received chloramphenicol eyedrops 1 day preoperatively. A 5%
povidone-iodine conjunctival wash was applied preoperatively. All patients
received a subconjunctival injection of gentamicin (0.2 ml of 80 mg/2 ml) and
dexamethasone (0.4 mg in 0.1 ml). Topical chloramphenicol-dexamethasone or
levofloxacin-prednisolone was used postoperatively in all patients
Endophthalmitis rate was 19/29539 in Group 1 and 2/20638 in Group 2
No conflict of interests reported
Observational study
Compares two different periods with different intracameral use of cefuroxime
Country and clinic: Cape Town, South Africa
Patients undergoing cataract surgery
Demogrpahics of study population: not reported
Patient assignment procedure: patients operated before july 2006 did not receive
intracameral cefuroxime, patients operated later did receive cefuroxime
Group 1: intracameral cefuroxime
Group 2: no intracameral cefuroxime
All patients received 2.5% povidone-iodine conjunctival wash
Endophthalmitis rate was 3/3971 in Group 1 versus 23/4219 in Group 2
Information regarding conflict of interests or funding was not provided
Observational study comparing different intracameral antibiotic use
Country and clinics: nationwide study in Sweden
Patients undergoing cataract surgery
Demographics of study groups was not reported
Patient assignment procedure: patients received intracameral cefuroxime
depending on the choise of the clinic where they were operated
Group 1: no intracameral antibiotic
Group 2: intracameral cefuroxime
Preoperative povidone-iodine was not used since it was not available in Sweden.
No information on topical postoperative antibiotics
Endophthalmitis rate was 15/6805 in Group 1 and 81/151874 in Group 2
Information regarding conflict of interests or funding was not provided
Retrospective, observational study
Compares two time periods with different use of intracameral antibiotic
Country and clinic: Sunderland Eye Infirmary, UK
Patients undergoing cataract surgery
Demographics of study groups was not reported
Patient assignment procedure: Before November 16, 2003 patients received
subconjunctival cefuroxime. After November 17, 2003 patients received
intracameral cefuroxime.
Group 1: no antibiotic
Group 2: intracameral cefuroxime (1 mg in 0.1 ml saline)
All patients received 5% povidone-iodine preoperatively. Prednisolone acetate
was used postoperatively
Endophthalmitis rate was 7/427 in Group 1 and 8/17318 in Group 2
No conflict of interests reported
Does not compare endophthalmitis rates in patients receiving/not receiving
intracameral antibiotics since no intracameral antibiotics was used
No information regarding the use of intracameral antibiotic was provided
Preliminary report of the same data included in the Garat 2009 publication
Does not provide information regarding the use of intracameral antibiotic
Does not provide information regarding the use of intracameral antibiotic
Does not provide information regarding the use of intracameral antibiotic
Does not provide information regarding the use of intracameral antibtiotic
Intracameral antibiotics was not used in any case during cataract surgery
All patients received the same intracameral injection of moxiflocaxin from a bottle
designed for topical use
Intracameral antibiotics was not used in any case during cataract surgery
Population-based study. Does not evaluate the effect of intracameral antibiotics
Retrospective study. No cases and no controls received intracameral antibiotics
Description of endophthalmitis cases. Does not compare outcome between
groups receiving/not receiving intracameral antibiotics
Description of endophthalmitis cases. Does not compare outcome between
groups receiving/not receiving intracameral antibiotics
Non-randomized trial evaluating the safety of intracameral cefuroxime
Experimental study examining the pharmacokinetics of intracameral vancomycin
Retrospective study. No cases or controls received intracameral antibiotics
Case-control study. Does not compare outcome between groups receiving/not
receiving intracameral antibiotics
Patient population is the same (but with smaller number) as in the included study
Romero-Aroca 2012
Case reports. Does compare outcome in patients receiving intracameral versus
no intracameral antibiotic
Intracameral antibiotic was not used in any patient
Case-control study
No information regarding the use of intracameral antibiotic was provided
Retrospective case-control. Does not evaluate the effect of intracameral
antibiotics
Intracameral antibiotic was used in any case during cataract surgery
Selection bias (biased allocation to interventions) due to inadequate generation of a randomised sequence
Not a randomized study. Group assignment depended on
when the patient had surgery: before January 2001 no
intracameral vancomycin was used, after January 2001 all
patients received intracameral vancomycin
Not a randomized study. Group assignment depended on
when the patient had surgery: before June 2006 no
intracameral cefuroxime was used, after June 2006 all
patients received intracameral cefuroxime
"A database permitting direct data entry in both the operating
theaters and the outpatient clinics was created... Patient
randomization was performed by the database"
Not a randomized study. Not described how the decision to
use or not use intracameral cefuroxime was made
Not a randomized study
Not a randomized study
Not a randomized study
Not a randomized study
Not a randomized study
Not a randomized study
Not a randomized study
Not a randomized study
Not a randomized study
Selection bias (biased allocation to interventions) due to inadequate concealment of allocations prior to assignment
Surgeons knew the date when the procedure (intracameral
vancomycin) changed
Surgeons knew the date when the procedure (intracameral
cefuroxime) changed
Not described
Not reported
Surgeons knew the date when the procedure (to give
intracameral cefazolin) changed
Surgeons knew the date when the procedure (to give
intracameral cefuroxime) changed
Not reported how the decision to use or not use
intracameral cefuroxime was made
Surgeons knew the date when the procedure
(intracameral cefazolin) changed
Long transition phase from not using intracameral antibiotic to
a consequent use of intracameral antibiotic in all patients
Surgeons knew the date when the procedure
(intracameral cefazolin) changed
Not reported how the decision to use/not use
intracameral antibiotic was made
Surgeons knew the date when the procedure
(intracameral cefuroxime) changed
Surgeons knew the data when the procedure
(intracameral cefuroxime) changed
Performance bias due to knowledge of the allocated interventions by participants and personnel during the study
Not a blinded study
Not a blinded study
"The use of cefuroxime was not masked, surgeons were
requested to give patients who had been randomly
allocated... the intracameral injection at the end of surgery"
Not a blinded study
Not a blinded study
Not a blinded study
Not a blinded study
Not a blinded study
Not a blinded study
Not a blinded study
Not a blinded study
Not a blinded study
Not a blinded study
Detection bias due to knowledge of the allocated interventions by outcome assessors
Not reported
Not reported
"While the study was running, the DMC would remain
masked. The treatment groups would be identified to the
DMC only by code"
Not reported
Not reported
Not reported
Not reported
Not reported
Not reported
Not reported
Not reported
Not reported
Not reported
Attrition bias due to amount, nature or handling of incomplete outcome data
Not likely to have occured
Not likely to have occured
"2% were lost to follow-up, 68 were omitted because they did
not have the planned surgery or withdrew their consent.. 240
patients could not be included... due to protocol deviation"
Not likely that endophthalmitis cases were missed
Not likely that cases of endophthalmitis was missed
Not likely that cases of endophthalmitis was missed
Not likely that cases of endophthalmitis missed
registration
Not likely that cases of endophthalmitis was missed
Not likely that cases of endophthalmitis were missed
Not likely that endophthalmitis cases were missed
Not likely that cases of endophthalmitis were missed
Not likely that cases of endophthalmitis were missed
Not reported how endophthalmitis cases were
identified
Reporting bias due to selective outcome reporting
Not likely that endophthalmitis cases should have been
missed. Important outcome is reported
Not likely that endophthalmitis cases should have been
missed. Important outcome is reported
Not likely. Important outcome is reported
Important outcome reported
Important outcome reported
Important outcome was reported
Important outcome reported
Important outcome reported
Important outcome reported
Important outcome reported
Important outcome reported
Important outcome reported
Important outcome reported
Bias due to problems not covered elsewhere in the table
Not likely in this study
Not likely in this study
Study was terminated before the intended number of patients
were included due to a large difference in number of
endophthalmitis rates between treatment groups.
Not likely in this study
Not likely in this study
Not likely in this study
Not likely in this study
Not likely in this study
Not likely in this study
Not likely in this study
Not likely in this study
Not likely in this study
Not likely in this study