Background Rhegmatogenous retinal detachment (RRD) is the separation of the retina from
the wall of the eye, where it is normally attached. It affects 10-18 people per 100,000
per year. Thus, in the UK, with a population of around 68 million, there are 6700-12,000
RRD per year.
RRD can lead to permanent central and peripheral vision loss (i.e. total blindness) if
not treated in a timely manner. Vitrectomy is the most common surgery done for RRD.
Retinal reattachment following one vitrectomy occurs in 80-88% of eyes, with 59-67%
gaining vision above 0.3 LogMAR, which is the vision required to fulfil UK driving
standards.
Data from the British and Eire Association of Vitreo-Retinal Surgeons (BEAVRS) showed
that most patients with RRD are phakic: they still have their natural lens (i.e. have not
had cataract surgery). Data from the BEAVRS database (unpublished, but accessible to the
COMBAT research group) showed that of 8436 vitrectomies done in non-high myopic phakic
eyes out of 13,900 vitrectomies performed at several sites in the UK, only 10% combined
vitrectomy with cataract surgery (called phacovitrectomy) at the time of the RRD repair.
Of 664 phakic eyes with at least 12 months follow-up in which combined surgery was not
performed at the time of RRD surgery, 320 (48%) were recorded to have a cataract or to
have had it already removed within three months of the original vitrectomy. Only 20% of
these eyes had their own clear natural lens at 12 months following vitrectomy. Therefore,
phacovitrectomy is done in a minority of patients at the time of RRD repair; and many
patients have either a cataract present or undergo separate cataract surgery soon after
the RRD surgery.
Previous studies have shown the potential benefits of doing combined RRD and cataract
surgery. If the cataract is removed when the RRD is repaired, the view of the back of the
eye may be clearer, because the surgeon will see through the clear artificial lens. The
purpose of the vitrectomy is to remove the vitreous, which is a gel-like structure that
fills the back of the eye. The vitreous is involved in the occurrence of the RRD and,
given that the vitreous is also present around the posterior part of the lens of the eye,
removing the lens allows a more thorough removal of the vitreous. This has the potential
to improve the anatomical outcome of the vitrectomy (i.e. reattachment of the retina) and
reduce the number of surgeries required because there would be less need for reoperations
(repeated vitrectomies) if the first vitrectomy fails to put the retina back in place.
Phacovitrectomy would also obviate the need to do the cataract surgery at a later time
following the vitrectomy. Adding a phaco to the vitrectomy will add approximately 8
minutes to the surgical time but could save 6000 cataract surgeries per year to the NHS.
Moreover, when cataract surgery is done after vitrectomy, the vitreous is no longer
present and there is less support to the lens. This may make the cataract surgery more
difficult and increase the risk of intraoperative complications.
However, some studies have suggested that postoperative inflammation, accumulation of
fluid in the centre of the retina (cystoid macular oedema), epiretinal membrane
formation, and posterior synechiae (adherence of the iris to the intraocular lens) may be
more frequent after phacovitrectomy. Calculation of the required artificial intraocular
lens, which is used to replace the natural lens of the eye during cataract surgery, may
be less accurate when doing combined phacovitrectomy in eyes where the macula (i.e. the
central retina) is detached. This is because the calculation of the intraocular lens
requires measuring the length of the eye and this is less accurate if the macula is
detached. However, errors appear to depend on the technique used for this calculation. A
less accurate measure of the intraocular lens would mean people may need glasses to
improve their vision following surgery.
Rationale for the Study Previous studies comparing vitrectomy with phacovitrectomy in
people with RRD have been mostly retrospective and small, with methodological weaknesses.
None have addressed patient reported outcomes, patient views and acceptability for one
surgery over the other, or costs.
Two systematic reviews of phacovitrectomy versus vitrectomy and subsequent cataract
surgery have been published recently. Farahvash et al. included studies in which
vitrectomy or phacovitrectomy were used for a variety of retinal disorders, including
RRD, macular hole, epiretinal membrane, and others. They included six retrospective
before and after studies and one prospective study on RRD but no randomised trials.
Results for the RRD group were not given separately. The Mirshahi et al. review included
studies specifically comparing surgeries in people with RRD. The review included five
retrospective before and after studies, one study defined as a "quasi randomised" and one
randomised trial. Only one of the studies included in the review by Mirshahi et al. was
in the Farahvash et al. review. The randomised trial in Mirshahi et al. review was a
small poorly reported study with 59 patients (59 eyes) randomised to vitrectomy alone
(n=29) or phacovitrectomy (n=30). Although not clearly stated, the trial seems to have
been a non-inferiority trial with a non-inferiority margin of 10% or less between groups
for the primary outcome (retinal reattachment with a single surgery at 6 months).
However, a 10% difference in single surgery success on retinal re-attachment seems
clinically important. The authors stated that, accepting a 5% type 1 error and 80% power,
a sample size of 50 eyes would be needed to detect differences between groups in the
primary outcome. Secondary outcomes included final best-corrected visual acuity (BCVA),
intraocular pressure, central macular thickness, and progression of cataract in the
vitrectomy only group. Mora et al. stated that all outcomes (primary and secondary) were
assessed at 1 week and 1, 3 and 6 months after surgery. The trial found no differences in
outcomes at six months postsurgery except progression of cataract in the vitrectomy
group. Patient reported outcomes and patient experience and acceptability, intra and
post-operative complications and costs were not systematically evaluated.
Authors of both systematic reviews called for randomised trials to determine the best
surgical approach for people with phakic RRD. COMBAT is such a trial.
Rationale for the Intervention Phacovitrectomy (i.e. phaco to remove the cataract and
insert an intraocular lens and vitrectomy to repair the retinal detachment) is performed
routinely throughout the UK to treat people with a idiopathic macular hole. However, it
is not the standard procedure for people with phakic, non-highly myopic RRD.
Rationale for the Comparator Current standard care in the UK is vitrectomy without
concomitant cataract extraction), as shown by data from the BEAVRS database.
Research Hypothesis Combined phacovitrectomy will be preferred by patients because it
will provide faster visual recovery and obviate the need for a second surgery (the
cataract surgery) following RRD repair, and possibly reduce the number of reoperations
related to failure to re-attach the retina. Phacovitrectomy may also reduce other health
inequalities because if cataract surgery is required after RRD repair, long waiting times
for this surgery in the NHS could adversely affect people with lower income who may be
unable to achieve visual rehabilitation as quickly as those with higher income who can
afford to have their cataracts removed privately.
Research Question In people with non-highly myopic phakic RRD (Population) in whom
vitrectomy is planned to repair their RRD, is phacovitrectomy (Intervention) equivalent
(equivalence margin +/- 7 ETDRS letters) to vitrectomy and subsequent cataract surgery
(phacoemulsification) if/when needed (Comparator) for improving vision following surgery
(primary Outcome) but superior for other outcomes in the 52 weeks (+/- 6 weeks) after
surgery.
Study aim To conduct a pragmatic randomised trial evaluating clinical and
cost-effectiveness, safety and acceptability and patient experience of combined
phacovitrectomy vs vitrectomy and subsequent cataract surgery (if/when needed) for
non-highly myopic phakic RRD and to undertake a structured analysis of strategies to
implement the phacovitrectomy pathway, if this is shown to be as good or better than
current standard care.
Study Objectives In people presenting with a phakic, non-highly myopic RRD, the
objectives of the COMBAT study are To determine if the clinical effectiveness of
phacovitrectomy is equivalent (or superior) to vitrectomy and subsequent cataract surgery
(if/when needed).
To compare the cost-effectiveness of phacovitrectomy vs vitrectomy and subsequent
cataract surgery (if/when needed).
To evaluate the participant experience and acceptability of phacovitrectomy compared with
vitrectomy and subsequent cataract surgery (if/when needed).
To evaluate the post-trial implementation and scalability of phacovitrectomy.
Study Design Pragmatic, allocation-concealed, single-masked (outcome assessors),
multicentre, randomised, equivalence trial with an internal pilot to assess feasibility
of recruitment.
Internal Pilot The internal pilot will take place during the first nine months of
recruitment, to assess feasibility of recruitment and determine if the study should
continue to a full trial. The progression criteria will be assessed nine months after the
first site is open to recruitment. The target recruitment will be an average of 0.5
participants per month per open site. With staggered opening of sites, the recruitment
target for the pilot is 63 participants (23% of the required sample size for the full
trial). Criteria for progression to the full trial are GREEN (Average recruitment
rate/site/month: 0.5; number of sites opened: 24; number of participants recruited: 63):
Progress to full trial.
AMBER (0.25-0.49; 12-23; 31-62): Discuss feasibility with the TSC and NIHR and develop a
recovery plan to reach the recruitment target, including evaluating options to improve
recruitment and considering number of eligible patients identified, percentage of these
patients randomised and reasons for non-randomisation, site recruitment performance, and
recruitment procedures.
RED (<0.25; <12; <31): Decision to progress will be made by the TSC and NIHR.
Study Setting Recruitment for COMBAT will take place in at least 30 hospital eye services
(HES) across the four UK nations, with catchment areas that cover diverse populations.
Sequence Generation When a potentially eligible person with a RRD is identified, they
will be informed about COMBAT. If they are willing to take part, informed consent will be
obtained, and eligibility confirmed. If eligible, the baseline visit will be undertaken
(as soon as possible and before the surgery) and participants will be randomised (1:1) to
receive phacovitrectomy or vitrectomy. An automated system with the allocation concealed
to the ophthalmologist recruiting the participant will be used to generate the random
allocation sequence. Minimisation will be used to balance allocation across treatment
groups for potentially important factors including centre, surgeon's experience
(consultant/vitreoretinal fellow), presenting BCVA (BCVA ≥ 69 ETDRS letters [Snellen
equivalent ≥ 20/40; logMAR ≥ 0.3], 24-68 ETDRS letters [Snellen equivalent ≤20/50-20/320;
logMAR 0.4-1.2], <24 ETDRS letters [Snellen equivalent <20/320]); macular status (macula
on & fovea on, macula off & fovea on, fovea off <7 days, fovea off 7-14 days, fovea off
15-30 days, fovea off >30 days); and presence/absence of PVR-C.
Allocation Concealment An automated system will be used to generate the random allocation
sequence. This will conceal the allocation from the ophthalmologist.
Allocation Implementation Each participant will be allocated their own unique trial
identifier during the randomisation process, which will be used throughout the study for
participant identification on all data collection forms and questionnaires. The patient's
allocation will be provided only to unmasked staff (e.g. the surgeon, theatre team).
Definition of Study Eye The unit of randomisation will be the participant, not the eye.
In participants with RRD in both eyes, if both eyes are eligible, the "study eye" will be
the eye with the best BCVA. If both eyes have the same BCVA, the study eye will be,
arbitrarily, the right eye. If both eyes are eligible, the fellow eye would be treated,
if possible, in the same way as the study eye. This would apply also if the fellow eye
does not have a RRD at presentation but develops it during the trial period (i.e. the
participant will be offered the same type of surgery in the fellow eye to that received
in the Study Eye, provided that this is pertinent and the participant agrees).
Recruitment Strategy COMBAT will include at least 30 sites. Considering that urgent
vitreoretinal cases and specifically RRD are currently done by vitreoretinal fellows in
the UK, vitreoretinal fellows will be allowed to take part in the trial. Potentially
eligible participants will be identified by members of the direct care team from urgent
optometry referrals, from Eye Casualty and from Retina and Vitreoretinal clinics at
participating sites. A member of the COMBAT team will then meet with potential
participants who express interest in taking part in COMBAT to provide verbal and written
details about the trial, including the Participant Information Leaflet (PIL). An audio
version of the PIL will be made available where required. If a participant is eligible
and provides informed consent, they will be enrolled in the trial. As part of the consent
process, patients will be asked if they agree to be invited, at a later date, to take
part in a series of 3 (maximum) interviews. Participants will also be asked if, following
completion of their participation in the COMBAT trial, data collected as part of their
standard care, can be used for future follow-up research.
Screening Procedure All screening data will be recorded via electronic data capture (EDC)
to document all participants screened for the study and all participants recruited.
Participants screened and not recruited to the study will be documented via EDC,
including reasons for not being enrolled. A minimal dataset will be recorded for
non-recruited patients, including Age, Sex at Birth, Sexual Orientation, Ethnicity,
Preferred language, Socioeconomic Status (e.g. Employment Status, Education level) and
partial post code (with the exception of those in Scotland) to determine if there are
differences with those willing to participate and recruited. Consent will not be
requested to collect this full minimal dataset as data is fully anonymised and is
partially collected as part of standard care. Where these data are not routinely
collected, participants will be provided with a brief questionnaire and consent will be
implied by its completion and return.
Informed Consent COMBAT will be conducted in accordance with the ethical principles that
have their origin in the Declaration of Helsinki. Eligible participants may only be
included in the trial after written informed consent is obtained. Informed consent must
be obtained prior to conducting any trial specific interventions.
Process Evaluation with the Purpose of Informing Implementation A process evaluation
focused on the provider perspective of phacovitrectomy will be carried out to assess how
the treatment pathways were delivered, barriers and facilitators to implementation, and
the influence of local service and wider contexts in implementing phacovitrectomy.
Firstly, semi-structured, theory-driven interviews will be carried out with
multidisciplinary groups of healthcare providers, including surgeons, nurses and service
managers. These will start 6-12 months after a site starts recruitment to COMBAT,
allowing them to develop a level of familiarisation with the service. At the start of
recruitment to the trial, these staff will be asked to record important items they needed
to put in place at their site to allow implementation of trial procedures (e.g. training
of staff to carry out biometry on people with macula-off RRD). Data collection will be
guided by the updated Consolidated Framework for Implementation Research (CFIR), which
allows systematic study of barriers and drivers to intervention implementation. Secondly,
a structured analysis of the required skills and resources and wider strategies used to
introduce phacovitrectomy and support implementation over time will be carried out by a
study researcher using the ERIC framework (Expert Recommendations for Implementing
Change). Together with analysis of barriers and facilitators, and patient data about
treatment experience and acceptability; this will assist the development of an
implementation blueprint for phacovitrectomy, which would be ready if the COMBAT results
favour phacovitrectomy. It is likely to include strategies related to training staff to
carry out biometry on people with macula-off RRD and to influence the attitude of
surgeons towards phacovitrectomy for non-highly myopic, phakic, RRD.
Sample Size COMBAT is powered to demonstrate treatment equivalence for the primary
outcome. Based on two one-sided t-tests at the 2.5% significance level, a 15.2-letter
standard deviation and an equivalence margin of +/- 7 ETDRS letters (with a difference of
7 or fewer letters being considered not clinically meaningful), 248 participants would be
required to be 90% sure that the 95% confidence interval will exclude a difference in
means of more than 7 ETDRS letters. Allowing for 10% dropout, this requires at least 276
participants. This will be sufficient to detect a difference of 16% in the rate of
retinal reattachment surgery between groups, which is clinically relevant.
Studies Within a Trial (SWATs) Cochrane reviews of recruitment and retention for
randomised trials found little evidence for strategies targeting underserved groups and
what limited evidence is available is methodologically poor. Greater inclusion of
underserved populations is also a NIHR Equality, Diversity and Inclusion (EDI) strategic
priority and meeting this need will require evidence of the effects of strategies that
might be used. One observational ("SWAT A") and one intervention SWAT ("SWAT B) will be
included in COMBAT to target recruitment and retention of underserved groups. SWAT A will
record the proportion in each of the demographic groups who are recruited and retained at
each site. Rewards will be offered to the best performing site teams over the recruitment
period. Information will be collected on how often the translated PIL are used and
whether people for whom these are used are recruited and retained. SWAT B will evaluate
whether an EDI-informed PIL increases the recruitment of underserved groups compared to
the standard PIL. COMBAT sites that are willing to take part in this SWAT will be asked
to use the modified or the standard PIL, in a random order generated by the study
statistician. Detailed procedures for the SWATs will be recorded outside of the study
protocol and made publicly available with a statistical analysis plan before analysis.
