Infection prevention and control factors associated with post-cataract surgery endophthalmitis - a review of the literature from 2010 – 2023

Summary

Patients undergoing cataract surgery are at risk of post-cataract surgery endophthalmitis (PCSE), a sight-threatening complication. Cataract surgery is a relatively straightforward and quick procedure often performed under local anaesthetic. It is therefore simple to scale up to reduce the currently long waiting times, but it is important to maintain patient safety when considering high throughput surgery. This literature review aimed to identify appropriate infection prevention and control (IPC) measures to support increased throughput of cataract surgery in Scotland. Database searches were conducted using Medline and Embase from 2010 to 2023. Further hand-searching was also performed. The organisms associated with PCSE and IPC factors relevant to PCSE were analyzed. A range of microorganisms was associated with PCSE, where outbreak reports were most associated with Gram-negative bacteria and fungi, whereas retrospective chart reviews were most associated with Gram-positive bacteria. IPC risk factors identified were related to the built environment and issues with sterilization. Specifically, the sources of outbreaks included failures in the ventilation system, as well as contaminated ophthalmic solutions, surgical instruments, and medications. The factors identified in this review should be considered when implementing high throughput cataract surgery to ensure that patient safety is maintained.

Introduction

With an estimated global prevalence of 17.2%, cataracts are the leading cause of blindness and the second highest cause of moderate or severe visual impairment globally. [1,2] It is the opacification of the lens which causes a loss of lens transparency. [3] The prevalence of cataracts increases with age and is associated with a plethora of negative health impacts amongst the elderly population, including a higher risk of dementia, falls, road traffic accidents, a significant reduction in quality of life, and a higher risk of mortality. [[3], [4], [5], [6]].

Surgery is the most effective treatment for cataracts and is one of the commonest elective surgical procedures in the world. [3] In a small proportion of cases, however, endophthalmitis may occur. Endophthalmitis is a sight-threatening inflammation of the inner layers of the eye caused by intraocular colonization by micro-organisms. [7,8] Post-cataract surgery endophthalmitis (PCSE) can be classified based on the interval between surgery and infection. It is categorized as acute when infection occurs shortly after the surgery – usually within one to two weeks but may occur up to six weeks after. [9,10] Chronic cases are those that manifest several weeks or months after surgery, usually after six weeks. [10,11] The use of antiseptic agents and other interventions has led to a huge decline in the incidence of endophthalmitis over the last few decades. [12].

As a result of the disruptions caused by the COVID-19 pandemic, there have been delays and increased waiting times for elective surgeries. As cataract surgery is a straightforward day surgery, measures can be implemented to increase throughput and rapidly decrease waiting times. However, given the potential risk of PCSE, there must be assurance that patient safety is maintained. [13] Although recent reviews of the incidence and aetiology of endophthalmitis have been published, they focused on outbreaks and clusters, therefore sporadic single cases were not captured. [14,15] Therefore, a literature review was commissioned by the Scottish National Cataract Short Life Working Group (SLWG) to evaluate the scientific literature to identify appropriate infection prevention and control (IPC) measures to support increased throughput of cataract surgery in Scotland. The following research questions were considered:

• i.Which organisms are associated with post-cataract surgery endophthalmitis (PCSE)?
• ii.What factors related to infection prevention and control are associated with post-cataract surgery endophthalmitis?

Methods

A draft review protocol was approved by the Scottish National Cataract SLWG, commissioned to address the delays in cataract surgery in Scotland. A database search was undertaken using two comprehensive search strategies on Medline and Embase (see Appendix 1). The search strategies were developed by a single author and peer-reviewed by a Librarian. Due to time constraints, all searches were limited to articles published between 1 January 2010 and 14 April 2023. Hand-searching of reference lists and a search of online resources was carried out to identify grey literature.

Titles and abstracts were screened by a single author who also conducted the full-text screen. Evidence was critiqued by a single reviewer using the SIGN50 principles however critical analysis tools were not used. [16].

The final version of the review was approved by the SLWG following consultation.

Inclusion criteria

Studies included are case studies, case series, outbreak reports and retrospective chart reviews published in the English language regardless of the country where they were published. Retrospective chart reviews assess all cases of PCSE in a particular location over a defined period irrespective of the causative organism. Only studies reporting post-cataract surgery cases of endophthalmitis with positive microbial culture were considered.

Exclusion criteria

Studies were excluded if they reported PCSE cases with negative patient microbial cultures or endophthalmitis secondary to trauma or a procedure other than cataract surgery. Also excluded were studies focused on non-human subjects, intervention bundles, or non-infection prevention and control (IPC) factors such as prophylaxis, surgical preparation, and intra-operative surgical practice and techniques.

Conference abstracts, and papers not published in the English Language were excluded.

Results

Study selection

A total of 390 papers were identified after deduplication. After title and abstract screening, 109 were retrieved for full-text review and 42 were considered appropriate for inclusion. Seven papers identified through hand searching were included.

Study characteristics

A total of 880 patients were described in the 49 included studies. Twenty-six of the studies were case reports/series, [[17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42]] 11 were outbreak studies [[43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [53]] and the remainder (n=12) were retrospective chart reviews [[54], [55], [56], [57], [58], [59], [60], [61], [62], [63], [64], [65]] (Table I). Over half of the studies were published in Asia (28/49) with seven each published in Europe and North America, respectively. There were two studies from the United Kingdom [33,34], three from South America [24,42,52], and one each from Africa [30] and Oceania [37].

Table I.

General characteristics of included studies

Study ID	Country/Territory	Study type	Number of patients	Age/mean age	Sampling
Agrawal 2022 [25]	India	Case series	10	-	Vitreous samples
Rammohan et al., 2021 [22]	India	Case series	4	-	Various, vitreous or both or scleral abscess
Dave et al., 2020 [19]	India	Case series	4	-	Vitreous biopsy/sample
Kannan et al., 2020 [18]	India	Case series	28	66.07 ± 8.6	Various including vitreous tap, aqueous aspirate, anterior chamber membrane, vitreous biopsy, IOL, scleral swab, corneal scraping, AS exudate
Sen et al., 2020 [21]	India	Case series	17	62.44 ± 9.6	Intra-ocular fluids - not specific
Hsu et al., 2018 [23]	Taiwan	Case series	9	69	Either aqueous humor or vitreous fluid or both
Mesnard et al., 2016 [17]	French West Indies	Case series	4	67.5	Aqueous humor
Mithal et al., 2015 [20]	India	Case series	8	55.75	Corneal scrapings, vitreous biopsy, and explanted intraocular
Williams et al., 2014 [42]	Argentina	Case series	3	80.7	Vitreous samples
Mattos et al., 2013 [24]	Brazil	Case series	7	-	Vitreous samples
Francomacaro et al., 2022 [28]	USA	Case study	1	60–69	Anterior chamber paracentesis and vitreous sample
Lam et al., 2022 [39]	USA	Case study	1	60	Vitreous sample
Ledesma et al., 2022 [41]	Spain	Case study	1	77	Vitreous sample
Babalola 2020 [30]	Nigeria	Case study	1	84	Vitreous sample
Dave et al., 2020 [29]	India	Case study	1	50	Vitreous biopsy/sample
Shah et al., 2020 [36]	India	Case study	1	39	Vitreous sample
Voon et al., 2019 [37]	New Zealand	Case study	1	46	Anterior chamber and a vitreous tap
Palioura et al., 2018 [27]	USA	Case study	1	62	Anterior chamber sample
Alvarez-Ramos et al., 2016 [38]	Spain	Case study	1	-	Vitreous humor
Garg et al., 2016 [33]	England	Case study	1	85	Anterious chamber and vitreous sample
Lodhi et al., 2016 [32]	India	Case study	1	50	Aqueous humor and Vitreous sample
Arici et al., 2014 [31]	Turkey	Case study	1	73	Vitreous humor, corneal scraping, aqueous humor
Amissah-Arthur et al., 2013 [34]	England	Case study	1	85	AC samples and intravitreal biopsy
Khan et al., 2013 [26]	India	Case study	1	50	Vitreous sample
Gupta et al., 2010 [35]	USA	Case study	1	80	Vitreous sample
Javey et al., 2010 [40]	USA	Case study	1	86	Vitreous humor, vitreous biopsy
Kim et al., 2023 [48]	South Korea	Outbreak study	103	65.4 ±10.8	Not provided
Arasaki et al., 2022 [45]	Japan	Outbreak study	2	63.4±8.5	Vitreous and IOL samples
Spilker et al., 2022 [43]	Norway	Outbreak study	6	75.5	Cultures of vitreous or anterior chamber fluid or implanted intraocular lenses and lens capsules from each patient
Cheraqpour et al., 2021 [49]	Iran	Outbreak study	10	69.3	Vitreous samples
Bawankar et al., 2019 [50]	India	Outbreak study	13	67	AC samples for 10 patients and vitreous samples for 3 patients
Ji et al., 2015 [53]	China	Outbreak study	14	64.6	Vitreous and aqueous fluid
Buchta et al., 2015 [46]	Czech Republic	Outbreak study	20	70.5	Mostly vitreous humor
Lalitha et al., 2014 [44]	India	Outbreak study	13	57.7	Vitreous samples
Guerra et al., 2012 [52]	Brazil	Outbreak study	26	-	Aqueous humor and vitreous samples
Ramappa et al., 2012 [51]	India	Outbreak study	5	-	Vitreous samples
Gungel et al., 2011 [47]	Turkey	Outbreak study	9	-	Aqueous or vitreous samples
Jiang et al., 2022 [63]	China	Retrospective Study	3	-	Some eyes were sampled using vitreous humor others by aqueous humor
Malmin et al., 2021 [61]	Norway	Retrospective study	6	-	Vitreous samples, anterior chamber samples, or both
Jeong et al., 2017 [65]	South Korea	Retrospective study	58	70.7	Vitreous samples
Artsi et al., 2016 [64]	Israel	Retrospective study	2	-	Not clearly stated
Kelkar et al., 2016 [54]	India	Retrospective study	30	-	Aqueous humor and vitreous samples
Yannuzzi et al., 2016 [55]	USA	Retrospective study	63	-	Vitreous sample and vitrectomy cassette
Sharma et al., 2014 [59]	India	Retrospective study	16	-	Vitreous samples
Yao et al., 2013 [57]	China	Retrospective study	25	-	Aqueous humor and vitreous samples
Friling et al., 2012 [56]	Sweden	Retrospective study	113	-	-
Rahimi et al., 2012 [62]	Iran	Retrospective study	33	65.04	AC samples and vitreous taps
Cheng et al., 2010 [60]	Taiwan	Retrospective study	34	-	-
Pijl et al., 2010 [58]	Netherlands	Retrospective study	166	74	Vitreous biopsy or a primary vitrectomy

Abbreviations: AC, anterior chamber; IOL, intraocular lens; -, not reported.

Organisms associated with post-cataract surgery endophthalmitis

The included studies were screened to identify organisms associated with PCSE. As shown in Table II, 40 different genera of micro-organisms were identified as associated with PCSE, including Gram-positive and Gram-negative bacteria, fungi, and amoeba. Gram-negative bacteria were identified in 23 studies, the majority being Pseudomonas aeruginosa [35,[49], [50], [51], [52],57,58,60,62,65]. Stenotrophomonas maltophilia was identified in five studies [42,53,57,60,65].

Table II.

Endophthalmitis related organisms and their sources

Study ID	Study type	Organisma	Organism group	Source	Genetic relatedness	IPC factors
Mesnard et al., 2016 [17]	Case series	α-hemolytic streptococcus	Gram-positive bacteria	-	-	-
Agrawal 2022 [25]	Case series	Trichosporon spp.	Fungi	Possible contamination of disposables with unsterile water but no credible link	-	-
Williams et al., 2014 [42]	Case series	Stenotrophomonas maltophilia	Gram-negative bacteria	Silicon surgical-reusable tube suspected.	Not tested. Isolates from all cases exhibited a similar spectrum of antibiotic sensitivity. However, antibiotic sensitivity of S. maltophilia isolated from the suspect reusable tube was not reported.	-
Mattos et al., 2013 [24]	Case series	Ochrobactrum anthropi	Gram-negative bacteria	Contaminated tubing of phaco-emulsification machine suspected	-	-
Kannan et al., 2020 [18]	Case series	Nocardia	Gram-positive bacteria	Not found	-	-
Hsu et al., 2018 [23]	Case series	Mycobacterium chelonae/Mycobacterium abscessus	NTM	-	-	-
Dave et al., 2020B [19]	Case series	Enterobacter spp.	Gram-negative bacteria	-	-	-
Mithal et al., 2015 [20]	Case series	Aspergillus terreus	Fungi	-	-	-
Sen et al., 2020 [21]	Case series	Aspergillus niger (5), A. flavus, A fumigatus, A. nidulans (2), A. terreus (2), Candida spp. (2), Fusarium spp. (2), unclassified dematiaceous fungi (2)	Fungi	-	-	-
Rammohan et al., 2021 [22]	Case series	Acanthamoeba culbertsoni	Protozoa	-	-	-
Ledesma et al., 2022 [41]	Case study	Wickerhamomyces anomalus	Fungi	-	-	-
Javey et al., 2010 [40]	Case study	Staphylococcus epidermidis	Gram-positive bacteria	-	-	-
Alvarez-Ramos et al., 2016 [38]	Case study	Rothia mucilaginosa	Gram-positive bacteria	Possible self-contamination or contaminated eye drops but no credible link reported	-	-
Lam et al., 2022 [39]	Case study	Rothia mucilaginosa	Gram-positive bacteria	iStent device suspected but no viable link was reported	-	-
Voon et al., 2019 [37]	Case study	Pseudozyma aphidis	Fungi	Not found	-	-
Shah et al., 2020 [36]	Case study	Pseudomonas stutzeri	Gram-negative bacteria	-	-	-
Gupta et al., 2010 [35]	Case study	Pseudomonas aeruginosa	Gram-negative bacteria	-	-	-
Amissah-Arthur et al., 2013 [34]	Case study	Prevotella spp.	Gram-negative bacteria	-	-	-
Garg et al., 2016 [33]	Case study	Penicillium citrinum	Fungi	-	-	-
Lodhi et al., 2016 [32]	Case study	Nocardia asteroids	Gram-positive bacteria	-	-	-
Arici et al., 2014 [31]	Case study	Fusarium solani	Fungi	-	-	-
Babalola 2020 [30]	Case study	Enterococcus faecium	Gram-positive bacteria	-	-	-
Dave et al., 2020 [29]	Case study	Curvularia spp.	Fungi	-	-	-
Francomacaro et al., 2022 [28]	Case study	Clostridium intestinale	Gram-positive bacteria	-	-	-
Palioura et al., 2018 [27]	Case study	Candida parapsilosis	Fungi	-	-	-
Khan et al., 2013 [26]	Case study	Burkholderia cepacia	Gram-negative bacteria	-	-	-
Ji et al., 2015 [53]	Outbreak study	Stenotrophomonas maltophilia	Gram-negative bacteria	Aspiration tube of a phaco emulsifier - tested positive for S. maltophilia	Not tested. Vitreous isolates from patients and the suspect aspiration tube had similar antibiotic sensitivity profiles.	-
Bawankar et al., 2019 [50]	Outbreak study	Pseudomonas aeruginosa	Gram-negative bacteria	Trypan blue solution	PFGE	Contaminated medical product
Cheraqpour et al., 2021 [49]	Outbreak study	Pseudomonas aeruginosa	Gram-negative bacteria	A contaminated phaco probe was used for all 10 patients without sterilization in between	Contaminated phaco probe tested positive for P. aeruginosa - no genetic testing was performed	-
Guerra et al., 2012 [52]	Outbreak study	Pseudomonas aeruginosa	Gram-negative bacteria	-	-	-
Ramappa et al., 2012 [51]	Outbreak study	Pseudomonas aeruginosa	Gram-negative bacteria	IOL and IOL suspension solution	ERIC-PCR	Contaminated medical product
Kim et al., 2023 [48]	Outbreak study	Fusarium spp.	Fungi	Viscoelastics	Direct sequencing	Contaminated medical product
Gungel et al., 2011 [47]	Outbreak study	Fusarium solani	Fungi	Contaminated BSS and/or cefuroxime solution suspected	-	-
Arasaki et al., 2022 [45]	Outbreak study	Fusarium oxysporum	Fungi	Not found	-	-
Buchta et al., 2015 [46]	Outbreak study	Fusarium oxysporum	Fungi	Suspected - viscoelastic solution. Not tested because the suspected batch was exhausted before the first case presented. Endophthalmitis was reported in 62.5% (n= 32) of patients who used suspected viscoelastic compared to 0 patients on whom it was not used.	-	Contaminated medical product
Lalitha et al., 2014 [44]	Outbreak study	Burkholderia cepacia	Gram-negative bacteria	Contaminated aesthetic eye drop	BOX-PCR	-
Spilker et al., 2022 [43]	Outbreak study	Burkholderia contaminans	Gram-negative bacteria	Ventilation system	MLST	Built environment contamination
Artsi et al., 2016 [64]	Retrospective study	Streptococcus viridans, Staphylococcus epidermidis	Gram-positive bacteria	-	-	-
Jiang et al., 2022 [63]	Retrospective study	Staphylococcus hominis (2), Streptococcus spp.	Gram-positive bacteria	-	-	-
Rahimi et al., 2012 [62]	Retrospective study	Staphylococcus epidermidis (4), Staphylococcus aureus (12), Streptococcus hemolyticus (2), Streptococcus pneumonia, Pseudomonas aeruginosa (4), Enterobacter spp. (4), E. coli (2), Acinetobacter spp., Proteus vulgaris, Haemophilus influenza, Candida Albicans	Mixed	-	-	-
Jeong et al., 2017 [65]	Retrospective study	Staphylococcus epidermidis (33) Enterococcus faecalis (11) Pseudomonas aeruginosa (14)	Mixed	-	-	-
Malmin et al., 2021 [61]	Retrospective study	Staphylococcus epidermidis (2), Streptococcus oralis (1), Enterococcus faecalis (3)	Gram-positive bacteria	-	-	-
Cheng et al., 2010 [60]	Retrospective study	Staphylococcus aureus (9), Enterococcus, Streptococcus pneumonaie, Paenibacillus glucanolyticus, Coagulase negative staphylococci, Pseudomonas aeruginosa (13), Proteus vulgaris (2), Stenotrophomonas maltophilia (2), Moraxella cataralis (2)	Mixed	-	-	-
Sharma et al., 2014 [59]	Retrospective study	Pseudomonas spp. (4), Staphylococcus spp. (2), Streptococcus (2), Bacillus licheniformis, Acremonium spp. (2), Aspergillus flavus, A. terreus, A. flavipes, Candida spp. (2)	Mixed	-	-	-
Pij et al., 2010 [58]	Retrospective study	Gram-positive coagulase negative Staphylococci (89), Staphylococcus aureus (20), Streptococcus pneumoniae (12), Viridans group Streptococci (11), B hemolytic Streptococcus x9, Enterococcus (3), Diphtheroid Gram-positive rods (3), Abiotrophia spp. (2), Gemella morbillorum, Peptostreptococcus spp., Propionibacterium acnes, Proteus mirabilis (3), Haemophilus influenzae (3), P. aeruginosa (2), Achromobacter xylosoxidans, Acinetobacter iwoffi, polymicrobial (4)	Mixed	-	-	-
Yao et al., 2013 [57]	Retrospective study	Gram positive coagulase negative (8), S. aureus (3), Enterococcus faecalis, Dry Corynebacterium, Streptococcus pyogenes, Pseudomonas maltophilia (8), P. paucimobilis, P. aeruginosa, Polymicrobial (Bacillus cereus + Streptococus viridians)	Mixed	-	-	-
Friling et al., 2012 [56]	Retrospective study	Enterococci (42), Coagulase-negative Staphylococci (35), Other Streptococci (9), Other Gram-positive species (8), Pseudomonas spp. (10), Enterobacteria spp. (7), other Gram-negative bacteria (2)	Mixed	-	-	-
Yannuzzi et al., 2016 [55]	Retrospective study	Coagulase-negative Staphylococcus (39), Streptococcus spp. (7) (S. salivarius, S. sanguinis, S. constellatus, S. mitis (2), S. viridans), Staphylococcus aureus (7), Staphylococcus spp. (3) (S. warneri, S. Lugdunensis (2), Enterococcus (3) (E. faecalis (3), Propionibacterium (2) (Propionibacterium acnes, P. granulosen), Pseudomonas spp., Serratia spp.	Mixed	-	-	-
Kelkar et al., 2016 [54]	Retrospective study	Coagulase negative Staphylococcus x8, MRSA (5), Staphylococcus aureus (3), Streptococcus pneumoniae (3), Propionibacterium acnes, Staphylococcus epidermidis (3), Streptococcus mitis, E. coli (3), Pseudomonas spp. (1), Klebsiella spp., Sphingomonas paucimobilis	Mixed	-	-	-

Abbreviations: ERIC PCR - enterobacterial Repetitive Intergenic Consensus Polymerase Chain Reaction; PFGE - Pulse-field gel electrophoresis; NTM – Nontuberculous Mycobacteria MLST – Multilocus sequence typing.

^aNumbers in brackets represent the number of cases for studies reporting more than one organism. In such studies, organisms with no number in the bracket mean a single case.

Gram-positive bacteria were identified in 16 papers, chief of which were Staphylococcus aureus [54,55,57,58,60,62,65] and Staphylococcus epidermidis [40,54,61,62,64].

Fungal endophthalmitis was described in 17 studies with Fusarium spp. being the most frequently reported (n = 6) [31,[45], [46], [47], [48]]. Others included Aspergillus spp. [20,59,65], Candida spp. [62,65], Curvularia spp. [29], Penicillium citrinum [33], Trichosporon spp. [25], Pseudozyma aphidis [37], Wickerhamomyces anomalus [41] and Acremonium spp [59].

Other organisms identified included Acanthamoeba culbertsoni and Mycobacterium spp. [22,23] Two studies reported cases where more than one organism was isolated from a single patient. [57,58].

Figure 1, Figure 2 show that most of the cases in the included retrospective chart reviews were associated with Gram-positive bacteria (79.1%), the predominant being Staphylococcus spp., followed by Gram-negative bacteria (19.3%) and fungi (1.6%), whereas outbreak studies were associated with either gram-negative bacteria or fungi.

Figure 1.

Comparison of organism types reported in outbreak studies and retrospective studies.

Figure 2.

Organisms reported by number of Cases in retrospective reviews.

IPC factors associated with post-cataract endophthalmitis

To understand the IPC factors that are associated with PCSE, sources of transmission described in the outbreak studies were analyzed. Only studies that demonstrated genetic relatedness or a strong epidemiological link were considered in the discussion about IPC factors.

To establish the source of the outbreaks, five studies employed various techniques to demonstrate genetic relatedness between the organisms isolated from patients and the environment. These included variants of repetitive element-based PCR (Rep-PCR) reported by two studies; BOX-A1R-based repetitive extragenic palindromic-PCR (BOX PCR), [44] and Enterobacterial Repetitive Intergenic Consensus Polymerase Chain Reaction (ERIC PCR). [51] Others include direct sequencing, [48] pulsed-field gel electrophoresis (PFGE) [50] and multi-locus sequence typing (MLST). [43].

The sources included contaminated ophthalmic solutions and medications (n = 4 studies), [44,48,50,51] contaminated surgical instruments (n = 1 study) [49] and the ventilation system (n = 1 study). [43] Behind some of these were IPC factors such as built environment contamination, [43] or poor sterilization practices (Table II). [48] The organisms associated with these outbreaks were either Gram-negative bacteria (P. aeruginosa, [[49], [50], [51]] and Burkholderia spp.), [43,44] or fungal (Fusarium spp.) (Table II). [48].

In one outbreak study, an epidemiological link was demonstrated between patient samples and the suspected source, even though typing was not done. It involved a single ophthalmologist who used the same phaco probe in uncomplicated cataract surgeries with IOL implantations for ten patients without sterilization between patients. All ten patients developed PCSE with vitreous samples yielding Pseudomonas aeruginosa, as did the phaco probe. [49].

Discussion

In the studies identified for this review, the IPC risk factors for PCSE were found to include environmental contamination, ineffective sterilization procedures, or a lack thereof. Outbreaks caused by ophthalmic solutions contaminated at the batch manufacturing level were also found. These issues relating to IPC should be considered, particularly when scaling up cataract surgery, to ensure that patient safety is maintained. Moreover, this review explored the pathogens associated with PCSE. These included Gram-positive and Gram-negative bacteria and fungi. Evidence compiled by this review suggested a difference in the type of pathogen according to case type (sporadic cases versus outbreaks). Outbreak studies were associated with Gram-negative bacteria and fungi, whilst retrospective chart reviews of sporadic cases were associated with Gram-positive organisms. The knowledge that fungi and Gram-negative pathogens are often associated with outbreaks should prompt IPC teams to consider and investigate the possibility of epidemiologically linked cases and their sources to limit additional cases of PCSE.

Organisms associated with post-cataract surgery endophthalmitis

A large variety of microorganisms are associated with PCSE. Staphylococcus spp. was associated with the most cases (32.4%) even though Pseudomonas spp. were reported by more studies (32.7%). The occurrence of the various pathogens is better understood when evidence is synthesized according to study type.

Gram-positive bacteria (especially Staphylococcus spp. and Streptococcus spp.) were most implicated in retrospective chart reviews. This is not surprising since aqueous contamination with skin commensal bacterial flora is considered the main pathogenesis for endophthalmitis [8].

While Gram-positive bacteria were associated with most of the cases in retrospective chart reviews, this was not the case in outbreak studies. Outbreak studies included in this review were found to exclusively involve fungal and Gram-negative organisms. Previous reviews that solely assessed outbreak studies or did not distinguish on study type support the latter and found Pseudomonas aeruginosa as the most frequently isolated pathogen in outbreaks and clusters. [12,14] This review suggests that samples positive for Fusarium spp., Burkholderia spp., and P. aeruginosa should be viewed with concern because they are not normal flora of the skin and are more likely to be associated with contaminated instruments or solutions. This is particularly concerning regarding Fusarium spp., which can be aerosolized and has often been associated with outbreaks in healthcare. [[66], [67], [68], [69]].

Factors related to infection prevention and control associated with post-cataract surgery endophthalmitis

Sterile ophthalmic solutions were reported to be the cause of PCSE outbreaks in four studies. [44,48,50,51] These solutions included anaesthetic eye drops, ophthalmic viscoelastic devices (OVDs), trypan blue solution, and intraocular lens (IOL) suspension solution. They were contaminated by P. aeruginosa, B. cepacia, and Fusarium spp. and were all genetically linked to the organisms isolated from patient samples. In all four studies, the contamination occurred at a manufacturing level as samples from unopened bottles yielded growth of the contaminating organisms [44,48,50,51]. A nationwide outbreak in South Korea was only resolved after the withdrawal of a particular brand of sodium hyaluronate viscoelastic materials. [48] Batchwise sampling of ophthalmic solutions may be considered as a strategy to reduce the likelihood of such outbreaks occurring.

Failures in the sterilization of surgical instruments were also reported as a probable source in one study, in which the surgeon used the same phaco probe for all 10 cases without sterilization between patients. [49].

A contaminated ventilation system was identified as the source of an outbreak of B. contaminans in a private single-physician clinic in Norway. [42] Seven samples from a particular air intake duct in the ventilation system yielded bacterial growth that tested positive using a Burkholderia-specific PCR assay, one being from pooled standing water and the other six being swabs from biofilms. Multi-locus sequence typing (MLST) analysis showed that all seven isolates had an identical allelic profile to those recovered from patient cultures. It was hypothesized that this contamination occurred due to water pooling in air intake ducts following flooding. However, air sampling was not performed to further investigate this as the transmission route.

A key limitation of the body of evidence included in this review is a potential for publication bias as many outbreak investigations are not published, hence the potential IPC factors related to PCSE may not have been identified. Another limitation is that screening and data extraction were performed by a single reviewer and there was no formal risk of bias assessment. It is however important to consider variations in symptom severity, medication effects, healthcare systems, and health-seeking behaviours across different countries that can affect this measure. Extraction of this data was challenging in certain studies, particularly retrospective chart reviews where cases of endophthalmitis following surgeries other than cataract surgeries or cases without positive microbial culture were included. Despite these limitations, this review demonstrates the differences in organisms associated with endophthalmitis outbreaks compared to those in sporadic cases, which may assist in prompt management and control of PCSE outbreaks.

Conflict of interest statement

None declared.

Funding statement

None.

Ethic statement

Not required.

Credits author statement

Ogheneochuko A. Saba: Methodology, Investigation, Writing - Original Draft, Visualization.

Yasmine Benylles: Conceptualization, Methodology, Writing - Review & Editing, Supervision.

Mireille H. Howe: Writing - Review & Editing, Supervision, Project administration.

Teresa Inkster: Conceptualization, Writing - Review & Editing.

Emma L. Hooker: Conceptualization, Writing - Review & Editing, Supervision.

Appendix 1. Search strategy

Database search on post-cataract surgery endophthalmitis.

Search performed on April 13, 2023.

Ovid MEDLINE

Line	Search term	Result number
1	∗Cataract Extraction/	18984
2	(cataract$ adj4 (surg$ or operat$ or extract$ or aspirat$ or excis$ or remov$ or emulsif$ or implant$)).ti,ab,kf.	33863
3	post?cataract.ti,ab,kf.	225
4	Phacoemulsification/	11634
5	(pha?oemulsif$ or phaco or phako).ti,ab,kf.	10897
6	1 or 2 or 3 or 4 or 5	45205
7	∗Endophthalmitis/	6416
8	endophthalmiti$.ti,ab,kf.	9421
9	ophthalmia.ti,ab,kf.	1989
10	7 or 8 or 9	12333
11	exp ∗Infection Control/	41325
12	exp ∗Cross Infection/	47255
13	exp ∗Disease Transmission, Infectious/	45086
14	exp ∗Decontamination/	3295
15	exp ∗Equipment Contamination/	7767
16	Postoperative Complications/pc [Prevention & Control]	50789
17	((infect$ or endophthalmiti$) adj3 (prevent$ or control$ or manag$)).ti,ab,kf.	130832
18	(cross infect$ or contamina$ or decontamina$ or sterili$ or disinfect$).ti,ab,kf.	380922
19	11 or 12 or 13 or 14 or 15 or 16 or 17 or 18	641568
20	6 and 10 and 19	564
21	limit 20 to english language	504
22	limit 21 to yr="2010 -Current"	280

Embase

Line	Search term	Result number
1	∗cataract extraction/	20593
2	(cataract$ adj4 (surg$ or operat$ or extract$ or aspirat$ or excis$ or remov$ or emulsif$ or implant$)).ti,ab,kf.	39488
3	post?cataract.ti,ab,kf.	261
4	phacoemulsification/	17987
5	(pha?oemulsif$ or phaco or phako).ti,ab,kf.	14329
6	1 or 2 or 3 or 4 or 5	53244
7	∗endophthalmitis/	6953
8	∗fungal endophthalmitis/Note: Embase has a separate subject heading for fungal endophthalmitis, so this has been included in order to ensure that this form of the infection is captured	308
9	endophthalmiti$.ti,ab,kf.	11615
10	ophthalmia.ti,ab,kf.	1432
11	7 or 8 or 9 or 10	13697
12	exp ∗infection control/	38258
13	exp ∗cross infection/	12282
14	exp ∗disease transmission/	40521
15	exp ∗medical device contamination/	401
16	exp ∗"prevention and control"/	741670
17	postoperative complication/pc [Prevention]	19072
18	((infect$ or endophthalmiti$) adj3 (prevent$ or control$ or manag$)).ti,ab,kf.	168932
19	(cross infect$ or contamina$ or decontamina$ or sterili$ or disinfect$).ti,ab,kf.	444514
20	12 or 13 or 14 or 15 or 16 or 17 or 18 or 19	1338291
21	6 and 11 and 20	725
22	limit 21 to english language	620
23	22 not conference∗.so,pt.	543
24	limit 23 to yr="2010 -Current"	329