Estimating the rate of severe visual loss (wipe-out) following cataract surgery, a British Ophthalmological Surveillance Unit (BOSU) study

Abstract

Background

A sudden, irreversible reduction in visual acuity (‘wipe-out’) is a feared complication of cataract surgery. Current literature on wipe-out is limited in quantity and quality, and largely predates modern cataract surgery and imaging techniques. The objectives of our study were to estimate the incidence of wipe-out and to identify potential risk factors.

Methods

We prospectively collated cases of wipe-out occurring in the UK during a 25-month study period using the British Ophthalmic Surveillance Unit reporting system. A total of 21 potential cases of wipe-out were reported, 5 of which met all inclusion and exclusion criteria.

Results

The estimated incidence of wipe-out during the study period was 0.00000298, or approximately 3 cases per million cataract operations. All cases of wipe-out occurred exclusively in patients with advanced glaucoma (mean deviation −21.0 decibels or worse in the operated eye), with an over-representation of black people (40%) in our case series. A prior diagnosis of retinal vein occlusion (60%) and elevated post-operative IOP (40%) were more common among individuals suffering from wipe-out compared to the general population, suggesting these factors may contribute to the pathogenesis of wipe-out.

Conclusions

Our study shows that wipe-out is a rare complication, affecting approximately 3 per million undergoing cataract surgery. Patients with advanced glaucoma, black patients, and those with previous retinal vein occlusions may be at greater risk of wipe-out. We hope that the findings of our study will be used to help inform treatment decision-making and the cataract surgery consent process.

Introduction

The overwhelming expectation of patients undergoing cataract surgery and surgeons performing cataract surgery is for vision to improve. Rarely, an unexpected sharp decline in vision occurs post-operatively [1]. This phenomenon has been referred to as ‘wipe-out’ (or ‘snuff-out’). These terms have been most commonly applied to label an unexplained, irreversible drop in vision in patients with advanced glaucoma after cataract surgery. Before this study, there have been no published estimates of the incidence of wipe-out following cataract surgery.

Cataract and glaucoma comorbidity is a common clinical scenario worldwide. Cataract surgery is one of the most frequently performed surgical procedures and glaucoma features prominently as a cause of registered blindness [2], becoming increasingly common with age. A recent summary of trends in certifications of vision impairment (CVI) in England and Wales shows that glaucoma is the second most common cause of blindness registration and that the number of certifications has remained stable over time [3]. Furthermore, glaucoma surgery with or without antimetabolite augmentation is well recognised to accelerate cataract development [4, 5]. Quoted rates for advanced glaucoma diagnosed at presentation vary from 4% [6] to 38% [7].

The risk of sudden central visual loss in patients with advanced glaucoma, alternatively known as wipe-out, following intraocular surgery has been debated since the introduction of glaucoma surgery by Von Graefe in 1856 [8], but existing literature pertaining to wipe-out is limited and clinicians may be largely informed about this entity via anecdotes and personal experience. Much of the literature predates the use of phacoemulsification for cataract surgery. In a previous study of 23 patients with advanced glaucoma undergoing cataract surgery, sudden loss of central vision to <6/60 occurred in 2 cases (8.7%) [9], while O’Connell et al. observed no cases of severe visual loss in 13 cases of cataract surgery for patients with advanced glaucoma [10]. Further reports have corroborated the finding that patients with advanced glaucoma are at risk of central vision loss following intraocular surgery [11, 12]. However, the veracity of selected cases as true wipe-out has since been challenged and it has been suggested that subtle macular pathology could have been overlooked prior to the widespread use of ocular imaging [13]. Conversely, Altemeyer and colleagues more recently reported an overall improvement in visual acuity and visual fields following cataract surgery in 12 eyes with advanced glaucoma, with no cases of severe visual loss [14] and so the risk is thought by some to be minimal with modern techniques [15].

Before the initiation of this study, our group canvassed the UK and Eire Glaucoma Society with a survey. We received 46 responses out of a potential 194, who collectively reported 40 cases of wipe-out occurring following cataract surgery for patients with advanced glaucoma in the preceding decade. Our survey also identified quoted rates for the occurrence of wipe-out when taking consent for cataract surgery for patients with advanced glaucoma, which ranged from quoting wipe-out in >1:100 cases to the omission of wipe-out as a potential risk, representing a wide variation in practice.

The objectives of this British Ophthalmological Surveillance Unit (BOSU) study are to estimate the incidence of wipe-out following cataract surgery and to identify potential risk factors. With no currently available estimates of the incidence of wipe-out in the context of improved, modern surgical and imaging techniques, the findings of our study provide important information to equip clinicians with relevant data to better inform surgical decision-making and consent processes.

Methods

BOSU study

Over a 25-month period from 2017 to 2019, all independent UK consultant and associate specialist ophthalmologists working in the National Health Service were sent a monthly report card to report any case of wipe-out following cataract surgery as part of the BOSU programme. All reports were followed up with questionnaires sent by the research team. BOSU methodology has previously been fully described [16].

Questionnaire—see supplementary file

Exclusion criteria

Patients undergoing cataract surgery combined with another procedure were excluded from the study. To exclude patients who had vision loss secondary to causes other than wipe-out, we included the following statement in the cover letter accompanying the questionnaire:

“Did your patient lose vision due to demonstrable ophthalmic pathology other than wipe-out (e.g. retinal vascular occlusion, endophthalmitis, suprachoroidal haemorrhage, non-glaucomatous ischaemic optic neuropathy, retinal detachment, vitreous haemorrhage, etc.)?”

Inclusion criteria

We used a previously published level of visual acuity loss to define cases of wipe-out following cataract surgery: [15]

• Vision loss to a level of 6/60 Snellen (or equivalent e.g. 1.0 LogMAR) or worse

OR

• If vision was 6/60 or less prior to the operation, a visual acuity of Counting Fingers (CF) or worse; or if vision was CF, a drop to hand movements (HM) or worse; or if vision was HM, a drop to perception of light (PL) or worse; or if vision was PL, a drop to no perception of light (NPL)

OR

• Visual acuity loss by an equivalent of 4 lines of Snellen visual acuity

Results

In the 25 months that ‘Wipe-out following cataract surgery’ was on the BOSU report card, there were 21 reports of potential cases to the surveillance unit. Of the 21 questionnaires that were sent out, there were 14 responses. Of the 14 responses, 3 were noted to be administration errors. Of 11 possible cases, 2 did not meet the inclusion criteria since both had retinal arterial occlusions. Of the remaining 9 cases, one did not meet the visual acuity reduction inclusion criterion and one case was a combined case with endoscopic cyclophotocoagulation and was therefore excluded as we had planned to exclude all combined cases prospectively.

We report data from the remaining 7 cases (Table 1). Surgery was performed outside the study window in two cases, which were hence excluded from incidence analysis. Two cases are reported but excluded from incidence and risk factor analysis (see cases 6 & 7). This leaves a total of 3 included cases of reported wipe-out following cataract surgery over the 25-month study period.

Table 1.

Summary of reported cases of wipe-out.

Case	Age	Sex	Ethnicity	Systemic comorbidities	POAG (MD dB, PSD dB)	Other ocular history	RVO	Pre-op BCVA	Post-op BCVA	Post-op IOP (days post-op)	Axial length	Anaesthesia	Intraoperative complication	CVI registration post-op	Other
1	75	F	White	HTN, DM	POAG (−26.89, 8.9)	CRVO	Yes	6/9	6/60	30 (8 days)	-	Topical with intracameral	No	No
2	85	F	White	HTN, DM	POAG (−21.95.0, 10.2)	BRVO	Yes	6/60	CF	18 (12 days)	23.33	Topical with intracameral	No	No
3	82	M	Black	HTN	POAG (−32.55, 3.7)	Nil	No	6/36	NPL	18 (14 days)	24.14	Subtenons (4 ml)	No	Yes
4	90	M	White	HTN	POAG (−21.93, 8.6)	AMD	No	5/60	HM	32 (19 days)	-	Unknown	Unknown	Yes	Surgery done at another centre
5	81	M	Black	DM	POAG (−28.63, 6.07)	HRVO	Yes	6/24	PL	11 (7 days)	24.09	Subtenons (3.5 ml)	Yes	No
6	67	M	White	Nil	Nil	Nil	No	6/24	1/60	N/A (13 days)	23.63	Peribulbar	No	No	?PAMM
7	76	F	White	Nil	Nil	Nil	No	6/9	HM	23 (1 day)	22.08	Topical with general	Unknown	No	?Sub-conjunctival preparation injected intracamerally

POAG primary open angle glaucoma, MD mean deviation, PSD pattern standard deviation, RVO retinal vein occlusion, BCVA best-corrected visual acuity, IOP intraocular pressure, CVI certificate of visual impairment, HTN systemic hypertension, DM diabetes mellitus, AMD age related macular degeneration, CRVO central RVO, HRVO hemi RVO, BRVO branch RVO, HM hand movements, CF count fingers, NPL no perception of light, PL perception of light, PAMM paracentral acute middle maculopathy, N/A not available.

Case 1

A 75-year-old white female with no history of ocular surgery and a diagnosis of advanced primary open angle glaucoma (POAG) (mean deviation (MD) of −26.89 decibels (dB) and pattern standard deviation (PSD) of 8.85 dB). The patient had previous bilateral central retinal vein occlusions with evidence of optic disc collaterals on images from 2012. Her past medical history includes hypertension and diabetes with no significant diabetic retinopathy. Prior to surgery, the best corrected visual acuity (BCVA) in the affected eye was 6/9 and the intraocular pressure (IOP) had been stable on medication with a pressure maintained around 15 mm Hg. Cataract surgery was performed under topical and intracameral anaesthesia and there were no intraoperative complications noted. At the first clinic visit 8 days following surgery, the visual acuity was 6/60 and the IOP was 30 mm Hg.

Case 2

An 85-year-old white female with known advanced POAG (MD −21.95 dB and PSD 10.19 dB) with no previous ocular surgery. The patient had been successfully treated with 7 bevacizumab injections for a macular branch retinal vein occlusion in the 11 months preceding cataract surgery. The macula was noted to be dry with a central retinal thickness of 220 µm pre- and post-operatively. The patient has a past medical history of hypertension and diabetes with no diabetic retinopathy. Her pre-operative BCVA was 6/60. Her cataract surgery was performed under topical and intracameral anaesthesia and was reported to be uncomplicated. The post-operative visual acuity in the operated eye 12 days following surgery was count fingers and the IOP was 18 mm Hg.

Case 3

An 82-year-old black male with a history of hypertension and advanced POAG (MD −32.55 dB and PSD 3.71 dB) with no previous ocular surgery. His pre-operative BCVA was 6/36. The cataract surgery was uncomplicated and performed with a 4 ml volume of subtenons anaesthesia. At the first post-operative visit 13 days following surgery, the visual acuity was NPL and the intraocular pressure was 18 mm Hg.

Case 4

A 90-year-old white male with a past medical history of treated hypertension and a past ocular history of age related macular degeneration (AMD) and POAG on maximum medical treatment. The MD on visual field testing pre-operatively was −21.93 dB with a PSD of 8.59 dB in the eye undergoing cataract surgery. Prior to surgery, the patient had a BCVA of 5/60 which dropped to HM following surgery. On the first post-operative visit, the IOP of the operated eye was noted to be raised at 32 mm Hg 19 days post-operatively. This patient’s surgery occurred outside the study window and therefore this case was excluded from incidence analysis.

Case 5

An 81-year-old black male with type 2 diabetes and a past ocular history of a hemiretinal vein occlusion in the operated eye, advanced POAG (MD −28.63 dB, PSD 6.07 dB) and no previous ocular surgery in the operated eye. He underwent cataract surgery under 3.5 ml of subtenons anaesthesia consisting of a mixture of bupivacaine and lignocaine. The surgery was complicated by an anterior capsular tear. His pre-operative BCVA was 6/24, reducing to PL 2 weeks post-operatively with an IOP of 11 mm Hg. This patient’s surgery occurred outside the study window and therefore this case was excluded from incidence analysis.

Case 6

A 67-year-old white male with no previous medical or ophthalmic history underwent uncomplicated cataract surgery under peribulbar anaesthesia and his BCVA reduced from 6/24 to 1/60 post-operatively. His IOP was recorded within normal limits during each clinic visit. His macula OCT showed macula thickening and the cause of the drop in his visual acuity was labelled at the time as transient retinal artery occlusion, but with no firm diagnosis. Due to the macula pathology, this case was excluded from incidence and risk factor analysis.

Case 7

A 76-year-old white female with no medical or ocular history underwent cataract surgery under general anaesthesia. Her BCVA reduced from 6/9 to HM post-operatively. There was concern following the case regarding whether or not a sub-conjunctival preparation of antibiotics was erroneously injected intracamerally and hence this case has been excluded from incidence and risk factor analysis.

Discussion

Incidence estimation

There were three cases of wipe-out that met the inclusion criteria during the 25-month study period. According to Eurostat, the statistical office of the European Union, there were 482,938 cataract operations performed in the UK in 2017 while this study was taking place [17]. Therefore, the estimated annual incidence of wipe-out following cataract surgery in the United Kingdom was 0.00000298, or almost 3 per million.

Identifying possible risk factors for wipe-out

Glaucoma

There were five cases returned that met inclusion and exclusion criteria. All five patients had POAG, compared to the background prevalence of approximately 9% in the over 80 age group [18]. Moreover, all five cases had a MD worse than −21.0 dB in the operated eye, indicative of advanced glaucoma, suggesting that advanced POAG may be a significant risk factor for wipe-out. Due to the low numbers in this study, it is difficult to accurately extrapolate but it follows that wipe-out might be as much as ten times more common than estimated here in the context of eyes with glaucoma as nearly 1 in 10 have glaucoma in this age group. Furthermore, all those included had advanced glaucoma, making the denominator smaller and therefore potentially increasing the incidence of wipe-out in the context of advanced glaucoma.

Retinal vein occlusion

Three patients (60%) experiencing wipe-out had previous retinal vein occlusions in the operated eye; one had a history of a branch retinal vein occlusion treated with anti-vascular endothelial growth factor (VEGF) injections, one had a history of a central retinal vein occlusion with evidence of collaterals on imaging pre-operatively, and one had a history of a hemiretinal vein occlusion treated with anti-VEGF injections. Data from an Australian cohort show that the background rate for retinal vein occlusions in people over 80 is only 1.6–4.6% [19], which suggests that a pre-existing retinal vein occlusion may be a risk for wipe-out following cataract surgery. A putative mechanism for wipe-out may involve reduced ocular perfusion during anaesthesia, intra-operatively, or post-operatively in the context of diseased vasculature and severe glaucomatous optic neuropathy, permanently compromising optic nerve function and leading to irreversible loss of visual acuity.

Raised IOP

Elevated IOP occurred in two cases (40%), at 32 mm Hg in case 1 and 30 mm Hg in case 2 at the first post-operative visit. It should be noted, however, that the most significant IOP rise following cataract surgery occurs in the first 24 h post operatively [20] and so maximal IOP may not have been captured in our series, since the first post-operative visits occurred at an mean of 12 days post operatively (range 7–19 days). The observation of high IOP in 40% of our cases suggests that raised IOP may be a factor in the development of wipe-out following cataract surgery. Again, the low number of cases in this report make it difficult to arrive at any firm conclusions.

Ethnicity

Two out of five patients (40%) reported to have wipe-out following cataract surgery were black, making black people vastly over-represented in this cohort when compared to 2011 census data showing that approximately 3.4% of people in England and Wales identify as black [21]. Again, due to the small sample size it is difficult to be conclusive about this observation. We also acknowledge that the perceived risk associated with ethnicity may be somewhat confounded since all five cases of wipe-out had existing advanced glaucoma, which is both more prevalent and severe among people of African ancestry [22].

Diabetes and older age

Three patients (60%) suffering wipe-out had diabetes. In the over 75 age group, an estimated 12–14% have diabetes and so this association may confer a risk for wipe-out [23]. Older age may also predispose to wipe-out as the median age of the patients in our cohort is 82 compared to 76.6, the median age of the National Ophthalmology Database (NOD) cohort for 2017 while this study was running [24]. Due to our limited sample size, it is difficult to be certain of the impact of diabetes and age on the incidence and risk of wipe-out.

Variables less likely to increase the risk of wipe-out following cataract surgery

Pre-operative factors

There is no significant difference in the sex distribution of wipe-out cases reported (3 men and 2 women). Four (80%) of the cases had systemic hypertension which is in keeping with the background rate of hypertension in the age group in question [25]. One patient had AMD which is proportionate to the background rate of AMD in this age group of 16.4% in people over 80 [26]. Axial length (AL) data were only supplied in three out of the five reported cases, with a mean AL of 23.85 mm (standard deviation [SD] 0.45), which is similar to the mean AL of 23.43 mm (SD 1.38) noted in the UK NOD study of 180,114 eyes [27]. The lack of significant differences in the sex distribution, axial lengths, systemic hypertension prevalence, and AMD prevalence within our cohort compared to population averages suggests that these factors may not contribute to the overall risk of wipe-out. However, our study may be underpowered to detect potentially significant variables of individually small effect.

Intraoperative factors

Details on the type of anaesthesia used were available for 4 cases; 2 had subtenons anaesthesia (3.5 ml and 4 ml volumes, respectively) and 2 had topical with intracameral anaesthesia. Intraoperative complication data were available for 4 cases, one of which had an anterior capsular tear during surgery for a white, dense cataract (Argentinian flag). No posterior extension was noted, and the lens was sited in the capsular bag. The significance of this in the presence of other putative risk factors for wipe-out is unknown.

Visual loss and its consequences

In all but one case, the visual acuity dropped to CF or worse, with the largest drop being from 6/9 to 6/60. Two out of five cases in our series were noted to be eligible for CVI registration following wipe-out. While the total number of cases in the study is low and care must be taken with data extrapolation, this finding potentially represents a more significant individual and societal burden resulting from wipe-out than first impressions may suggest. Glaucoma is a bilateral disease and all the patients in this series had advanced glaucomatous optic neuropathy, with likely similar or potentially worse visual deficits affecting their contralateral eyes. It is common practice to offer cataract surgery for the eye with less severe glaucoma first in order to maximise the visual rehabilitation achievable with a single procedure. If wipe-out were to occur in a patient’s better-seeing eye, this could have devastating consequences for the patients concerned. It therefore behoves us to further our understanding of wipe-out and we hope that further research follows this study.

Excluded cases

In our study, 7 out of 21 questionnaires were not returned. One possible explanation for this is that wipe-out is not simple to define and so the inclusion and exclusion criteria were not succinct. Some respondents may have latterly discovered that their cases did not meet the criteria and therefore did not proceed to return questionnaires. A further letter was sent to these 7 cases as a follow-up, none replied.

Case 6 had no past ophthalmic history and suffered a significant reduction in visual acuity following cataract surgery. Investigation revealed subtle macula thickening on OCT and this was diagnosed by the treating team as a possible transient retinal artery occlusion, but with no firm diagnosis made. It is possible that this was case of retinal capillary ischaemia, or what is now often termed paracentral acute middle maculopathy (PAMM). PAMM was first reported in 2013 [28] and cataract surgery was performed in 2014 in this case, so it may reasonably be assumed that PAMM was not considered as a differential diagnosis. Furthermore, PAMM is more easily detected on higher-resolution OCT than may have been available at the time, and imaging may not have been performed in a time window when the signs would have been most obvious. PAMM is now a well-documented cause of visual loss post-cataract surgery [29]. The presence of PAMM does not fit with our definition of wipe-out which requires there to be no new post-operative pathology. As a result, this case was excluded from analysis.

Case 7 also had no ocular comorbidities and developed a significant drop in visual acuity following cataract surgery. On further investigation, it was noted that a sub-conjunctival antibiotic preparation may have been injected intracamerally during surgery which may account for the reduction in vision post-operatively. Case 7 was therefore excluded from analysis.

General discussion

Fortunately, wipe-out following cataract surgery appears to be rare; our study suggests an incidence of approximately 3 per million. It is possible that unreported cases of wipe-out occurred during the study timeframe, resulting in an underestimation of the incidence. It is likely that wipe-out is more common in patients with glaucoma, especially in those with advanced glaucoma, as evidenced by the cases we report.

One aim of our study is to shed light on the risk of wipe-out to help inform the consent process for cataract patients. While the overall incidence of wipe-out is low, those with glaucoma (especially advanced glaucoma), black patients, and patients with existing retinal vein occlusions could be counselled to have a higher risk. There is possibly a role for additional ocular antihypertensives immediately post-operatively in patients with advanced glaucoma, and it may be prudent to review patients with both glaucoma and existing retinal vein occlusions in the immediate post-operative period to monitor IOP closely.

Limitations

In the study of rare phenomena such as wipe-out, small cohorts increase uncertainty regarding the significance of potential risk factors, as outlined in our paper. It is possible in a future study to use these cases to build a case control study. In addition, inaccurately reported or unreported cases also have the potential to substantially influence the estimated incidence. We acknowledge that the approach taken for prospective surveillance in our study is susceptible to the underreporting of cases, which may be due to clinicians failing to receive or return report cards. If all 7 of the reported but with no questionnaire cases had returned a positive case, the rate might be closer to 1 in 100,000.

Conclusion

Our study suggests that wipe-out is a rare complication with an approximate incidence of 3 per million cases of cataract surgery. Wipe-out occurred exclusively in patients with advanced POAG with an overrepresentation of black people in our case series. An existing diagnosis of retinal vein occlusion was more prevalent among individuals suffering from wipe-out compared to the general population. Elevated IOP post-operatively may play a role in the pathogenesis of wipe-out. We hope that the findings of our study will be used to help quantify the risk of this serious complication of cataract surgery for patients and surgeons.

Summary

What was known before

• Little was known about wipe-out following cataract surgery, especially in the phacoemulsification era.

• Until now, many surgeons have based their assumptions about wipe-out on anecdotes.

What this study adds

• Wipe-out following cataract surgery is very rare.

• In this study, wipe-out after cataract surgery only occurred in the context of advanced glaucoma.

• Retinal vein occlusion, black ethnicity and high post-operative intraocular pressure may predispose to wipe-out following cataract surgery.

Author contributions

Conception: CR, YS, RK, AV Concept Development: BF Data Collection: CR Data Analysis: CR, CB Manuscript Preparation: CR, YS, RK, CB, BF, AV.

Funding

The research was funded by Moorfields Eye Charity through a grant, GR000135. CB’s post is part funded by the NIHR BRC at the Royal Marsden NHS Foundation Trust and the Institute of Cancer Research, London. The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41433-023-02606-9.