To study the effect of the duration from initial evaluation to repair on outcomes for fovea-sparing rhegmatogenous retinal detachment (RRD).
Retrospective, single-surgeon, consecutive case series.
Medical records were reviewed for preoperative and intraoperative factors possibly associated with visual and anatomic outcomes for all patients undergoing scleral buckling procedure (SBP) for fovea-sparing, primary RRD between 1989 and 2004.
Fifty-five percent of 199 patients had symptoms for ≤ 7 days, 83% had best-corrected visual acuity (BCVA) ≥ 20/40, and 33% had a RRD that had extended to within the macular arcade vessels. Eighty-five percent were operated within 3 days, including 56% within 24 hours. One case progressed to fovea-off status before surgery 4 days after initial evaluation (0.5%). The single-operation success rate was 88% and final anatomic success was 99.5% (1 patient refused reoperation). Eighty-six percent were examined postoperatively for at least 2 months; 73% had ≥ 20/40 vision. The strongest predictor of postoperative BCVA was initial BCVA (r = 0.47; P < .001). There was no statistically significant difference in postoperative BCVA or single-operation success rate at any point within 3 days of initial examination. No statistically significant correlation was found between postoperative BCVA and duration of symptoms, RRD location, direction of the closest approach of the RRD to the fovea, or need for reoperation.
Progression to fovea-off status was rare in this series when a selectively urgent, but not strictly emergent, surgical approach was employed for fovea-sparing RRD.
Rhegmatogenous Retinal Detachment (RRD) is a serious ocular disorder that may result in severe visual loss. Many treatment options have been established, but the scleral buckling procedure (SBP) has been a standard treatment for RRD for over 50 years. While many preoperative and intraoperative prognostic factors have been studied, the strongest and most consistent predictors of visual outcome have been preoperative visual acuity and foveal detachment. The rationale for prompt surgery in eyes with a fovea-sparing RRD is to prevent foveal detachment. Many consider a fovea-sparing RRD of recent onset to be a surgical emergency, but few studies have evaluated this presumption.
The current study investigates a series of primary, fovea-sparing RRD managed with SBP and assesses visual and anatomic outcomes, as well as progression and complications. The effect of the duration of time from initial patient evaluation to surgical repair was the primary focus of the study.
Patients and Methods
The study design was a retrospective, nonrandomized, consecutive case series. The medical records of all patients who underwent primary SBP (without vitrectomy) for primary fovea-sparing RRD performed by a single surgeon (W.E.S.) at Bascom Palmer Eye Institute between July 1989 and April 2004 were reviewed. The surgeon’s treatment algorithm was to attempt scleral buckling alone in primary cases, reserving vitrectomy techniques for exceptional cases with more advanced media opacities, epiretinal membranes, or large retinal breaks. The general clinical bias was to operate more quickly for RRD that had extended more posteriorly, for RRD that were superior or temporal in location, and for patients with a shorter duration of symptoms.
Preoperative data collected included patient age, gender, lens status, time and date of first documentation of RRD, duration and quality of symptoms, refractive error, previous eye surgery, other eye diseases, location and extent of RRD, direction of the closest approach of the RRD to the fovea, and best-corrected visual acuity (BCVA). The time of initial documentation of RRD (time of initial patient evaluation) was most commonly determined as the time that the intraocular pressure measurement was recorded at the initial clinical evaluation. Since all patients in this study had fovea-sparing RRD, the preoperative BCVA was usually excellent; if it was < 20/40, the reason was recorded. Patients with giant retinal tears, failed prior pneumatic retinopexy, retinal dialyses, or severe, acute trauma with other concurrent ocular injury were excluded. Intraoperative data collected included date and time of surgical repair, location and number of retinal breaks, drainage of subretinal fluid, gas injection, scleral buckle configuration, and complications. The extent of any progression in the distribution of the RRD was not systematically measured except to verify whether the macula had become detached since the preoperative examination had been performed.
Postoperative data collected included reoperations; 2-month, 6-month, and 12-month BCVA; retinal attachment status at postoperative visits; and additional surgical management. Patients referred for treatment of acute-onset, primary RRD were commonly returned to their comprehensive ophthalmologist or home locality as soon as possible. Hence, there was a follow-up period < 2 months for 27 eyes (14%); for many others the final follow-up examination took place at about 2 months postoperatively and clinical experience is that the risk of recurrent RRD after 2 months is low.
All eyes underwent SBP using standard surgical techniques. Cryotherapy was used in all patients to treat retinal breaks. Encircling silicone bands, circumferentially oriented silicone tires, or radially oriented sponges were used in different combinations; generally, all pseudophakic or aphakic eyes were encircled, solitary tears occurring posterior to the equator were commonly supported on radial sponges, and broader circumferential elements were used if multiple tears at or posterior to the equator were present or if other tractional pathology was detected. Drainage of subretinal fluid was commonly performed in the following situations: when a large-volume exoplant was used; when highly bullous detachments jeopardized accurate localization of break(s); or when there were multiple retinal breaks, inferior breaks, chronic retinal detachments, or uncertain break location. Normal intraocular pressure after drainage was usually restored with air; sulfur hexafluoride or perfluoropropane were more commonly used if the breaks were posterior or were highly elevated over the buckled retinal pigment epithelium, especially in nondrainage cases. No special decision making or techniques were invoked because of proximity of the detached retina to the fovea or the timing of the procedure.
Descriptive statistics are presented for demographic and ocular characteristics. Categorical variables were analyzed using the χ 2 test or the Fisher exact test and continuous variables were analyzed using the Student t test. Kaplan-Meier survival analysis was performed to compute the rate of reoperation. Analysis of variance (ANOVA) methods were used to study the differences in visual acuity outcomes between groups of patients based on the duration from initial evaluation to surgical repair. Visual acuity was treated as a continuous variable after applying a logarithm of the minimal angle of resolution (logMAR) transformation for the ANOVA. All statistical analyses were performed using SPSS version 17 (SPSS, Chicago, Illinois, USA).
Primary, fovea-sparing RRDs were diagnosed and treated by scleral buckling without vitrectomy in 199 patients ( Table 1 ). The mean age was 54 years and 111 patients (56%) were male. There were 109 eyes (55%) that had undergone prior surgical or laser interventions, including cataract surgery in 77 eyes (39%), yttrium-aluminum-garnet (YAG) capsulotomy in 12 eyes (16% of pseudophakic eyes), refractive surgery in 6 eyes (3%), glaucoma surgery in 4 eyes (2%), retinal tears treated with cryopexy or laser in 34 eyes (17%), and other procedures in 14 eyes (7%). Symptoms reported included floaters in 142 eyes (71%), visual field defect in 107 eyes (54%), photopsias in 76 eyes (38%), and loss of vision in 50 eyes (25%). Duration from onset of symptoms to initial evaluation was less than 1 week for 109 eyes (55%) and more than 1 month for 22 eyes (11%).
|Operated eye, right||114 (57%)|
|Age, mean years||54 (SD 15, range 16–87)|
|Gender, male||111 (56%)|
|Follow-up, median months||17 (range 0–190)|
|Previous ocular surgery or laser||109 (55%)|
|Lens status, n (%)|
|Refractive error, n (%)|
|Myopia < −6D||88 (44)|
|Myopia > −6D||42 (21)|
|Hyperopic, plano, or unknown||69 (35)|
|RRD extent, n (%)|
|1 quadrant||108 (55)|
|2 quadrants||72 (36)|
|3 or 4 quadrants||18 (9)|
|RRD posterior extension, n (%)|
|Only anterior to the equator||11 (6)|
|Posterior to the equator but not within the macular arcades||123 (62)|
|Within the macular arcades||65 (33)|
|Direction of closest approach of the RRD to the fovea, n (%)|
Preoperative BCVA was ≥ 20/25 in 108 eyes (54%), ≥ 20/40 in 166 eyes (83%), and ≤ 20/50 in 33 eyes (17%) ( Table 2 ). Median preoperative visual acuity was 20/25 (range 20/15 to hand motions). The clinical explanation for vision < 20/40 at initial evaluation included 1 or more of the following: macular disease in 9 eyes (27%), cataract in 8 eyes (24%), vitreous hemorrhage in 8 eyes (24%), amblyopia in 5 eyes (15%), and corneal opacity in 3 eyes (9%). The reason for initially reduced vision was not identified in 8 eyes (24%). The RRD involved 1 quadrant in 108 eyes (55%), 2 quadrants in 72 eyes (36%), 3 quadrants in 17 eyes (8.5%), and 4 quadrants in 1 eye (0.5%). The posterior extent of the RRD extended within the macular arcade vessels in 65 eyes (33%), were posterior to the equator but not within the macular arcade vessels in 123 eyes (62%), and remained anterior to the equator in 11 eyes (6%). The direction of the closest RRD approach to the fovea was inferior in 59 eyes (30%), superior in 56 eyes (29%), temporal in 50 eyes (25%), and nasal in 31 eyes (16%).
|Preoperative n = 199||Postoperative n = 172||Postoperative Vision According to Time From Initial Evaluation to Surgery a|
|All Patients||All Patients||< 12 h (n = 46)||12–24 h (n = 47)||1–3 d (n = 53)||> 3 d (n = 26)|
|≥ 20/25||108 (54%)||77 (45%)||24 (52%)||20 (43%)||22 (42%)||11 (42%)|
|20/30-20/40||58 (29%)||48 (28%)||12 (26%)||17 (36%)||15 (28%)||4 (15%)|
|≤ 20/50||33 (17%)||47 (27%)||10 (22%)||10 (21%)||16 (30%)||11 (42%)|
a There was no statistical difference among postoperative visual acuity outcomes by time from initial evaluation to surgery. P = .20 (by median visual acuity, Kruskal-Wallis test); P = .33 (by visual acuity groupings, χ 2 test).
Surgery was performed on a weekend or holiday for 12 patients (6%) and during the week for the remaining 187 patients (94%). Time from initial evaluation to the operating room was recorded and available for 187 eyes (94%); it was less than 24 hours in 104 eyes (56%) and less than 72 hours in 159 eyes (85%) ( Figure 1 ). Three clinical variables were statistically correlated with a shorter interval from initial patient evaluation to surgery ( Table 3 ): a shorter duration of symptoms ( P < .001); patients with a RRD whose closest approach to the fovea was superior or temporal were operated sooner (median 21.0 hours; n = 106), compared to patients with a RRD whose closest point to the fovea was inferior or nasal (median 29 hours; n = 90) ( P = .02); patients with superior retinal breaks (9-o’clock to 3-o’clock position) were operated sooner (median 23.0 hours; n = 130), compared to patients with inferior retinal breaks or no identified retinal breaks (median 40 hours; n = 69) ( P = .009). There was no statistically significant correlation of the duration from initial evaluation to surgery and the number of quadrants involved ( P = .49) or posterior extent of the RRD ( P = .22).
|Variable||Time to Surgery a|
|Symptom duration||P < .001|
|Direction closest to fovea b|
|Superior or temporal||21 (n = 106)|
|Inferior or nasal||29 (n = 90)|
|Location of retinal breaks c|
|Superior (9 to 3 o’clock)||23 (n = 130)|
|Inferior or none identified||40 (n = 69)|
Patients initially evaluated on Friday or Saturday (n = 49, 25%) had a longer time to surgery (median 68 hours) than for other days (median 22 hours; P < .001) ( Table 4 ). Furthermore, patients initially evaluated on Friday (n = 38, 19%) had the longest median time to surgery, 71.4 hours (mean 70 hours, SD 58 hours). Of the Friday patients, however, there was a bimodal distribution, with certain patients operated within 24 hours (n = 10, 26%) and others operated 3 days later (n = 22, 58%). A small fraction were operated beyond 3 days (n = 6, 16%). In particular, Friday patients with a superior or temporal RRD (n = 15) were operated sooner if their duration of symptoms was ≤ 3 days (median 6 hours, range 2 to 68 hours; n = 4) versus 4 to 7 days (median 69 hours, range 4 to 125 hours; n = 6) or > 30 days (median 110 hours, range 75 to 148 hours; n = 4; P = .03). In contrast, time to surgery for patients with superior or temporal RRD initially evaluated on other days was not related to duration of symptoms.
|Initial Evaluation||Number (%) of Patients||Time to Surgery, Median Hours a||Final VA (n) b|
|Friday or Saturday||49 (25%)||68||20/30 (45)|
|Sunday-Thursday||150 (75%)||22||20/30 (127)|