Denniston Crabb BJO 2017

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

 

             

City Research Online

City, University of London Institutional Repository

Citation: Denniston, A. K., Chakravarthy, U., Zhu, H., Lee, A. Y., Crabb, D. P. ORCID:
0000-0001-8754-3902, Tufail, A., Bailey, C., Akerele, T., Al-Husainy, S., Brand, C., Downey,
L., Fitt, A., Khan, R., Kumar, V., Lobo, A., Mahmood, S., Mandal, K., Mckibbin, M., Menon,
G., Natha, S., Ong, J. M., Tsaloumas, M. D., Varma, A., Wilkinson, E., Johnston, R. L. and
Egan, C. A. (2017). The UK Diabetic Retinopathy Electronic Medical Record (UK DR EMR)
Users Group, Report 2: real-world data for the impact of cataract surgery on diabetic
macular oedema. British Journal of Ophthalmology, 101(12), pp. 163-1678. doi:
10.1136/bjophthalmol-2016-309838

This is the published version of the paper.

This version of the publication may differ from the final published
version.

Permanent repository link: http://openaccess.city.ac.uk/21799/

Link to published version: http://dx.doi.org/10.1136/bjophthalmol-2016-309838

Copyright and reuse: City Research Online aims to make research


outputs of City, University of London available to a wider audience.
Copyright and Moral Rights remain with the author(s) and/or copyright
holders. URLs from City Research Online may be freely distributed and
linked to.

City Research Online: http://openaccess.city.ac.uk/ publications@city.ac.uk


Clinical science

The UK Diabetic Retinopathy Electronic Medical

Br J Ophthalmol: first published as 10.1136/bjophthalmol-2016-309838 on 9 May 2017. Downloaded from http://bjo.bmj.com/ on 11 May 2018 by guest. Protected by copyright.
Record (UK DR EMR) Users Group, Report 2: real-
world data for the impact of cataract surgery on
diabetic macular oedema
Alastair K Denniston,1,2 Usha Chakravarthy,3 Haogang Zhu,4 Aaron Y Lee,5
David P Crabb,4 Adnan Tufail,2 Clare Bailey,6 Toks Akerele,7 Sahar Al-Husainy,8
Christopher Brand,9 Louise Downey,10 Alan Fitt,11 Rehna Khan,12 Vineeth Kumar,13
Aires Lobo,14 Sajjad Mahmood,15 Kaveri Mandal,16 Martin Mckibbin,17
Geeta Menon,18 Salim Natha,19 Jong Min Ong,20 Marie D Tsaloumas,1 Atul Varma,21
Elizabeth Wilkinson,22 Robert L Johnston,23 Catherine A Egan,2 on behalf of the UK DR
EMR Users Group

►► Additional material is Abstract such as prostaglandins, thromboxane A, nitric acid


published online only. To view Aim  To assess the rate of ’treatment-requiring diabetic and various cytokines in addition to vascular endo-
please visit the journal online
(http://d​ x.​doi.o​ rg/​10.​1136/​ macular oedema (DMO)’ in eyes for the two  years before thelial growth factor (VEGF).4–7 Patel et al4 showed
bjophthalmol-​2016-​309838). and after cataract surgery. that some of these changes are present within  1 day
Methods  Multicentre national diabetic retinopathy (DR) of surgery and do not return to normal for several
For numbered affiliations see database study with anonymised data extraction across weeks. Retinovascular changes also occur, including
end of article.
19 centres from an electronic medical record system. vascular hyperpermeability and leukostasis.4–6 8
Inclusion criteria: eyes undergoing cataract surgery in Despite these recognised changes in the ocular
Correspondence to
Catherine A Egan, Moorfields patients with diabetes with no history of DMO prior to microenvironment, clinical studies are incon-
Eye Hospital NHS Foundation study start. The minimum dataset included: age, visual clusive as to the effect of cataract surgery on
Trust, 162 City Road, London, acuity (all time-points), injection episodes, timing of the onset of diabetic macular oedema (DMO).
EC1V 2PD, UK; C ​ atherine.​ cataract surgery and ETDRS grading of retinopathy and Furthermore there is quite marked variation in the
Egan@​moorfields.​nhs.u​ k
maculopathy. Main outcome measure: rate of developing rates of DMO reported in such studies, which may
Received 31 October 2016 first episode of treatment-requiring DMO in relation to reflect study design, variable definition of DMO,
Revised 14 March 2017 timing of cataract surgery in the same eye. and the population sampled. Most studies in this
Accepted 19 March 2017 Results  4850 eyes met the inclusion criteria. The rate of area are small with fewer than 100 patients.9–13 An
Published Online First developing treatment-requiring DMO in this cohort was
9 May 2017 important exception is ETDRS Report 25 which
2.9% in the year prior to surgery versus 5.3% in the year analysed the subset of 270 eyes (205 patients) who
after surgery (p<0.01). The risk of ’treatment-requiring underwent cataract surgery during the course of
DMO’ increased sharply after surgery, peaking in the 3–6 the prospective ETDRS randomised controlled
months’ period (annualised rates of 5.2%, 6.8%, 5.6% trial (RCT). These patients with non-proliferative
and 4.0% for the 0–3, 3–6, 6–9 and 9–12 months’ diabetic retinopathy (NPDR) had mild to severe
post-operative time periods respectively). Risk was or early proliferative diabetic retinopathy (PDR)
associated with pre-operative grade of retinopathy: risk and/or macular oedema. This was a post-hoc anal-
of DMO in the first year post-operatively being 1.0% (no ysis of patients who happened to undergo cataract
DR pre-operatively), 5.4% (mild non-proliferative diabetic surgery during the study period without stan-
retinopathy; NPDR), 10.0% (moderate NPDR), 13.1% dardisation of timing or perioperative manage-
(severe NPDR) and 4.9% (PDR) (p<0.01). ment. With regard to macular oedema the defined
Conclusions  This large real-world study demonstrates endpoint was clinically significant macular oedema
that the rate of developing treatment-requiring DMO (CSME), which was defined without the use of
increases sharply in the year after cataract surgery for optical coherence tomography (OCT) imaging and
all grades of retinopathy, peaking in the 3–6 months’ was based on the clinical endpoint of stereoscopic
postoperative period. Patients with moderate and severe
fundus photography. This study showed no signif-
NPDR are at particularly high risk.
icant increase in the rate of CSME with 29% of
patients having CSME at the study visit prior to
cataract surgery and 31% having CSME post-sur-
Introduction gery.9 A more recent prospective study looking at
Visually significant cataract is a common problem unilateral cataract surgery in 132 eyes (with the
To cite: Denniston AK, in people with diabetes, occurring earlier and with fellow eye acting as control) noted a trend towards
Chakravarthy U, Zhu H, et al. a prevalence of 2–5 times that of the population a higher rate of CSME in the operated eye versus
Br J Ophthalmol without diabetes.1–3 Cataract surgery is known to the fellow eye (6.1% vs 4.5%). In this study all
2017;101:1673–1678. lead to increased levels of inflammatory mediators patients had 3 monthly fluorescein angiography;
Denniston AK, et al. Br J Ophthalmol 2017;101:1673–1678. doi:10.1136/bjophthalmol-2016-309838 1673
Clinical science
OCT was not performed in all patients but only those with treatment, but the term ‘treatment-requiring’ is used to indi-

Br J Ophthalmol: first published as 10.1136/bjophthalmol-2016-309838 on 9 May 2017. Downloaded from http://bjo.bmj.com/ on 11 May 2018 by guest. Protected by copyright.
suspected CSME.12 cate that the severity of DMO was considered sufficient by the
The use of electronic medical record (EMR) systems which attending clinician to institute treatment. The EMR mandated
routinely gather clinically relevant data provides the opportunity structured assessment only requires recording of the location of
to analyse larger study cohorts than would be practical within retinal thickening, not the causative factor(s). Recorded post-op-
a clinical trial. In the UK, the most widely-used ophthalmic erative episodes of retinal thickening in patients with diabetes
EMR has implemented a nationally-defined minimum dataset may therefore include some cases of pseudophakic macular
for diabetic retinopathy (DR) that mandates recording of the oedema (post-operative, cystoid macular oedema). In line with
minimum clinical signs necessary to allow an algorithm within the study aims of providing a ‘real world’ analysis, patients were
the software to automatically calculate a precise proxy-ETDRS/ included even if they had additional pre-operative risk factors
International Clinical Grading System retinopathy and maculop- (such as history of uveitis) or intra-operative complications.
athy grade after each consultation (as described in UK DR EMR
Report 1  appendix 1).14 All interventions including laser proce- Analysis
dures, intravitreal injections and ophthalmic operations are The primary outcome measure was the rate of developing first
recorded, providing the opportunity to assess whether the rate episode of treatment-requiring DMO in relation to timing of
of treatment-requiring DMO is influenced by cataract surgery cataract surgery in the same eye, described both as incident risk
in a large ‘real-world’ cohort. In the UK, the National Health and as cumulative risk over each successive year. Subsidiary
Service (NHS) use of ranibizumab is regulated by the National analyses included effect of level of retinopathy on the rate of
Institute for Health and Care Excellence (NICE), with treat- developing treatment-requiring DMO in the same eye in relation
ment only being funded for DMO where it is demonstrated on to cataract surgery. Survey of practice included analysis of the
OCT to be associated with a central macular thickness (CMT) provision of type of treatments used for DMO and their timing.
of over 400 microns.15 It is important to note that CMT is not
precisely defined as, for example, subfoveal thickness or central Results
grid thickness. Random audits of ranibizumab use by commis-
Baseline characteristics
sioners have the potential to withdraw funding, so there is a
Participants
strong motivation for physicians to comply with the guidance.
Data were extracted on 123 968 eyes of 61 984 patients with
It is also important to note that there is no visual acuity level
diabetic retinopathy grades (including a grade of no diabetic reti-
specified in the NICE guidance. Due to this unique regulatory
nopathy) in the EMR. There were 33 967 male patients, 28 002
framework, the assessment of ‘Ranibizumab-requiring DMO’
female patients and 15 cases of unrecorded gender.
in the UK therefore provides a surrogate for an OCT-defined
The inclusion criteria were met for 4850 eyes from 3837
threshold of DMO severity.
patients, namely eyes undergoing cataract surgery in patients
with diabetes, with retinopathy grade(s) recorded in the  2 years
Methods pre-sugery and post-surgery, and who had not developed DMO
Ethical approval requiring treatment prior to study entry (ie, 2 years pre-surgery).
The lead clinician and Caldicott Guardian at each centre gave Treatments for DMO included macular laser, injection of intra-
written approval for extraction of anonymised data. The study vitreal anti-VEGF therapies (ranibizumab, bevacizumab, afliber-
protocol was approved by the head of research governance at cept) and injection of intravitreal corticosteroids (triamcinolone,
the lead clinical centre. This study was conducted in accordance Iluvien and Ozurdex).
with the Declaration of Helsinki and the UK Data Protection The study group comprised 1719 eyes with no apparent
Act. DR, 1034 with mild NPDR, 1527 with moderate NPDR, 165
with severe NPDR, and 405 with proliferative DR.
Data collection
Anonymised data were remotely extracted from 19 centres using Rate of developing first episode of treatment-requiring DMO
the same EMR system (Medisoft Ophthalmology, Medisoft, in relation to cataract surgery
Leeds, UK). Each site is the only NHS provider of DMO care The rate of developing first episode of treatment-requiring DMO
to their local population and very few patients switch between in the whole cohort was 3.1% and 2.9% in each of the 2 years
providers or access care privately. Data were extracted through prior to surgery versus 5.3% and 4.8% in each of the 2  years
the EMR compulsory DR structured assessment module. The after surgery (p<0.01; see table 1). More detailed analysis by
minimum dataset included: age, visual acuity at baseline and 3-month intervals demonstrated that the risk of ‘treatment-re-
at all subsequent visits, injection episodes, timing of cataract quiring DMO’ increased sharply after surgery, peaking in the
surgery and proxy-ETDRS/International Clinical Grading system 3–6 months post-operative period (figure 1). The cumulative
grading of retinopathy and maculopathy. risk rose from 6.2% for the 24-month pre-operative period to
14.7% for the full 48-month peri-operative period (see table 2).
Inclusion criteria
Eyes undergoing cataract surgery from patients with diabetes, Effect of retinopathy grade on the risk of developing first
with retinopathy grade(s) recorded in the 2 years pre-surgery episode of ‘Treatment-requiring DMO’ in relation to cataract
and post-surgery, and who had not developed DMO requiring surgery
treatment prior to study entry (ie, 2 years pre-surgery). For the The risk of developing ‘Treatment-requiring DMO’ was signifi-
purposes of this study we analysed the development of DMO in cantly associated with pre-operative grade of retinopathy. In
terms of the requirement for treatment, that is DMO of sufficient the first year after cataract surgery the risk of developing first
severity to warrant an intervention; this is referred to as ‘treat- episode of ‘Treatment-requiring DMO’ was 1.0% (no DR seen
ment-requiring DMO’ throughout the study. It is recognised that pre-operatively), 5.4% (mild NPDR), 10.0% (moderate NPDR),
some of these cases might have spontaneously resolved without 13.1% (severe NPDR) and 4.9% (PDR) (table 1, figure 2). For
1674 Denniston AK, et al. Br J Ophthalmol 2017;101:1673–1678. doi:10.1136/bjophthalmol-2016-309838
Clinical science

Br J Ophthalmol: first published as 10.1136/bjophthalmol-2016-309838 on 9 May 2017. Downloaded from http://bjo.bmj.com/ on 11 May 2018 by guest. Protected by copyright.
Table 1  Risk of Developing First Episode of Treatment-Requiring diabetic macular oedema (DMO)
Time in relation to cataract surgery (months) −24 to −12 −12 to 0 0 to +12 +12 to +24
Risk of DMO (%) All eyes (n=4850) 3.1 2.9 5.3 4.8
Stratified by ETDRS grade*
No DR (n=1719) 0.0 0.0 1.0 0.8
Mild NPDR (n=1034) 2.1 2.6 5.4 5.6
Moderate NPDR (n=1527) 5.4 5.5 10.0 8.9
Severe NPDR (n=165) 11.5 11.0 13.1 11.5
PDR (n=405) 5.9 4.2 4.9 4.2
*Stratification is by proxy-ETDRS/International Clinical Grading system grade of retinopathy at most recent assessment prior to cataract surgery (mild, moderate and severe
NPDR and PDR retinopathy). Time periods are up to but not including the upper boundary, ie, −24 to −12 equates to any time point from −24 months to 1 day prior to −12
months. An event occurring at exactly −12 months is included in the −12 to 0 month group. Similarly, patients treated on the day of surgery are included in the 0–12 month
group. Since development of DMO is a censoring event, the number of eyes at risk reduces over time as follows: 4850 (−24 to −12 months), 4702 (−12 to 0 months), 4565 (0
to +12 months), 4323 (+12 to +24 months).
NPDR, non-proliferative diabetic retinopathy; PDR, proliferative diabetic retinopathy.

mild and moderate grades of retinopathy this represented an Treatment modalities for DMO
approximate doubling of the risk of DMO compared with the Of the 4850 eyes included in this study, 622 received treatment
year prior to surgery; the relative increase in risk was less great for the first episode of ‘Treatment-requiring DMO’ during the
for severe NPDR and for PDR, but it is noteworthy that the risk study period (ie, up to 2 years post cataract surgery). Macular
of DMO in the severe NPDR group was more than twice that of laser was the predominant therapy used both before and
the PDR group. after cataract surgery, comprising 84.6% vs 68.5% of all first
treatments for DMO in the 2 years pre-surgery versus 2 years
Cumulative risk of developing Treatment-Requiring DMO over post-surgery (figure 3). The proportion of cases treated with
time ranibizumab increased over eightfold from 2.5% to 20.8% of all
The cumulative risk of developing ‘Treatment-requiring DMO’ first treatments for DMO; the use of ‘other treatments’ showed
was calculated according to the level of retinopathy prior to little change, being 13% vs 10.7% (pre-surgery versus post-sur-
surgery. The cumulative risk varied according to retinopathy gery respectively).
grade and time in relation to surgery. It should be noted that as
per the study inclusion criteria no patients had ‘treatment-re-
Discussion
quiring DMO’ at their baseline (defined as 2 years prior to cata-
This real-world dataset collated from a large number of centres
ract surgery). Cumulative risk of developing first episode of
across the UK demonstrates the increased rate at which patients
‘Treatment-requiring DMO’ at 4 years of follow-up (ie, 2 years
with diabetes develop DMO in the period after cataract surgery.
post cataract surgery) was 14.7% for the whole cohort, but
The percentage of eyes developing first episode of ‘treatment-re-
varied significantly according to retinopathy being 1.4% (no DR
quiring DMO’ increased from 2.9% in the year immediately
seen pre-operatively), 13.4% (mild NPDR), 29.5% (moderate
prior to surgery to 5.3% in the first year and 4.8% in the second
NPDR), 59% (severe NPDR) and 19.1% (PDR) (table 2,
year after surgery; the highest risk period was between 3 and 6
figure 2).
months after surgery with 1.7% of eyes developing first episode
of ‘treatment-requiring DMO’ during that 3-month period,
equating to an annualised risk of 6.8%. The percentage of
‘ranibizumab-requiring DMO’ which in the UK is based on an
OCT definition of DMO with a minimum CMT of 400 microns,
increased from 0.1% for the 2-year period prior to surgery to
1.5% of eyes for the 2-year period post-surgery.
Our findings provide additional data to suggest that there
is a real increase in treatment-requiring DMO after cataract
surgery. This has been a matter of considerable debate, espe-
cially since the ETDRS Report 25 failed to show a significant
increase with 29% patients being noted to have CSME at their
study visit occurring prior to cataract surgery and 31% having
CSME at their study visit occurring post-surgery.9 It should
be noted however that this study may now be less relevant to
current practice due to the modernisation of cataract surgery
(including the widespread adoption of phacoemulsification since
that time) and advances in the care of diabetic eye disease. Addi-
tionally the population analysed in ETDRS Report 25 differed
from most studies (including ours) in that it had a very high
Figure 1  Risk of developing first episode of ‘Treatment-requiring proportion of patients with more advanced retinopathy: PDR
diabetic macular oedema (DMO)’ in relation to cataract surgery. Risk of and/or vitreous haemorrhage was present in 40% of those who
developing first episode of ‘Treatment-requiring DMO’ in patients with had a gradeable fundus photograph from the visit immediately
diabetes in the 2 years before and after cataract surgery. Percentages preceding their cataract operation. A more recent prospective
given are of risk of ‘treatment-requiring DMO’ in that 3-month period. study looking at unilateral cataract surgery in 132 eyes (with the
Denniston AK, et al. Br J Ophthalmol 2017;101:1673–1678. doi:10.1136/bjophthalmol-2016-309838 1675
Clinical science

Br J Ophthalmol: first published as 10.1136/bjophthalmol-2016-309838 on 9 May 2017. Downloaded from http://bjo.bmj.com/ on 11 May 2018 by guest. Protected by copyright.
Table 2  Cumulative risk of developing first episode of treatment-requiring diabetic macular oedema (DMO)
Time in relation to cataract surgery (months) −24 to −12 −24 to 0 −24 to +12 −24 to +24
Cumulative Risk of DMO All eyes 3.1 6.2 12.1 14.7
(%) Stratified by ETDRS grade*
No DR (n=1719) 0.0 0.0 1.0 1.4
Mild NPDR (n=1034) 2.2 4.8 10.7 13.4
Moderate NPDR (n=1527) 5.6 11.8 23.9 29.5
Severe NPDR (n=165) 12.6 26.7 46.0 59.0
PDR (n=405) 6.2 10.9 16.5 19.1
*Stratification is by ETDRS grading of retinopathy at most recent assessment prior to cataract surgery (mild, moderate and severe NPDR and PDR retinopathy). See previous table
legend regarding upper boundaries of time periods.
NPDR, non-proliferative diabetic retinopathy; PDR, proliferative diabetic retinopathy.

fellow eye acting as control) noted a trend towards higher rate of It is important to note that our study is not directly compa-
CSME in the operated eye versus the fellow eye (6.1% vs 4.5%) rable to earlier studies in that we are not assessing CSME, but
within the first 6 months.12 It should be noted that the difference rather ‘treatment-requiring DMO’. ‘Treatment-requiring DMO’
between rates of CSME (both pre-operative and post-operative) is a clinically relevant group which, while subjective, provides
may also reflect the inclusion of patients with milder levels of an important metric of the burden of disease on the patient and
retinopathy in this later study. The findings of our study indi- the clinical service. The ‘ranibizumab-requiring DMO’ subgroup
cate that this previously observed trend towards higher levels provides an objectively defined disease threshold for DMO due
of treatment for DMO in the post-operative phase, is indeed to the regulatory requirements in the UK requiring an OCT-de-
significant and provides, for the first time, evidence that there fined threshold of DMO with a CMT of >400 microns before
is a ‘high-risk’ period in the post-operative phase which peaks treatment with ranibizumab is funded. This avoids the subjec-
between 3 and 6 months. This dataset is taken using a single data tivity of clinically-defined CSME, which may show significant
capture platform across a large population in the UK, reflecting a inter-observer variability. Both the ‘treatment-requiring DMO’
wide ethnic-mix from urban and rural environments. The EMR group and the ‘ranibizumab-requiring DMO’ subgroup show the
mandates a structured assessment of DR resulting in an ETDRS same significant increase in treatment for DMO the post-oper-
grading of retinopathy and maculopathy based on the presence ative phase, and both peak between 3 and 6 months post-oper-
or absence of clinical signs at each clinic visit. atively. It is recognised that in this analysis  we have considered
neither the actual level of CMT nor which instrument it has been
measured on (which may influence the threshold)16; these OCT
data are not currently available for our dataset but may be avail-
able in future data extractions.
The relatively modest use of ranibizumab as first line treatment
in this cohort probably reflects the fact that most UK centres
only adopted its use routinely after the NICE guidance in 2013.

Figure 2  Effect of retinopathy on risk of developing first episode of


‘Treatment-requiring diabetic macular oedema (DMO)’ in relation to
cataract surgery. Risk of developing first episode of ‘Treatment-requiring
DMO’ in patients with diabetes in the 2 years before and after cataract
surgery. Percentages given are of risk of ‘treatment-requiring DMO’ in
that 3-month period. Stratification is by ETDRS grading of retinopathy
at most recent assessment prior to cataract surgery (mild, moderate and
severe non-proliferative diabetic retinopathy (NPDR) and proliferative Figure 3  Treatment selection for diabetic macular oedema (DMO)
diabetic retinopathy (PDR) retinopathy). The variability of the ‘Severe over time in relation to cataract surgery. Treatment selection for first
NPDR’ cohort is likely to be a function of its significantly smaller size episode of ‘Treatment-requiring DMO’ in patients with diabetes in
(n=165) compared with the other cohorts (n=1719 for no apparent DR, the 2 years before and after cataract surgery. ‘Other’ comprises other
n=1034 for mild NPDR, n=1527 for moderate NPDR, and n=405 for intravitreal treatments including alternative anti-vascular endothelial
proliferative DR). growth factor (VEGF) agents and intravitreal corticosteroid therapies.
1676 Denniston AK, et al. Br J Ophthalmol 2017;101:1673–1678. doi:10.1136/bjophthalmol-2016-309838
Clinical science
The increased availability of this drug at later time points of the very expensive and generally deliberately exclude large sections

Br J Ophthalmol: first published as 10.1136/bjophthalmol-2016-309838 on 9 May 2017. Downloaded from http://bjo.bmj.com/ on 11 May 2018 by guest. Protected by copyright.
study must be borne in mind when interpreting any increase of the population for whom the treatment is clinically indicated.
in its usage over time, such as the fact that the proportion of ‘Real-world’ studies are of immense value in a number of situ-
cases treated with ranibizumab increased from 2.5% for the ations. First, they provide some indication of how a treatment
2 years prior to surgery to 20.8% for the 2 years after surgery. In effect seen in the ideal but artificial environment of an RCT
contrast any decline in usage (such as after the observed peak in actually translate into benefits (and adverse events) in a much
the 4 to 6 month period post-surgery) can be assumed to be real more heterogeneous group of patients in the less controlled
(since availability would have increased over this period). The environment of clinical practice. Second, they provide a mecha-
treatment data therefore reliably inform us that the risk of devel- nism of identifying and investigating rarer conditions, for which
oping DMO exceeding 400 microns peaks in the 4 to 6 month a suitably scaled RCT would be unaffordable.
post-operative period, which is also in line with the incidence Data from ‘real world’ studies can still be quality-assured,
of cases of DMO overall. Ideally a future study would take a although not to the same level as an RCT. In our study, the design
new cohort with study inception after 2013. This would show of the EMR system ensures that the data entry is structured,
a much higher proportion of ranibizumab (and more recently key fields are compulsory, and we have demonstrated previ-
aflibercept) usage both pre-surgery and post-surgery, and would ously low rates of missing non-compulsory fields.17 Fields have
provide a more complete estimate of the proportion of ‘treat- value cut-offs to stop major data entry errors; most minor errors
ment-requiring DMO’ which is associated with a macular thick- may occur in either direction and so with a large sample size it
ness of greater than 400 microns. is anticipated that these would become negligible. In summary,
A possible alternative explanation when assessing studies we would argue that any additional ‘noise’ in our study brought
reporting progression of diabetic retinopathy and/or macu- about by its real world, retrospective design, will be compen-
lopathy after cataract surgery is whether this simply reflects sated for by the large sample size and multiple physicians and
improved visualisation of the fundus. An advantage of our hospital.
study is that the ranibizumab endpoint reflects an OCT-defined One of the key questions arising from this study is whether
measure in a subset of our cohort, which is usually possible this new evidence regarding the risk of post-operative DMO
pre-surgery and post-surgery. We would however recognise should alter practice, particularly regarding frequency (and
two limitations. First we do not know how rigorously clinicians location) of follow-up in the post-operative phase supported
obeyed the >400 micron limit before offering ranibizumab to by the use of OCT to screen for DMO. The follow-up of
their patients as we have not extracted the OCT data. Second, patients with diabetic retinopathy after cataract surgery is very
OCT was not compulsory at all visits so there may be cases of variable, being primarily dependent on the level of diabetic
DMO with CMT >400 microns which were missed and not retinopathy. In the UK and many other countries, patients
treated. It may be argued that this was more likely to happen with mild NPDR may only receive the same level of hospital
pre-operatively due to worse visualisation. We do not think that follow-up as patients without diabetes. Depending on local
this was a major issue however, as most patients with worsening practice they receive one (or no) hospital reviews before being
vision and diabetic retinopathy being considered for cataract discharged to their community optometrist for post-operative
surgery are likely to have had an OCT, although we cannot refraction. Unless either the patient or the optometrist report
confirm the exact rates of pre-operative OCT scanning in the specific concerns, patients may not receiveany further assess-
current data extract. ment of their fundus until their annual photographic review
It may be difficult to distinguish pseudophakic (postopera- within the National Diabetic Eye Screening Service. It should
tive, cystoid macular oedema), macular oedema from DMO be noted that this assessment is based on the grading of two-di-
post-operatively; the EMR mandated structured assessment only mensional fundus photographs and the recorded level of visual
requires recording of the location of retinal thickening. Standard acuity, and does not include stereo-photographs or OCT assess-
clinical practice in the UK would have been to use topical steroids ment which would assist in the detection of DMO. Our study
and/or non-steroidal drops as first line treatment if the clini- suggests however that the highest risk for the development
cian believed the problem to be pseudophakic macular oedema of DMO occurs in the period after the typical timing of the
followed by intravitreal or sub-Tenon’s triamcinolone acetonide first post-operative visit and occurs at a significant rate even in
as second line treatment. Analysis of our study suggests that this those patients that have been assessed as having mild NPDR. It
would only account for a small proportion of the cases seen. may however be possible to further stratify this risk. A D ​ RCR.​
Intravitreal and sub-Tenon’s triamcinolone together accounted net study found that at sixteen weeks postoperatively, OCT-de-
for around one-third of the ‘Other Treatment’ category (specif- fined centre-involving ME developed in 0% of eyes with no
ically 12% intravitreal, 20% sub-Tenon’s), and did not increase preoperative DMO versus 10% if eyes had had pre-operative
significantly after surgery with ‘Other Treatments’ comprising non–centre-involving DMO and 12% if eyes had been assessed
13% vs 10.7% of total treatments (pre-surgery versus post-sur- as ‘possible’ centre-involving DMO.18  In this context it is also
gery respectively). interesting to note that those patients with PDR appeared to
The lack of compulsory pre-operative and post-operative be relatively protected from developing DMO in the post-op-
OCT and fundus fluorescein angiography (FFA) in all cases is erative phase when compared with NPDR. This is interesting
one example of the differences between ‘real-world’ studies but should be treated with caution as the numbers in the PDR
and prospective RCTs. An RCT can provide stringently quali- group were significantly lower than in the other groups.
ty-assured data, taken on a group of patients with the condi- It may be argued that most patients will be aware of a decline
tion of interest, accompanied by minimal co-pathology and in visual function, and so should be able to self-report visually
with carefully scheduled visits allowing treatment to be deliv- significant DMO; indeed the NICE (UK) threshold is, in part,
ered and monitored under closely controlled conditions. Such based on this equating to a visually significant difference. It is
studies have many advantages, and are valuable particularly for well known however that many patients may ignore the changes
licensing studies as they may provide a high signal:noise ratio in one eye, particularly where reasonable vision is retained in the
and, if well-designed, have a low risk of bias. They are however other eye.19 There may be therefore  an argument for providing
Denniston AK, et al. Br J Ophthalmol 2017;101:1673–1678. doi:10.1136/bjophthalmol-2016-309838 1677
Clinical science
a further OCT-based assessment, whether in the community or Competing interests  RLJ is the Medical Director of Medisoft Limited, the

Br J Ophthalmol: first published as 10.1136/bjophthalmol-2016-309838 on 9 May 2017. Downloaded from http://bjo.bmj.com/ on 11 May 2018 by guest. Protected by copyright.
in the hospital eye clinic, at around 4 months post-operatively to Electronic Medical Record software provider from which data were extracted.
screen for the development of DMO, and to provide the oppor- Provenance and peer review  Not commissioned; externally peer reviewed.
tunity for early treatment. Data sharing statement  All data have been published within this article and the
In summary, this analysis of a real world dataset achieved at accompanying manuscripts.
a large number of centres across the UK indicates an increased © Article author(s) (or their employer(s) unless otherwise stated in the text of the
rate of developing visually significant DMO in the post-operative article) 2017. All rights reserved. No commercial use is permitted unless otherwise
phase, with a particularly high risk from 3 to 9 months. The UK expressly granted.
requirement for an OCT-defined threshold of DMO to access
ranibizumab treatment, and the presence of a clear post-operative
References
1 Klein BE, Klein R, Moss SE. Incidence of cataract surgery in the Wisconsin
DMO spike in this cohort provides further evidence that this is Epidemiologic Study of Diabetic Retinopathy. Am J Ophthalmol
a real increase in DMO and not just an issue of visualisation. We 1995;119:295–300.
would recommend that assessment prior to cataract surgery for 2 Klein R, Klein BE, Moss SE, et al. Prevalence of diabetes mellitus in Southern
any patient with diabetes should include a record of DR severity Wisconsin. Am J Epidemiol 1984;119:54–61.
3 Javadi MA, Zarei-Ghanavati S. Cataracts in diabetic patients: a review article. J
status and a macular OCT of both eyes, while recognising that the Ophthalmic Vis Res2008;3:52–65.
lens opacities may limit assessment in some eyes. There may also 4 Patel JI, Hykin PG, Cree IA. Diabetic cataract removal: postoperative progression
be a role for increased monitoring of patients with even relatively of maculopathy--growth factor and clinical analysis. Br J Ophthalmol
mild diabetic retinopathy during the post-operative phase to help 2006;90:697–701.
5 Johnson MW. Etiology and treatment of macular edema. Am J Ophthalmol
detect DMO and enable earlier treatment. 2009;147–11–21.
Author affiliations 6 Leal EC, Manivannan A, Hosoya K, et al. Inducible nitric oxide synthase isoform
1
University Hospitals Birmingham NHS Foundation Trust, Birmingham, UK is a key mediator of leukostasis and blood-retinal barrier breakdown in diabetic
2
The NIHR Biomedical Research Centre at Moorfields Eye Hospital NHSFT and UCL retinopathy. Invest Ophthalmol Vis Sci 2007;48:5257–65.
Institute for Ophthalmology, London, UK 7 Dong N, Xu B, Chu L, et al. Study of 27 aqueous humor cytokines in type 2 Diabetic
3
Centre for Experimental Medicine, Queens University, Belfast, UK patients with or without Macular Edema. PLoS One2015;10:e0125329.
4
City University, London, UK 8 Guex-Crosier Y. The pathogenesis and clinical presentation of macular edoema in
5
Department of Ophthalmology, University of Washington, Seattle, USA inflammatory diseases. Doc Ophthalmol 1999;97:297–309.
6
Bristol Eye Hospital, University Hospitals Bristol NHS Foundation Trust, Bristol, UK 9 Chew EY, Benson WE, Remaley NA, et al. Results after Lens extraction in patients with
7
Department of Ophthalmology, Hinchingbrooke Health Care NHS Trust, Huntingdon, diabetic retinopathy: early treatment diabetic retinopathy study report number 25.
UK Arch Ophthalmol 1999;117:1600–6.
8
Department of Ophthalmology, Heart of England NHS Foundation Trust, 10 Dowler JG, Hykin PG, Hamilton AM. Phacoemulsification versus extracapsular cataract
Birmingham, UK extraction in patients with diabetes. Ophthalmology 2000;107:457–62.
9
Department of Ophthalmology, Sheffield Teaching Hospitals NHS Foundation Trust, 11 Ostri C, Lund-Andersen H, Sander B, et al. Phacoemulsification cataract surgery in
Sheffield, UK a large cohort of diabetes patients: visual acuity outcomes and prognostic factors. J
10
Department of Ophthalmology, Hull Royal Infirmary, Hull, UK Cataract Refract Surg 2011;37:2006–12.
11
Department of Ophthalmology, Peterborough and Stamford Hospitals NHS 12 Romero-Aroca P, Fernández-Ballart J, Almena-Garcia M, et al. Nonproliferative
Foundation Trust, Peterborough, UK diabetic retinopathy and macular edema progression after phacoemulsification:
12
Department of Ophthalmology, Calderdale and Huddersfield NHS Foundation Trust, prospective study. J Cataract Refract Surg 2006;32:1438–44.
Huddersfield, UK 13 Flesner P, Sander B, Henning V, et al. Cataract surgery on diabetic patients. A
13
Department of Ophthalmology, Wirral University Teaching Hospital NHS Foundation prospective evaluation of risk factors and complications. Acta Ophthalmol Scand
Trust, Wirral, UK 2002;80:19–24.
14
Moorfields Eye Unit, Bedford Hospitals NHS Trust, Bedford, UK 14 Egan C, Zhu H, Lee A, et al. The united kingdom diabetic retinopathy electronic
15
Manchester Royal Eye Hospital, Manchester, UK medical record users group, report 1: baseline characteristics and visual acuity
16
Department of Ophthalmology, Warrington and Halton Hospitals NHS Foundation outcomes in eyes treated with intravitreal injections of ranibizumab for diabetic
Trust, Warrington, UK macular oedema. Br J Ophthalmol 2017;101:75–80.
17
Department of Ophthalmology, Leeds Teaching Hospitals NHS Trust, Leeds, UK 15 NICE. Technology appraisal 274 ranibizumab for treating diabetic macular oedema
18
Department of Ophthalmology, Frimley Park Hospital, Frimley, UK (rapid review of technology appraisal guidance 237). Nice (Published Online First: 27
19
Department of Ophthalmology, Wrightington, Wigan and Leigh NHS Foundation February 2013).
Trust, Wigan, UK 16 Kidess A, Salloukh AE, Dutt P, et al. Measurement bias between optical coherence
20
Department of Ophthalmology, Cambridge University Hospitals NHS Foundation tomography instruments can affect access to treatment: a new lottery. Clin
Trust, Cambridge, UK Ophthalmol2013;7:2299–302.
21
Department of Ophthalmology, Mid Yorkshire Hospitals NHS Trust, Yorkshire, UK 17 The neovascular age-related macular degeneration database: multicenter
22
Department of Ophthalmology, Northern Devon Healthcare NHS Trust, Barnstaple, study of 92 976 ranibizumab injections: report 1: visual acuity. Ophthalmology
UK 2014:121:1092–101.
23
Gloucestershire Hospitals NHS Foundation Trust, Cheltenham, UK 18 Baker CW, Almukhtar T, Bressler NM, et al. Macular edema after cataract surgery
in eyes without preoperative central-involved diabetic macular edema. JAMA
Collaborators  The full list of authors and affilliations comprising the UK DR EMR Ophthalmol 2013:131:870–9.
Users Group is provided in Appendix 1. 19 Mitchell P, Bressler N, Tolley K, et al. Patient-reported visual function
outcomes improve after ranibizumab treatment in patients with vision
Contributors  All the authors have contributed to the planning, conduct and impairment due to diabetic macular edema: randomized clinical trial. JAMA
reporting of the work described in the article. Ophthalmol2013;131:1339–47.

1678 Denniston AK, et al. Br J Ophthalmol 2017;101:1673–1678. doi:10.1136/bjophthalmol-2016-309838

You might also like