Article Text

Download PDFPDF

Rate and risk factors for the conversion of fovea-on to fovea-off rhegmatogenous retinal detachment while awaiting surgery
  1. Andreas Kontos,
  2. Tom H Williamson
  1. Department of Ophthalmology, St Thomas’ Hospital, London, UK
  1. Correspondence to Andreas Kontos, Department of Ophthalmology, St Thomas' Hospital, Westminster Bridge Road, London SE1 7EH, UK; akontos{at}


Background/aims Progression of a fovea-on to a fovea-off rhegmatogenous retinal detachment (RRD) while awaiting surgery is rare.

Methods A retrospective review of patient records to identify patients in whom a fovea-on retinal detachment at presentation was found to be fovea off at surgery.

Results We identified 10 cases over 14 years which converted from fovea-on to fovea-off RRD while awaiting surgery. This represented 1.1% of RRDs that were fovea on at presentation (n=930). Nine out of 10 patients had superotemporal RRDs extending to at least the vascular arcade at presentation and all had superotemporal breaks within detached retina, which was significantly higher than the rate for other retinal detachments (100% vs 63%, p=0.02). There were 2.4 breaks per eye, similar to other retinal detachments. Six of the 10 patients converted to a fovea-off retinal detachment within a few hours and the rest by the following day. Visual outcomes were good, with eight patients maintaining their presenting visual acuity and two losing one Snellen line.

Conclusions Offering same-day surgery to high-risk fovea-on RRDs may not significantly influence visual outcomes and would only prevent about half of the conversions to fovea off. Superotemporal retinal detachments extending to near the arcades are most at risk and might warrant posturing to limit spread of the detachment in the preoperative period.

  • Retina
  • Macula
  • Treatment Surgery

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Rhegmatogenous retinal detachment (RRD) converting from a retinal detachment with an attached fovea (fovea on) to a retinal detachment with a fovea which is detached (fovea off), while the patient awaits surgery, is regarded as a detrimental event. It is assumed that this has an effect on the visual outcome for the patient. Great efforts are therefore made by vitreoretinal units to try to avoid this complication. However, the frequency with which this event occurs is uncertain and whether it is avoidable is unclear. There is considerable pressure on services to offer an early surgery for fovea-on retinal detachments to avoid this. However, it is still unknown whether these patients suffer significant harm as a result. The characteristics of retinal detachments are highly variable,1 and the factors that would allow stratification of risk of conversion to a fovea-off detachment for individual patients are not yet established. One relatively small study of only 199 patients, with a relaxed protocol of urgency of surgery of a few days, has shown a risk of conversion of 0.5% (one patient).2 No study has provided information on whether there are features of the retinal detachment itself, which increase the risk of progression to foveal detachment.

In this study, we wished to determine the frequency of conversion of fovea-on retinal detachment to fovea-off detachment. In addition, the study was designed to allow an assessment of the characteristics of the retinal detachment and to relate these to progression to fovea-off retinal detachment.


A retrospective case note review of patients in our unit who converted from a fovea-on to a fovea-off retinal detachment over a period of 14 years was undertaken (January 2001–December 2014 inclusive). The project was registered with Guy's and St Thomas' NHS Foundation Trust research and development department who waived the need for ethical approval in view of the methodology. This research project abides by the ethical principles underpinning the World Medical Association’s Declaration of Helsinki and relevant good practice guidelines in the conduct of research.

We defined conversion to fovea off as any patient presenting to our unit with fovea-on retinal detachment as confirmed in the paper medical notes by a vitreoretinal surgeon (a consultant or fellow grade), who was then found preoperatively or intraoperatively to have a fovea-off or bisecting retinal detachment, as confirmed by the electronic operation records and drawings. Patients who were referred as fovea-on detachments from other units but were found to have fovea-off/bisecting detachments at presentation to our unit were not included.

Patients were identified using our unit's electronic vitreoretinal surgical patient record database (VITREOR; AxSys Technology, Glasgow, UK) and the case notes as follows. A search was performed on the database to identify patients with fovea-off retinal detachment (as recorded intraoperatively) with duration of central vision loss of ≤3 days prior to surgery. This duration was chosen as for medicolegal reasons, the policy in our unit has been to operate on acute retinal detachments on the next available operating list or out of hours if the fovea was on or recently became off, and at the latest within 3 days of presentation. It would be extremely unlikely for a fovea-on detachment presenting to our unit to be scheduled for surgery >3 days from presentation, therefore this duration was considered to give the best balance between the number of notes to be reviewed and identifying almost all cases of fovea on to off conversion. The medical records were reviewed to identify the clinical entry of the presentation to the department and whether the fovea was recorded to be on or off at presentation.

The duration of vision loss had been entered prospectively at the time of surgery as reported by the patient. There were patients in the database with recorded fovea-off detachments where no duration of central vision loss had been entered. For these patients, a review of the medical records was undertaken to establish the foveal status at presentation.

The database was also searched to establish the number of consecutive primary fovea-on and fovea-off retinal detachments undergoing surgery (pars plana vitrectomy or scleral buckling) over the same time period.

Cases were identified with a clearly documented attached fovea at presentation, which were recorded in the operation notes to be detached or bisected. For these cases, a detailed review of the medical records was undertaken, taking note of the clinical drawings at presentation and at surgery, and the location and number of retinal breaks. The final visual acuity, lens status and follow-up duration for each of these cases were recorded.

Previously vitrectomised eyes or eyes that had other treatment for the retinal detachment such as laser retinopexy were excluded.

A statistical analysis was performed using Analyse-it (Cambridge, UK). The present group was compared with 847 eyes of 847 patients from this time period, which have been extensively investigated and described elsewhere.1 ,3 In brief, the group consisted of 509 (60.1%) men and 338 (39.9%) women, 471 (55.6%) right and 376 (44.4%) left eyes, 477 (56.3%) fovea-on RRDs and 360 (43.7%) fovea-off RRDs. Mean age of the patients was 62.2 years (range 40–97 years). In cases where both eyes were treated for retinal detachment, only the first eye was included in the analysis. Proportions in the two groups were compared using Fisher's exact test, and means were compared using the Mann-Whitney U test.


Ten cases of conversion of fovea on to off were identified over the 14-year period. There were five men and five women with four right eyes and six left eyes. The mean age of the patients was 57 years (range 39–77 years, SD 13 years). Overall, 1982 eyes of 1982 patients were operated on for RRD in this period with 920 recorded as fovea on at surgery and 1062 as fovea off. As 10 of these cases initially presented as fovea on, the total number of fovea-on RRDs at presentation was 930. The rate of progression from fovea on to fovea off was therefore 1.1% (95% CI 0.4 to 1.7).

Retinal detachment characteristics

The clinical drawings for all patients are shown in figure 1. All patients had superior detachments, with nine patients having a predominantly superotemporal retinal detachment (all 3 clock hours detached) and one having a predominantly superonasal detachment. All patients but one had extension of the retinal detachment to at least the edge of the superotemporal vascular arcades at presentation. All were found to have superotemporal retinal breaks and detached retinal breaks in the superotemporal quadrant, features that were significantly higher than in other RRDs from our unit4 (100% vs 69%, p=0.04 and 100% vs 63%, p=0.02, respectively, using Fisher's exact test). There was a mean of 2.4 breaks per eye (range 1–5 breaks, SD 1.4 breaks), which was similar to other retinal detachments (p=0.943, Mann-Whitney U test). All were early retinal detachments with a mean of 3.7 clock hours of extension (range 3–5 hours, SD 0.8 clock hours). A more detailed comparison of the characteristics of the fovea-on to and fovea-off retinal detachments and other detachments is shown in table 1.

Table 1

Comparison between patients with fovea on to off and controls

Figure 1

Clinical drawings. For each patient (1–10), clinical drawings made at the time of presentation in the medical notes were replicated (top drawings), together with the intraoperative drawings (bottom). These highlight the observation that all detachments had detached superotemporal breaks and almost all had subretinal fluid extending to at least the superotemporal vascular arcade.

Timing of surgery

Six of the 10 patients had surgery on the same day they presented to our department. Two patients had surgery on the following day, one patient had surgery at 2 days and one patient had surgery originally planned for the following day that was postponed for 5 days after presentation due to concurrent pneumonia. In both cases operated beyond 1 day, the fovea was documented to be detached at the clinical review the following day after presentation.

Visual outcomes

Most patients (8 out of 10) retained or improved their preoperative visual acuity. Only two patients lost a line of acuity on the Snellen chart. No assessment of metamorphopsia was undertaken. The final visual outcomes and lens status for all patients are shown in table 2 together with demographics and data described above.

Table 2

Patient characteristics and outcomes


In this study, progression from fovea on to fovea off was detected in 10 patients or 1.1% of all patients presenting to the service with their fovea on. In one previous study of only 199 patients, only 1 patient progressed from fovea on to fovea off (0.5%), and in another study 13% of patients showed some progression of subretinal fluid (SRF) in RRDs at a rate of 1.8 disc diameters per day.2 ,5 In the present study, all patients who progressed from fovea-on to fovea-off RRD had superior detachments with 9 out of 10 having superotemporal detachments with detached superotemporal retinal breaks. Nine out of 10 patients had an extension of the SRF to the vascular arcades superotemporally at presentation. The rate of detached breaks in the superotemporal quadrant was significantly higher than the rate for other retinal detachments from our unit as previously described.4 The increased frequency of this feature in patients in this study suggests that it is a relative risk factor for progression to fovea off especially if the retina is already elevated close to the vascular arcades at presentation.

Despite detachment of the fovea, all patients had good visual outcomes with only two patients losing a line of visual acuity on the Snellen chart when compared with their presenting visual acuity, suggesting that at least for visual acuity any loss of function was insignificant. There was, however, no assessment of metamorphopsia or aniseikonia that could potentially affect such patients.6 From the findings of this study, the authors doubt if there is any detrimental effect on the vision from the conversion of a fovea-on to fovea-off retinal detachment of such short duration. The authors have previously shown in autofluorescence studies that it is possible to convert a fovea-on retinal detachment to fovea-off retinal detachment during surgery (or in the postoperative period) as evidenced by shift of the retina postoperatively.7 ,8 This suggests that surgeons inadvertently create progression of fovea-on to fovea-off RRD during vitrectomy. In addition, in a large cohort study the authors have previously demonstrated that the difference in visual acuity outcomes is very small between fovea-on and fovea-off retinal detachment of ≤3 days.3

The patients were all operated on promptly after presentation to the vitreoretinal service. All patients in this series had surgery within 1 day of presentation to the department apart from one patient who was operated at 2 days and one who was too unwell to have surgery. With both these cases, the fovea became detached the day after the patients presented when they were re-examined. The progression rate may be higher in services in which there is further delay to surgery.

It is of interest both from a health economics and medicolegal perspective whether services should be adapted to offer same-day surgery for fovea-on RRDs to prevent the fovea from detaching. Our data suggest that, if this was the case in our unit, it would have only prevented foveal detachment in 4 out of the 10 cases (reducing the rate from 1.1% to 0.6%). Furthermore, all our patients had good visual outcomes, which are comparable to the outcomes following fovea-on retinal detachment repair.3 All these patients had a detached fovea for a very short time, and this may offer a favourable outcome compared with fovea-off detachments of a longer duration.3 ,9 However, there may be some visual morbidity in these patients, which has not been measured using visual acuity alone.

Although beyond the scope of this study, the additional resources required to implement same-day surgery for all fovea-on retinal detachments may be difficult to justify. Stratifying risk however and offering same-day surgery to high-risk patients as previously discussed may be cost-effective. Alternatively, strict posturing of patients has been shown to reduce the volume of SRF in patients with superior RRD and could be used in high-risk cases to try to reduce the risk of extension under the fovea.10 ,11

This study has some limitations. Due to the retrospective design, we may not have identified every patient with conversion of fovea on to off, because not all clinical notes were available for review or because documentation was inadequate. Also, cases may have been missed by using the cut-off of 3 days of visual loss as an inclusion criterion. Therefore, the rate of conversion we present may be slightly underestimated. As already discussed, there is no assessment of the quality of vision in these patients apart from Snellen visual acuity, which does not allow thorough evaluation of the degree of disability these patients may experience.

In conclusion, offering same-day surgery to high-risk fovea-on RRDs would be likely to only reduce the rate of progression to fovea off from 1.1% to 0.6%; however, the benefit in terms of visual outcome for patients is likely to be negligible. It would appear that superotemporal retinal detachments extending to near the retinal arcades are most at risk and might warrant early surgery or posturing to limit spread of the retinal detachment in the preoperative period.



  • Contributors AK designed, collected and analysed the data and prepared the manuscript. THW designed the study, analysed the data and reviewed and edited the manuscript.

  • Competing interests THW: Bausch and Lomb Insights Panel, Software development for Axsys Technology, Glasgow, UK, and a visiting lecturer for Alcon, UK.

  • Ethics approval Guy's and St Thomas’ NHS Foundation Trust research and development department.

  • Provenance and peer review Not commissioned; externally peer reviewed.