More Than 40% of the PERK patients went on to be hyperopic and the hyperopic shift never stops!

"Several years earlier, when radial keratotomy (RK) a procedure to change the shape of the cornea and allow nearsighted people to “throw away” glasses was new, Waring had pioneered a study to be sure it would be safe for patients to undergo. Then and now, any surgeon in the country could try any new operation he wanted, as long as the patient could be talked into being a guinea pig and the hospital allowed it. Federal regulations for evaluating new surgical procedures do not exist as they do for new drugs.

Even though RK was considered cosmetic in nature and was not covered by insurance policies, Waring succeeded in obtaining a nation wide moratorium on RK procedures until he and his academic colleagues at Emory and centers around the nation could study the operation. They would perform the operation in a uniform manner and follow the patients carefully, noting the success rate as well as complications. If RK proved to be safe and effective, then any surgeon trained in the technique would be able to go ahead, but not until the study was completed. A lover of acronyms, Waring dubbed the study “PERK”Prospective Evaluation of Radial Keratotomy.

A number of pioneer surgeons in private practice who had already begun profiting from performing RK, sued Waring personally, Emory, and other academic centers for restraint of trade."

Dr Tom Harbin, 'Waking up Blind' - 2009

"In recognition of the need to evaluate the efficacy, safety, predictability and stability of radial keratotomy, the National Eye Institute funded the Prospective Evaluation of Radial Keratotomy (PERK) study, a five-year collaborative study with nine centers that included both university professors and private practitioners as investigators. Four hundred thirty-five patients were followed under a standardized protocol that specified eight incisions and a varied diameter of the clear zone based on the spherical equivalent of the refractive error. The effect of age was not known and thus not considered. 

The outcome of the first eye was used to plan the surgery for the second eye, so that if the first eye was overcorrected or undercorrected by more than D, the diameter of the clear zone for the second eye was increased or decreased by 0.5 mm, respectively. Repeated surgeries (enhancements) were discouraged. Corticosteroids were not used postoperatively."

Incisional Keratotomy, Dr Charles Casebeer-1995

"However, it should be remembered that PERK surgery is not aggressive radial keratotomy surgery, and therefore may not be representative of the operation as it is widely practiced today. As a corollary, if problems arise in PERK patients, the significance will be duly amplified in other groups of patients more aggressively treated."

Dr Irving Schwab, Refractive Keratoplasty - 1987

Prospective Evaluation of Radial Keratotomy (PERK)

Study Start Date:

April 1981

Study Completion Date:

October 1983

Detailed Description:

Approximately 11 million Americans have myopia that can be corrected with eyeglasses or contact lenses. Some of these people may also be candidates for radial keratotomy (RK), a procedure that aims to correct or reduce myopia by surgery that flattens the corneal curvature.

Keratotomy was first performed by surgeons in Europe and the United States in the late 1800s, and the basic optical and mechanical principles of the operation were defined in the 1940s and 1950s by the Japanese doctors T. Sato and K. Akiyama, who used anterior and posterior corneal incisions. The posterior incisions damaged the cornea, and the procedure was modified in the Soviet Union by doctors Fyodorov and V. Durnev to include incisions in only the anterior cornea. Since its introduction into the United States in 1978, numerous ophthalmologists have modified the procedure by introducing technical and surgical improvements such as ultrasonic methods to measure the thickness of the cornea and the use of diamond-bladed micrometer knives to make the incisions.

However, scientific assessment of RK lagged behind growing public and professional interest in the procedure. In 1980, in response to widespread concern about the long-term safety and efficacy of RK, a group of ophthalmic surgeons approached the National Eye Institute with a proposal for a multicenter clinical trial that would evaluate the potential benefits and risks of this procedure.

The Prospective Evaluation of Radial Keratotomy study, involving 435 patients and 99 pilot patients, was a clinical trial designed to evaluate the short- and long-term safety and efficacy of one technique of radial keratotomy. The procedure was evaluated by comparing a patient's refractive error and uncorrected vision before and after surgery. The more myopic eye received surgery first. Patients were required to wait 1 year before having the operation on the second eye.

The surgical technique was standardized, consisting of eight centrifugal radial incisions made manually with a diamond micrometer knife. The diameter of the central, uncut, clear zone was determined by the preoperative spherical equivalent cycloplegic refraction (-2.00 to -3.12 D = 4.0 mm; -3.25 to -4.3 D = 3.5 mm; -4.50 to -8.00 D = 3.0 mm). The blade length, which determined the depth of the incision, was set at 100 percent of the thinnest of four intraoperative ultrasonic corneal thickness readings taken paracentrally at the 3-, 6-, 9-, and 12-o'clock meridians just outside the mark delineating the clear zone. The incisions were made from the edge of the trephine mark to the limbal vascular arcade and were spaced equidistantly around the cornea.

Patients were examined preoperatively and after surgery at 2 weeks, 3 months, 6 months, annually for 5 years, and at 10 years. Examinations in the morning and evening of the same day were done at 3 months, 1 year, 3 years, and 11 years in a subset of the patients to test for diurnal fluctuation of vision and refraction.

The primary outcome variables measured at each visit was the uncorrected and spectacle-corrected visual acuity and the refractive error with the pupil dilated and undilated. The corneal shape was measured with central keratometry and photokeratoscopy. Endothelial function was evaluated using specular microscopy. A slit-lamp microscope examination was made to check for complications from the incisions. Contrast sensitivity was tested in a subset of patients. Patient motivation and satisfaction were studied with psychometric questionnaires at baseline, 1 year, 5-6 years, and 10 years.

Three Year Results on Radial Keratotomy Released      

Study Population

The study design specified a sample size of approximately 400 eyes in order to estimate the percentage of patients with visual acuity of 20/40 or better with a precision of about 5%. Surgery was done on the first eye of 435 patients between March 1982 and October 1983. Each of the nine clinical centers entered 43 to 50 patients into the study.

Of the 435 patients, 10 patients (2.3%) were lost to follow-up and eight additional patients (1.8%) missed a three year examination, leaving 417 patients (96%) examined approximately three years after surgery on the first operated eye. The three year follow-up interval ranged from three months before to six months after the three year anniversary after surgery; 90% of patients were examined within two months of the three-year anniversary and eight patients (2%) were examined four to six months after the three year anniversary.

We accounted for the 18 patients without a three year visit; their last follow-up visit after surgery was two weeks (2 patients), one year (8 patients), and two years (8 patients). Two of these 18 patients are no longer being sought, one who died and another who had further radial keratotomy surgery in both eyes by a physician not associated with the PERK study.

The refractive, keratometric, and visual acuity results are presented in this paper for the first operated eyes of 401 of the 417 patients examined at three years after surgery. We excluded 16 eyes (3.7%) because of surgery inconsistent with the protocol or because of events that altered the outcome of the surgery. These 16 eyes included: (1) eight eyes that received a clear zone inconsistent with the surgical protocol, seven receiving a clear zone 0.50 mm smaller than specified and one receiving a 0.50 mm larger clear zone; (2) two eyes that received a repeated radial keratotomy before six months after the initial surgery; (3) three eyes that were fitted with hard contact lenses; (4) one eye that underwent a cataract extraction before the three year follow- up visit; (5) one eye that was treated for glaucoma with timolol and had large changes in intraocular pressure and refraction; [13] and (6) one eye of a patient who received hyperbaric oxygen treatment to speed healing of a foot ulcer and who experienced a marked change in refraction.

In the presentation of complications, we have included all eyes that had surgery in the PERK study.

For the 57 eyes that had a repeated operation, we used refractive, keratometric, and visual acuity results from the follow-up visit preceding the reoperation six months (39 eyes), one year (15 eyes), and two years (3eyes) as described previously.  We reported the results in these eyes previously. There were 16 eyes excluded from the presentation of the refractive and uncorrected visual acuity outcomes. To ensure that no relevant information remains unreported,we have presented this data elsewhere.

Change in astigmatism

The change in astigmatism ranged from a decrease of 1.25 D to an increase of 2.75 D. In 41 eyes (10%), astigmatism increased by 1.00 D or more and in four eyes (1 %) the increase was 2.00 D or more. For seven eyes (2%) the astigmatism decreased by 1.00 D or more. For the 305 eyes with astigmatism at baseline and three years, the axis of the cylinder changed 0 to 20 degrees in 56%, 21 to 40 degrees in 14%, 41 to 60 degrees in 7%, and 61 to 90 degrees in 23%.

Decrease in best corrected visual acuity

Since a loss of best corrected vision is a serious complication, we have included all 435 eyes, regardless of their postoperative course, using data from follow-up visits before the three year visit for the 18 patients who missed that visit. We have interpreted the gain or loss of one Snellen line of vision as a possible normal variation and therefore considered a change of 2 lines or more as noteworthy.

At baseline, all eyes had a visual acuity of 20/20 or better with the best manifest refraction. At three years, nine eyes saw worse than 20/20 (eight saw 20/25 and one saw 20/40). In six of these eight eyes the vision decreased by one line.

Five eyes (1 %) lost 2 lines and one eye lost 3 lines of best corrected visual acuity. The eye losing three lines to 20/40 developed glaucoma after surgery; other details of this case are described in a later section. Interestingly, 16 eyes (4%) gained 2 lines in best corrected acuity after the surgery. In general there appeared to be little change in best corrected acuity three years after radial keratotomy.

Irregular astigamatism

On the keratographs of the majority of eyes, the central two or three circles were regular because they overlaid the central clear zone and the peripheral six or seven circles exhibited a slight irregularity, because they overlaid the corneal scars. We think that this irregularity is caused by slight elevation of each corneal scar with a relative slight depression both between and in the flattened center of the cornea. This topography is the anatomic basis for the creation of the stellate iron line. One patient complained of bothersome monocular diplopia that persisted five years after surgery because of a diffusely gray hypertropic scar in the 7:30 axis. The keratograph demonstrated focal pleating and irregularity over the scar. The patient insisted on treatment for his diplopia, so five and one-half years after his initial surgery, he had revision of the scar under topical anesthesia. Surface irregularity and monocular diplopia were less severe, but persistent.

Effect of intraocular pressure on refraction

Busing and colleagues reported a PERK patient whose intraocular pressure increased from 12 to 26 mm Hg two years after surgery, with flattening of the central cornea and a change in refraction from +2.00 D to +5.75 D. Topical timolol decreased the intraocular pressure to 10 mm Hg, with steepening of the cornea and lessening of the overcorrection to +1.75 D. After subsequent cessation of timolol, the intraocular pressure rose to 20 mm Hg and the refractive error again increased to +5.00 D.


Linberg and colleagues documented the appearance of ptosis in approximately 105 of eyes after radial keratotomy at one PERK center. The cause of the ptosis was not clearly determined, but compression of the levator palpebrae aponeurosis by a heavy eyelid speculum was suspected. Carroll and colleagues reported a similar case at another PERK center.

Stability of refraction

The unsutured radial keratotomy wounds in the avascular cornea heal slowly, requiring at least four to five years to completely eject the epithelial plug and to remodel the stroma adjacent to the incision scar (personal observations). Thus, at three years, we expected these wounds to produce some instability of refraction.

We have reported elsewhere persistent diurnal fluctuation of vision three years after surgery.  Fifty-two patients were examined before 10:00 a.m and after 5:30 p.m. on the same day. Between the morning and evening examinations, 31 % of the eyes had an increase in minus power of 0.50 D to 1.50 D, and 19% had a decrease in uncorrected visual acuity of 2 to 5 Snellen lines.

Wyzinski demonstrated that the largest daily changes in refraction occur soon after awakening.

Between one and three years after the surgery, 12% of the eyes changed by 1.00 D to 2.50 D, 9% showing a continued decrease in minus power, that is, a continued effect of the surgery sometimes called "progressive hyperopia". This is a much lower percentage than reported by Deitzand Sanders, sub 1.00 D or more in 31 of eyes between one and three to four years after surgery.

Other authors have presented stability results only in terms of the average change in the minus power of the refraction: Sawelson and Marks,  0.17 between 18 months and 3 years, and Arrowsmith and Marks,  0.23 D between one and two years. The average change in the PERK study was a decrease in minus power of 0.23 D. Deitz and Sanders sub reported an average decrease in minus power that was greater than in these other studies, 0.53 D between one and four years.

All evidence indicates that there is not a loss of the effect of radial keratotomy during the first years after surgery; whether there is a clinically meaningful, systematic progression of effect in individual eyes needs further study.

Ten-Year Results on Radial Keratotomy Released

National Institutes of HealthNational Eye Institute

October 13, 1994

After a decade of patient followup, a National Eye Institute (NEI)-supported study reported today that radial keratotomy (RK) remained a reasonably safe and effective technique to improve distance vision.

However, the study found that more than 40 percent of RK-operated eyes continued to have a gradual shift toward farsightedness. This finding suggests that some people who have RK may need glasses at an earlier age for poor close-up vision, a common problem after age 40, than if they had chosen not to have the surgery.

"Based on these findings, it may be that some people will be pleased with their vision shortly after having RK, but their opinion may change five, ten, or fifteen years down the road," said Peter J. McDonnell, M.D., of the Doheny Eye Institute at the University of Southern California and the study's co-chairman.

Today's findings, published in Archives of Ophthalmology, were issued from the Prospective Evaluation of Radial Keratotomy (PERK). The PERK study is the first large, well-designed clinical study to evaluate the long-term effects of radial keratotomy on the eye and vision.

RK is performed to improve poor distance vision, called myopia, which affects millions of Americans. For some people with myopia, RK offers the prospect of good distance vision without the need for glasses or Contact lenses.

The surgery changes the shape of the cornea, the clear, rounded tissue at the front of the eye. It is performed by making spoke-like, partial-thickness incisions into the healthy cornea. These wounds cause the cornea to flatten, producing clearer distance vision.

Today, about 250,000 RK surgeries are performed annually in the United States, up from 30,000 operations just five years ago. However, eye care professionals still have little scientific information about the procedure's long-term effects on the cornea and vision.

To provide these data, PERK clinicians periodically examined the eyes of the 435 participants since the study began in the early 1980s. Based on these examinations, researchers have published occasional reports in medical journals, including the results issued today.

At the PERK's 10-year mark, researchers reported that RK effectively reduced but did not completely eliminate myopia in all patients. They found that 53 percent of the RK-operated eyes registered 20/20 vision, while 85 percent of the eyes had 20/40 uncorrected vision or better (required for a driver's license in most states). Approximately 70 percent of study participants said they did not wear corrective lenses for distance vision at the 10-year mark.

Interestingly, the PERK scientists reported that 43 percent of the RK-operated eyes continued to have a gradual change toward farsightedness, called hyperopic shift. In fact, 36 percent of the eyes had become farsighted at the 10-year point.

"Typically, people who are myopic need reading glasses later in life than those who are not myopic," said Dr. McDonnell. "It may be that some people will reduce their need for glasses to see at a distance, but will need to wear reading glasses earlier than they otherwise would have needed to."

But the hyperopic shift was beneficial in some cases. In those who remained nearsighted after surgery, this gradual change caused an improvement in their vision, moving their sight closer to 20/20.

According to the researchers, this shift was detected in some affected patients as soon as six months after surgery and continued to progress a decade later. They said they do not know when and if this change will cease in the future.

The scientists noted that the shift in vision was not related to the patient's age or post-surgical outcome. They added that they could not predict based on the PERK data which patients will develop the hyperopic shift. They did note, however, that the shift was more common in those who had RK surgery using longer incisions in the cornea, a common technique in younger and/or more myopic patients.

"This clinical study provides ophthalmologists with scientifically validated information regarding the safety and effectiveness of radial keratotomy, said Carl Kupfer, M.D., director of the NEI, part of the National Institutes of Health. "With these results, prospective patients can have the best informed consent when considering radial keratotomy."

The National Eye Institute is the Federal government's lead agency for vision research.


  • Waring GO 3rd, Lynn MJ, McDonnell PJ. Results of The Prospective Evaluation of Radial Keratotomy (PERK) Study 10 years After Surgery. Arch Ophthalmol. 1994 Oct. PubMed

October 1994