From the Department of Ophthalmology (ML, AN, JMS), University of California–San Francisco; San Francisco Veterans Affairs Medical Center (AN); and San Francisco General Hospital (JMS), San Francisco, California.
The authors have no financial or proprietary interest in the materials presented herein.
The authors thank Kathryn Ray and Travis Porco of the Francis I. Proctor Foundation, University of California–San Francisco for providing statistical assistance.
Supported by That Man May See, Inc., San Francisco, California, and an unrestricted grant from Research to Prevent Blindness, Inc., New York, New York.
Address correspondence to Jay M. Stewart, MD, Department of Ophthalmology, University of California–San Francisco, 10 Koret Way, Room K301, Box 0730, San Francisco, CA 94143-0730.
Rhegmatogenous retinal detachments occur when fluid from the vitreous compartment enters the subretinal space through breaks in the retina. Such retinal tears are often caused by posterior vitreous detachments and tend to produce symptoms of dark floaters or flashing lights.1 High myopia, trauma, peripheral lattice degeneration, and pseudophakia or aphakia all predispose patients to tractional retinal breaks.2,3 Prophylactic retinopexies, usually accomplished with argon laser photocoagulation, surround these precursor lesions and associated subretinal fluid with rows of contiguous thermal adhesions to prevent subsequent rhegmatogenous retinal detachments. However, rhegmatogenous retinal detachments occasionally occur despite prophylactic treatment by experienced ophthalmologists, with reported incidences ranging from 3% to 6% at late follow-up (mean: 5 to 10 years).4–7
Although ophthalmology residents also routinely perform retinopexies, it remains unclear whether residents are adequately trained to perform these procedures effectively. One recent study examined the success rate of resident-performed retinopexies of 100 predominantly symptomatic patients presenting to the emergency department. A high proportion of these patients (24%) required further treatment at short follow-up (mean: 7 weeks), including eight patients who required cryotherapy and scleral buckling and three patients who progressed to frank rhegmatogenous retinal detachment.8 Although this study raises questions about the current practice of resident-performed retinopexy, further investigations are warranted to assess its suitability in the clinic setting. Our study aimed at determining the effectiveness of retinopexies performed by residents at our institution in preventing rhegmatogenous retinal detachment. We evaluated the rate of additional prophylactic treatment as a secondary measure of procedure adequacy.
Patients and Methods
Permission was first obtained from the Committee of Human Research at the University of California–San Francisco and from the Institutional Review Board of San Francisco General Hospital and the Research & Development Committee at the San Francisco Veterans Affairs Medical Center. We included all patients who had undergone prophylactic thermal retinopexy by an ophthalmology resident between January 1997 and December 2006 at either San Francisco General Hospital or the San Francisco Veterans Affairs Medical Center. Although all patients receiving retinopexy were instructed to return for repeat examination, only patients who later visited the same ophthalmology clinic remained eligible (see Discussion), with last follow-up taken as the final dilated fundus examination of the peripheral retina by indirect binocular ophthalmoscopy.
Exclusion criteria were previous vitreoretinal surgery in the same eye, senior/attending physician participating directly in the procedure, advanced diabetic retinopathy requiring panretinal photocoagulation, no documented follow-up, and inadequate charting. Patients were identified from laser log books and by using billing codes designating prophylactic retinopexy procedures and diagnoses of retinal breaks or degenerations. We also reviewed surgeries coded as retinal detachments in an effort to identify any cases of failed primary retinopexy.
Risk Factors and Lesion Characteristics
Posterior vitreous detachment was defined as a detached posterior hyaloid with or without associated pre-papillary glial ring or tissue. Designations of a tear associated with subretinal fluid accumulation versus limited rhegmatogenous retinal detachment versus rhegmatogenous retinal detachment were made based exclusively on the documented clinical assessment. Myopia was considered to be greater than -5 diopters of refractive error determined by manifest or autorefraction at the time of presentation. Eyes with multiple retinal breaks treated simultaneously were categorized for analysis based on the lesion carrying the greatest risk for subsequent rhegmatogenous retinal detachment (horseshoe tear > operculated hole > atrophic hole), although all lesions were included in the analysis.
The peripheral retinas of all patients were examined with 360° of scleral depression to identify all lesions needing treatment. Procedures were generally performed either in consultation with or under the guidance of an attending physician. In most cases, the argon laser barrier was made using indirect ophthalmoscopy as several contiguous rows 360° around the lesion at its junction with the unaffected retina, delivered with a laser intensity of 100 to 800 mW for a duration of 0.1 to 0.2 second per spot. In other cases, argon laser photocoagulation was accomplished using contact lens condensing systems. Cryopexy was occasionally necessary when it was deemed that extensive subretinal fluid would prevent laser burns from taking. Because the majority of laser procedures were accomplished using indirect ophthalmoscopy, cases with equivocal records were assumed to have been performed by this methodology.
All statistics are expressed with indicated units as mean ± standard error of the mean unless otherwise indicated, with analyses performed using Intercooled Stata software (StataCorp LP, College Station, TX). Mean ages and follow-up lengths at the San Francisco Veterans Affairs Medical Center and San Francisco General Hospital were compared using the two-sided Student’s t test, whereas all other characteristics were compared using the two-sided Fisher’s exact test. The confidence interval for incidence of retinal detachment was estimated using a Poisson-distributed variable.
Baseline characteristics of the patient population treated for retinal breaks and ruptures at San Francisco General Hospital and the San Francisco Veterans Affairs Medical Center are detailed in Table 1. One-hundred sixty-six eyes from 157 subjects (mean age: 59.8 ± 1.0 years) met the inclusion criteria during the 10-year study period. These eyes were observed with repeat examinations following retinopexy for an average of 2.4 ± 0.2 years; 135 were observed for more than 3 months after treatment and 96 for at least 1 year. Approximately 50 additional eyes were excluded for reasons listed in the Methods section. No additional cases of prophylactic retinopexy were identified on examining more than 100 consecutive surgical cases coded for retinal detachment. Table 2 summarizes best-corrected visual acuities of treated eyes at presentation and at last follow-up. Less than half of the patients were symptomatic, presenting with floaters, flashes, shadows, or subjective decreases in vision. Treated lesions were variably associated with risk factors for subsequent rhegmatogenous retinal detachment, including peripheral lattice degeneration (24.1%), myopia (22.3%), a history of orbital trauma (16.9%), a previous rhegmatogenous retinal detachment in the fellow eye (7.8%), pseudophakia (7.2%), and a family history of rhegmatogenous retinal detachment (2.4%). Comparing the experiences at both institutions, veteran patients were significantly older (64.5 ± 1.1 vs 54.3 ± 1.4 years; P < .0001) and had longer follow-up (3.0 ± 0.3 vs 1.6 ± 0.2 years; P = .0002) than San Francisco General Hospital patients. Among other relevant baseline characteristics, only pseudophakia was found to differ in prevalence between hospitals (n = 11 vs 1 comparing San Francisco Veterans Affairs Medical Center to San Francisco General Hospital; P = .006).
Table 1: Patient Characteristics
Table 2: Best-Corrected Visual Acuities
Lesion characteristics, pooled from both institutions, are summarized in Table 3. In total, 69.9% of cases involved solitary lesions. Although the majority of retinas were treated for horseshoe tears (n = 105), operculated holes (n = 21), and atrophic holes (n = 18), a sizable fraction of barrier treatments were for limited rhegmatogenous retinal detachment, retinoschisis, retinal dialysis, or lattice degeneration without discrete breaks (n = 22 combined). The type of lesion was not strongly predicted by clock hour location in the peripheral retina. On dilated fundus examination, posterior vitreous detachments and localized accumulations of subretinal fluid were noted in 40.4% and 34.3% of cases, respectively. In contrast, 23% of the asymptomatic breaks that were treated (n = 22) lacked any recognized risk factors in the patient history or on examination; none of these lesions were noted to have pigmented retinal changes at their borders.
Table 3: Lesion Characteristics
As outlined in Table 4, treatments consisted of circumferential argon laser photocoagulation by indirect or slit-lamp–delivered argon laser, cryopexy, or a combination of techniques. Table 5 presents the major long-term retinal complications in treated eyes. Retinal detachments requiring vitreoretinal surgery occurred at a rate of 0.50 per 100 person years of follow-up (95% confidence interval of 0.061 to 1.8), developing in two eyes (1.2%). When removing from analysis the 22 cases of retinal dialysis, retinoschisis, limited rhegmatogenous retinal detachment, and lattice degeneration without discrete breaks, retinal detachments occurred 1.4% of the time (see Discussion). New epiretinal membranes also developed in the maculas of 9 eyes (5.4%) during the follow-up period, including in 1 of the 2 eyes that progressed to rhegmatogenous retinal detachment. Finally, 15% of patients were retreated with at least one subsequent retinopexy procedure, which was also performed by a resident. The average time at first retreatment, the reasons for retreatment, and the method of retreatment are all detailed in Table 6.
Table 4: Retinopexy Methodology
Table 5: Complications Following Retinopexy
Table 6: Retreatment Rates and Methodology
This study examined the effectiveness of resident-performed prophylactic treatment of retinal breaks in preventing progression to rhegmatogenous retinal detachment. Rhegmatogenous retinal detachments developed in only 2 of 166 treated eyes during follow-up, with a low incidence of 0.50 per 100 person years of follow-up. One of these complications, a limited rhegmatogenous retinal detachment, arose in a patient with lattice holes who remained asymptomatic from initial presentation through last follow-up. Of note, we did not detect a single instance of rhegmatogenous retinal detachment arising from a treated tear; both rhegmatogenous retinal detachments were attributed to new tears at sites remote from the initial treatment. Indeed, new retinal tears have been implicated in most retinal detachments following prophylactic treatment.2,9 This demonstrated that success of resident-performed prophylactic therapy should enable healthcare providers at teaching institutions to provide more reassurance to patients when discussing risks and benefits of retinopexy performed by a resident.
Our study also highlights other important considerations in the management of retinal tears. First, we found that both rhegmatogenous retinal detachments that evolved during follow-up (after 0.5 and 2.1 years, both at San Francisco General Hospital) involved a patient with prior rhegmatogenous retinal detachment in the other eye. Both fellow eye detachments were treated successfully during follow-up with cryotherapy and scleral buckling. Reported incidences of fellow eye rhegmatogenous retinal detachments have ranged from 4% to 14%.10 Although prophylaxis in the fellow eye has been shown to dramatically reduce this rate,11 these patients remain at heightened risk.9 Also of note, Gratton et al. demonstrated laser adhesions to be a successful treatment alternative to scleral buckling for limited rhegmatogenous retinal detachments,12 a conclusion supported by the successful laser treatments of the 6 cases of limited rhegmatogenous retinal detachment in our population.
The rate of retreatment was studied as a secondary measure of failed prophylactic treatment. Our rate of 15% compares favorably to that of 24% reported by Ghosh et al.,8 especially given the substantially longer mean follow-up period and the inclusion only of cases followed up with repeat examinations in our population. Here, the primary lesion was actually only retreated due to incomplete barrier in 9% of cases. Ghosh et al. argue that lack of familiarity with indirect argon laser photocoagulation among trainees, and their resulting reliance on slit-lamp methods, limited the ability to adequately surround tears at their most anterior edge. Among the 15 eyes in our population requiring additional procedures based on inadequate barrier coverage, the method of delivery did not predict the likelihood of retreatment. Differences in rates among institutions may be due in part to the subjective threshold for retreatment. The high retreatment rate reported in the previous study may reflect a higher risk population, consisting almost exclusively of patients with acute and symptomatic retinal tears.
A low proportion of our treated population was symptomatic (43.4%) because many of the lesions were identified incidentally at optometry or general ophthalmology appointments, prompting referral to resident-staffed clinics. Although symptomatic tears are generally treated, evidence is lacking to guide the approach to asymptomatic lesions13,14; only fellow eye detachments clearly predispose such breaks to progression.15 The presence of additional risk factors often directs treatment, whereas old pigmentation at the lesion border is thought to be protective against progression.
Retinoschisis and lattice degeneration in the absence of discrete breaks carry a particularly low risk of progression to detachment. Excluding the 11 representative cases from our cohort would not have changed the study conclusions (see Results), but it should be noted that 4 of 5 cases of retinoschisis and 1 of 6 cases of lattice degeneration lacked symptoms and any identifiable risk factors for rhegmatogenous retinal detachment.
In 9 instances, new epiretinal membranes were identified following laser retinopexy, highlighting a potential long-term complication of intraocular laser procedures.16,17 Retinal pigment epithelial migration through retinal breaks, tears, and holes promotes epiretinal membrane formation. As such, epiretinal membrane formation cannot be attributed exclusively to laser treatment in these cases. Indeed, the applied laser energies preceding epiretinal membrane appearance were on average not higher than those used in the remainder of our cohort, in contrast to what was reported by Mester et al.16 The only factor predictive of subsequent epiretinal membrane formation in our study was the presence of posterior vitreous detachment at presentation (seen in 8 of 9 patients).
It is incumbent on the ophthalmology resident to proceed judiciously with treatments, weighing risk factors for rhegmatogenous retinal detachment against potential harmful effects of retinopexy. Besides contributing to epiretinal membrane formation, laser treatment lesions might actually precipitate new remote breaks by accelerating progression of partial posterior vitreous detachments.10 In our cohort, asymptomatic patients without clear risk factors were often treated based on the suboptimal follow-up anticipated in these treatment settings.
Aggressive treatment practices and the overall low rates of progression of asymptomatic tears, retinoschisis, and lattice degeneration with tears might contribute to our low measured rate of rhegmatogenous retinal detachment. Frequent losses after only short follow-up periods limited the ability to detect rhegmatogenous retinal detachments in many eyes. Furthermore, the large number of patients excluded due to lack of follow-up may also have introduced some bias, underestimating the incidence of rhegmatogenous retinal detachment. However, although patients in both situations could have been subsequently treated for symptomatic rhegmatogenous retinal detachments at other hospitals, many of these patients, especially those at San Francisco General Hospital, lacked the financial resources to access care elsewhere. The retrospective study design and heterogeneous mix of patients represent additional limitations. Although no absolute standard exists to judge the performance of residents in referral-based clinics, our results provide evidence supporting the standard practice of resident-performed prophylactic retinal coagulation in hospital-based referral clinics.
- van Overdam KA, Bettink-Remijer MW, Klaver CC, Mulder PG, Moll AC, van Meurs JC. Symptoms and findings predictive for the development of new retinal breaks. Arch Ophthalmol. 2005;123:479–484. doi:10.1001/archopht.123.4.479 [CrossRef]
- Byer NE. Rethinking prophylactic therapy of retinal detachment. In: Stripe M, ed. Advances in Vitreoretinal Surgery. New York: Ophthalmic Communications Society; 1992:399–411.
- The Eye Disease Case-Control Study Group. Risk factors for idiopathic rhegmatogenous retinal detachment. Am J Epidemiol. 1993;137:749–757.
- Kovacevic D, Loncarek K. Long-term results of argon laser retinal photocoagulation for retinal ruptures [article in Croatian]. Acta Med Croatica. 2006;60:149–152.
- Manys-Kubacka K, Krause A, Finke S, Switek-Tyma B, Kociecki J. Results of prevention of retinal detachment by using photocoagulation. Klin Oczna. 1991;93:315–316.
- Schroeder W, Baden H. Retinal detachment despite preventive coagulation [article in German]. Ophthalmologe. 1996;93:144–148.
- Pollack A, Milstein A, Oliver M, Zalish M. Circumferential argon laser photocoagulation for prevention of retinal detachment. Eye. 1994;8:419–422.
- Ghosh YK, Banerjee S, Tyagi AK. Effectiveness of emergency argon laser retinopexy performed by trainee doctors. Eye. 2005;19:52–54. doi:10.1038/sj.eye.6701416 [CrossRef]
- Chauhan DS, Downie JA, Eckstein M, Aylward GW. Failure of prophylactic retinopexy in fellow eyes without a posterior vitreous detachment. Arch Ophthalmol. 2006;124:968–971. doi:10.1001/archopht.124.7.968 [CrossRef]
- Mastropasqua L, Carpineto P, Ciancaglini M, Falconio G, Gallenga PE. Treatment of retinal tears and lattice degenerations in fellow eyes in high risk patients suffering retinal detachment: a prospective study. Br J Ophthalmol. 1999;83:1046–1049. doi:10.1136/bjo.83.9.1046 [CrossRef]
- Avitabile T, Bonfiglio V, Reibaldi M, Torrisi B, Reibaldi A. Prophylactic treatment of the fellow eye of patients with retinal detachment: a retrospective study. Graefes Arch Clin Exp Ophthalmol. 2004;242:191–196. doi:10.1007/s00417-003-0783-9 [CrossRef]
- Gratton I, Gazocchi M, Simonini F, Fattori CM, Citroni M. Argon laser photocoagulation in the management of retinal detachments and predisposing lesions. Lasers Surg Med. 1984;4:337–344. doi:10.1002/lsm.1900040406 [CrossRef]
- Kreis AJ, Aylward GW, Wolfensberger TJ. Prophylaxis for retinal detachment: evidence or eminence based?Retina. 2007;27:468–472. doi:10.1097/01.iae.0000243069.11708.f6 [CrossRef]
- Wilkinson C. Interventions for asymptomatic retinal breaks and lattice degeneration for preventing retinal detachment. Cochrane Database Syst Rev. 2005;1: CD003170.
- Byer NE. What happens to untreated asymptomatic retinal breaks, and are they affected by posterior vitreous detachment?Ophthalmology. 1998;105:1045–1049. doi:10.1016/S0161-6420(98)96006-7 [CrossRef]
- Mester U, Volker B, Kroll P, Berg P. Complications of prophylactic argon laser treatment of retinal breaks and degenerations in 2,000 eyes. Ophthalmic Surg. 1988;19:482–484.
- Saran BR, Brucker AJ. Macular epiretinal membrane formation and treated retinal breaks. Am J Ophthalmol. 1995;120:480–485. Erratum in: Am J Ophthalmol. 1996;121:334.
|No. of eyes||77||89||166|
|Right eye/left eye||36/41||47/42||83/83|
|Mean patient age ± SD (years)b||54.3 ± 1.4||64.5 ± 1.1||59.8 ± 1.0|
|Follow-up length ± SD (years)b||1.6 ± 0.2||3.0 ± 0.3||2.4 ± 0.2|
|Lattice degeneration||19 (24.7%)||21 (23.6%)||40 (24.1%)|
|Myopia (> −5 diopters)||21 (27.3%)||16 (18.0%)||37 (22.3%)|
|History of orbital trauma||13 (16.9%)||15 (16.9%)||28 (16.9%)|
|Previous RRD in fellow eye||7 (9.1%)||6 (6.7%)||13 (7.8%)|
|Pseudophakiab||1 (1.3%)||11 (12.4%)||12 (7.2%)|
|Family history of RRD||3 (3.9%)||1 (1.1%)||4 (2.4%)|
|Symptomatic||40 (51.9%)||32 (36.0%)||72 (43.4%)|
| Floaters||36 (46.8%)||26 (29.2%)||62 (37.3%)|
| Flashes||14 (18.2%)||14 (15.7%)||28 (16.9%)|
| Decreased visual acuity||7 (9.1%)||4 (4.5%)||11 (6.6%)|
| Shadows||7 (9.1%)||3 (3.4%)||10 (6.0%)|
|Asymptomatic, risk factors||29 (37.7%)||43 (48.3%)||72 (43.4%)|
|Asymptomatic, no risk factors||8 (10.4%)||14 (15.7%)||22 (13.2%)|
Best-Corrected Visual Acuitiesa
|Visual Acuity||Presentation||Last Follow-up|
|≥20/40||131 (78.9%)||138 (83.1%)|
|20/50 to 20/200||25 (15.1%)||19 (11.4%)|
|20/300 to LP||5 (3.0%)||7 (4.2%)|
|NLP||0 (0%)||0 (0%)|
|Unknown||5 (3.0%)||3 (1.8%)|
|Single lesion||116 (69.9%)|
|Most serious lesion treated|
| Horseshoe tear||105 (63.3%)|
| Operculated hole||21 (12.7%)|
| Atrophic hole||18 (10.8%)|
| Limited RRD||6 (3.6%)|
| Retinoschisis||5 (3.0%)|
| Dialysis||5 (3.0%)|
| Lattice degeneration||6 (3.6%)|
|Posterior vitreous detachment||67 (40.4%)|
|Subretinal fluid||57 (34.3%)|
|Pigmented borders||39 (23.5%)|
|Vitreous or subretinal hemorrhage||20 (12.0%)|
| Inferior||16 (9.1%)|
| Inferonasal||7 (4.0%)|
| Inferotemporal||43 (24.6%)|
| Superior||25 (14.3%)|
| Superonasal||16 (9.1%)|
| Superotemporal||46 (26.3%)|
| Nasal||4 (2.3%)|
| Temporal||18 (10.3%)|
|Indirect laser delivery||136 (81.9%)|
|Slit-lamp laser delivery||23 (13.9%)|
|Slit-lamp + indirect laser delivery||3 (1.8%)|
|Slit-lamp laser delivery + cryopexy||1 (0.6%)|
Complications Following Retinopexya
|Retinal detachment||2 (1.2%)b|
|New epiretinal membrane||9 (5.4%)|
Retreatment Rates and Methodologya
|Repeat retinopexy procedure||25 (15.0%)|
|Reason for retreatment|
| Inadequate barrier||15 (9.0%)|
| New and remote lesion||10 (6.0%)|
|Weeks before retreatment ± standard error of the mean||17.0 ± 4.3|
| Indirect laser||22 (84.6%)|
| Slit-lamp laser||3 (11.5%)|
| Cryopexy||1 (3.8%)|