The Drainage of Subretinal Fluid: a Randomized Controlled Clinical Trial

Home , Drain (surgery), Von Graefe knife

THE DRAINAGE OF SUBRETINAL FLUID: A RANDOMIZED CONTROLLED CLINICAL TRIAL

BY George F. Hilton, MD

INTRODUCTION AMONG THE MANY CONTROVERSIES IN RETINAL DETACHMENT SURGERY, NONE HAS been more persistent than the unresolved question of drainage versus nondrainage. The controversy has persisted for over 20 years, and more than 60 papers have been written on the subject; however, to date there have been no controlled studies. The ongoing interest in this problem is illustrated by a recent Jules Gonin Club symposium on the drainage of subretinal fluid. After numer- ous papers on the subject the final summary acknowledged: "We are still challenged by the question: To drain or not to drain.'"1 The interest in this question is enhanced by the fact that most retinal surgeons regard the procedure ofsubretinal fluid drainage as a potentially hazardous step. Martin2 has defined it as "the most dangerous part of a retinal detachment operation." Ferguson3 referred to it as "the most crucial point" in the operation, and Norton4 observed that "fluid drainage is the one aspect ofthe surgical procedure over which the surgeon has the least control and although complications are unusual, they can be disastrous." Not all authors are equally impressed with the potential complications ofdrainage. Chawla5 regarded this surgical step as "only one danger among equals," and Schepens6 wrote that "the rate of complications from correctly performed perforation for the release of subretinal fluid is less than 1%." This question was recently brought into focus by a pair of papers representing the two major schools of thought on the issue. Schepens,6 a major spokesman for the drainage school, stated that all retinal detachments should be managed with fluid drainage "except a few selected cases." Lincoff,7 a major spokesman for the nondrainage school, contend- ed that nearly nine out of ten detachments could be adequately repaired without surgical drainage of subretinal fluid.

TR. AM. OPHTH. Soc. vol. LXXIX, 1981 518 Hilton The present study was undertaken with the hypothesis that equally good results, both anatomic and visual, could be obtained with either technique.

HISTORICAL REVIEW The question ofsubretinal fluid drainage can best be appreciated against a background of the historical development of the entire field of retinal detachment surgery. The evolution of this operation is one of the most remarkable chapters in the history of ophthalmology. Gonin's operation for the repair of the detached retina ranks with Daviel's cataract extrac- tion and von Graefe's iridectomy for acute glaucoma as one of three great surgical treatments for blindness. The entity of retinal detachment was first recognized by Saint-Yves (1722)8 who reported the gross pathologic examination of an eye with this condition. The first clinical description did not appear until a report by Beer (1817).9 Apparently this was a very bullous detachment immediately behind the crystalline lens as he was able to make his observation without the benefit of an ophthalmoscope. With the invention of the ophthalmoscope by Helmholtz (1851), 10 a rapid succession of ophthalmoscopic ob- servations of detachment followed. Shortly after, Coccius (1853)" reported the ophthalmoscopic detection of breaks in the retina. von Graefe (1854)12 theorized that retinal detachment was due to a serous effusion from the choroid into the subretinal space. When he observed a retinal break, he assumed that it was secondary and represented an attempt by the eye to cure itself. He supposed that with the development of such a break, the subretinal fluid could pass from the subretinal space into the vitreous cavity resulting in retinal reattachment. With this conviction in mind, he treated detachments with deliberate incision of the retina. Predictably his therapeutic approach was short- lived. Taking a point of view opposite to von Graefe's, de Wecker (1870)13 proposed that the retinal break was the cause ofthe detachment, with the passing of fluid vitreous through the break into the subretinal space. Unfortunately this correct observation was not generally accepted at that time. Leber (1882)14 reported the observation ofretinal breaks in 14 out of27 detachments, and he implicated vitreous traction in the pathogenesis of retinal tears. This accurate observation, however, was later repudiated by Leber. Subretinal Fluid 519 Gonin's15 operation involved the "searing" of the sclera, choroid, and retina at the retinal break site with a red hot probe, a procedure referred to as "ignipuncture." He reported his results initially to the French Ophthalmological Society in 1923 and the Heidelberg Congress in 1925. The great potential ofthese initial reports was generally overlooked by his colleagues. 16 Later, general interest was aroused by his paper presented before the International Congress of Ophthalmology in Amsterdam in 1929 wherein he reported surgical cure of 20 out of 30 cases of recent retinal detachment. However, in detachments of greater than three months duration, he reported a cure in only 15 out of 35. To Gonin goes the credit not only for developing his operation but for promoting and disseminating the new gospel. He did this with 34 papers, published in Belgian, German, English, American, Swiss and French journals. In each he emphasized the essential role of the retinal break and the necessity of closing all breaks with ignipuncture, a process which involved the drainage of subretinal fluid. What had previously been a uniform cause of blindness responded to Gonin's operation in about 60% ofcases. In 1934, he published his classic text "Retinal Detachment."'17 To Weve (1932)18 goes the credit for developing the use ofdiathermy, a marked improvement over the red hot searing technique of Gonin. Lindner (1933)19 introduced the technique of full thickness scleral resection and continued to utilize fluid drainage. Jess (1937)2° gave us the first description of scleral buckling but his brief mention was not followed up and the ophthalmic world had to wait almost two decades for the full development of scleral buckling. Rosengren (1938)21 added the valuable technique of intravitreal air injection for the closure of retinal breaks and this is still used in con- junction with the drainage of subretinal fluid. Schepens (1947)22,23 modified and developed the binocular indirect ophthalmoscope, initially described by Girard-Teulon in 1861. Schepens' instrument is now widely used and this important step forward was further augmented by the supplemental use of scleral depression, a technique originally introduced by Trantas (1926).24 With this technique the periphery of the retina could be brought more clearly into view, and of even greater importance, peripheral lesions could be viewed in dynamic profile as they were gently massaged by the scleral depressor. Both the indirect ophthalmoscope and the scleral depressor are widely used in selecting the site for subretinal fluid drainage. Shapland (1951)25 supplanted the full thickness scleral resection of Lindner with lamellar scleral resection in which only a partial thickness 520 Hilton band of sclera was removed. With the surgical closure of this wound, a moderate scleral buckling was achieved by the infolding of the partially thinned sclera and choroid. He also utilized perforation for the drainage of subretinal fluid. It is to Custodis (1952-1956)26-28 that we are indebted for the re- introduction and development of scleral buckling. He used a polyviol plomb which was sutured to the surface of the globe and thereby created a prominent segmental scleral buckle. Moreover he made the remarkable contribution of demonstrating that many retinal detachments may be cured without the drainage of subretinal fluid. He noted that retinal breaks do not necessarily have to be closed at the time of surgery and that they will usually close during the postoperative period if the buckle has been properly placed. The manner by which the buckle promotes the absorption of subretinal fluid is still not fully understood. The segmental scleral buckle of Custodis was extended and amplified by the development of an encircling scleral buckle by Schepens (1957). 29 He used a polyethylene tube which encircled the eye. Unfortunately this rather rigid encirclement led to the late complication of erosion of the sclera, choroid, and retina in some eyes. Therefore, it was replaced by a softer material, silicone rubber. The technique of subretinal fluid drainage was summarized by Schwartz (1959).30 He identified the two most critical points in retinal detachment surgery as the localization of the retinal breaks and the drainage of subretinal fluid. He described two techniques: perforating Pischel pins, and the Schepens method ofsclerotomy followed by perforation of the choroid with either an electrode or a punctum dilator. Three years later Ferguson3 published a very comprehensive review of the subject of fluid drainage and emphasized the two step technique of sclerotomy followed by perforation of the exposed choroid. The drainage versus nondrainage controversy crystallized when Cus- todis visited the United States in 1960.3' He presented his results with the nondrainage method to a large symposium of retinal surgeons who were greatly interested in this unusual paper. Shortly thereafter, Lincoff visited the clinic of Custodis, returned to America with the zeal of a new convert, and published a series of papers popularizing the Custodis technique. 123' Lincoff and associates disseminated the nondrainage technique, but modified the Custodis procedure by replacing diathermy with cryotherapy and by replacing the polyviol sponge with silicone sponge. They reported satisfactory results in many types ofdetachments including those which were long standing, those with large breaks, posterior breaks, or inferior dialyses, as well as aphakic detachments. Subretinal Fluid 521 Their enthusiasm notwithstanding, the Lincoff group clearly specified that drainage was required in certain selected cases.40 While they were able to repair detachments in 87% of a large series with the nondrainage method, they encouraged the use of drainage in certain cases. Eyes selected for drainage included those with: (1) atrophic choroid which they felt would not readily absorb subretinal fluid; (2) massive periretinal proliferation; (3) giant tears; (4) uncertain localization of posterior retinal breaks; (5) glaucoma; (6) thin sclera and a few other rare cases such as macular breaks and eyes with very recent cataract surgery. In a series of 1,000 detachments, 752 were managed without drainage and 87% ofthose were cured by the first operation.38 Segmental buckles were generally employed but the authors noted that an encircling buckle could also be utilized without the drainage ofsubretinal fluid.4' This encircling method has also been successfully employed by others.42'43 A number of other surgeons adopted the Custodis-Lincoff procedure and reported satisfactory results. Included in these reports on nondrainage were the papers of O'Connor, Weidenthal,48 Chignell and Talbot,49'50 Chester et al,5 Friedman,52 and Scott.5 O'Connor (1976)46 indicated that it was possible to utilize intravitreal air injection with the Custodis procedure. He proposed the use of a freezing test to determine which eyes could be suitably selected for nondrainage.47 If cryotherapy produced retinal whitening, completely surrounding the break with one or more applications with a firmly applied cryoprobe, in a time interval consistent with probe temperature and scleral thickness, then O'Connor anticipated 97% success with a nondrainage buckling procedure. Weidenthal (1967)48 proposed that a detachment would be suitable for nondrainage ifthe buckle produced a definite change in the contour ofthe retina at the break site such that the retina was relatively close to the retinal pigment epithelium over the surface of the buckle. Humphrey et al, (1972)54 reported on a large series from the Massachu- setts Eye and Ear Infirmary which included 5,973 cases operated with the drainage of subretinal fluid. Complications from the drainage of subretinal fluid were noted in 7.1%. Of those, 11.2% were ultimate failures as a direct result of the complications of fluid drainage. Verdaguer (1970)55 considered this controversy and recommended omitting the drainage of subretinal fluid if the retina reattached with preoperative ocular rest or if the retinal break could be closed without drainage at the time of surgery. Most nondrainage procedures have been done with the use of episcleral silicone sponges but other materials can be used, such as solid silicone 522 Hilton or retinal gelatin.56'57 Tablante and Tolentino (1972)58 cured all oftheir 20 cases buckled with gelatin and no fluid drainage. Straatsma et al, (1972)59 proposed that the retinal pigment epithelium may act as a diffusion barrier causing slow absorption of the subretinal fluid in old detachments. Thus, the breakdown ofthe pigment epithelium with photocoagulation, diathermy, or cryotherapy might aid in the absorption ofthe fluid. They injected a peroxidase tracer into the vitreous of rabbits and subsequently examined the tissue with electron microscopy. They noted that the peroxidate passed through the retina but could not pass through the pigment epithelium. However, the peroxidate did pass through the pigment epithelium after treatment with photocoagulation. Johnson, Okun and Boniuk (1975)6 contributed an extensive review on the why, when, where and how ofsubretinal fluid drainage. They emphasized that the nondrainage procedure, which nearly always causes marked increase in the intraocular pressure, should not be used in eyes potentially intolerant ofhigh pressure such as those with recent wounds or staphy- lomata. Robertson (1979)61 studied the problem of delayed absorption of subretinal fluid after scleral buckling procedures. He called attention to the preoperative finding of large clumps of cells on the undersurface of the detached retina, "subretinal precipitates." He reported an association between this finding and a delay in the absorption of subretinal fluid in nondrainage cases. He concluded that one in four patients in whom the precipitates are seen preoperatively will have subretinal fluid persisting more than six weeks after retinal reattachment surgery.

SUBJECTS AND METHODS The purpose ofthe study was to compare two randomly selected groups of patients: one managed with surgical drainage of subretinal fluid and the other without. The major endpoint for comparison was the visual result with secondary consideration of the anatomic cure rate, number of reoperations, and the incidence of operative and postoperative complications. A detailed protocol was written after the format of Bearman.62 The protocol included the following sections: Hypothesis, Personnel, Study Site, Rationale and Background, Objectives, Design, Inclusion Criteria, Exclusion Criteria, Treatment Procedures, Methodology, Methods of Assuring the Quality of the Data Collected, Response Variables, Proce- dures for Handling Problem Cases, Maintenance of Uniform Procedures on Factors that are to be Followed Longitudinally, Informed Consent, Responsibility for Medical Management of Subjects, and Forms for Re- Subretinal Fluid 523 cording of Data. The protocol was reviewed by two statisticians and three retinal surgeons and they made several helpful suggestions which were subsequently incorporated in the final protocol. Using random permutations, the statistician prepared a randomized treatment allocation plan in which four ofeach eight sequentially entering patients received drainage, and the other four, no drainage. One hundred and twenty envelopes were prepared to designate the assignment to the two groups. The envelopes were arranged according to the pattern of randomized numbers and each envelope was then numbered consecu- tively. Each patient seen with rhegmatogenous retinal detachment was examined in a standardized manner. The examination included: ocular history, general medical history, visual acuity-corrected and uncorrected, external examination, motility, pupils, refraction (generally taken from the patient's spectacles), intraocular pressure, biomicroscopy, and bilateral fundus drawings with binocular indirect ophthalmoscopy and scleral depression. Optional tests included three mirror biomicroscopy, gonios- copy, and perimetry. The data were then evaluated for inclusion and exclusion criteria and a decision was made regarding the admissibility of the patient to the study.

All rhegmatogenous detachments were included in the study except the following: (a) detachments which should not be operated, such as those with no light perception, phthisis bulbi, or marked disorganization from far ad- vanced massive periretinal proliferation. (b) detachments that required scleral buckling but were too flat to consider drainage. (c) special complex detachments where it is generally agreed that the drainage of subretinal fluid is mandatory. (1) massive periretinal proliferation (2) giant tears (3) very bullous detachments with posterior breaks where it is not possible to make an accurate localization. (4) detachments where there is an inferior break which could not be closed with a scleral depressor, in patients that are over 65 years ofage or in patients with more than six diop- ters of myopia (Lincoff4 has proposed that the pigment epithelium and choroid in older or myopic eyes may not absorb the subretinal fluid readily). 524 Hilton (5) eyes with very weak sclera which could not tolerate elevation ofintraocular pressure, such as those with recent cataract surgery. (6) primary retinal detachments which appeared to be of the rhegmatogenous type but where no breaks were seen (min- ute, inapparent breaks were assumed to be present). (7) detachments caused by macular breaks. Patients meeting the admission criteria were given an explanation of the study and were invited to participate. The surgery was performed in a standard manner and differed only in the inclusion or exclusion of drainage. The surgical procedure started with routine preparation and draping. The lid margins were covered with an adhesive plastic drape, and the lids were separated with a wire speculum. A 360 degree conjunctival peritomy was made at the limbus with two radial incisions ofthe conjunctiva and Tenon's capsule at 3 o'clock and 9 o'clock meridians. The rectus muscles were secured with traction sutures and the sclera was examined. The fundus was then examined with binocular indirect ophthalmoscopy and scleral depression, and all findings were compared with the preoperative drawing. The retinal breaks were localized on the surface of the sclera with a pressure type scleral marker, and the mark was made permanent with the light touch of an ophthalmic cautery on the surface of the sclera. All breaks were treated with transcleral cryotherapy as monitored with indirect ophthalmoscopy. The end point for freezing was the first appearance of the ice ball in the fundus; the ball was allowed to enlarge for no more than one second. Episcleral silicone was used for buckling; the protocol allowed the option of either segmental or encircling buckles. Either solid or sponge silicone was utilized. The envelope was opened in the operating room and those eyes that were randomized for drainage were managed with the two-step technique: sclerotomy was generally made at a site that could be covered by the silicone buckle but occasionally it was placed posteriorly. A 3 mm sclerotomy exposed the choroid which was then examined with transillumination and magnification. A + 20 diopter condensing lens was used as a magnifying loupe. An attempt was made to avoid apparent choroidal vessels with the perforating needle. A small cataract suture needle was used. Care was taken to avoid any undue elevation of the intraocular pressure from traction sutures, the speculum or retractor. Ifpossible, the sclerotomy was placed just superior or inferior to one of the horizontal rectus muscles and care was taken to avoid the area of the vortex veins. After drainage the sclerotomy was either covered by the silicone buckle Subretinal Fluid 525 or closed with a suture. The protocol allowed for the injection ofbalanced salt solution or air into the vitreous cavity as indicated. The main indi- cation for such an injection was marked hypotony often accompanied by partial collapse ofthe globe. The episcleral space ofall four quadrants was irrigated with 10 ml of neomycin, gramicidin and polymyxin B. The wound of Tenon's capsule and the conjunctiva was closed with several interrupted 5-0 plain gut sutures. All surgery was done with general anesthesia. The details of the operative procedure were recorded on the appropri- ate form. This included an estimation of the depth of the subretinal fluid at the break site at the end of the operation. The eye was examined daily for three days and the patient was generally discharged from the hospital on the fourth postoperative day. On each day, a record was made of the clarity of the ocular media, the amount of subretinal fluid, and whether or not the retinal breaks were closed. Ifone or more breaks were still open an estimate of the depth of the subretinal fluid at the break site was made. Any complications were recorded. After discharge, the patient had follow-up examinations at ten days, three weeks, seven weeks, three months, and six months postoperatively. On each occasion the following data were recorded: interval history, visual acuity (with old glasses and pinhole), intraocular pressure, biomicroscopy, and ophthalmoscopy. The final visual acuity was determined six months postoperatively in a double masked fashion. Neither the patient nor the refractionist knew which eyes had been drained. The refractionist did not know which eye had been operated on. The visual acuity was measured with the American Optical Company Illiterate E Projecto Chart. The corrected visual acuity, as determined with refraction, allowed for two incorrect responses on lines 20/15 to 20/60. One incorrect response was allowed on the 20/70 and 20/80 lines and no incorrect responses were allowed on the 20/100, 20/200, and 20/400 lines. Visual acuity less than 20/400 was determined with a 20/200 E held at the distance from the patient at which it could barely be discerned, recording it as 8/200, 6/200, 4/200, 2/200 or 1/200. Vision below 1/200 was recorded as either hand movements, light perception or no light perception. The statistical analysis primarily utilized independent sample tests for comparing the two treatment groups. Fisher's exact test and the T-test for independent samples were employed. For the analysis of the Complica- tions Score, the pairs of the total score values from the ten clinical consultants were analyzed by the paired sample T-test. 526 Hilton Ifthe retina was not reattached with the initial procedure, one or more reoperations were utilized if the detachment was still judged to be oper- able. For reoperations, the protocol allowed the surgeon to utilize any indicated technique which included either drainage or nondrainage. The protocol specified that in cases initially operated with nondrainage, that a reoperation would be indicated if the retinal break was still definitely open and if there was no decrease in the volume of subretinal fluid after the first four postoperative days. If some decrease of subretinal fluid had been noted, then further observation was warranted. Because complications are infrequent and because they vary in their significance, a new method of evaluation was utilized for complications. Our consultant in biostatistics recommended that we determine a "Com- plications Score" for each eye that had any complications. A grading system, on a scale of zero to six was established. A description of all complications for each study patient was distributed to a panel of ten expert consultants who had extensive background in retinal detachment surgery. Each consultant was asked to assign a score for each eye and a total score was then determined for the dr-ainage group and the nondrainage group. The average score from the ten consultants was then compared

TABLE I: THE COMPLICATIONS SCORE (DEFINITIONS USED BY THE PANEL OF EXPERT CONSULTANTS IN ASSIGNING A NUMERICAL SCORE TO EACH EYE THAT HAD COMPLICATIONS) SCORE DEFINITION EXAMPLE 0 Any "complication" which requires no Small subretinal hemorrhage near action on the part of the surgeon, the equator has no effect on the postoperative course and does not influence the final anatomic or visual result 1 Any complication that requires reme- Needle perforation of the retina dial action at the time of the ini- managed with cryotherapy tial operation but which has no and scleral buckling detrimental effect during the postoperative period 2 Any complication which requires post- Glaucoma, diplopia, or a leak- operative medical, optical or photo- ing retinal break that is coagulation treatment but no fur- almost closed ther surgery 3 Any complication requiring reop- Persistent detachment due to eration misplaced buckle 4 Any complication which decreases the Macular pucker, cataract or un- inal visual result resolved vitreous hemorrhage 5 Any complication which results in Massive periretinal pro- anatomic failure despite reoper- liferation ation 6 Any complication which requires enu- Intractable neovascular cleation glaucoma Subretinal Fluid 527

TABLE II: RHEGMATOGENOUS RETINAL DETACHMENTS (NOVEMBER 1977 TO AUGUST 1979) Admitted to study 120 Not admitted 87 Total 207 for the two groups. The definitions utilized in scoring are shown in Table I. The diagnosis ofeach complication was made with the concurrence ofat least two observers, the principal investigator and one additional col- league on the Retina Service. The final reattachment rate, or "cure," was defined as reattachment of the retina for at least six months follow-up. This is in accordance with the definition introduced by the committee on Retinal Detachment Surgery appointed by the American Academy of Ophthalmology and Oto- laryngology. 63

RESULTS The time span of the study extended from November, 1977 to August, 1979. During this period, a total of 207 patients with rhegmatogenous retinal detachment were referred for examination. By applying the provisions of the protocol, 87 cases were excluded leaving 120 cases for randomization (Tables II & III).

TABLE III: RHEGMATOGENOUS DETACHMENTS NOT ADMITTED TO THE STUDY (IN ACCORDANCE WITH THE PROVISIONS OF THE PROTOCOL) A. Mandatory Drainage (28 eyes): 1. Massive periretinal proliferation 14 2. No apparent breaks 5 3. Giant breaks 3 4. Bullous detachment with posterior breaks 3 5. Recent cataract surgery 1 6. Retinal and choroidal detachment 1 7. Inferior break, age 72 1 B. Mandatory Nondrainage Very shallow detachment 20 C. Patient refused to be randomized 1 D. Required vitrectomy as the initial procedure 20 E. Examined but not operated (18 eyes): 1. Referred for opinion only 14 2. Inoperable massive periretinal proliferation 3 3. Patient refused surgery 1 Total 87 528 Hilton

Twenty-eight of the excluded eyes met the criteria for mandatory drainage and included such cases as massive periretinal proliferation, primary detachments without apparent breaks, giant breaks, or bullous detachments with posterior breaks (Table III). There were 20 cases with very shallow subretinal fluid requiring nondrainage procedures and they were therefore excluded. Twenty ofthe cases were unusually complicated with extensive intravitreal and/or preretinal membranes; these were managed with vitrectomy as the initial procedure. Eighteen of the eyes were examined but not operated upon. Most ofthese were referred for an opinion only. How good was the process of randomization? To answer this question, the two groups were compared for 22 preoperative features (Table IV). These included such variables as age, sex, ocular trauma, duration of detachment, aphakia, pseudophakia, high myopia, and vitreous hemorrhage. The incidence of each of these features in the two groups was analyzed statistically and in no case was there a significant difference (p < 0.05). It was concluded that the process of randomization was quite adequate in yielding two similar groups for comparison.

TABLE IV: PREOPERATIVE FEATURES DRAINAGE GROUP NONDRAINAGE SIGNIFICANT 60 Eyes GROUP DIFFERENCE? 60 Eyes 1. Median age 62 60 no (p = 0.31) 2. Male 37 34 no (p = 0.35) 3. Direct trauma 4 1 no (p = 0.18) 4. Detachment duration < 7 days 23 26 no (p = 0.35) 5. Detachment duration > 30 days 17 20 no (p = 0.35) 6. Aphakia 19 25 no (p = 0.17) 7. Intraocular lens 2 4 no (p > 0.20) 8. Previous retinal surgery 1 3 no (p > 0.20) 9. Myopia > 6 D 8 9 no (p > 0.20) 10. Vitreous hemorrhage 3 7 no (p = 0.16) 11. One quadrant detachment 6 9 no (p = 0.29) 12. Four quadrant detachment 10 7 no (p = 0.30) 13. Macula detached 39 36 no (p > 0.20) 14. Macula detached 1-3 days 12 11 no (p > 0.20) 15. Macula detached 4-13 days 17 10 no (p = 0.09) 16. Macula detached 14-30 days 4 7 no (p > 0.20) 17. Macula detached > 30 days 6 8 no (p > 0.20) 18. Star folds 6 3 no (p > 0.20) 19. Demarcation lines 11 8 no (p > 0.20) 20. Average number of breaks 1.8 1.9 no (p = 0.62) 21. Breaks larger than 1 clock hour 5 9 no (p = 0.19) 22. Breaks : 16 mm posterior to limbus 9 11 no (p > 0.20) Subretinal Fluid 529 The anatomic results are summarized in Table V. Of major importance is the reattachment rate after just one operation. This is the true comparison of the two techniques because the surgeon had the option to use drainage in any reoperations. Fifty-two of the 60 eyes operated on with drainage were reattached with the initial procedure and 50 of the 60 eyes operated on without drainage were reattached. There is no statistically significant difference between these two reattachment rates. The final result, after reoperations, was 97% reattachment for both groups.

TABLE V: ANATOMIC RESULTS NUMBER OF OPERATIONS EYES REATTACHED DRAINAGE NONDRAINAGE GROUP GROUP SIGNIFICANT 60 Eyes 60 Eyes DIFFERENCE? 1 52 (87%) 50 (83%) no (p = 0.61) 2 5 (8%) 7 (12%) 3 1 (2%) 1 (2%) Total 58 (97%) 58 (97%) no (p = 1.00)

The most important cause offailure of the initial operation was postoperative massive periretinal proliferation, three cases in each group. Less important causes are summarized in Table VI. The cause of final failure in the four eyes that did not respond to reoperation(s) was also massive periretinal proliferation.

TABLE VI: CAUSES OF FAILURE OF THE INITIAL OPERATION DRAINAGE GROUP NONDRAINAGE GROUP 1. Persistent fluid with open break 2 4 2. Persistent fluid without open break 1 3 3. New spontaneous breaks 2 0 4. Massive periretinal proliferation 3 3 Total 8 10

The protocol defined postoperative visual acuity as the major endpoint for comparison. The acuity was studied from three different points of view. First, the visual acuity was measured six months after surgery in all reattached eyes. In each group, acuity in the range of 20/20 to 20/50, 530 Hilton occurred in 31 out of 58 eyes. Vision less than 20/400 was found in nine eyes of the drainage group and seven of the nondrainage group. There were no significant differences between the two groups (Table VII).

TABLE VII: POSTOPERATIVE VISUAL ACUITY (ALL REATTACHED EYES) DRAINAGE GROUP NONDRAINAGE GROUP SIGNIFICANT DIFFERENCE? 20/20 to 20/50 31 31 no (p = 1.00) 20/60 to 20/100 10 12 20/200 to 20/400 8 8 Less than 20/400 9 7 no (p = 0.39) Total 58 58

Secondly, the visual results were analyzed for all eyes in which the macula had been detached preoperatively and in which the retina was subsequently reattached (Table VIII). Fourteen of 39 such eyes in the drainage group obtained 20/20 to 20/50 vision compared to ten of 36 eyes in the nondrainage group. Nine eyes in the drainage group were less than 20/400 compared to eight in the nondrainage group. The differences between the two groups were again found not to be significant.

TABLE VIII: POSTOPERATIVE VISUAL ACUITY (ALL EYES WITH PREOPERATIVE DETACHMENT OF THE MACULA AND POSTOPERATIVE REATTACHMENT OF THE RETINA.) DRAINAGE GROUP NONDRAINAGE GROUP SIGNIFICANT DIFFERENCE? 20/20 to 20/50 14 10 no (p = 0.31) 20/60 to 20/100 8 12 20/200 to 20/400 8 6 Less than 20/400 9 8 no (p = 0.58) Total 39 36

Thirdly, the postoperative visual acuity was compared in all reattached eyes which had a history of good visual acuity before the onset of the detachment (Table IX). This group excluded all cases of pre-existent amblyopia, cataract, macular disease and so forth. Visual acuity of20/20 to 20/50 occurred in 28 of 49 such eyes in the drainage group and 28 of 47 eyes in the nondrainage group. Seven ofthe nondrainage group obtained final vision ofless than 20/400 compared to five in the nondrainage group. These differences were not statistically significant. Subretinal Fluid 531

TABLE IX: POSTOPERATIVE VISUAL ACUITY (ALL REATTACHED EYES WITH A HISTORY OF GOOD VISUAL ACUITY BEFORE THE ONSET OF THE DETACHMENT [EXCLUDES CASES OF PREDETACHMENT AMBLYOPIA, CATARACT, MACULAR DISEASE, ETC.]) DRAINAGE GROUP NONDRAINAGE GROUP SIGNIFICANT DIFFERENCE? 20/20 to 20/50 28 28 no (p = 0.80) 20/60 to 20/100 8 9 20/200 to 20/400 6 5 Less than 20/400 7 5 no (p = 0.41) Total 49 47

Seventeen types of complications were observed. The incidence in each group is summarized in Table X. These data, together with supplemental details regarding outcome, were distributed to ten expert consultants who assigned a numerical score for each eye that had complications. The average Complications Score for the drainage group was 33.0 compared to 29.5 for the nondrainage group (Table XI). This difference was found to be statistically significant (p = 0.005).

TABLE X: COMPLICATIONS COMPLICATIONS DRAINAGE GROUP NONDRAINAGE GROUP 1. Subretinal hemorrhage, peripheral 5 1 2. Vitreous hemorrhage 1 0 3. Anterior chamber hemorrhage 0 1 4. Hemorrhagic choroidal detachment 1 0 5. Subretinal suture 1 0 6. Retinal perforation with suture needle 2 1 7. Retinal break at drainage site 2 0 8. Small dehiscence of recent cataract wound 1 0 9. Unplanned drainage with suture needle 0 3 10. Avulsion of medial rectus 1 0 11. Choroidal detachment 3 1 12. Macular pucker 2 1 13. Massive periretinal proliferation 4 3 14. Transient glaucoma 1 0 15. Optic atrophy 1 0 16. Enophthalmos 1 0 17. Diplopia, persistent 1 2

In addition to the results given above, which were called for by the protocol, it is possible to answer an additional question by further analysis ofthe data. What was the effect ofthe depth ofthe subretinal fluid, at the break site, at the end ofthe operation, on the anatomic reattachment rate 532 Hilton

TABLE XI: COMPLICATIONS SCORE (AS JUDGED BY A PANEL OF TEN EXPERT CONSULTANTS) DIFFERENCE: DRAIN- CONSULTANT DRAINAGE GROUP NONDRAINAGE GROUP AGE-NONDRAINAGE 1 33 28 5 2 38 30 8 3 32 26 6 4 33 30 3 5 30 28 2 6 35 30 5 7 31 30 1 8 30 24 6 9 33 35 -2 10 35 34 1 Total 330 295 35 Average ± S.D. 33.0 + 2.5 29.5 ± 3.3 3.5 ± 3.0 The average Complication Score of the Drainage Group is significantly higher than the average score of the Nondrainage Group (p = 0.005). in the 120 eyes? As noted in Table XIIA, the cure rate was 86% in those cases where the break was completely closed at the time of surgery. If there was 1 mm offluid between the retinal break site and the underlying buckle, the cure rate was found to be 88%. With 2 mm offluid it was 79%, with 3 mm offluid it was 75%, and the one case in which the retinal break was separated from the underlying buckle by 4 mm of fluid was complete-

TABLE XII: ANATOMIC RESULTS (THE EFFECT OF THE DEPTH OF THE SUBRETINAL FLUID, AT THE RETINAL BREAK SITE, AT THE END OF THE INITIAL OPERATION.) A. Raw data Estimated depth of the subretinal fluid 0 1 MM 2MM 3 MM 4MM TOTAL Eyes operated 66 25 24 4 1 120 Eyes reattached 57 22 19 3 1 102 Percent reattached 86% 88% 79% 75% 100% 85% B. Consolidated data for testing of significance with the Fisher Exact Test Estimated depth of the subretinal fluid

<2MM - 2 MM TOTAL Reattached 79 (87%) 23 (79%) 102 Not reattached 12 (13%) 6 (21%) 18 Eyes operated 91 (100%) 29 (100%) 120 There is no significant difference in the two groups (p = 0.24). Subretinal Fluid 533 ly reattached after 24 hours. In order to test for possible significance, the data was grouped to compare all eyes with less than 2 mm of fluid at the break site with those which had 2 or more mm of fluid (Table XIIB). The cure rate for the first group was 87% and for the second it was 79%. When analyzed with the Fisher Exact Test, this difference was not statistically significant (p = 0.24). Excellent patient cooperation was noted throughout. Only one patient refused to be randomized, and all 120 patients were available for complete follow-up examinations for six months. The average duration of the drainage operation was 65 minutes and the average for the nondrainage operation was 67 minutes.

DISCUSSION The findings reported in this study,are applicable to the great majority of scleral buckling cases. Out of the total of 207 eyes which were examined, a buckling procedure was indicated in 169 (Table III). Ofthese, 120 (71%) met the admission criteria. Therefore the results of this study apply to approximately three out of four cases. There is little or no controversy regarding the criteria for exclusion. Even the ardent believers in frequent drainage acknowledge that the hazards of drainage increase in very flat detachments and that the nondrainage method is preferred in these cases. Ofcourse, the proponents of drainage agree with the Lincoff school that the previously mentioned complex cases should have drainage. This type of study would be very difficult to conduct with a group of surgeons. The intersurgeon variation of skill would introduce an unmea- surable variable. This problem was eliminated by having all surgery performed by one surgeon who had experience with both techniques. Many different techniques for the drainage of subretinal fluid have been previously described. The most widely used technique is a two-step procedure consisting of a sclerotomy followed by perforation of the choroid with a pin or needle. An infrequently used technique is the use of multiple drainage pins as described by Pischel.30 The Cibis technique, drainage with ophthalmoscopic observation of the tip of the drainage instrument as seen through the detached retina, has been modified and perpetuated by Wilder, and evaluated by Okun and Johnson.'6465 The internal drainage technique, with a trans-pars plana needle, was originally described by Cibis and has been recently popularized by Charles.64 6 534 Hilton A variety of instruments have been utilized for the perforation of the choroid; they include a hot penetrating diathermy pin, a cold diathermy pin, a catholysis needle, a lacrimal punctum dilator, a disposable ophthalmic cautery, a suture needle, and a von Graefe knife. Freeman et al67 68 have emphasized the use of transillumination and diathermy of the exposed choroid in an effort to minimize intraocular hemorrhaging. Bovino69 has proposed the use of a scleral spreading forcep to facilitate exposure of the choroid. Gartner70'7' has reported on a rather extensive experience with drainage of subretinal fluid with the aid of an operating microscope and he also utilizes a sclerotomy retractor. Rodriguez72 has reviewed the use of intraocular injections of balanced salt solution and/or sterile air as an aid in the release of subretinal fluid. Spalter73 has described the technique of "traction release of subretinal fluid" in which a twist grip device is used to moderately elevate the sclera as the choroid is perforated. Several rather comprehensive reviews of the subject of fluid drainage have been contributed by Schwartz,30 Ferguson,3 Johnson et al,60 and Eifrig. 74 From this great variety of potential techniques, we selected a simple two-step procedure for the present study. This is the method unan- imously described in eight contemporary textbooks of retinal detachment surgery.56'75-81 The technique consists of a careful incision of the sclera down to the surface of the choroid, exposing two or three millimeters of the choroid which is subsequently perforated with a sharp device such as a diathermy pin or a suture needle. The present study makes use of a new method for the evaluation of complications. The statistician proposed the utilization of a "Complica- tions Score" in an attempt to deal with the fact that the many types of complications vary considerably in their impact, some being virtually negligible and others disastrous. The assessment ofthe complications was done by a panel of ten expert consultants. They were drawn from the roster of the Retina Society and represented an extensive fund of experience in retinal detachment surgery. The relatively small standard devi- ation for the mean scores reflects the fact that the consultants were generally in rather close agreement with one another (Table XI). Virtually all the consultants agreed that a small subretinal hemorrhage near the equator should be given a score ofzero and persistent massive periretinal proliferation should warrant a score of 5. The most serious complications, causing more than the expected amount of permanent visual deficit, included massive periretinal proliferation, macular pucker, and optic atrophy. There were seven (5.8%) cases of massive periretinal proliferation and this is comparable to that Subretinal Fluid 535 noted in other series.',82 There were three cases of macular pucker and just one case ofoptic atrophy, also in agreement with previously reported experience. Francois and Verbraeken' have suggested that fluid drainage increases the incidence ofmacular pucker, but our data failed to show any specific relationship. In addition to the complications of drainage noted in the present study, others have reported retinal incarceration, the loss of formed vitreous and endophthalmitis. A frequently stated concern about the complications ofnondrainage is that this operation may require more reoperations with increased hospitalization, morbidity, and expense. The present study failed to reveal a significant increase in reoperations. There were also no instances of other possible complications from nondrainage, such as occlusion of the central retinal artery from marked elevation of intraocular pressure, infection associated with the epibulbar silicone sponge, and tearing ofthe sclera by sutures stressed by elevated intraocular pressure. The anatomic reattachment rate for the initial operation, 87% and 83% for the drainage and nondrainage groups respectively, is comparable to that in Qther series. 1,82 The excellent final cure rate after reoperations, 97% for both groups, is what might be expected with the elimination of such problems as massive periretinal proliferation and giant breaks. Analysis ofTable VI indicates that only four of the failures in the initial nondrainage operation could possibly be attributed to the lack of drainage. In each there was persistent fluid with an open retinal break. Drain- age, of course, does iot guarantee that all breaks will be closed per- manently and a simlar problem of persistent fluid with open breaks was found in two eyes of the drainage group. The other reasons for failure noted in Table VI do not appear to be directly related to the presence or absence of the drainage step. Intuitively one might feel that eyes in which the breaks are closed, or almost closed (less than 2 mm of subretinal fluid at the break site) should achieve a better cure rate than eyes in which the breaks are still wide open (2 or more mm of fluid at the break site). The cure rate for the "closed, or almost closed" group was 87%, compared to 79% for the group with "wide open" breaks. This difference suggests a trend, but it is not statistically significant (Table XII). The protocol stipulated that the major criterion for the comparison of the two techniques should be best corrected visual acuity measured six months after the operation. The visual results were analyzed from three different points of view. In each case it was noted that there was no significant difference in the visual results between the two groups. 536 Hilton Machemer84'85 and associates hypothesized that better visual acuity might be anticipated with a nondrainage procedure where the foveal cones were not forced back into a possibly distorted position by the act of drainage. We were unable to support this contention. A larger series, limited to cases with macular detachment, would be required for a more definitive answer to this interesting question. The experimental technique of the randomized controlled clinical trial was introduced by Hill86 in the late 1940's. He has been subsequently knighted for his contribution. His technique has now become widely recognized as an excellent method to minimize the biases inherent in clinical research.87'88 The randomized clinical trial has raised some ethical issues, but it is now generally agreed that it is more ethical to randomized patients while seeking the preferred treatment, rather than to persist indefinitely in using controversial and unproven methods. One cannot ethically investigate a clinical problem if one has a strong conviction or bias as to the preferred treatment. This was not a problem in the present study because the investigator was genuinely uncertain as to the preferred technique.

SUMMARY A randomized controlled clinical trial was conducted on a series of 120 consecutive cases of rhegmatogenous retinal detachment. One group of 60 eyes, randomly selected, was managed with a scleral buckling operation which included the drainage of subretinal fluid. The other group of 60 had a similar operation without drainage of subretinal fluid. A new method for evaluating complications, the "Complications Score," was devised. The score was significantly higher for the drainage group but this did not decrease the final anatomic or visual results. Eighty-seven percent of the drainage group were reattached with one operation, a figure not significantly different from the 83% reattached in the nondrainage group. The final cure rate after reoperations was 97% for both groups. The best corrected visual acuity, measured six months postoperatively, was not significantly different for the two groups. This study concludes that equally good results can be obtained with either a drainage or nondrainage operation.

REFERENCES 1. Boke W: Discussion on release of subretinal fluid. Mod Probl Ophthalmol 15:219-225, 1975. Subretinal Fluid 537 2. Martin B: Controlled release of subretinal fluid. Mod Probl Ophthalmol 15:149-153, 1975. 3. Ferguson EC: Drainage of subretinal fluid in scleral buckling. Int Ophthalmol Clin 2:181-205, 1962. 4. Norton EWD: Complications of retinal detachment surgery, in Symposium on Retina and Retinal Surgery. Trans New Orleans Acad Ophthalmol, St Louis, CV Mosby Co, 1969, pp 222-234. 5. Chawla HB: Reflections on subretinal fluid release. Trans Ophthalmol Soc UK 97:627-629, 1977. 6. Schepens CL: How should retinal breaks be closed?, in Brockhurst RJ, et al (eds): Controversy in Ophthalmology. Philadelphia, WB Saunders Co, 1977, pp 566-581. 7. Lincoff HA: Should retinal breaks be closed at the time ofsurgery?, in Brockhurst RJ, et al (eds): Controversy in Ophthalmology. Philadelphia, WB Saunders Co, 1977, pp 582-598. 8. Saint-Yves C: Les maladies des yeux. Paris, Pierre-Augustin le Mercier, 1722. 9. Beer GJ: Lehre Von Den Augenkrankheiten. Wien, Heubner und Volke, 1817. V 2, p 495. 10. Helmholtz J: cited by Friedenwald H: The history of the invention and the development of the ophthalmoscope. JAMA 38:549, 1902. 11. Coccius EA: Ueber Die Anwendung Des Augenspiegels, Nebst Angabe Eines Neuen Instrumentes. Leipzig, Muller, 1853, p 125. 12. von Graefe A: Notiz uber die ablosungen der netzhaut von der chorioidea. Graefe's Archiv Fuer Ophthalmologie 1:362-371, 1854. 13. Wecker L de: Traite des maladies du fond de l'oeil et atlas d'ophtalmoscopie. Paris, A Delavaye, 1870. 14. Leber T: Ueber die entstehung der netzhautablosung. Berichte Uber Der Versamm- lung Der Ophthalmologie Geselschaft (Stuttgart) 14:18, 1882. 15. Gonin J: The treatment of detached retina by searing the retinal tears. Arch Ophthal- mol 4:621-625, 1930. 16. Arruga H: Gonin. Am l Ophthalmol 60:659-662, 1965. 17. Gonin J: Retinal Detachment. Lausanne, Payot et Cie, 1934. 18. Weve H: Zur Behandlung der Netzhautablosung Mittels Diathermie. Berlin, S Karger, 1932. 19. Lindner K: Shortening of the eyeball for retinal detachment. Arch Ophthalmol 42:634-645, 1949. 20. Jess A: Temporare skleraleindellung als hilsfmittel bei der operation der netzhautablosung. Klin Monatsbl Augenheilkd 99:318, 1937. 21. Rosengren B: Uber die behandlung der netzhautablosung mittelst diathermie und luftinjektion in den glaskorper. Acta Ophthalmologica 16:3-42, 1938. 22. Schepens CL: A new ophthalmoscope demonstration. Trans Am Acad Ophthalmol Otolaryngol 51:298-301, 1947. 23. Girard-Teulon: Ophthalmoscopie binoculaire ou s'exercant par le concours des deux yeux associes. Ann Oculist 45:233-250, 1861. 24. Trantas A: Ophthalmoscopie diaphanoscopique totale par pression sur la scl6rotique. Archives D'Ophtalmologie 43:149-169, 1926. 25. Shapland CD: Scleral resection-lamellar. Trans Ophthalmol Soc UK 71:29-51, 1951. 26. Custodis E: Beobachtungen bei der diathermischen behandlung der netzhautablosung und ein hinweis zur therapie der amotio retinae. Ber Deutsch Ophth Gesellsch 57:227-230, 1952. 27. Custodis E: Bedeutet die plombenaufniihung auf die sklera einen fortschritt in der operativen behandlung der netzhautablosung? Ber Deutsch Ophth Gesellsch 58:102-105, 1953. 28. Custodis E: Die behandlung der netzhautablosung durch umschriebene diathermie-ko- agulation und einer mittels plombenaufniihung erzeugten eindellung der sklera im bereich des risses. Klin Monatsbl Augenheilkd 129:476-495, 1956. 538 Hilton 29. Schepens CL, Okamura ID, Brockhurst RJ: The scleral buckling procedures. I. Surgical technique and management. Arch Ophthalmol 58:797-811, 1957. 30. Schwartz A: Drainage of subretinal fluid in retinal detachment surgery. Am J Ophthal- mol 47:98-107, 1959. 31. Custodis E: Scleral buckling without excision and with polyviol implant, in Schepens CL (ed): Importance of the Vitreous Body with Special Emphasis on Reoperations. St Louis, CV Mosby Co, 1960, pp 175-182. 32. Lincoff HA, McLean J, Nano H: Cryosurgical treatment of retinal detachment. Trans Am Acad Ophthalmol Otolaryngol 68:412-432, 1964. 33. Lincoff HA, Baras I, McLean JM: Modifications to the Custodis procedure for retinal detachment. Arch Ophthalmol 73:160-163, 1965. 34. Lincoff HA, McLean JM: Modifications to the Custodis procedure. II. A new silicone implant for large tears. Am J Ophthalmol 64:877-879, 1967. 35. Lincoff HA, Kreissig I: Modifications of the Custodis procedure. VI. The treatment of retinal detachment without drainage of subretinal fluid. Trans Am Acad Ophthalmol Otolaryngol 76:1221-1233, 1971. 36. O'Connor PR, Lincoff HA: Modifications of the Custodis procedure. V. Radial scleral folding. Am J Ophthalmol 74:1110-1112, 1972. 37. Lincoff HA, Kreissig I: The treatment of retinal detachment without drainage of subretinal fluid. Trans Am Acad Ophthalmol Otolaryngol 76:1221-1232, 1972. 38. Lincoff H, Kreissig I, Goldbaum M: Reasons for failure in nondrainage operations. Mod Probl Ophthalmol 12:40-48, 1974. 39. Lincoff HA, Kreissig I, LaFranco F: Large retinal tears. Am J Ophthalmol 84:501-507, 1977. 40. Lincoff HA, Kreissig I, Goldbaum M: Selection ofpatients for nondrainage operations, in Pruett RC, Regan CDJ (eds): Retina Congress. New York, Appleton-Century-Crofts, 1972, pp 397-412. 41. Lincoff HA, Ramirez V, Kreissig I, et al: Encircling operations without drainage of subretinal fluid. Mod Probl Ophthalmol 15:188-196, 1975. 42. Siam AL: Encircling silicone rod without drainage for retinal detachment with giant breaks. Br J Ophthalmol 57:537-541, 1973. 43. Mortada A: Non-drainage of subretinal fluid with the encircling silastic 3 mm band in retinal detachment surgery. Bull Ophthalmol Soc Egypt 69:361-368, 1976. 44. O'Connor PR: Absorption of subretinal fluid after external scleral buckling without drainage. Am J Ophthalmol 76:30-34, 1973. 45. External buckling without drainage for selected detachments in aphakic eyes. Am J Ophthalmol 82:358-364, 1976. 46. Intravitreous air injection and the Custodis procedure. Ophthalmic Surg 7: 86-89, 1976. 47. : External buckling without drainage. Int Ophthalmol Clin 16:107-126, 1976. 48. Weidenthal DT: Retinal reattachment without release of subretinal fluid. Am J Oph- thalmol 63:108-112, 1967. 49. Chignell AH: Retinal detachment surgery without drainage of subretinal fluid. Trans Ophthalmol Soc UK 93:355-362, 1973. 50. Chignell AH, Talbot J: Absorption of subretinal fluid after nondrainage retinal detachment surgery. Arch Ophthalmol 96:635-637, 1978. 51. Chester GH, Leaver PK, Saunders SH: Factors influencing absorption of subretinal fluid. Mod Probl Ophthalmol 18:348-356, 1977. 52. Friedman M: The surgical treatment of retinal detachment without drainage of subretinal fluid. Isr J Med Sci 8:1443, 1972. 53. Scott JD: Retinal detachment surgery without drainage. Trans Ophthalmol Soc UK 90:57-77, 1970. 54. Humphrey WT, Schepens CL, Elzeneiny IL, et al: The release of subretinal fluid and its complications, in Pruett RC, Regan CDJ (eds): Retina Congress. New York, Apple- ton-Century-Crofts, 1972, pp 383-390. Subretinal Fluid 539 55. Verdaguer J: To drain or not to drain subretinal fluid. Highlights of Ophthalmol 13: 121-128, 1970. 56. Hilton GF, McLean EB, Norton EWD: Retinal Detachment. San Francisco, Am Acad Ophthalmol 1979, pp 83-87, 112. 57. Wilson RS: Experience with episcleral gelatin as a buckling technique. Presented at the Retina Society Meeting, Houston, TX, Nov 10, 1978. 58. Tablante RT, Tolentino Fl: The use of retinal gelfilm in detachment surgery without subretinal fluid drainage, in Pruett RC, Regan CDJ (eds): Retina Congress. New York, Appleton-Century-Crofts, 1972, pp 355-368. 59. Straatsma BR: Objectives of subretinal fluid drainage. Highlights Ophthalmol 13: 112-120, 1972. 60. Johnson GD, Okun E, Boniuk I, et al: Drainage ofsubretinal fluid: Why, when, where, and how. Mod Probl Ophthalmol 15:197-206, 1975. 61. Robertson DM: Delayed absorption of subretinal fluid after scleral buckling procedures. Am J Ophthalmol 87:57-64, 1979. 62. Bearman JE: Writing the protocol for a clinical trial. Am J Ophthalmol 79:775-778, 1975. 63. Post LT: Introduction to first edition and definitions of terms, in Pischel DK (ed): Retinal Detachment. Omaha, Douglas Printing Co, 1965, p 9. 64. Cibis PA: Vitreoretinal Pathology and Surgery In Retinal Detachment. St Louis, CV Mosby Co, 1965. 65. Okun E, Johnson GP: Drainage of subretinal fluid, a comparison of external and internal observation techniques. Doc Ophthalmol 25:181-188, 1980. 66. Charles S: Fluid-gas exchange in the vitreous cavity. Ocutome Newsletter 2:1-2, 1977. 67. Freeman HM, Humphrey WT, Schepens CL, et al: Transillumination and choroidal diathermy in the drainage of subretinal fluid, in Pruett RC, Regan CDJ (eds): Retina Congress. New York, Appleton-Century-Crofts, 1972, pp 391-396. 68. Freeman HM, Schepens CL: Innovations in technique for drainage of subretinal fluid: Transillumination and choroidal diathermy. Trans Am Acad Ophthalmol Otolaryngol 78:829-836, 1974. 69. Bovino JA: Scleral spreading forceps for drainage of subretinal fluid. Am J Ophthalmol 87:721, 1979. 70. Gartner J: Release ofsubretinal fluid with the aid ofthe microscope, report of 100 cases. Mod Probl Ophthalmol 15:127-133, 1975. 71. Gartner J: Microsurgical technique for drainage of subretinal fluid with the aid of a sclerotomy retractor. Mod Probl Ophthalmol 18:344-347, 1977. 72. Rodriguez A: Intraocular injection of balanced salt solution and/or air as aids in the release of subretinal fluid. Mod Probl Ophthalmol 15:158-162, 1975. 73. Spalter HF: Traction release of subretinal fluid. Mod Probl Ophthalmol 15:163-168, 1975. 74. Eifrig DE: Techniques for drainage of subretinal fluid. Surv Ophthalmol 17:348-356, 1973. 75. Havener WH, Gloeckner S: Atlas of Diagnostic Techniques and Treatment of Retinal Detachment. St Louis, CV Mosby Co, 1967, pp 126-133. 76. Chawla HB: Retinal Detachment, The Essentials ofManagement. Edinburgh, Churchill Livingstone, 1974, pp 31-32. 77. Ming ALS: Fison's Retinal Detachment Surgery. Singapore, PG Medical Books, 1977, pp 33-36. 78. Morse PH: Vitreoretinal Disease. Chicago, Year Book Medical Publishers, 1979, pp 180-186. 79. Bonnet M (adapted by Brucker AJ): Microsurgery of Retinal Detachment. New York, Masson Publishing USA, 1980, pp 41-44. 80. Benson WE: Retinal Detachment, Diagnosis and Management. Cambridge, Harper & Row, 1980, pp 119-126. 540 Hilton 81. Chignell AH: Retinal Detachment Surgery. Berlin, Springer-Verlag, 1980, pp 123-124. 82. Rachal WF, Burton TC: Changing concepts offailures after retinal detachment surgery. Arch Ophthalmol 97:480-483, 1979. 83. Francois J, Verbraeken H: Relationship between the drainage of the subretinal fluid in retinal detachment surgery and the appearance of macular pucker. Ophthalmologica, Basel 179:111-114, 1979. 84. Machemer R, Aaberg TM, Norton EWD: Giant retinal tears. II. Experimental produc- tion and management with intravitreal air. Am J Ophthalmol 68:1022-1029, 1969. 85. Machemer R: Letter to the Editor. Choroidal folds. Am J Ophthalmol 87:252, 1979. 86. Hill AB: Statistical Methods in Clinical and Preventive Medicine. London, Churchill Livingstone, 1962. 87. Ederer F, Kupfer C, Chalmers TC, et al: The randomized controlled clinical trial. AmJ Ophthalmol 79:752-789, 1975. 88. Ederer F: The randomized clinical trial, in Phillips CI, Wolfe JN (eds): Clinical Practice and Economics. Bath, Avon, The Pitman Press, 1977, pp 3-11.